Multivariate analysis of soil particle size distribution and Spatial correlation with soil moisture characteristics in different vegetation types of Mu Us Sandy Land

In this paper, the fractal mechanism of soil improvement by vegetation was revealed by analyzing the soil characteristics under four typical vegetation types: Salix cheilophila, Caragana korshinskii, Hippophae rhamnoides, and Corethrodendron fruticosum in Mu Us Sandy Land. The results showed that (1) the soil of each vegetation type was mainly composed of sand (> 90%), and the content of clay and silt was less than 10%. CF had the least heterogeneity of soil particles in the 0-120 cm soil layer, HR was uniformly distributed in the 120-200 cm soil layer, and the change trend of particle composition in the 140-200 cm soil layer was consistent. (2) Soil fractal dimension (D) was positively correlated with soil clay and silt content, and negatively correlated with sand content. The generalized dimension spectrum D(q) decreases in the "S" type, and the sensitivity was higher in the region of q < 0, indicating that the particle distribution in the sparse area is more susceptible to disturbance. (3) HR had the best water holding capacity in the surface layer (0-40 cm) and deeper layer (120-200 cm), and the average soil water storage in 0-200 cm reached 791.61 mm (13.76% higher than CF). All vegetation had soil water deficit in specific soil layers, among which SC had water deficit in 20-30 cm, CK in 0-10 cm, HR in 60-80 cm, and CF in 180-200 cm soil layers (0.91 mm, 0.90 mm, 0.92 mm, and 0.93 mm). (4) HR significantly affected soil bulk density and porosity by increased silt and sand (P < 0.05). The fractal parameters of SC were significantly correlated with soil water content (P < 0.05). The pH of CK was a significant correlation with soil water storage (P < 0.01). CF soil particle composition and fractal parameters were significantly correlated with soil moisture content, bulk weight, and capillary porosity (P < 0.05). The study showed that HR achieves the best water retention and soil modification effect by optimizing soil structure and is the preferred vegetation for ecological restoration of the Mu Us Sandy Land.