The study area is situated in the Gander Mountain Core Protection Area of T. mongolica, located in the Hainan District of Wuhai City, Inner Mongolia, China (106°87′–106°89′E, 39°52′–39°55′N). This region falls within a temperate zone and experiences a continental climate characterized by cold winters, hot summers, minimal rainfall, strong winds, sandy terrain, and high thermal energy. The landscape is predominantly harsh and desert like.

In August 2019, three distinct desert plant communities in a protected area were selected as research sites. These included the community of T. mongolica, Reaumuria songarica, Salsola passerine, and Stipa capillata (Tm_Rs_Sp_S); T. mongolica and S. capillata (Tm_S); and the community dominated by T. mongolica (Tm) ( Supplementary Figure S1 ). A large sample plot measuring 100 × 100 m was established for each plant community type, with approximately 1.2 km separating each plot. Within each large sample plot, five small plots (10 × 10 m) were arranged in a five-point pattern. All T. mongolica plants in the small plots were selected for soil collection. After removing 3 cm of topsoil, root zone soil samples (Rz_Soil) were collected at a radial distance of 5 cm from the plants at depths of 0–20 cm in four cardinal directions: southeast, southwest, northwest, and northeast. Simultaneously, a bare soil samples (B_soil) was collected from the area closest to T. mongolica that was devoid of roots. All the samples from the same small plot were combined into a single sample. Thirty soil samples were collected at each sampling site. All samples were stored on ice in sterile zip bags until arrival at the laboratory. Each sample was divided into two portions: one was stored at -80°C for molecular biology research, and the other was air-dried and sieved to determine its physical and chemical properties.

pH was measured using a calcium chloride extraction pH-sensitive electrode. Soil organic matter (SOM) content was measured using the potassium dichromate external heating method. Organic carbon (OC) was quantified using the potassium dichromate oxidation-external heating method. Total nitrogen (TN) content was determined via Kjeldahl nitrogen determination after the elimination of concentrated sulfuric acid and hydrogen peroxide. Organic nitrogen (ON) was measured using the hydrochloric acid hydrolysis-distillation method. Ammonium-nitrogen (NH₄-N) was analyzed using the indigo blue colorimetric method after extracting the potassium chloride solution. Nitrate-nitrogen (NO₃-N) was determined using the dual-wavelength colorimetric method after the extraction of the potassium chloride solution. The total phosphorus (TP) content was determined using the molybdenum-antimony resistance colorimetric method after digestion with concentrated sulfuric acid and hydrogen peroxide. The available phosphorus (AP) content was determined using the molybdate blue colorimetric method after extraction with a 0.5 M NaHCO₃ solution. The total potassium (TK) content was measured using sodium hydroxide melt-flame spectrophotometry, and the available potassium (AK) was evaluated using ammonium acetate extraction-flame spectrophotometry. Water-soluble Ca²⁺, Mg²⁺, Na⁺, and K⁺ were quantified using deionized water extraction-flame spectrophotometry. Carbonate (CO₃ ^2-) was determined via deionized water extraction, followed by acid-alkali neutralization titration. Sulfate (SO₄ ^2-) was measured using deionized water extraction-barium sulfate turbidimetry, and chloride (Cl^-) was determined through deionized water extraction-silver nitrate titration. These standard methods were followed, as described by Bao (2000) and Lu (2000).

Total microbial genomic DNA was extracted from approximately 200 mg of soil using the E.Z.N.A.^® soil DNA Kit (Omega Bio-tek, Norcross, GA, US. PCR amplification of bacterial 16S rRNA and fungal ITS genes was conducted using Illumina-overhang-added primer pairs targeting the bacterial V4 region (515FmodF: 5’-GTGYCAGCMGCCGCGGTAA-3’ and 806RmodR: 5’-GGACTACNVGGGTWTCTAAT-3’) (Sampson et al., 2016; Walters et al., 2016) and fungal ITS1 region (ITS1F: 5’-CTTGGTCATTTAGAGGAAGTAA-3’ and ITS2R: 5’-GCTGCGTTCTTCATCGATGC-3’) (Sampson et al., 2016).

