*3.4. Bacterial β-Diversity*

ADONIS analysis showed significant differences in vegetation and bacterial community structure between distances at the same elevation, as well as between elevations (Table S2). From the results of NMDS analysis(Figure 9), it can be seen that there are significant differences in the structure of soil bacterial community among the samples at different altitudes (Stress value is 0.084), and the effect of altitude on bacterial community structure was more significant than that of distance. In different regions, significant differences in soil bacterial community structure were revealed among different distances (Stress values were 0. 058, 0.045, and 0. 043), which indicated that the soil bacterial community structure in the highway disturbed zone had changed significantly.

**Figure 9.** Non-metric multidimensional scaling (NMDS) analysis of (**A**) overall bacterial in 3 altitudes and bacterial in 3 sites of (**B**) 4000 m, (**C**) 4600 m, and (**D**) 5200 m.

According to the RDA results, soil bacterial community structures in three sites could be distinguished by plant and soil factors (Figure 10). In the site of 4000 m, AN had the greatest impact on the bacterial community. The first and second axes explained 54.70% and 18.53% of the variation, respectively. AK, SOC, and Zn are the main correlation factors of axis 1. Associated with axis 2 are mainly AN and AK (Figure 10A). In the 4600 m site, the most important factor is distance, followed by TN and coverage (Figure 10B). The first axis (mainly related to distance and TN) and the second axis (mainly related to SM and AP) described 35.70% and 33.37% of the variation, respectively. In the 5200 m site, SOC and pH had the greatest impact on the bacterial community, followed by distance (Figure 10C). The first axis, mainly related to SOC, pH, and AN explained 60.96% of the variation. At that same time, the second axis was mainly related to Cd and AN explained 9.95% of the variation.

#### *3.5. Relative Effects of Environmental Factors on Bacterial Communities*

Pearson correlation coefficient analysis results between bacterial abundance and environmental factors are shown in Figure 11. At the phylum level, the abundance of *Proteobacteria* and *Fibrobacteres* was positively correlated with AK, SM, pH, and Cu (*p* < 0.01), and negatively correlated with SOC and TN (*p* < 0.05). *Actinobacteria* abundance was positively correlated with AK, SM, Zn, and Cu (*p* < 0.05). *Bacteroidetes* were negatively correlated with AK, SM, pH, Zn, and Cu (*p* < 0.05) and positively correlated with SOC, TN, and AN (*p* < 0.01). At the genus level, the abundance of *Lactobacillus* and *Enterococcus* was positively correlated with Cu, Zn, pH, and SM (*p* < 0.01), and negatively correlated with SOC (*p* < 0.05). The abundance of *Sphingomonas* and *Rubrobacter* was positively correlated with AK, Cu, pH, and SM (*p* < 0.001), and negatively correlated with SOC, TN, Density, and AN (*p* < 0.01). *Faecalibacterium*, *Prevotella* 9, and *Blautia* were negatively correlated with AK, SM, pH, and Cu (*p* < 0.01), and positively correlated with SOC and TN (*p* < 0.001). The abundance of *Bacteroides* was negatively correlated with AK, Zn, pH, and SM (*p* < 0.05). In addition, *Gemmatimonas* was positively correlated with Zn (*p* < 0.05) and negatively

correlated with SOC (*p* < 0.01). These results show that soil properties, especially SOC, TN, AN, AK, SM, pH, Cu, and Zn have significant effects on soil bacterial communities.

**Figure 10.** Ordination biplots of redundancy analysis (RDA) analysis of bacteria community structure, physical and chemical properties of soil and plant between treatments at 5, 20, 50, 100, and 400 m from the curb G109 highways at altitudes of (**A**) 4000 m, (**B**) 4600 m, and (**C**) 5200 m.

**Figure 11.** Pearson correlation heat map with correlation coefficient and significance levels based on the relative abundance of bacteria at the genus level and environmental factors. \* Means *p* < 0.05 for significance test; \*\* means *p* < 0.01 for significance test; \*\*\* means *p* < 0.001 for significance test.

#### *3.6. Bacterial Functional Community*

Effect of road traffic on soil bacterial function FAPROTAX function prediction results (Figure 12) showed that chemoheterotrophy, fermentation, animal parasites or symbionts, and nitrate reduction are the main functional groups along the highway. Among the top 30 functional predicted relative abundances, 25 species showed significant differences (*p* < 0.05): chemoheterotrophy, human pathogens, predatory or exoparasitic, aerobic ammonia oxidation, ureolysis, cellulolysis. The functional groups of ligninolysis, sulfate respiration and aromatic compound degradation increased in the disturbed area. The functional groups showing a downward trend in the highway disturbance area included fermentation, animal symbionts, photoheterotrophy, chlorate reducers and other functional groups. In addition, soil nitrifying bacteria such as ammonia oxidation, nitrite oxidation and nitrification showed an increasing trend in the disturbed area from the road compared with the control. Nitrate reduction and nitrogen fixation related to soil denitrification were significantly decreased in the sites within 100 m from the road.

**Figure 12.** Ecological functional diversity of soil bacterial community in different altitudes and distances. \* *p* < 0.05.
