Land use changes induced soil organic carbon ... - Springer Link
J Soils Sediments (2013) 13:981–988 DOI 10.1007/s11368-013-0684-4
SOILS, SEC 1 • SOIL ORGANIC MATTER DYNAMICS AND NUTRIENT CYCLING • RESEARCH ARTICLE
Land use changes induced soil organic carbon variations in agricultural soils of Fuyang County, China Lefeng Qiu & Jinxia Zhu & Yuanhong Zhu & Yang Hong & Ke Wang & Jinsong Deng
Received: 6 May 2012 / Accepted: 15 March 2013 / Published online: 29 March 2013 # Springer-Verlag Berlin Heidelberg 2013
Abstract Purpose The purpose of this study is to understand spatial and temporal variations of soil organic carbon (SOC) under rapid urbanization and support soil and environmental management. Materials and methods SOC data in 1979 and 2006, of 228 and 1,104 soil samples respectively, were collected from surface agricultural lands in Fuyang County, East of China. Land use data were also gathered at the same time. Results and discussion The mean SOC was 17.3 (±4.6) g/kg for the 1979 data and 18.5(±5.8) g/kg for 2006. There was a significant difference in SOC between the 2 years according to the t test result. Geostatistical analysis indicated that SOC had a moderate spatial correlation controlled by extrinsic anthropogenic activities. The spatial distribution of SOC, derived from ordinary kriging, matched the distribution of
industry and urbanization. Using a six-level SOC classification scheme (23.2 g/kg) created by Zhejiang Province, approximately 15 % of soil had SOC increase from low to high levels from 1979 to 2006. Conclusions The main cause of SOC variation in the study area was land use change from agriculture to industrial or urbanized uses. The increasing SOC trend near most towns may be attributed to use of organic manure, urban wastes, sewage sludge, and chemical fertilizers on agricultural land.
Responsible editor: Gilbert C. Sigua
The carbon (C) cycle in global terrestrial ecosystems has received increasing attention in recent years due to its association with climate change. On the global scale, about 1,500– 2,000 Pg C is stored in the soil organic carbon (SOC), which represents three times higher than the amount of C within plant biomass and twice the amount of C in the atmosphere (Janzen 2004). Moreover, changes in the soil organic carbon pool can influence the concentration of CO2 in the atmosphere (Smith et al. 2008). Revealing the variation of SOC and understanding its distribution is crucial for developing effective management approaches for reducing atmospheric CO2 concentrations. Many factors impact the biogeochemical cycling of SOC and, consequently, impact the distribution of SOC (Dai and Huang 2006; Eynard et al. 2005; Harradine and Jenny 1958). The effects of agricultural land use change on SOC storage have been the focus of numerous studies (Celik 2005; Kong et al. 2009; Solomon et al. 2000), while few investigations cover the influence of transitioning from the agricultural to industrial or urbanized uses on SOC in urban fringe areas. This
L. Qiu (*) Institute of Rural Development, Zhejiang Academy of Agricultural Sciences, Hangzhou 310021, China e-mail: [email protected] J. Zhu Institute of Economic and Social Development, Zhejiang University of Finance and Economics, Hangzhou 310018, China K. Wang : J. Deng (*) Institute of Remote Sensing and Information System Application, Zhejiang University, Hangzhou 310029, China e-mail: [email protected] Y. Zhu Department of Crop and Soil Sciences, The Pennsylvania State University, University Park, PA 16802, USA Y. Hong School of Civil Engineering and Environmental Sciences, University of Oklahoma, National Weather Center ARRC Suite 4610, Norman, OK 73019, USA
Keywords Distribution pattern . Geostatistics . Land use . Urbanization
1 Introduction
982
transition process, however, has resulted in significant human effects on soil quality (Chen 2007) and therefore, should not be neglected in SOC studies. As the world’s largest rapidly developing country, China has experienced a dramatic and unprecedented rate of urbanization since the initiation of economic reform in 1978. There has been a massive transfer from agricultural land use to various other land uses in eastern China due to the rapid urban expansion. At the same time, overwhelming human activity produced a huge amount of urban waste, which is rich in organic matter (Garcia et al. 1992; Pascual et al. 1997). The urban waste, if properly applied to soil, can directly modify the soil’s physical, chemical and biological properties (Lax et al. 1994). Application of industrial and municipal sewage can also change soil properties. Li et al. (2003) found that organic matter content of urban sewage sludge could be as high as 696 g/kg, with an average of 384 g/kg in China. Using conventional statistics and geostatistics, this study investigated the spatial and temporal variability of SOC in Fuyang County during the last three decades. The objectives of this study were to assess spatial and temporal SOC changes,
J Soils Sediments (2013) 13:981–988
to explore the influence of agricultural land uses, and to investigate the transition from agricultural to industrial or urbanized uses on SOC variation. The information gathered would provide a scientific basis for land management to enhance C storage and conservation in urban–rural transition areas.