Purified amplicons were pooled in equimolar amounts and paired-end sequenced on an Illumina MiSeq PE300 platform (Illumina, San Diego, CA, USA) following the standard protocols established by Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China). The sequence reads were deposited in the NCBI Sequence Read Archive (SRA) database under accession numbers PRJNA1183314 and PRJNA1183552.

Bacterial and fungal OTUs were annotated using the Greengenes database (v13.8) (DeSantis et al., 2006) and the UNITE database (v8.2) (Nilsson et al., 2019), respectively, and reads that were not classified as bacterial or fungal were excluded. Bioinformatic analysis of soil bacteria and fungi was conducted using the Majorbio Cloud platform (https://cloud.majorbio.com). Microbial alpha diversity metrics, including Sobs, Shannon index, Simpson index, Heip evenness, Ace, and Chao1 richness, were estimated using Mothur v1.30.1 (Schloss et al., 2009). Both bacterial and fungal beta diversity analyses were performed using non-metric multidimensional scaling analysis (NMDS) with QIIME 2 (Bolyen et al., 2019) to calculate the distance matrix and the Vegan v2.5-3 package for analysis and plotting. Student’s t-test was used to evaluate differences in the relative abundance of microbial taxonomic groups. Venn diagram was created using the Venn diagram package in R (v1.6.20). Co-occurrence networks were constructed to explore internal community relationships across the samples (Barberán et al., 2012). A correlation between two nodes was considered statistically robust if Spearman’s correlation coefficient was >0.6 or <-0.6, with a P-value of <0.05. The functions of bacterial communities were predicted using the FAPROTAX tool (Louca et al., 2016). The functions of fungal communities were predicted using FUNGuild (http://www.funguild.org/). Student’s t-test was used to evaluate the differences in the functions of the microbial communities. Canonical correspondence analysis (CCA) and redundancy analysis (RDA) were performed using the Vegan v2.5-3 package to investigate the effects of soil physicochemical properties on the structure of the soil bacterial and fungal communities. The non-parametric permutational multivariate analysis of variance (PERMANOVA) test was used to assess the percentage of variation explained by the treatment, along with its statistical significance, using the vegan v2.5-3 package.

Various physicochemical characteristics of the samples were measured in the Rz_soil of T. mongolica in the three different plant communities ( Table 1 ). In these communities, the soil pH in the T. mongolica Rz_soil decreased, whereas AK increased when compared with B_soil, which was sampled from areas without plant roots closest to T. mongolica plants. In both plant communities, SOM, OC, TN, nitrogen-to-phosphorus ratio (N/P), and K⁺ increased with the establishment of the T. mongolica plantations. Conversely, NH₄-N and Na⁺ levels decreased in the T. mongolica plantations. NO₃-N, Mg²⁺, and CO₃ ^2- levels increased in T. mongolica plantations in only one plant community. ON, AP, and organophosphorus (OP) decreased in T. mongolica plantations in only one community. Across all three plant communities, the T. mongolica plantation had no significant effects on TP, TK, carbon-to-nitrogen ratio (C/N), carbon-to-phosphorus ratio (C/P), or Cl^-. Ca²⁺ and SO₄ ^2- levels increased in one community, decreased in the other, and showed no significant differences in the third community.

The indices of α-diversity, including Sobs, Shannon, Chao, Pielou, and Coverage, were calculated to quantify the diversity, richness, evenness, and sequencing depth of the microbial communities in the three plant communities. The coverage index for all the microbial communities was approximately 0.9, indicating that the sequencing capability was acceptable ( Table 2 ). T. mongolica significantly improved the Sobs and Ace diversity indices, as well as the Chao richness index, of bacterial and fungal communities in the Rz_soil compared with the B_soil. Shannon diversity and Heip evenness indices of the bacterial communities in the Rz_soil were significantly higher in the two plant communities, whereas the fungal communities showed significantly higher indices in only one plant community. However, there were no significant effects on the Simpson diversity index for either the bacterial or fungal communities across the three plant communities.

