Chemical properties analysis of the soil samples
The chemical properties of 24 soil samples with varying degrees of salinization were analyzed and are presented in Supplementary Table 2. In all of the soil samples studied, the pH value was 8.27–10.51, and the CV was 6.89%. The CVs of AN, AP, AK, SOM, HCO3−, CO32−, Na+, K+, and EC were 45.37%, 59.36%, 42.06%, 41.575%, 42.69%, 94.53%, 79.86%, 49.79%, and 69.86%, respectively. The CV of Cl−, SO42−, Ca2+, and Mg2+ were 139.27%, 132.03%, 154.56%, and 119.54%, respectively. The results indicated that each index of the soil samples had obvious variation, except for the pH (Table 1), which might have affected the microbial community structure.
Table 1
The coefficient of chemical characteristics of the collected soil samples
Indicators | Maximum value | Minimum value | Average | Standard deviation | Variation coefficient (%) |
pH | 10.51 | 8.27 | 8.96 | 0.62 | 6.89 |
AN | 73.05 | 2.68 | 36.40 | 16.52 | 45.37 |
AP | 59.14 | 6.28 | 23.81 | 14.14 | 59.36 |
AK | 365.85 | 87.79 | 206.83 | 87.00 | 42.06 |
SOM | 9.04 | 0.36 | 5.18 | 2.15 | 41.57 |
Cl− | 4.48 | 0.13 | 0.66 | 0.91 | 139.27 |
SO42− | 7.55 | 0.13 | 1.61 | 2.13 | 132.03 |
HCO3− | 0.79 | 0.14 | 0.34 | 0.14 | 42.69 |
CO32− | 0.71 | 0.08 | 0.17 | 0.16 | 94.53 |
Ca2+ | 2.70 | 0.60 | 0.49 | 0.76 | 154.46 |
Mg2+ | 0.79 | 0.01 | 0.18 | 0.22 | 119.54 |
Na+ | 1.49 | 0.05 | 0.41 | 0.33 | 79.86 |
K+ | 0.25 | 0.03 | 0.12 | 0.06 | 49.79 |
EC | 4.10 | 0.20 | 1.37 | 0.96 | 69.86 |
Bacterial and fungal community analyses
The effects of soil chemical factors on the bacterial and fungal communities were assessed by using the ABT model (Fig. 2). The largest driving force in different saline-alkali soil bacteria was EC, which accounted for 22.80%, followed by AK (19.42%) and SOM (13.20%). Among the fungi, EC emerged as the predominant driving force (21.30%), closely followed by AK (17.27%) and SOM (12.11%). The samples were initially categorized into three groups based on their EC level, namely the low-salinity group (L), medium-salinity group (M), and high-salinity group (H). The EC for L processing ranged from 0 to 1 ms/cm (n = 10), the EC for M processing ranged from 1 to 2 ms/cm (n = 8), and the EC for H processing exceeded 2 ms/cm (n = 6). The PCoA test results revealed a significant correlation between bacteria (R = 0.5835, p < 0.001) and fungus (R = 0.4121, p < 0.001). However, the bacterial and fungal communities in different saline-alkali soils exhibited clear isolation (Fig. 3), indicating that the L, M, and H groups, which were categorized based on the EC levels, were highly representative.
Changes in the microbial alpha diversity
The alpha diversity analysis revealed significant variations in the composition of the top ten bacterial and fungal species across different saline–alkaline soils at the phylum level (Fig. 4). The abundance of bacterial species was higher, whereas the fungal diversity was lower. In the Hetao irrigation area of the Inner Mongolia region, Proteobacteria, Actinobacteriota, and Chloroflexi are the predominant bacterial groups in various saline–alkaline soils. The relative abundance of Proteobacteria was 22% in L, 26% in M, and 30% in H (Supplementary Table 3). The relative abundance of Actinobacteriota was 19% in L, 19% in M, and 17% in H. The relative abundance of Chloroflexi was 18% in L, 16% in M, and 11% in H. The top three fungal communities at the phylum level were Ascomycota, Basidiomycota, and Mortierellomycota. The prevalence of Ascomycota was significantly higher in various saline-alkali soils, accounting for 83% in H, 86% in M, and 76% in L. The number and species diversity of bacterial and fungal communities in various saline-alkali soils were analyzed. The Chao, Shannon, and Simpson indices of the bacteria and fungi exhibited a negative correlation with increasing salinity (Fig. 5), and the treatments exhibited notable variations (p < 0.05). The species and quantity of bacteria and fungi indicated the highest level in L, whereas H exhibited the lowest level. The findings revealed that the abundance and diversity of bacteria and fungi were the lowest in the high saline-alkali soil.
Microbial co-occurrence networks
The microbial communities in the Hetao saline-alkali soil of Inner Mongolia exhibit a high degree of complexity and diversity, thus warranting the selection of the top 200 bacterial and fungal species (Fig. 6). The results of network correlation analysis revealed significant variations in the network parameters of bacteria and fungi across various saline-alkaline soils (p < 0.05). The bacterial network of L treatment had 177 points and 1,173 edges, and the modularity was 0.398 with 8 modules (Table 2). The M treatment had 192 points and 1,324 edges, and the modularity was 0.457 with 8 modules. The H treatment had 199 points and 2,371 edges, and the modularity was 0.668 with 6 modules. The fungal network of L treatment had 185 points and 631 edges, and the modularity was 0.551 with 10 modules. The M treatment had 191 points and 1,249 edges, and the modularity was 0.564 with 8 modules. The H treatment had 194 points and 1,253 edges, and the modularity was 0.729 with 6 modules. The network points, network edges, and modularity coefficients exhibit an upward trend with increasing salinity, while the number of modules shows a decreasing pattern.
