Fine Scale Variation of Soil Organic Carbon in Different Land Use and Land Cover Types in Florida.

Land use and land cover (LULC) have been well documented to be major drivers of soil organic carbon (SOC) variation. Yet, our understanding of SOC variation at fine (field) scale under various LULC types is still limited. The aims of this study were to characterize and compare the fine scale spatial variation of SOC in prominent LULC types in Florida using statistical and geostatistical methods. Five dominant LCLU types in Florida were targeted in this study, i.e., pineland, dry prairie, improved pasture, mesic mixed forest, and xeric upland forest, which in total account for approximately half of the acreage of all LCLU found in Florida. In each LCLU, 108 soil samples were taken using the optimized unbalanced spatially nested sampling design in which soil sample spaces varied at four hierarchical levels - 2m, 7m, 23m, and 67m in order to efficiently capture the fine scale SOC variation. SOC was measured using dry combustion with a Shimatzu TOC analyzer. Results show that SOC varied among different LULC significantly. The ANOVA analysis for the spatially nested design indicates that the major variation of SOC at fine scale in different LULC came from different hierarchical levels. The spatial structures of the SOC in the five LCLU types also differed among each other in terms of the spatial autocorrelation range, sill, and nugget variances. These results advance our understanding of fine scale variation of SOC in the major LULC types in Florida, improve soil sampling design in support of digital soil mapping of SOC at fine scale in the future, and facilitate to incorporate knowledge of fine scale SOC variation in regional digital soil models.