All samples from Rz_soil and B_soil were positioned in distinct regions along the coordinate axis in the NMDS figure based on the Bray-Curtis method (stress <0.2). The bacterial and fungal communities in Rz_soil were separated from those in B_soil along the x-axis across the three plant communities ( Figure 1 ). This finding demonstrates that T. mongolica plantations significantly altered the bacterial and fungal communities. However, the bacterial communities in the two soil types and the fungal communities in the Rz_soil were not distinctly different, and only the fungal communities in the B_soil exhibited a clear differentiation. This suggests that although the fungal communities in B_soil were distinct, the bacterial communities in B_soil showed less variation. In the T. mongolica plantation, both the bacterial and fungal communities in the Rz_soil exhibited minimal differences.

T. mongolica did not significantly affect the number and diversity of bacterial and fungal phyla with a relative abundance of >1% in the B_soil and Rz_soil across the three plant communities. Actinobacteria was the dominant phylum in all samples. The relative abundances of Proteobacteria, Chloroflexi, Acidobacteria, Planctomycetes, Gemmatimonadetes, Verrucomicrobia, and Bacteroidetes were ranked from the second-most dominant phylum to the subsequent phyla ( Supplementary Figure S2A ). However, the relative abundance of these phyla varied between B_soil and Rz_soil in the three plant communities. Notably, the relative abundance of Bacteroidetes was significantly higher in the Rz_soil than in the B_soil across all three plant communities. Conversely, the relative abundance of Actinobacteria decreased, whereas the relative abundance of Proteobacteria increased significantly between Rz_soil and B_soil in Tm_Rs_Sp_S and Tm plant communities. These changes were not significant in the Tm_S community of the T. mongolica plantation. In addition, the relative abundance of Planctomycetes was greater in Rz_soil than in B_soil; however, this was only observed in the Tm_Rs_Sp_S plant community. In contrast, the relative abundance of Gemmatimonadetes was lower in Rz_soil than in B_soil, specifically in the Tm_S plant community ( Supplementary Figures S2B–D ).

Ascomycota was the dominant phylum in all samples. Basidiomycota, unclassified_k:fungi, and Mortierellomycota exhibited higher relative abundance ( Supplementary Figure S3A ). The relative abundance of phyla also varied between B_soil and Rz_soil across the three plant communities. Specifically, the relative abundance of Ascomycota increased, whereas the relative abundance of unclassified_k: fungi decreased significantly between Rz_soil and B_soil in the plant communities Tm_Rs_Sp_S and Tm. However, this change was not significant in the Tm_S plant community of the T. mongolica plantation. The relative abundances of Glomeromycota and Calcarisporiellomycota were lower in Rz_soil than in B_soil; however, this was only observed in the plant communities Tm_S and Tm ( Supplementary Figures S3B–D ).

T. mongolica did not alter the dominant bacterial genera in the Rz_soil and B_soil across the three plant communities. Rubrobacter and norank_c:Actinobacteria were the dominant genera in all the soil samples from the three plant communities. However, the relative abundances of these two genera also changed between the B_soil and Rz_soil ( Figure 2A ). The relative abundances of Rubrobacter and norank_c:Actinobacteria declined significantly in the Rz_soil compared with those in the B_soil across the three plant communities ( Figures 2B–D ).