2 Materials and methods 2.1 Site description The study site is located in Fuyang County, northern Zhejiang Province at the east of China (Fig. 1). This county (119°25′ 00″–120°19′30″ E, 29°44′45″–30°11′58.5″ N) has an area of 1,831 km2, with a landscape characterized by a mountain and valley topography, with elevation varying from 6 to 1068 m above sea level. The flat plains and low hilly regions, with a relative elevation of less than 150 m, were selected as our study area to minimize the influence of different topography. The study region has a total area of 860 km2. Currently in the study area, the land use types include five agricultural land use
Fig. 1 Location of study area, samples, and spatial distribution of urbanization
J Soils Sediments (2013) 13:981–988 Table 1 Summary statistics of soil organic carbon (SOC, gram per kilogram) in study area
983
Year
n
1979 2006
228 1,104
Mean ± SD
Median
Minimum
Maximum
CV
K-S test
17.3±4.6 18.5±5.7
17.7 18.0
0.7 0.6
29.8 43.9
0.27 0.31
0.207 0.085
types (paddy field, dryland, vegetable field, forests, and orchard), built-up land, water body, and vacant land. The subtropical climate (average temperature of 16.1 °C) and high precipitation (1,441.9 mm, annually) make the study area a typical rice paddy production region. The dominant soil types include clay red earth and paddy soil (Zhang et al. 2009). During the past three decades since the economic reform in 1978, urbanization and industrialization have occurred at an unprecedented pace. The urban population in the county had increased from 38,000 in 1980 to 210,000 in 2006. Urban areas had expanded from 7 km2 in 1978 to 334.6 km2 in 2006. The gross domestic product increased from 1.7 billion Chinese yuan in 1978 to 238.4 billion Chinese Yuan in 2006. Fuyang County has become one of the top 100 economically developed counties in China. Paper mills, garment plants, smelting factories, and handicraft workshops have been well-developed and played a very important role in the county’s environment (Zhang et al. 2009). The distribution of urban area is presented in Fig. 1. 2.2 Data collection Soil samples were collected in agricultural land across the study area in 1979 and 2006 to determine the temporal changes in SOC over the 27-year period. The SOC data for 228 samples taken in 1979 and records on sampling locations were obtained from the Bureau of Agriculture of Fuyang County. In 2006, the SOC was measured again for 1,104 soil samples taken in the study area (see Fig. 1). The sampling locations were collocated with the sampling locations from 1979 as exactly as possible. The land use types of sampling locations were also considered. Global positioning systems were used to precisely locate every sampling location. A total of five sampling points were collected, to a depth of up to 20 cm, within a 5-m radius of a specific sampling location; then the samples were mixed and air-dried at room temperature. Stone and plant residue in the soil samples were manually removed, and then the samples were ground to pass a 2mm sieve. SOC was determined according to the Walkley and Black method (Walkley and Black 1934).
