3.1 Structure and composition of Ferula sinkiangensis rhizosphere soil bacterial community.
All samples contained 4,634,264 raw tags in total, filtering out the low quality tags and removing the chimeric sequences to obtain the final 4,405,120 effective sequences. On average, these reads were grouped into 4,017 bacterial OTUs and contained a total of 75,570 taxon tags per sample (Supplementary Table S3).
Species sequence analysis identified 60 phyla and 901 genera of bacteria. The most abundant phyla were (Figure 1): Actinobacteria (22.7%), Proteobacteria (18.6%), Acidobacteria (14.0%), Gemmatimonadetes (10.1%), Cyanobacteria (7.9%), Bacteroidetes (6.9%), Planctomycetes (3.9%), Verrucomicrobia (3.5%), Firmicutes (3.4%), and Chloroflexi (3.2%). The most abundant genera were (Supplementary Figure S2): RB41 (3.03%), Sphingomonas (1.64%), Rubrobacter (1.43%), Gaiella (1.10%), Pseudarthrobacter (1.01%), Solirubrobacter (0.85%), Gemmatimonas (0.71%), Bacteroides (0.63%), Acinetobacter (0.57%), Adhaeribacter (0.57%), Haliangium (0.47%), Staphylococcus (0.43%), Rhodococcus (0.42%), Pseudomonas (0.40%), Bacillus (0.38%); the remaining 76.59% include genera outside the top 30 of the annotated list and the parts not annotated. Moreover, multivariate analysis of variance revealed significant differences between Shannon and Chao1 indexes at the top and bottom of the slope. The diversity and richness of the bacterial community were affected by the slope position, as illustrated by a higher bacterial diversity and richness at the top of the slope compared to the bottom. In addition, there were significant differences in Shannon and Chao1 indices at different soil depths and different rhizosphere regions (rhizosphere and non-rhizosphere). The diversity and richness of bacterial communities are thus affected by soil depth and rhizosphere regions. The diversity and richness of bacterial communities at soil depths of 0-10 cm were significantly higher than that in 10-25 cm and 25-40 cm. Furthermore, the diversity and richness of bacteria were significantly higher in the rhizosphere than in the non-rhizosphere (Supplementary Table S4).
3.2 Relationships between root zone (rhizosphere and non-rhizosphere), slope position, soil depth and dominant bacteria (phyla and genera).
Multivariate analysis of variance and LDA effect size analysis showed that Cyanobacteria, Actinobacteria, Acidobacteria, Chloroflexi and Firmicutes were significantly changed in the rhizosphere and non-rhizosphere. The specific abundance of Actinobacteria, Chloroflexi and Acidobacteri in the rhizosphere was significantly higher than that observed in the non-rhizosphere. However, the relative abundance of Cyanobacteria and Firmicutes in the non-rhizosphere was significantly higher than that observed in the rhizosphere. Moreover, Firmicutes, Planctomycetes and Verrucomicrobia were significantly affected by the slope position, as shown by a significantly higher relative abundance of Firmicutes at the bottom of the slope compared to the top of the slope. The opposite was observed for Verrucomicrobia. In addition, the relative abundance of Planctomycetes in the middle and top of the slope was significantly higher than at the bottom of the slope. Soil depth had a significant effect on the relative abundance of Cyanobacteria, Bacteroidetes, Acidobacteria and Planctomycetes. The relative abundance of Planctomycetes and Acidobacteria at a 0-10cm soil depth was significantly higher than at 25-40cm, and Bacteroidetes were more abundant at 0-10cm than at a 10-25cm soil depth. However, the relative abundance of Cyanobacteria at 10-40cm soil depth was significantly higher than that observed at 0-10cm soil depth (Figure 2 and Supplementary Table S5). Figure 2 also shows specific differences in the distribution of different bacterial genera.
3.3 Relationship between slope position, soil depth and Ferula sinkiangensis rhizosphere soil physicochemical properties.
