Akademik Veri Yönetim Sistemi
Plant and Soil, cilt.516, sa.1, ss.217-239, 2025 (SCI-Expanded, Scopus)
Aims: Understanding the relationships between water-soluble and exchangeable elements in drylands is valuable to understand several important processes such as soil–plant-interactions, fertilizers recommendations, and leaching potential. However, few studies have been conducted to examine such relationships worldwide. This study investigates the relationships between water-soluble potassium (WSK) and exchangeable potassium (ExchK) in drylands of Sudan. Methods: The random forest (RF) algorithm and statistical model based on environmental covariates (ECOVs) representing climate, biota, topography, and selected properties was utilized to develop pedotransfer functions (PTFs) examining the relationships between WSK and ExchK. Results: The statistical model indicated decreases in WSK-ExchK relationships with increasing pH, CaCO3, CEC, and clay content. The PTF of RF-based ECOVs was best for predicting ExchK (MAE: 0.83 ± 0.10; RMSE: 1.020 ± 0.15; NRMSE: 21%), with precipitation seasonality and mean temperature of coldest quarter being the most important ECOVs. RF-PTFs based on soil data were the most effective PTFs for WSK prediction, with pH and clay content being the most importance predictors. Conclusions: The findings of this study are important to overcome the difficulties of building national soil databases for large-scale modeling using open access digital data with easily measurable soil properties. Further regional studies should investigate the potential for upscaling WSK and ExchK-PTFs considering local factors such as land uses, which can then be integrated into K fertilization guidelines. We recommend future studies expand investigations of the relationships between WSK and ExchK as functions of new ECOVs which would benefit efforts to model K storage in soils and assist predictions of K sorption.