Land use data for 2006, with information on industrial types, was obtained from Bureau of Land and Resources of Fuyang, the land administrator in Fuyang County. The land use map for 1979 was produced by photo-interpretation of 1979 Landsat MSS photographs with a spatial resolution of 30 m (Zhong et al. 2011). 2.3 Data analysis An important contribution of geostatistics is to assess the uncertainty of unobserved interpolated values (Goovaerts 1997). The semivariogram could be used to quantify the spatial variation of SOC between two points, x and x+h, as a function of their distance h: gðhÞ ¼
NðhÞ 1 X ½Z ðxi Þ Z ðxi þ hÞ 2 2N ðhÞ i¼1
ð1Þ
Where Z(xi) and Z(xi +h) represent the measured value for SOC at location xi andxi +h, y(h) is the variogram for a lag distance h, and N(h) is the number of data pairs separated by h. All semivariogram analysis was carried out using GS+® (Version 7). Based on the semivariogram results, ordinary kriging, which is the most common interpolation method, was used to estimate the unobserved SOC values. Ten percent of data points were randomly selected to test the prediction error by the mean prediction error (ME) and rootmean-square standardized prediction error (RMSSE). ME ¼
n 1X Zðxi Þ z* ðxi Þ n i¼1
ð2Þ
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi n 1X Zðxi Þ z* ðxi Þ RMSSE ¼ n i¼1 σ ð xi Þ
ð3Þ
Where Z(xi) and z*(xi) represent the measured value and predicted value (respectively) at location of xi, and σ(xi) is the standard error of prediction at location of xi.
Table 2 Semivariogram models for soil organic carbon (SOC) in study area Year
Model
1979 2006
Spherical Spherical
Range (km)
Angle direction
Nugget (C0)
Sill (C0 +C)
C0/Sill
ME
13.4 4.5
48.5 58.5
13.6 18.4
23.3 35.2
0.583 0.523
−0.035 0.012
RMSSE 0.948 0.953
984 Fig. 2 Distribution maps of soil organic carbon (SOC) content in 1979 (a) and 2006 (b) in study area, changes in SOC between 1979 and 2006 (c)
J Soils Sediments (2013) 13:981–988
J Soils Sediments (2013) 13:981–988
985
Table 3 Land areas and percentages in different soil organic carbon classes in 1979 and 2006 Time
SOC class (g/kg)
1979
Area (km2) Percentage (%) Area (km2) Percentage (%)
2006
I >23.2
II 17.4–23.2
III 11.6–17.4
IV 5.8–11.6
– – 55.8 6.5
442.9 51.4 511.6 59.4
416.2 48.3 287.4 33.4
2.9 0.3 6.5 0.7
Interpolated results were exported as a raster format and consequently resulted in the spatial distribution maps of SOC 1979 and 2006, respectively. Then, the raster calculator was used to get the distribution map of the difference in SOC between 1979 and 2006. Other spatial analyses included overlay analysis, zonal statistics, and buffer analysis. All spatial analyses were done in ArcGIS® (Version 9.2). Descriptive statistics (mean, median, minimum, maximum, and coefficient of variation) for 1979 and 2006 SOC data were obtained to get a preliminary understanding of the datasets. Normality of the datasets was analyzed using the Kolmogorov–Smirnov (K-S) test. The difference of measurement data was compared with analysis of variance and the t test. All statistical analyses were performed using SPSS® (Version 16.0).
