Diversity of 16S/18S rDNA genes in different regions

16S/18S rDNA genes sequencing was used to study the population characteristics of rhizosphere bacteria and eukaryotes in wild apples from different regions of Xinjiang. After filtering, denoising, and removing chimeras, the average number of bacterial sequences per sample was 54,992, with an effective rate of 79.26%. The average number of eukaryotic sequences was 108,548, with an effective rate of 84.79%. The bacterial sequences were clustered into 15,789 operational taxonomic units (OTUs), and eukaryotic sequences were clustered into 16,853 OTUs. The coverage of all samples was > 0.99, and with the rarefaction and Shannon curves converged. This indicates that sampling was reasonable, and that the sequencing was sufficient to characterize the diversity of the study sites. A plot shows (Fig. 2) that there were 47 core OTUs of bacteria and 184 core OTUs of eukaryotes, respectively, in wild apples rhizosphere soil samples from eight regions in Xinjiang. The number of bacterial OTUs in regions C, D, G, and H was greater than that of eukaryotic OTUs, whereas the number of bacteria in regions A, B, E, and F was lesser than that of eukaryotic OTUs. The results showed that the number of OTUs in each soil sample varied across the eight regions. The bacterial Chao 1, ACE, Shannon and Simpson indices in regions C, D, G, and H were higher than the eukaryotic biodiversity indices, whereas the bacterial diversity indices in regions A, B, E, and F were lower than the eukaryotic biodiversity indices (Fig. 3). The α-diversity analysis showed that species diversity in different regions was quite different. The bacterial species diversity in regions C, D, G, and H was higher than in other regions whereas regions A, B, E, and F had higher species diversity than in other regions.

Fig. 2

Petal plot based on OTUs: (a) bacterial (b) eukaryotic. Different colored petals represent the number of OTUs of samples from different regions, and the number of cores represents the number of OTUs common to all samples. A to H represent Daxigou in Huocheng County, Wild Fruit Forest in Xinyuan County, Resource Nursery, Damorhu in Gongliu County, Xiaomorhu in Gongliu County, Gongliu County Nazi Work Team, Laofengkou Guozigou in Tuoli County, and Yeguolin Scenic Area in Emin County

Fig. 3

Cluster analysis results in sequence number and diversity/richness index of 97%. (a) bacterial (b) eukaryotic. Different colors represent regions A to H

PCoA was used to rank at OTU level to reveal similarities or differences in community structure between different regional groups. The first and second axes of the bacterial community structure contributed 41.3% and 13.4% of the explanation, respectively, and 16.4% and 13.7% for eukaryotes, respectively. (Fig. 4a,b). In general, the majority of samples from each group clustered together, indicating significant differences in the community composition of bacterial and eukaryotic species.

Fig. 4

OTU PCoA based on Bray–Curtis distance method: (a) bacterial (b) eukaryotic. Different colors represent samples from different regions, A to H represent different regions

Rhizosphere bacteria species composition in different regions

Compared to the Silva138 database, the proportion of phylum, class, order, family, genus, and species in bacteria was 97.78%, 97.24%, 95.8%, 94.24%, 84.63%, and 32.39% respectively. The 16S rRNA gene sequences were divided into phyla, and the nine most abundant phyla were Firmicutes, Proteobacteria, Actinobacteriota, Bacteroidota, Acidobacteriota, Verrucomicrobiota, Chloroflexi, Planctomycetota, and Methylomirabilota, unclassified (Fig. 5a). Firmicutes were the most dominant bacterial phylum in regions of A (56.47%), C (24.46%), D (32.45%), E (57.58%), and F (56.43%), and Proteobacteria were the most dominant bacterial phylum in regions B (40.82%), G (54.70%), and H (46.23%). The community composition in regions A, E, and F was less; in particular, the relative abundance of Acidobacteriota, Verrucomicrobiota, Planctomycetota, Chloroflexi, and Methylomirabilot was very low. The relationships between the sequential datasets were visualized by non-metric multidimensional scale (NMDS), and the samples in each region were basically clustered together, but the distance between groups was large, with a P ≤ 0.001 (Fig. 5b), indicating differences in the composition of bacterial communities in different regions.

Fig. 5

Composition of 16S rDNA species communities. a Relative abundance of bacterial communities at the phylum level. b Nonmetric multidimensional scale (NMDS) ordering of ASV-level data. c Clade map of LEfSe analysis of major bacterial differential microbiota in eight regions. Each small circle on a different classification level represents a classification at that level, and the diameter of the small circle is proportional to the relative abundance. The coloring principle is to uniformly color the species with no significant difference as yellow and the other species with differences according to the species. The most abundant groupings are colored. d Bacterial association network analysis diagram. The nodes represent individual family, and the size of the nodes represents the average relative abundance in the sample. The connections between the nodes indicate a correlation between the two family, the red line indicates a positive correlation, and the blue line indicates a negative correlation. The thickness of the line is proportional to the correlation between family, and the thicker the line, the stronger is the correlation. At the same time, the more connections through a node, the more closely related the family are to other members of the flora

Linear discriminant analysis effect size (LEfSe) analysis was used to identify microorganisms explicitly enriched in bacteria from the phyla-to-species level in different regions. The results showed that there were 41 phylum-to-species differences in the eight regions. In the bacterial species community, the various species in region A were Bifidobacteriaceae, Actinobacteria, and Coriobacteriales. The different communities in region B were Bacteroidaceae and uncultured_bacteria. Chloroflexi, Bacillus_luciferensis, and uncultured bacteria were the different species in region C. The main communities in region D included Vicinamibacteraceae, Blastocatellia, and Pyrinomonadaceae. There were no apparent species differences in regions E and F, but uncultured_bacteria were different in region G. Prevotella and Bacteroidia were different in region H (Fig. 5c). Figure 5d shows that the rhizosphere bacterial family of wild apples co-occurred. The results showed that the rhizosphere bacterial communities of wild apples in different regions cooperated, and a few competed.

