Recalcitrant C Source Mapping Utilizing Solely Terrain-Related Attributes and Data Mining Techniques

Agricultural practices affect arbuscular mycorrhizal fungal (AMF) hyphae growth and glomalin production, which is a recalcitrant carbon (C) source in soil. Since the spatial distribution of an interesting issue for agronomists terms sequestration, digital maps are cost-free useful approach. For this study, set 120 soil samples was collected from experimental area 310 km2 Sarab region Iran. Soil total (TG) easily extractable (EEG) were determined via ELISA using monoclonal antibody 32B11. organic (OC) also measured. The ratios TG/OC EEG/OC as glomalin–C quotes OC calculated. A 17 terrain-related attributes derived elevation model (DEM) used static environmental covariates mapping (DSM) three predictive models, including multiple linear regression (MLR), random forests (RF), Cubist (CU). major findings follows: (a) DSM facilitated interpretation C variation; (b) RF outperformed MLR CU models predicting available covariates; (c) best accuracy predictions EEG, followed by EEG/OC, TG, TG/OC.