Spearman correlation analysis showed that most variables including total phosphorus content (TP), ammonium nitrogen content (AN), nitrate nitrogen content (NN) and available phosphorus content (AP) are associated with altitude (Table 1). Specifically, altitude was negatively correlated with AN and NN. On the other hand, a significant positive correlation was observed between altitude and TP or AP. Total organic carbon content (TOC), total nitrogen content (TN), and total potassium content (TK) were positively correlated with AP. TOC was positively correlated with TN, total salt content (TS) and AP. TS was negatively correlated with pH, and TP was positively correlated with altitude. However, TP was negatively correlated with AN (Table 1). In addition, multivariate analysis of variance showed that slope position was significantly correlated with most soil physicochemical properties including TOC, TP, TK, NN, AN, AP, and TS. However, only a few soil physicochemical properties, including AP, TS and pH, had a significant correlation with depth. Specifically, TOC, TP, TK, AP, and TS were significantly higher at the top than the middle of slope. The contents of TP and AP at the top of slope were significantly higher than those at the bottom and middle of slope. Interestingly, the content of AN was significantly higher at the bottom than at the top and middle of slope. In terms of depth, the content of AP and pH in 0-10 cm soil were significantly higher than those in 25-40 cm soil (Supplementary Table S6).
3.4 pH largely explained the variation of Ferula sinkiangesis rhizosphere bacterial community structure
Across all samples, soil AP showed a significant positive correlation with α-diversity of the bacterial community(r=0.538, p<0.01). Soil TP and TK and the α-diversity displayed a significant positive correlation (Shannon index, r = 0.495 and 0.405, respectively, p < 0.05, Table 2). In addition, distance-based redundancy analysis (db-RDA) also showed a correlation between soil physicochemical properties and the distribution of bacterial communities in the rhizosphere of Ferula sinkiangesis (Figure 3). All the soil physicochemical factors explained 29.81% of the variation in the rhizosphere bacterial community structure of Ferula sinkiangensis. The pH explained 5.58% of variation, altitude explained 5.53%, TS 5.21%, TP 4.90%, NN 3.89% and AP explained 3.60%. Among them, pH, altitude, TS and TP explained the largest proportion of the variation of the bacterial community structure in the rhizosphere of Ferula sinkiangensis (Supplementary Table S7).
3.5 Spearman correlation analysis between relative abundance of dominant bacteria (phyla and genera) and soil physicochemical properties.
Although the relative abundance of bacterial phylum has a significant correlation with slope position, soil depth and root zone, the relative abundance of bacterial also has a significant correlation with soil Physicochemical Properties. Table 3 shows the relationship between the top ten dominant bacteria phyla and soil Physicochemical Properties. Specifically, TN, AP and TS have a significant positive correlation with the relative abundance of Actinobacteria; AP and TS showed significant positive relationships with Chloroflexi; pH showed significant positive relationships with Bacteroidetes; Altitude have significant positive relationships with Gemmatimonadetes; AN have significant positive relationships with Verrucomicrobia. Conversely, NN and TS significantly negative correlated with the relative abundances of Gemmatimonadetes; TN and AP showed significant negative relationships with Cyanobacteria. Interestingly, Firmicutes, Proteobacteria, Planctomycetes and Acidobacteria were not significantly related to each variable (Table 3). Moreover, the relationship between the relative abundance of the first thirty-five bacterial phyla and the Physicochemical Properties of the soil is shown in Supplementary Figure 3. In addition, correlation analysis between relative abundance of bacteria genera and soil physicochemical properties showed that pH and altitude were significantly correlated with most bacteria genera. For example, pH has significant positive correlation with the relative abundance of Adhaeribacter, Altererythrobacter, Gemmatimonas, Hymenobacter, Massilia, Opitutus, Rubellimicrobium and Sphingomonas (Figure 4A). Altitude has significant negative correlation with Blastococcus, Opitutus, RB41, and Rubellimicrobium (Figure 4B).