Distribution mapping of soil profile carbon and nitrogen with laboratory imaging spectroscopy

Abstract

Conversion of arable cropland to forage crops has been proposed as a potential method to increase soil organic carbon (SOC) stocks to sequester carbon and improve soil quality. In this study, intact soil cores were collected from long-term boreal forest soil research plots established in 1980 consisting of: a mixed arable crop and forage agroecological rotation (AE), continuous forage (CF), and continuous grain (CG) rotations. These cores were analyzed using a SisuROCK automated hyperspectral imaging system in a laboratory setting collecting shortwave infrared reflectance data. Samples were then analyzed for SOC and total nitrogen (TN) contents by dry combustion to prepare a training data set. Predictive models were successfully built for SOC and TN using a combination of wavelet analysis and Bayesian Regularized Neural Nets. The CF rotation was found to have the highest SOC and TN contents compared to AE rotation for only the top 3 and 4 cm, respectively. These two rotations had comparable contents for both parameters for the rest of the topsoil, which was greater than the SOC and TN contents in the CG rotation to depths of approximately 12 cm. Increases in both SOC and TN were associated with increased spatial aggregation at fine spatial scales. These results indicate that adding forages to rotations in boreal forest soils increases SOC and TN, however these changes were concentrated in the surface depths.