Relationships between surface soil organic carbon pool and site variables

Abstract

Surface soil organic carbon pool (SOCP) and turnover time are particularly sensitive to site variables and land use change. An understanding of influences of site variables on surface SOCP is important to planning future land use for C sequestration and sustaining agricultural production. This study was conducted to clarify relationships between the SOCP in the upper 30-cm mineral soil and site variables (soil taxon, texture, drainage class, slope gradient and elevation) under cropland, grassland and forestland in Ohio, USA. The data of the SOCPs and site variables of 2137 pedons were derived from the Ohio Soil Survey Database and analyzed using SAS. The results showed that mean SOCP values significantly varied with soil order within each land use system: mollisols were 9.4 (kg m −2), followed by inceptisols 7.2, entisols 5.6, alfisols 5.4 and ultisols 4.6. Generally, the SOCP was highly correlated with all site variables ( p<0.001) and the strongest correlation was present in cropland and the weakest one in forestland. Poor drainage and heavy texture favored SOC sequestration in all land uses. The ANOVA factorial model estimated that the contribution of site variables to the variation of SOCP was 54.9% in cropland, 50.7% in grassland and 30.3% in forestland, which was 2.6%, 2.5% and 7.8% higher than that described by canonical redundancy analysis, respectively. The significance of individual site variables was generally in the order of soil taxon>drainage>texture≫slope>elevation, of which elevation was proven ignorable under cultivation but significant under forests. Relatively weak correlations and less explainable variance observed in forestland suggest that actual controlling factors over SOCP differ from those in cultivated lands.