Spatial Variation And Influential Factors Of Soil Organic Carbon In Differentecological Areas

This research studies the spatial distribution characteristics and influencing factors of farmland soil’s organic carbon content from typical counties in different ecological regions of Sichuan province including Wenjiang county in plain area, Renshou county in hill area, Mabian county in basin around and hill area, Xichang city in hill area of southwest Sichuan and Li county in plateau area of northwest Sichuan by adopting the combination method of high-accuracy surface modeling multiple linear regression. And it calculates the organic carbon density and reserves of each area further and reaches the following main conclusions.(1) Based on the conventional statistical analysis to farmland soil organic carbon content of each area the organic carbon content of Wenjiang area and Xichang city is normally distributed and the one of Renshou county, Mabian county and Li county shows a logarithmic normal distribution. The average farmland soil organic carbon content of Li county is highest among five areas with32.78g·kg-1and followed by Xichang city, Wenjiang area, Mabian county and Renshou county with average farmland soil organic carbon content of17.29g·kg-1,15.81g·kg-1,12.41g·kg-1and9.17g·kg-1respectively. The semi-variance functions are conducted to the organic carbon content in five areas and it shows the following results. The organic carbon space correlation of Renshou county in hill area in the middle of Sichuan and Mabian county in basin around and hill area is weak and its space variation is mainly affected by stochastic factors; the organic carbon of Wenjiang area in Chengdu plain and Li county in plateau area of northwest Sichuan has medium spatial correlation whose spatial variation is affected by stochastic factors and structural factors together; and the organic carbon content of Xichang city in hill area of southwest Sichuan has strong spatial correlation and its space variation is mainly affected by structural factors. The spatial distribution of organic carbon content in each area is gotten by adopting ordinary Kriging interpolation and the soil organic carbon content of Renshou county, Xichang county, Wenjiang area, Mabian county and Li county is7.8-10.0g·kg-1,15.6-22.1g·kg-1,11.7-16.3g·kg-1,13.8-31.3g·kg-1and26.4-41.8g·kg-1 (2) According to the specific characteristics of different areas, different influence factors in five areas are chosen as modeling indicators. After conducting correlation analysis between modeling factors of each area and organic carbon content, indicators whose correlations reaching to be significant or extremely significant are chosen to establish regression equation. Among them, Renshou county chose10indicators including elevation, slope, cumulative flow, texture, parent material, crop rotation system, planting system, total nitrogen content, ph value and average annual rainfall; Wenjiang area selected seven indicators covering altitude, accumulated flow, texture, total nitrogen content, ph value, crop rotation system and planting system; Xichang city chose10indicators including elevation, cumulative flow, annual average temperature, annual average rainfall, soil parent material, texture, ph value, total nitrogen content, planting system and crop rotation system; Mabian county chose elevation, cumulative flow, annual average temperature, annual rainfall, soil parent material, texture, ph value, total nitrogen content, planting system and crop rotation system; and Li county chose six indicators including elevation, rainfall, annual average temperature, total nitrogen content, ph value and texture. The regression equation decision coefficient between organic carbon content and the related influencing factors of five areas is0.166(Renshou county, lowest),0.333(Li county),0.570(Wenjiang area),0.733(Mabian county) and0.785(Xichang city, highest). The relatively accurate organic carbon content distribution diagram is obtained by curved surface modeling and its distribution trend is relatively consistent with the result of ordinary Kriging but it is more accurate in detail simulation and reflects the influence of terrain, climate, soil type and human activities on organic carbon content better. After conducting error test to the simulation results of each model, the average absolute error is1.23g·kg-1(Wenjiang area),1.45g·kg-1(Renshou county),1.85g·kg-1(Xichang city),2.59g·kg-1(Mabian county) and3.45g·kg-1(Li county) and the average relative errors are controlled below30%.(3) Adopting soil type method and taking soil as a unit, it is gotten that the organic carbon average density of in plateau area of northwest Sichuan is highest reaching5.77kg.m-2followed by Xichang city in hill area of southwest Sichuan, Wenjiang area in Chengdu plain and Mabian county in basin around and hill area with average density of 4.45kg.m-2,3.95kg.m-2and3.32kg.m-2respectively. The lowest organic carbon average density is Renshou county in hill area of middle Sichuan with2.28kg.m-2only. On the basis of the organic carbon density of different soil types in each area, the organic carbon reserves of each is received. Among them, Renshou county with largest cultivated land areas has the biggest total reserves of organic carbon with18.7×105tons followed by Xichang city, Wenjiang area and Mabian county with13.1×105tons,4.87×105tons and4.78×105respectively. And Li county with least cultivated land areas has the least amount of organic carbon reserves with1.38×105tons.(4)The soil organic carbon content is affected differently by various factors in different areas. Affected by altitude, the temperature, rainfall, vegetation and human activities in the elevation range are different. Among them the organic carbon content has extremely negative correlation with temperature, which shows that low temperature is conducive to organic carbon accumulation and it has no significant correlation with rainfall. In comparison with rivers and urban buffer areas, basin around and hill area, hill area of southwest Sichuan and plateau area of northwest Sichuan show a consistent trend, that is, the organic carbon content is rising obviously with the increase of buffer distance. However, plain and hill areas show the opposite trend at the same time. In terms of land use way, the influence to organic carbon content of different land use ways is also affected by local farmland distribution status.