T. mongolica altered the dominant fungal genera in the Rz_soil and B_soil across three plant communities. Unclassified p:Ascomycota was the dominant genus in B_soil in Tm_Rs_Sp_S and Tm_S plant communities. Unclassified f_Ceratobasidiaceae emerged as the dominant genus in the B_soil of the Tm plant community. Fusarium was the predominant genus in all Rz_soils of the three plant communities ( Figure 3A ). The relative abundances of Fusarium, unclassified f:Pleosporales_fam_Incertae_sedis, unclassified f:Nectriaceae, and unclassified o:Hypocreales were significantly higher in the Rz_soil than in the B_soil in the Tm_Rs_Sp_S plant community. Conversely, the relative abundances of unclassified p:Ascomycota, Monosporascus, unclassified f_Ceratobasidiaceae, unclassified k:Fungi, and Darksidea were lower in the Rz_soil than in the B_soil. In the Tm_S plant community, the relative abundances of Penicillium, Acremonium, unclassified f:Pleosporales_fam_Incertae_sedis, and unclassified o:Hypocreales were also higher in the Rz_soil than in the B_soil.

The relative abundances of Fusarium, Penicillium, and Beauveria were higher, whereas those of unclassified_k:Fungi, Kalaharituber, Monosporascus, and Darksidea were lower in Rz_soil than in B_soil from the Tm plant community ( Figures 3B–D ). At the phylum level, the relative abundance of Ascomycota increased, whereas that of unclassified_k _fungi decreased significantly in the Rz_soil and B_soil across the two plant communities. At the genus level, the relative abundances of Fusarium and Penicillium were higher, whereas those of Monosporascus and Darksidea were lower in the Rz_soil than in the B_soil in both plant communities.

In all three plant communities, the specific OTUs of bacteria and fungi in the Rz_soils were more abundant than those in the B_soils. Thus, the T. mongolica plantation enhanced the OTUs in the Rz_soil compared with the B_soil. In the bacterial communities, OTUs were more common than specific OTUs in both Rz_soil and B_soil. However, in the fungal communities, common OTUs were higher in the B_soil and lower in the Rz_soil ( Supplementary Figure S4 ).

Network analysis results indicated that the number of nodes, total edges, average degree, and density of the bacterial co-occurrence network in T. mongolica Rz_soil were lower than those in B_soil within Tm_Rs_Sp_S and Tm_S plant communities. However, in the Tm plant community, the co-occurrence network parameters in T. mongolica Rz_soil were higher than those in the B_soil. In contrast, for the fungal co-occurrence network, the parameters in T. mongolica Rz_soil exceeded those in B_soil across all three plant communities ( Figures 4 , 5 ; Supplementary Table S1 ). For the bacterial co-occurrence network, the percentage of negative edges was similar to that of the positive edges. However, for the fungal co-occurrence network, the percentage of negative edges was considerably lower than that of positive edges. Therefore, T. mongolica appeared to decrease soil bacterial co-occurrence networks and enhance soil fungal co-occurrence networks.

Functional roles of the bacterial community were predicted using the FAPROTAX tool ( Figure 6 A ). T. mongolica enhanced the functional capabilities of bacterial communities, including denitrification (nitrate respiration, nitrate denitrification, nitrite denitrification, and nitrous oxide denitrification), ureolysis, dark hydrogen oxidation, methylotrophy (methylotrophy and methanol oxidation), organic matter decomposition (xylanolysis, chitinolysis, and cellulolysis), aromatic compound degradation, and photoconversion (photoheterotrophy, anoxygenic photoautotrophy, and anoxygenic photoautotrophy S oxidizing) ( Figures 6B–D ). Consequently, T. mongolica facilitated soil functions related to energy acquisition through denitrification, methylotrophy, hydrogen oxidation, and photoconversion, while also being influenced by organic matter decomposition, aromatic compound degradation, and photoconversion.

The function of the fungal community was predicted using the FUNGuild tool. T. mongolica increased the relative abundance of undefined saprotrophs, animal pathogens, endophytes, lichen parasites, plant pathogens, soil saprotrophs, and wood saprotrophs whereas decreasing the relative abundance of plant pathogens within fungal communities. Thus, T. mongolica promoted the growth of saprotrophic fungi and reduced the prevalence of pathogenic fungi ( Figure 7A ). However, these changes were not significantly different between Rz_soil and B_soil across the three plant communities ( Figures 7B–D ).

