Abstract:

Based on field survey data of soil volumetric water content, topographic factors and soil physical and chemical properties from typical stands in the Loess hilly region of Ningxia, the 0-300cm soil profile was stratified into surface (0–20cm), shallow (20–80cm), middle (80–180cm) and deep (180–300cm) layers. Using correlation analysis, principal component analysis (PCA) and Random forest model, with natural grassland as the control, the differences and underlying mechanisms of soil water storage across 10 typical stands during the leaf-fall period of 2024 in the Yangwa (Zhongzhuang) watershed, Pengyang county, Guyuan city were analyzed. The stands included pure plantations of Robinia pseudoacacia L., Picea asperata Mast., Prunus sibirica L., Prunus davidiana (Carrière) Franch., Hippophae rhamnoides L. subsp. sinensis Rousi, and Caragana korshinskii Kom., as well as mixed stands of Robinia pseudoacacia×Picea asperata, Prunus sibirica×Hippophae rhamnoides subsp. sinensis, Prunus sibirica×Caragana korshinskii and Prunus davidiana×Caragana korshinskii. The relative importance of environmental factors, including slope, aspect, slope position, elevation, clay content, silt content, sand content, soil organic carbon, soil bulk density, tree height, diameter at breast height and biomass, was quantified on soil water storage and identified the core driving factors, so as to provide a reference for the sustainable utilization of regional soil water resources. The results showed that: (1) in 2024, the average soil volumetric water content in the 0–300cm soil layer was significantly higher in P. asperata pure forest (17.47%) and P. davidiana pure forest (17.40%) than in natural grassland (15.95%). P. davidiana pure forest had the highest soil water storage in the surface (31.59mm), shallow (122.27mm), and middle layers (177.65mm), while natural grassland showed the highest deep-layer soil water storage (227.21mm). Soil water storage across 0–300cm was highest in P. asperata pure forest, reaching 524.22mm. (2) Soil bulk density was the core driving factor of soil water storage. It was significantly positively correlated with water storage in the middle, deep and whole soil layer (r =0.567, 0.471 and 0.527, respectively; all P<0.01). In the Random forest model, the importance values were 26.238%, 9.426% and 23.715% for the middle, deep and total soil layers, respectively. The slope was the key factor affecting surface soil water storage, with an importance of 6.959%. Elevation and vegetation factors were key regulators of shallow soil water storage: elevation was significantly positively correlated with shallow soil water storage (r=0.345, P<0.05), and the importance values of diameter at breast height, tree height and biomass were 13.350%, 9.190% and 9.102%, respectively. (3) The first principal component, PC1 (25.0%), represented the synergistic effects of soil texture and soil organic carbon, while the second principal component, PC2 (21.1%), highlighted the influence of vegetation factors. In conclusion, under the current afforestation pattern in Pengyang county, P. asperata and P. davidiana pure forests exhibited stronger soil water storage capacity. Soil water storage across different depths is affected by distinct environmental factors. To enhance regional soil water conservation, particularly deep-layer water storage, priority should be given to maintaining or improving soil bulk density.

Key words: Soil water storage, Stand type, Driving mechanism, Southern Ningxia, Loess plateau