Rhizosphere eukaryotic species composition in different regions

The proportion of the eukaryotic phylum level was 83.17%, that of the class level was 75.23%, that of the order-level was 69.39%, that of the family level was 61.79%, that of the genus level was 53.09%, and that of the species-level was 30.05%. The 18S rRNA gene sequences were divided into phylum levels, among which the nine most abundant phyla were Ascomycota, Phragmoplastophyta, Basidiomycota, Cercozoa, Ochrophyta, Ciliophora, Mucoromycota, Chytridiomycota, and Chlorophyta, unclassified (Fig. 6a). Eukaryotes were dominated by Ascomycota, Phragmoplastophyta, and Basidiomycota. Ascomycota was the most dominant phylum in regions A (44.71%), B (23.23%), D (38.59%), E (39.98%), F (29.58%), G (37.94%) and H (42.70%), and Phragmoplastophyta was the most dominant phylum in region C (40.42%) regions. NMDS analysis showed that the samples in region C were far apart, and the samples in each area were clustered together, with a P ≤ 0.001 (Fig. 6b), indicating differences in the eukaryotic community composition in different regions.

Fig. 6

Composition of 18S rDNA species communities. a Relative abundance of eukaryotic communities at the phylum level. b Nonmetric multidimensional scale (NMDS) ordering of ASV-level data. c Clade map of LEfSe analysis of the significant differential microbiota of eukaryotes in the eight regions. Each small circle on a different classification level represents a classification at that level, and the diameter of the small circle is proportional to the relative abundance. The coloring principle is to uniformly color the species with no significant difference as yellow, and the other species with differences according to the species. The most abundant groupings are colored. d Diagram of eukaryotic association network analysis. The nodes represent individual species, and the size of the nodes represents the average relative abundance in the sample. The connections between the nodes indicate a correlation between the two species, the red line indicates a positive correlation, and the blue line indicates a negative correlation. The thickness of the line is proportional to the correlation between species, and the thicker the line, the stronger is the correlation. At the same time, the more connections through a node, the more closely related the species are to other members of the flora

LEfSe analysis showed significant differences in the leading 41 phyla-to-species in the eight different regions. In the eukaryotic community, the main differential species in region A were Trichoderma, Nectriaceae, and Phallus_hadriani. The differential species in region B was Cercomonadidae, and there was no apparent difference in fungal species. In region C, there was also no significant difference in species. Pleosporales and Dothideomycetes were the differential species in region D. Solicoccozyma and Piskurozymaceae were the differential species in region E. The differential species in region F included Filobasidiales. Helotiales, Incertaesedis, Leotiomycetes, and Geminibasidium were the differential species in region G (Fig. 6c). Figure 6d shows that the rhizosphere eukaryotic community of wild apples co-occured, and the network structure was different from that of bacterial. The results showed that the rhizosphere eukaryotic communities of wild apples in different regions cooperated with each other, and a few competed.

Relationship between β-diversity and environmental factors

The impact of geographic distance, climatic distance and soil pH on the composition of rhizosphere bacterial and eukaryotic communities in wild apples was examined by mantel analysis. It was found that the correlation between climatic distance and β-diversity was greater than that between geographic distance and soil pH (Fig. 7a). Geographic and climatic distances correlate were more strongly correlated with bacterial β-diversity than with eukaryotic β-diversity. Geographic distance remained positively correlated with eukaryotic β-diversity even after controlling for climatic distance and/or soil pH (Fig. 7b). The results indicated that geographic and climatic differences were important predictors for microbial β-diversity. Soil pH and climatic distance were positively with bacterial β-diversity and insignificantly with eukaryotic β-diversity when geographic distance was controlled (Fig. 7b). Climate distance was negatively with microbial β-diversity when the soil pH was controlled. The converse was also true. In summary, there was a correlation between geographical distance, climatic distance and soil pH with microbial β-diversity.

Fig. 7

Relationships between environmental factors and bacteria and eukaryotic β-diversity. a Bivariate associations between β-diversity and environmental factors (geographic, climatic, and soil pH distances). Lines represent the trends of the bivariate associations. Spearman correlation coefficients (rho) and the lower and upper 95% confidence intervals of simple Mantel tests are shown. b Partial Mantel tests for the bivariate associations between environmental factors and β-diversity. Symbol | represents partial Mantel tests. Variables in the left of | represent the independent variables, and in the right of | represent the controlling independent variables. Points represent the Spearman correlation coefficients (rho) and error bars 95% confidence intervals. Geo—geographic distance; Clim—climatic distance; pH—soil pH distance. Geographic distance has a unit of km, and the other distance metrics are unitless