2.2 Experimental design
The ecological restoration experiment was conducted from July to November 2020. According to the previous field survey data of highway, railway and other construction areas in Nyingchi(Soil gravel content between 10%~50%), three classes of soils with gravel contents (>2 mm) of 10% (low), 30% (medium) and 50% (high) were selected as ecological restoration test materials. Their physical and chemical properties are summarized in Table 1. In general, as gravel content increased, the contents of sand, silt, clay, total porosity, total nitrogen and phosphorus, and organic carbon decreased. Soil with medium gravel content had the largest bulk density. These studied soils were neutral with a pH of 6.87~6.98. These soils were packed into 1 m× 3 m test plots with a thickness of 60 cm according to native soil compactness.
Table 1 Physical and chemical properties of studied soils
Three plant species of Festuca elata (Fe,indigenous specie), Medicago sativa (Ms, introduced specie) and Elymus dahuricus(Ed, introduced specie) were selected as experimental plant materials according to the commonly used plants in ecological restoration projects and the natural vegetation of Qinghai-Tibet Plateau, Among them, Festuca elata and Medicago sativa are ecological restoration plants introduced in Qinghai Province, andElymus dahuricus is a native plant in Tibet. Considering the ecological restoration effect of plants and the more obvious competition effect between plants (same or different species), the method of close planting is adopted(Huang et al., 2021; Leinauer et al., 2021), seven typical and widely used plant species communities were set on studied soils, that is, Festuca elata (seeding rate of 200 kg hm-2)), Elymus dahuricus (seeding rate of 200 kg hm-2), Medicago sativa seed (seeding rate of 100 kg hm-2), Festuca elata (seeding rate of 200 kg hm-2) plus Elymus dahuricus (seeding rate of 200 kg hm-2), Festuca elata (seeding rate of 200 kg hm-2) plus Medicago sativa(seeding rate of 100 kg hm-2), Elymus dahuricus(seeding rate of 200 kg hm-2) plus Medicago sativa (seeding rate of 100 kg hm-2), andFestuca elata (seeding rate of 200 kg hm-2) plus Elymus dahuricus (seeding rate of 200 kg hm-2) plus Medicago sativa (seeding rate of 200 kg hm-2). These plant species treatments are separately denoted as Fe , Ed , Ms ,Fe +Ed (1:1), Fe +Ms (2:1), Ed+Ms (2:1), and Fe +Ed +Ms (2:2:1). Three replicates were set for each treatment, and the same planting method of evenly spreading and covering 0.2-0.5cm soil was adopted. After sowing, no additional field management (such as fertilization and watering) were conducted during the growth periods (94 days), and the experiments ended in November 2020.

2.3. Sample collection and measurement

Before the ecological restoration experiments, composited and undisturbed soil samples were taken from 0-10 cm and 10-20 cm soil layers, and then brought back to the laboratory for soil physical and chemical analysis. At the end of plant growth, plant traits including coverage, natural height, above-ground biomass and root biomass were measured. Rhizosphere soils (about 3 mm soil fractions around the roots) were collected.
Soil particle size composition was determined by the sieving-pipette method after ultrasonic dispersion; field water content by weight method after oven-dried; bulk density and total porosity by weight method on undisturbed soil cores (100 cm3); organic carbon of soils and plant samples by potassium dichromate oxidation and titration with ferrous sulfate; total nitrogen and phosphorus contents of plant and soil samples by Smartchem200 automatic discontinuous chemical analyzer after acid digestion (Tang et al.,2022; Institute of Soil Science.,1978).