Limits...
Three-Dimensional Mapping of Soil Organic Carbon by Combining Kriging Method with Profile Depth Function.

Chen C, Hu K, Li H, Yun A, Li B - PLoS ONE (2015)

Bottom Line: The results showed that the exponential equation well described the vertical distribution of mean values of the SOC contents.The SOC contents showed significant positive correlations between the five different depths and the correlations of SOC contents were larger in adjacent layers than in non-adjacent layers.Soil texture and land use type had significant effects on the spatial distribution of SOC.

View Article: PubMed Central - PubMed

Affiliation: Department of Soil and Water Sciences, China Agricultural University, No 2, Yuan Ming Yuan Xi Lu, Beijing 100193, China.

ABSTRACT
Understanding spatial variation of soil organic carbon (SOC) in three-dimensional direction is helpful for land use management. Due to the effect of profile depths and soil texture on vertical distribution of SOC, the stationary assumption for SOC cannot be met in the vertical direction. Therefore the three-dimensional (3D) ordinary kriging technique cannot be directly used to map the distribution of SOC at a regional scale. The objectives of this study were to map the 3D distribution of SOC at a regional scale by combining kriging method with the profile depth function of SOC (KPDF), and to explore the effects of soil texture and land use type on vertical distribution of SOC in a fluvial plain. A total of 605 samples were collected from 121 soil profiles (0.0 to 1.0 m, 0.20 m increment) in Quzhou County, China and SOC contents were determined for each soil sample. The KPDF method was used to obtain the 3D map of SOC at the county scale. The results showed that the exponential equation well described the vertical distribution of mean values of the SOC contents. The coefficients of determination, root mean squared error and mean prediction error between the measured and the predicted SOC contents were 0.52, 1.82 and -0.24 g kg(-1) respectively, suggesting that the KPDF method could be used to produce a 3D map of SOC content. The surface SOC contents were high in the mid-west and south regions, and low values lay in the southeast corner. The SOC contents showed significant positive correlations between the five different depths and the correlations of SOC contents were larger in adjacent layers than in non-adjacent layers. Soil texture and land use type had significant effects on the spatial distribution of SOC. The influence of land use type was more important than that of soil texture in the surface soil, and soil texture played a more important role in influencing the SOC levels for 0.2-0.4 m layer.

No MeSH data available.


The mean SOC contents at five depths and the corresponding fitting model.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4457849&req=5

pone.0129038.g003: The mean SOC contents at five depths and the corresponding fitting model.

Mentions: The descriptive statistics and analysis of variance showed that the SOC contents between five different soil depths were significantly different (P < 0.05) (Table 1). The mean values of SOC contents gradually decreased from the surface along the soil profile. The mean value of surface (0–20 cm) SOC content was the highest (8.25 g kg-1) while the lowest value was measured in the bottom depth (80–100 cm). This variation would be related to the long-term human activities such as tillage, fertilization and crop straw returns that increased the SOC level in the topsoil. The coefficients of variation (CV) of SOC data for all layers ranged from 0.26 to 0.43 and the CV values gradually increased from the surface to the bottom in the soil profile. However, the studies from the Loess Plateau showed that the CV values of SOC data in topsoil were higher than those in subsoil [31, 32]. This difference could be due to the complex textural layers in the alluvial plain in Quzhou County [11]. The results of the K-S test showed that the SOC contents were normally distributed at five soil depths (P > 0.05). The analysis of variance showed that the SOC contents were significantly different among soil depths above 0.6 m, while there was no difference in SOC values between soil depths below 0.6 m. The fitting performances (RMSE = 0.075 g kg-1, R2 = 1) showed that the equation [p(dj) = 9.42exp(-6.28dj) + 3.23] could give a reasonable prediction for the means of SOC contents at all five soil depths (Fig 3).


Three-Dimensional Mapping of Soil Organic Carbon by Combining Kriging Method with Profile Depth Function.

Chen C, Hu K, Li H, Yun A, Li B - PLoS ONE (2015)

The mean SOC contents at five depths and the corresponding fitting model.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4457849&req=5

pone.0129038.g003: The mean SOC contents at five depths and the corresponding fitting model.
Mentions: The descriptive statistics and analysis of variance showed that the SOC contents between five different soil depths were significantly different (P < 0.05) (Table 1). The mean values of SOC contents gradually decreased from the surface along the soil profile. The mean value of surface (0–20 cm) SOC content was the highest (8.25 g kg-1) while the lowest value was measured in the bottom depth (80–100 cm). This variation would be related to the long-term human activities such as tillage, fertilization and crop straw returns that increased the SOC level in the topsoil. The coefficients of variation (CV) of SOC data for all layers ranged from 0.26 to 0.43 and the CV values gradually increased from the surface to the bottom in the soil profile. However, the studies from the Loess Plateau showed that the CV values of SOC data in topsoil were higher than those in subsoil [31, 32]. This difference could be due to the complex textural layers in the alluvial plain in Quzhou County [11]. The results of the K-S test showed that the SOC contents were normally distributed at five soil depths (P > 0.05). The analysis of variance showed that the SOC contents were significantly different among soil depths above 0.6 m, while there was no difference in SOC values between soil depths below 0.6 m. The fitting performances (RMSE = 0.075 g kg-1, R2 = 1) showed that the equation [p(dj) = 9.42exp(-6.28dj) + 3.23] could give a reasonable prediction for the means of SOC contents at all five soil depths (Fig 3).

Bottom Line: The results showed that the exponential equation well described the vertical distribution of mean values of the SOC contents.The SOC contents showed significant positive correlations between the five different depths and the correlations of SOC contents were larger in adjacent layers than in non-adjacent layers.Soil texture and land use type had significant effects on the spatial distribution of SOC.

View Article: PubMed Central - PubMed

Affiliation: Department of Soil and Water Sciences, China Agricultural University, No 2, Yuan Ming Yuan Xi Lu, Beijing 100193, China.

ABSTRACT
Understanding spatial variation of soil organic carbon (SOC) in three-dimensional direction is helpful for land use management. Due to the effect of profile depths and soil texture on vertical distribution of SOC, the stationary assumption for SOC cannot be met in the vertical direction. Therefore the three-dimensional (3D) ordinary kriging technique cannot be directly used to map the distribution of SOC at a regional scale. The objectives of this study were to map the 3D distribution of SOC at a regional scale by combining kriging method with the profile depth function of SOC (KPDF), and to explore the effects of soil texture and land use type on vertical distribution of SOC in a fluvial plain. A total of 605 samples were collected from 121 soil profiles (0.0 to 1.0 m, 0.20 m increment) in Quzhou County, China and SOC contents were determined for each soil sample. The KPDF method was used to obtain the 3D map of SOC at the county scale. The results showed that the exponential equation well described the vertical distribution of mean values of the SOC contents. The coefficients of determination, root mean squared error and mean prediction error between the measured and the predicted SOC contents were 0.52, 1.82 and -0.24 g kg(-1) respectively, suggesting that the KPDF method could be used to produce a 3D map of SOC content. The surface SOC contents were high in the mid-west and south regions, and low values lay in the southeast corner. The SOC contents showed significant positive correlations between the five different depths and the correlations of SOC contents were larger in adjacent layers than in non-adjacent layers. Soil texture and land use type had significant effects on the spatial distribution of SOC. The influence of land use type was more important than that of soil texture in the surface soil, and soil texture played a more important role in influencing the SOC levels for 0.2-0.4 m layer.

No MeSH data available.