3 Results and discussion 3.1 Descriptive statistics Descriptive statistics for SOC are shown in Table 1. A significant difference in SOC between 1979 and 2006 was detected according to the t test results (P0.05) showed that SOC in Fuyang County was normally distributed in 1979 and 2006, which is a requirement for simple geostatistical analysis. The SOC data were moderately varied, shown by the relatively small coefficients of variation (CV). Table 4 Land use changes from 1979 to 2006 in study area (square kilometer)
Year
Builtup
1979 2006
20.4 129.2
Paddy field 461.5 205.0
V 3.5–5.8
VI
SOILS, SEC 1 • SOIL ORGANIC MATTER DYNAMICS AND NUTRIENT CYCLING • RESEARCH ARTICLE
Land use changes induced soil organic carbon variations in agricultural soils of Fuyang County, China Lefeng Qiu & Jinxia Zhu & Yuanhong Zhu & Yang Hong & Ke Wang & Jinsong Deng
Received: 6 May 2012 / Accepted: 15 March 2013 / Published online: 29 March 2013 # Springer-Verlag Berlin Heidelberg 2013
Abstract Purpose The purpose of this study is to understand spatial and temporal variations of soil organic carbon (SOC) under rapid urbanization and support soil and environmental management. Materials and methods SOC data in 1979 and 2006, of 228 and 1,104 soil samples respectively, were collected from surface agricultural lands in Fuyang County, East of China. Land use data were also gathered at the same time. Results and discussion The mean SOC was 17.3 (±4.6) g/kg for the 1979 data and 18.5(±5.8) g/kg for 2006. There was a significant difference in SOC between the 2 years according to the t test result. Geostatistical analysis indicated that SOC had a moderate spatial correlation controlled by extrinsic anthropogenic activities. The spatial distribution of SOC, derived from ordinary kriging, matched the distribution of
industry and urbanization. Using a six-level SOC classification scheme (23.2 g/kg) created by Zhejiang Province, approximately 15 % of soil had SOC increase from low to high levels from 1979 to 2006. Conclusions The main cause of SOC variation in the study area was land use change from agriculture to industrial or urbanized uses. The increasing SOC trend near most towns may be attributed to use of organic manure, urban wastes, sewage sludge, and chemical fertilizers on agricultural land.
Responsible editor: Gilbert C. Sigua
The carbon (C) cycle in global terrestrial ecosystems has received increasing attention in recent years due to its association with climate change. On the global scale, about 1,500– 2,000 Pg C is stored in the soil organic carbon (SOC), which represents three times higher than the amount of C within plant biomass and twice the amount of C in the atmosphere (Janzen 2004). Moreover, changes in the soil organic carbon pool can influence the concentration of CO2 in the atmosphere (Smith et al. 2008). Revealing the variation of SOC and understanding its distribution is crucial for developing effective management approaches for reducing atmospheric CO2 concentrations. Many factors impact the biogeochemical cycling of SOC and, consequently, impact the distribution of SOC (Dai and Huang 2006; Eynard et al. 2005; Harradine and Jenny 1958). The effects of agricultural land use change on SOC storage have been the focus of numerous studies (Celik 2005; Kong et al. 2009; Solomon et al. 2000), while few investigations cover the influence of transitioning from the agricultural to industrial or urbanized uses on SOC in urban fringe areas. This
L. Qiu (*) Institute of Rural Development, Zhejiang Academy of Agricultural Sciences, Hangzhou 310021, China e-mail: [email protected] J. Zhu Institute of Economic and Social Development, Zhejiang University of Finance and Economics, Hangzhou 310018, China K. Wang : J. Deng (*) Institute of Remote Sensing and Information System Application, Zhejiang University, Hangzhou 310029, China e-mail: [email protected] Y. Zhu Department of Crop and Soil Sciences, The Pennsylvania State University, University Park, PA 16802, USA Y. Hong School of Civil Engineering and Environmental Sciences, University of Oklahoma, National Weather Center ARRC Suite 4610, Norman, OK 73019, USA
Keywords Distribution pattern . Geostatistics . Land use . Urbanization
1 Introduction
982
transition process, however, has resulted in significant human effects on soil quality (Chen 2007) and therefore, should not be neglected in SOC studies. As the world’s largest rapidly developing country, China has experienced a dramatic and unprecedented rate of urbanization since the initiation of economic reform in 1978. There has been a massive transfer from agricultural land use to various other land uses in eastern China due to the rapid urban expansion. At the same time, overwhelming human activity produced a huge amount of urban waste, which is rich in organic matter (Garcia et al. 1992; Pascual et al. 1997). The urban waste, if properly applied to soil, can directly modify the soil’s physical, chemical and biological properties (Lax et al. 1994). Application of industrial and municipal sewage can also change soil properties. Li et al. (2003) found that organic matter content of urban sewage sludge could be as high as 696 g/kg, with an average of 384 g/kg in China. Using conventional statistics and geostatistics, this study investigated the spatial and temporal variability of SOC in Fuyang County during the last three decades. The objectives of this study were to assess spatial and temporal SOC changes,
J Soils Sediments (2013) 13:981–988
to explore the influence of agricultural land uses, and to investigate the transition from agricultural to industrial or urbanized uses on SOC variation. The information gathered would provide a scientific basis for land management to enhance C storage and conservation in urban–rural transition areas.