The effects of soil chemical properties on the bacterial and fungal communities were assessed using CCA and RDA. Among the 19 chemical properties examined, pH, TK, NH₄-N, OP, AK, and K⁺ had a more significant effect on the bacterial community than the other properties. Specifically, pH, TK, NH₄-N, and OP were positively correlated with B_soil whereas AK and K were negatively correlated with Rz_soil. OM, pH, TK, TN, AK, AP, OC, and K⁺ significantly affected the fungal community. pH was the most influential factor, showing a positive correlation with B_soil and a negative correlation with Rz_soil ( Figure 8 ).

Plants have a significant influence on the soil. This influence is multifaceted, encompassing the enhancement of the physical, chemical, and biological properties of the soil as well as the cycling and utilization of nutrients and soil microbes. Collectively, these factors promote the stability and sustainable development of ecosystems (Kranz et al., 2020; Liu et al., 2020; Furey and Tilman, 2021). This study demonstrated that T. mongolica reduced soil pH by decreasing Na⁺ and K⁺ content, increased AK, SOM, OC, and TN in desert soil. As a keystone species in desert communities, T. mongolica plays an irreplaceable role in maintaining local ecosystem stability and protecting the ecological environment through “plant-soil feedback.”

During plant growth, both acidic and alkaline substances are released, which can influence the acidity or alkalinity of soil. For instance, certain plants release organic acids during their growth process, resulting in increased soil acidity, whereas other plants may enhance soil alkalinity through root exudates (Tibbett et al., 2019). Yan et al. (2020) observed that tea plantations contribute to soil acidification, with pH levels decreasing by 0.47 to 1.43. In the fruit and vegetable systems, the pH decreased by 0.40 to 1.08, and in the cereal systems, it decreased by 0.30 to 0.89. Zhou et al. (2017) reported the pH was 4.58 in 78-year-old forest plantations of slash pine, 5.74 in Hoop pine, 6.01 in Kauri pine, and 4.49 in Eucalyptus. Jin et al. (2022) demonstrated that hickory plantations increased the pH of relatively acidic soils, but decreased the pH of relatively alkaline soils. The T. mongolica plantation also reduced the pH by 0.5 across the three plant communities. The environment in which T. mongolica thrives is the irrigation area in the middle and upper reaches of the Yellow River, characterized by long-term dryness, minimal rainfall, and high evaporation rates. Soil salinity is primarily composed of sulfates and chlorides. An arid climate and improper irrigation practices have led to the secondary salinization of the soil. The concentrations of water-soluble neutral salts, such as NaCl and Na₂SO₄, in the soil exceed 0.1%, whereas the concentrations of alkaline salts, such as Na₂CO₃ and NaHCO₃, exceed 5% (Li et al., 2023). Consequently, the soil is classified as severely saline-alkali. Our findings indicate that T. mongolica significantly decreased soil salinity ( Table 1 ). Na⁺ concentration decreased significantly in the T. mongolica plantation across both plant communities. However, the concentrations of Cl^-, SO₄ ^2-, and CO₃ ^2- were not reduced. Therefore, T. mongolica lowered soil pH by decreasing Na⁺ concentrations.

