Poster Number 515
Wednesday, 8 October 2008
George R. Brown Convention Center, Exhibit Hall E
Understanding the effect of land use on soil organic carbon, particularly at landscape scale, is critical for developing land management strategies to enhance soil carbon sequestration. In Florida the major soil orders are Spodosols, Entisols, Ultisols and Alfisols covering 24.9, 21.9, 18.2 and 12.7% of the state, respectively (based on available soil survey data from NRCS Soil Data Mart). Soil carbon is stored in the top soil as well as in subsoil, particularly in soil orders like Spodosols. The spodic horizons, which contain significant carbon content, contribute to the overall carbon budget of Florida’s soilscape. Therefore it is essential to determine how deep in the soil profile the land use effect can be observed. The objectives of this study were 1) to measure the maximum depth to which land use affected soil carbon content and 2) to identify interactions between land use and different environmental attributes. We used a comprehensive dataset collected on soils in Florida covering 8,235 horizons and 1,296 soil profiles in 58 counties in Florida. The carbon stock for each 20cm depth increment (up to 1m) was calculated from the carbon concentration and bulk density values using a profile reconstruction technique. The effect of land use on soil carbon was determined by generalized linear model with land use, soil suborder, drainage class and other factors as fixed effects as well as the regression tree technique. As expected, soil carbon was found to be significantly affected by land use, but the effects varied with suborder and drainage class of the soils under consideration. The effect of land use (as shown by R-square values) increased up to 40 to 60 cm and then diminished, but the effect of land use was significant even at the deeper horizons. Urban land use, however, exhibited extreme variability in the soil carbon values, probably due to the inherently diverse nature of this land use class.