2 Materials and methods 2.1 Site description The study site is located in Fuyang County, northern Zhejiang Province at the east of China (Fig. 1). This county (119°25′ 00″–120°19′30″ E, 29°44′45″–30°11′58.5″ N) has an area of 1,831 km2, with a landscape characterized by a mountain and valley topography, with elevation varying from 6 to 1068 m above sea level. The flat plains and low hilly regions, with a relative elevation of less than 150 m, were selected as our study area to minimize the influence of different topography. The study region has a total area of 860 km2. Currently in the study area, the land use types include five agricultural land use
Fig. 1 Location of study area, samples, and spatial distribution of urbanization
J Soils Sediments (2013) 13:981–988 Table 1 Summary statistics of soil organic carbon (SOC, gram per kilogram) in study area
983
Year
n
1979 2006
228 1,104
Mean ± SD
Median
Minimum
Maximum
CV
K-S test
17.3±4.6 18.5±5.7
17.7 18.0
0.7 0.6
29.8 43.9
0.27 0.31
0.207 0.085
types (paddy field, dryland, vegetable field, forests, and orchard), built-up land, water body, and vacant land. The subtropical climate (average temperature of 16.1 °C) and high precipitation (1,441.9 mm, annually) make the study area a typical rice paddy production region. The dominant soil types include clay red earth and paddy soil (Zhang et al. 2009). During the past three decades since the economic reform in 1978, urbanization and industrialization have occurred at an unprecedented pace. The urban population in the county had increased from 38,000 in 1980 to 210,000 in 2006. Urban areas had expanded from 7 km2 in 1978 to 334.6 km2 in 2006. The gross domestic product increased from 1.7 billion Chinese yuan in 1978 to 238.4 billion Chinese Yuan in 2006. Fuyang County has become one of the top 100 economically developed counties in China. Paper mills, garment plants, smelting factories, and handicraft workshops have been well-developed and played a very important role in the county’s environment (Zhang et al. 2009). The distribution of urban area is presented in Fig. 1. 2.2 Data collection Soil samples were collected in agricultural land across the study area in 1979 and 2006 to determine the temporal changes in SOC over the 27-year period. The SOC data for 228 samples taken in 1979 and records on sampling locations were obtained from the Bureau of Agriculture of Fuyang County. In 2006, the SOC was measured again for 1,104 soil samples taken in the study area (see Fig. 1). The sampling locations were collocated with the sampling locations from 1979 as exactly as possible. The land use types of sampling locations were also considered. Global positioning systems were used to precisely locate every sampling location. A total of five sampling points were collected, to a depth of up to 20 cm, within a 5-m radius of a specific sampling location; then the samples were mixed and air-dried at room temperature. Stone and plant residue in the soil samples were manually removed, and then the samples were ground to pass a 2mm sieve. SOC was determined according to the Walkley and Black method (Walkley and Black 1934).