Plants can loosen soil, increase soil porosity, reduce soil compaction, and enhance soil physical properties via root activity. Plant residues and root exudates can increase the organic matter content of soil (Oleghe et al., 2019). Changes in soil acidity or alkalinity can influence the solubility and availability of mineral elements, thereby affecting their absorption and utilization by plants (Zhao et al., 2021). Our findings indicate that T. mongolica increased SOM and OC in the Rz_soil of T. mongolica compared with B_soil. We hypothesized that T. mongolica enhances C content in the soil through root sediments, root secretions, and litter. TN also increased; however, ON and NH₄-N levels decreased. We speculate that T. mongolica contributes to the N content in the soil through nitrogen fixation and recovery. In addition, T. mongolica appeared to accelerate the decomposition of ON and absorption of NH₄-N; however, it did not affect TP or AP transformation. In only one plant community was P absorption enhanced, leading to reduced AP levels. Although, T. mongolica did not affect TK, it increased AK by accelerating K release.

The organic matter produced by the T. mongolica plantation not only provides abundant carbon and energy sources for soil microorganisms but also enhances soil structure and increases soil biodiversity. In this study, we found that the bacterial and fungal communities in Rz_soil were distinct from those in B_soil across the three plant communities within the T. mongolica plantation. T. mongolica significantly improved the diversity index (Sobs) and the richness index (Chao) of both bacterial and fungal communities in the three plant communities. In this study, Rubrobacter and norank_c_Actinobacteria were the dominant genera across all the soil samples collected from the three plant communities. The presence of T. mongolica did not alter the dominant bacterial genera in the Rz_soil or B_soil of these communities. However, T. mongolica plantations affected the relative abundances of certain genera. Specifically, the relative abundance of Rubrobacter and norank_c_Actinobacteria significantly declined in the Rz_soil compared with that in the B_soil across the three plant communities. Both Rubrobacter and norank_c_Actinobacteria belong to the phylum Actinobacteria, which is known for its potential to resist plant diseases, promote plant growth, and enhance salt tolerance in plants (Passari et al., 2019; Xiong et al., 2019). The small perennial shrub T. mongolica decreases the abundance of beneficial bacteria in long-term plantations, a phenomenon commonly observed in crop cultivation (Schlatter et al., 2017; Lee et al., 2021).

In this study, we found that the unclassified phylum Ascomycota was the dominant genus in B_soils in Tm_Rs_Sp_S and Tm_S. The unclassified family Ceratobasidiaceae emerged as the dominant genus in the B_soil from the Tm plant community. Fusarium was the predominant genus in all Rz soils across all three plant communities. Fusarium is a potential fungal pathogen (van Agtmaal et al., 2017). The increased abundance of Fusarium, a potential plant fungal pathogen, requires the recruitment of antagonistic bacteria to maintain plant health. Additionally, T. mongolica altered the relative abundance of the dominant bacterial genera Rubrobacter and norank_c_Actinobacteria in both Rz_soils and B_soils in the three plant communities. The relative abundances of Rubrobacter and norank_c_Actinobacteria were significantly lower in Rz_soil than in B_soil across the three plant communities. Consequently, T. mongolica poses a high risk for root disease outbreaks.

The high-throughput sequencing technology employed in this study is capable of delineating the species and relative abundance of microbial community structure. However, it is crucial to acknowledge that relative abundance alone may not provide a comprehensive reflection of microbial quantities, particularly when assessing microbial community. To address this limitation, absolute abundance can be suggested. Jiang et al. (2019) observed that the relative abundance of Proteobacteria remained statistically unchanged in fertilizer application experiments conducted on tomato crops in coastal saline-alkali soil. Nonetheless, when absolute abundance quantification methods can be utilized, a significant escalation in the absolute abundance of Proteobacteria was detected. Similar findings have been documented in the studies by Yang et al. (2018) and Lou et al. (2018). Consequently, absolute quantitation of microbiota abundance is essential.

