Mapping of Hulun Buir steppe topsoil bulk density based on different sets of predictors

Abstract

Climate change and human activities have contributed to extensive grassland loss worldwide. Soil bulk density is a critical indicator of grassland ecosystem health, and understanding the spatial distribution patterns of its changes is essential for mitigating further grassland degradation. Using distinct types of predictors, we developed corresponding random forest models to predict topsoil bulk density in the Hulun Buir Steppe. Our analysis revealed that the combined variable model demonstrated the highest predictive performance. The key variables impacting topsoil bulk density include soil type, surface soil temperature, surface soil moisture, wind speed, water vapor pressure, and NDVI. Under the interaction of soil type with climatic factors and vegetation indices, the topsoil bulk density values remained essentially constant and low with meadow soils. In contrast, the saline meadow soil type exhibited a notable increase in topsoil bulk density, with higher values being observed. Under the interaction of other soil properties with climatic factors and vegetation indices, the topsoil bulk density tends to be higher in situations where surface soil moisture and NDVI are low, and water vapor pressure and surface soil temperature are high. The Hulun Buir Steppe’s topsoil bulk density ranges from 0.46 to 2.00 g · cm⁻³. The spatial distribution predicted by different predictor sets of models indicated a general trend of lower in the eastern meadow steppe area and higher in the central and western typical steppe, which were influenced by natural conditions and degradation. This study offers insights into soil property changes during grassland degradation and guidance for grassland conservation and restoration.