The 10 nodes with the largest number of edges were defined as network hubs, which were active mediators in the bacterial community network (Supplementary Table 4). The Acidobacteriota, Proteobacteria, and Chloroflexi taxa represent the central nodes of bacterial communities in diverse saline-alkaline soils, including 70% network hubs at the phylum level. Half of the network hubs in the L treatment belonged to Acidobacteriota and Proteobacteria, such as ASV198, ASV692, ASV209, ASV149, ASV1596. The hubs of M treatment had similar phylogenetic classifications to the L treatment, with half assigned to Chloroflexi and Proteobacteria, such as ASV1411, ASV90, ASV1423, ASV2436, ASV9. However, at the highest salinity, the 40% hubs were classified as Proteobacteria, for instance, ASV2325, ASV81, ASV47, ASV608. The findings demonstrate that Acidobacteriota thrives in low to moderate salinity environments, whereas the majority of Proteobacteria species exhibit adaptability to high salinity conditions. The Ascomycota and Mortierellomycota taxa represent the central nodes of fungal communities in diverse saline-alkaline soils, including 90% network hubs at the phylum level. The Ascomycota community comprised 66.67% of the overall network hub. Therefore, at the three saline-alkali treatment, more than half hubs were classified as Ascomycota. The findings suggest that Ascomycota exhibits adaptability to diverse soil environments with varying salinity and alkalinity levels, and it represents the predominant fungal community in saline-alkali soils of Hetao Plain in Inner Mongolia.
Table 2
Topological characteristics of co-occurrence networks of soil microbial communities in various saline-alkaline soils. (corresponding to Fig. 6).
Toplogical features | L | M | H |
Bacteria Network metrics | | | |
Nodes | 177 | 192 | 199 |
Edges | 1324 | 1173 | 2371 |
Modularity | 0.398 | 0.457 | 0.668 |
Network diameter | 7 | 6 | 6 |
Average degree | 14.96 | 12.219 | 23.829 |
Graph density | 0.085 | 0.064 | 0.12 |
Average path length | 2.838 | 2.925 | 2.97 |
Average clustering coeffcient | 0.483 | 0.435 | 0.761 |
Fungi Network metrics | | | |
Nodes | 185 | 191 | 194 |
Edges | 631 | 1249 | 1253 |
Modularity | 0.551 | 0.564 | 0.729 |
Network diameter | 9 | 10 | 9 |
Average degree | 6.822 | 13.079 | 12.918 |
Graph density | 0.037 | 0.069 | 0.067 |
Average path length | 3.66 | 3.427 | 4.024 |
Average clustering coeffcient | 0.402 | 0.537 | 0.673 |
Relationship between soil factors and soil microbial communities
We investigated the relationship between environmental factors and the core flora by analyzing heat maps of bacterial and fungal communities and the soil characteristics in different saline–alkaline soils (Fig. 7). The abundance of bacteria and fungi exhibited a significant correlation with the chemical properties of the soil (p < 0.5). The bacterial dominant phylum of Acidobacteriota, Proteobacteria, and Chloroflexi exhibited significant correlations with AP, AK, Cl−, SO42−, HCO3−, CO32−, Ca2+, Mg2+, Na+, and EC (Supplementary Fig. 3). The fungal dominant phylum of Ascomycota, Basidiomycota, and Mortierellomycota exhibited significant correlations with pH, AP, AK, HCO3−, CO32−, Ca2+, Mg2+, Na+, K+, and EC (Supplementary Fig. 4). The variation distribution analysis (VPA) revealed that the predominant drivers of bacterial and fungal community changes in different saline–alkaline soils were the soluble salt ion components, accounting for 12.36% (bacterial) and 22.92% (fungal), respectively.
Contrasting determinants of bacterial and fungal beta diversities
The SEM was employed to elucidate the potential mechanisms of soil factors influencing the spatial distribution of the microbial community structure (Fig. 8). The data of the model was χ2/df = 2.807, p = 0.207, CFI = 0.891, GFI = 0.915, and RMSEA = 0.048. These findings demonstrated a significant impact of fungal and bacterial richness in various saline-alkali soils (p < 0.001). The negative correlation coefficients of EC with bacterial and fungal communities in different saline-alkali soils were 0.91 and 0.86, respectively. The positive correlation coefficients of SOM with the bacterial and fungal communities were the largest at 1.09 and 0.84, respectively. The test further demonstrated that EC exerted the greatest influence on the structure of bacterial and fungal communities, thereby exhibiting a significant negative correlation on the bacterial and fungal community composition. The bacterial and fungal communities of the saline-alkali land in the Inner Mongolia Hetao Plain were directly influenced by soil salt and alkali ions, whereas the soil nutrients indirectly affected these communities.