Methods of absolute quantitation of microbiota abundance encompass a variety of techniques, including the use of spike-in bacteria as documented by Stammler et al. (2016), the quantification of bacterial DNA through Flow Cytometry (FCM) as outlined by Vandeputte et al. (2017), and the integrated high-throughput absolute abundance quantification (iHAAQ) approach, which merges high-throughput sequencing with quantitative Polymerase Chain Reaction (qPCR), as introduced Lou et al. (2018). Additionally, Tkacz et al. (2018) presented an absolute quantitation method utilizing synthetic spike DNA. Guo et al. (2020) employed host-associated quantitative abundance profiling (HA-QAP) to delineate variations in the microbial load within the root microbiome. Additionally, Zhang et al. (2022) assessed two absolute microbiome profiling (AMP) methods in soil microbiota quantitative research, spike-AMP and qPCR-AMP, concluding that qPCR-AMP is superior for quantitative assessment of soil microbial communities. Wang et al. (2020) proposed an “amplification-selection” model for rhizosphere microbiome assembly, which utilizes synthetic chimeric spikes in plasmids for microbiome profiling. They hypothesized that plant root exudates can ‘feed’ soil microorganisms, nutrition-poor bulk soil is seen as “countryside”, while rhizosphere soil is as “metropolis”, where microbial communities can find more opportunities for growth and reproduction. The microorganisms in rhizosphere soil undergo amplification prior to the selection by the root system.

In this study, we observed T. mongolica notably enhanced the diversity indices (Sobs and Ace), as well as the richness index (Chao), of both bacterial and fungal communities across three plant communities. Moreover, the relative abundance of Bacteroidetes and Ascomycota in the Rz_soils was significantly higher than that in the B_soils. These findings aligned with “two-step or multiple-step selection” model, as inferred from the relative abundance data obtained through amplicon-based high-throughput sequencing. Therefore, absolute quantitation of microbiota abundance should be recommended in microbial ecology.

In this study, pH, TK, NH₄-N, OP, AK, and K had a more significant effect on the bacterial community in the soil than the other chemical properties. The pH, TK, NH₄-N, and OP were positively correlated with B_soil, whereas AK and K were negatively correlated with Rz_soil. For the fungal community, pH had the most substantial effect, showing a positive correlation with B_soil and a negative correlation with Rz_soil. Zhou et al. (2017) asserted that pH and vegetation are the primary factors influencing soil bacterial diversity and composition in the chronosequence of rubber trees (Hevea brasiliensis) plantations. Lauber et al. (2009) reported that soil pH significantly affects the structure of soil bacterial communities on a continental scale. Different microorganisms thrive within specific pH ranges, which are conducive to their growth and reproduction. Consequently, soil pH can influence the activity of soil microorganisms; if the pH of the soil solution falls outside the appropriate range, microbial activity is inhibited. Furthermore, the pH of the soil solution can alter the solubility of minerals, thereby affecting the nutrient availability in the soil and the activity of soil microorganisms (Philippot et al., 2024).

Soil enzymes play a crucial role in soil organic matter degradation, mineralization, and nutrient cycling. Their activities significantly influence soil nutrient content (Burns et al., 2013; Dotaniya et al., 2019). By measuring the soil enzyme activity, we can gain a deeper understanding of how T. mongolica affects soil properties from a protein perspective. Unfortunately, the soil enzyme activity between Rz_soil and B_soil in soil samples from three plant communities. It is one of further research work.

Microorganism in soil is another biological factor that affects soil nutrition (Zhang et al., 2021; Coban et al., 2022; Philippot et al., 2024). The high-throughput sequencing technology is common method for reveal microorganism in soil (Nkongolo and Narendrula-Kotha, 2020). In this study, throughput sequencing technology based on amplicon was used to determine microbial communities. However, metagenomic sequencing technology, which is superior to amplicon sequencing, can provide insights into bacteria and fungi involved in biogeochemical cycles, not only soil microbial community composition, but also their functional genes and metabolic pathways. This is particularly for the C cycle (including CO₂ fixation and respiration), N cycle (encompassing nitrification, denitrification, and N₂ fixation), P cycle, and S cycle (including sulfur assimilation, anaerobic sulfate respiration, and sulfide oxidation) (Simon and Daniel, 2011; Scholz et al., 2012). Further research in this area is required.