Abstract: This study aims to investigate the quantitative relationship between resistivity and the physical and mechanical properties of soil in different types of herbaceous slopes in the alpine arid and semi-arid loess area. The research is conducted in the self-built test area of Changlinggou Basin in Xining Basin. Five types of slopes, including Elymus nutans Griseb., Elymus sibiricus Linn., Agropyron trachycaulum Linn. Gaertn., Festuca arundinacea Schreb., and bare slopes are selected as the research objects. These slopes have been planted for 3 years. The study compares the effects of different herbaceous roots on the physical and mechanical properties of the soil by conducting tests of soil density and water content, and direct shear test on the soils with and without root systems. Based on these tests, a quantitative relationship between the physical and mechanical properties of different slope soils and resistivity data is established using 2D electrical resistivity tomography. The results show that: (1) Compared with the bare slope without planting, the maximum increase of soil moisture content in the upper layer (0–10 cm) of the Elymus sibiricus Linn. slope is 26.53%. The average soil density of the upper layer (0–10 cm) of the Festuca arundinacea Schreb. slope was 18.30% lower than that of the bare slope. The maximum added value of soil cohesion in the upper layer (0–10 cm) of the Elymus nutans Griseb. slope is 2.75 times that of the bare slope. (2) The resistivity characteristics of five types of slopes are affected by root distribution and slope position factors, and the resistivity value decreases with the increase of depth. The soil resistivity value of the four herbaceous slopes is larger than that of the bare slope at 0–20 cm, which is the approximately range of root distribution. (3) There are fitting equations between the physical and mechanical properties and resistivity data of five kinds of slope soils (with correlation coefficients R² ranging from 0.48 to 0.77), and the Pearson correlation analysis shows that the cohesion c value of the slope soil has the highest correlation with resistivity, with an R² value of 0.765. The results of this study demonstrate that 2D resistivity tomography technology can reflect the physical and mechanical properties of slope soil, as well as the distribution characteristics of plant roots. This provides a theoretical basis and practical guidance for effectively preventing and controlling soil erosion, shallow landslides, and other disasters in the study area and its surrounding areas.

Key words: alpine arid and semi-arid regions, Xining Basin, 2D electrical resistivity tomography, resistivity of slopes, soil physical and mechanical properties, Pearson correlation analysis