Land use data for 2006, with information on industrial types, was obtained from Bureau of Land and Resources of Fuyang, the land administrator in Fuyang County. The land use map for 1979 was produced by photo-interpretation of 1979 Landsat MSS photographs with a spatial resolution of 30 m (Zhong et al. 2011). 2.3 Data analysis An important contribution of geostatistics is to assess the uncertainty of unobserved interpolated values (Goovaerts 1997). The semivariogram could be used to quantify the spatial variation of SOC between two points, x and x+h, as a function of their distance h: gðhÞ ¼
NðhÞ 1 X ½Z ðxi Þ Z ðxi þ hÞ 2 2N ðhÞ i¼1
ð1Þ
Where Z(xi) and Z(xi +h) represent the measured value for SOC at location xi andxi +h, y(h) is the variogram for a lag distance h, and N(h) is the number of data pairs separated by h. All semivariogram analysis was carried out using GS+® (Version 7). Based on the semivariogram results, ordinary kriging, which is the most common interpolation method, was used to estimate the unobserved SOC values. Ten percent of data points were randomly selected to test the prediction error by the mean prediction error (ME) and rootmean-square standardized prediction error (RMSSE). ME ¼
n 1X Zðxi Þ z* ðxi Þ n i¼1
ð2Þ
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi n 1X Zðxi Þ z* ðxi Þ RMSSE ¼ n i¼1 σ ð xi Þ
ð3Þ
Where Z(xi) and z*(xi) represent the measured value and predicted value (respectively) at location of xi, and σ(xi) is the standard error of prediction at location of xi.
Table 2 Semivariogram models for soil organic carbon (SOC) in study area Year
Model
1979 2006
Spherical Spherical
Range (km)
Angle direction
Nugget (C0)
Sill (C0 +C)
C0/Sill
ME
13.4 4.5
48.5 58.5
13.6 18.4
23.3 35.2
0.583 0.523
−0.035 0.012
RMSSE 0.948 0.953
984 Fig. 2 Distribution maps of soil organic carbon (SOC) content in 1979 (a) and 2006 (b) in study area, changes in SOC between 1979 and 2006 (c)
J Soils Sediments (2013) 13:981–988
J Soils Sediments (2013) 13:981–988
985
Table 3 Land areas and percentages in different soil organic carbon classes in 1979 and 2006 Time
SOC class (g/kg)
1979
Area (km2) Percentage (%) Area (km2) Percentage (%)
2006
I >23.2
II 17.4–23.2
III 11.6–17.4
IV 5.8–11.6
– – 55.8 6.5
442.9 51.4 511.6 59.4
416.2 48.3 287.4 33.4
2.9 0.3 6.5 0.7
Interpolated results were exported as a raster format and consequently resulted in the spatial distribution maps of SOC 1979 and 2006, respectively. Then, the raster calculator was used to get the distribution map of the difference in SOC between 1979 and 2006. Other spatial analyses included overlay analysis, zonal statistics, and buffer analysis. All spatial analyses were done in ArcGIS® (Version 9.2). Descriptive statistics (mean, median, minimum, maximum, and coefficient of variation) for 1979 and 2006 SOC data were obtained to get a preliminary understanding of the datasets. Normality of the datasets was analyzed using the Kolmogorov–Smirnov (K-S) test. The difference of measurement data was compared with analysis of variance and the t test. All statistical analyses were performed using SPSS® (Version 16.0).
3 Results and discussion 3.1 Descriptive statistics Descriptive statistics for SOC are shown in Table 1. A significant difference in SOC between 1979 and 2006 was detected according to the t test results (P0.05) showed that SOC in Fuyang County was normally distributed in 1979 and 2006, which is a requirement for simple geostatistical analysis. The SOC data were moderately varied, shown by the relatively small coefficients of variation (CV). Table 4 Land use changes from 1979 to 2006 in study area (square kilometer)
Year
Builtup
1979 2006
20.4 129.2
Paddy field 461.5 205.0
V 3.5–5.8
VI
