Soil is an important reservoir of carbon in both regional and global C budgets, and there is much interest in estimating soil organic carbon (SOC) stocks at these scales. Sampling of SOC across large areas requires the collection of many samples to fully account for both small and large scale sources of variation. This translates into high analytical costs using modern quantitative methods for estimating percent carbon content. To address this issue we propose a ‘double sampling’ strategy that allows for the prediction of percent SOC content based on estimates of soil organic matter content from loss on ignition (LOI) analysis. LOI is a low-cost alternative to more precise measurements such as CHN analysis. Utilizing both in a double sampling approach maximizes spatial coverage while keeping total sampling cost at a minimum.

                Our study area is New Jersey’s outer coastal plain, a largely forested region dominated by pine and mixed pine/oak canopy types. The underlying soils are sandy, well to poorly drained Ultisols and Spodosols. They are low in carbonate content and accumulate moderate amounts of soil organic matter. We sampled soil organic carbon at three depth intervals (0-10, 10-20, & 20-30 cm) at forested plots across the region. Sampling was stratified based on both drainage class and forest canopy type. All samples were analyzed for percent organic matter using loss on ignition (sample N₁) and a smaller subset was also analyzed for percent carbon using CHN analysis (sample N₂). The relationship between SOC estimates from LOI and CHN in sample N₂  was used to calibrate the LOI estimates from the large sample (N₁₎ . Preliminary results suggest that this model provides statistically powerful predictions of percent carbon content based on measurements of soil organic matter from loss on ignition.