Heavy Grazing Reduces the Diversity of Soil Microbial Communities in Meadow Grassland Under Long-Term Grazing

Background: Soil microorganisms are an important part of the grassland ecosystem and promote material transformation and energy ow in the entire ecological environment. Moreover, Hulun Buir grassland is the material basis for the development of animal husbandry. Therefore, it is of great scientic signicance to study the changes of soil microbial community caused by grazing in Hulunbuir grassland for the sustainable and stable development of grassland ecosystem. Methods: The present research used high-throughput sequencing of soil microorganism (bacteria and fungi) genes to compare microbial communities in 6 levels of grazing intensity (0.00, 0.23, 0.34, 0.46, 0.69, and 0.92 Au ha -1 ) under the Hulun Buir structure and the diversity characteristics of Leymus chinensis meadow steppe. Results: The 0-10 cm soil layer of the G0.34 test area had the highest content, and the content of the G0.92 test area was lower than the other grazing treatments. With increasing depth, the carbon and nitrogen contents of microorganisms decreased. The diversity of soil bacteria in the light grazing test area (0.23Au ha -1 ) was higher than the heavy grazing area, and the diversity of fungi in the non-grazing area was higher than the specic grazing areas. Most bacterial species were enriched in the G0.00 grazing areas, and the other grazing intensities were less abundant. The underground biomass (P = 0.039) signicantly inuenced the bacterial community structure, and pH (P =0.032), total nitrogen (P =0.011) and litter (P =0.007) signicantly inuenced the fungal community. Conclusions: In conclusion, the structures of bacterial and fungal communities are very sensitive to grazing and varied with grazing intensity. Our ndings demonstrated that a grazing intensity of approximately 0.23 Au ha -1 was the most appropriate for the grassland of the meadow in Hulun Buir. area (0–10 cm soil layers) were collected Simultaneously. Soil moisture (SM) was measured by the oven-drying method, and the gravimetric soil water content was determined by drying the sampled soil at 105 ℃ for 24 h. Soil pH was measured in a 1:2.5 soil/water mixture by a multi-parameter water quality analyser. Soil total nitrogen (TN) concentration was measured the semi-micro Kjeldahl determination. The dichromate oxidation and sodium hydroxide alkali-molybdenum-antimony colorimetric methods were used to determine the soil organic carbon (SOC) concentration, and soil total phosphorus (TP) concentration. A spectrophotometer was used to measure soil total potassium (TK) concentration, soil available phosphorus (SAP) concentration and soil available potassium (SAK) concentration. Soil ammonium nitrogen and nitrate nitrogen (NH 4+ -N and NO 3− -N) were determined using a ow injection auto-analyzer (FIAstar 5000 Analyzer, Foss Tecator, Denmark). Soil microbial biomass carbon (MBC) and nitrogen (MBN) were measured using fumigation-volumetric analysis and the fumigation-ninhydrin colorimetric method, respectively. SAP. BGB. signicant correlations Fusarium, Mortierella and SBD were signicantly positively correlated (P < 0.05). Gibberella signicantly correlated with SM (P < 0.05), SBD (P < 0.05), TK (P < 0.01), NH4+ -- N (P < 0.05), Mull (P < 0.001), and coverage (P < 0.01). These results showed a signicant correlation between Penicillium and SM (P < 0.05), SBD


Introduction
Soil microorganisms are largest underground repository of grassland ecosystems and play an important role in the grassland ecological system in the promotion of the ow of material and energy transfer in the entire environment and the maintenance of the grassland ecological system [1]. Soil microorganisms are extremely sensitive to changes of environment [2]. Grazing is the main use of grassland ecological systems [3], and long-term unreasonable grazing affects the vegetation community structure and soil characteristics [4][5][6], which change the soil microbial community. Changes in the soil microbial community structure affect the vegetation in the community, which impacts the entire ecosystem [7]. Soil microbial diversity is an important index for evaluations of community characteristics and stability, which re ect the functional evolution of ecosystems and environmental changes. In general, greater microbial diversity indicates stronger soil biological activity, which is more conducive to plant growth [8]. Therefore, it is of great signi cance to examine the relationship between different grazing intensities and microbial community structure and diversity of grassland and soil and vegetation characteristics to reveal the mechanisms of succession direction of grazing grasslands.
To perform clear and reasonable grazing management of grasslands, increasing studies were performed in recent years on the impact of grazing on the soil microbial community, including soil microbial number [9], soil microbial quantity [10], soil microbial enzyme activity [11], soil microbial community structure [12] and soil microbial diversity [13]. These studies showed that light grazing signi cantly increased the soil microbial population and microbial biomass [14], but overgrazing reduced the soil microbial biomass carbon [15]. Moderate grazing changed the living environment of soil microorganisms, which was conducive to their growth [16], improved microbial diversity, changed the microbial community structure, improved soil quality and maximized grassland productivity [17]. Zhalnina et al. [18] found that soil pH determined microbial diversity and composition, and bacterial diversity was higher in neutral soil and weakly acidic soil. Sun Yifei et al. [19] showed that changes in usable nitrogen content caused by grazing led to signi cant changes in ammonia-oxidizing microbial community structure. Xun et al. [20] showed that grazing changed the microbial composition of meadow grasslands from slow-growing and fungi to fast-growing and bacteria-dominated communities. For the Qinghai-Tibet Plateau, the research results showed that soil bacterial diversity was the highest in moderate grazing areas, and the migration of soil bacterial diversity grazing changed [21][22][23]. Grazing may affect the distribution of alpine grassland microorganisms on a small spatial scale and changed the impact of water on the microbial community [24]. These results also re ect the inconsistent response mode of soil microorganisms to grazing due to different geographical locations and seasonal climate [25][26], the complex relationship between grazing and soil microorganisms, and different changing trends under the in uence of environmental factors.
As an important part of the Eurasian steppe, Hulunbuir steppe is an important material basis for the development of animal husbandry and the basic means of production for the survival of farmers and herdsmen. Therefore, to study the relationship between grazing and microorganisms in this region and establish a reasonable grazing system is vital importance. In this study, high-throughput sequencing was used to sequence some bacterial 16S R RNA and fungal 18S R RNA genes, to evaluate the structure of soil microbial community under six grazing intensities.The microbial biomass carbon and nitrogen and soil physical and chemical properties were measured under different grazing intensities.The objectives of this study were to investigate grazing-induced changes in the composition, diversity, microbial biomass carbon and nitrogen, and their interaction with the physical and chemical properties of vegetation and soil.

Study site and sampling
The research area is located in the Inner Mongolia Autonomous Region, China. The experiment was based on the National Field Scienti c Observation and Research Station of Hulun Buir Grassland Ecosystem (N 49°19 '349 " ~ 49°20' 173", E 119°56 '521 " ~ 119°57' 854", altitude of 666 ~ 680 m), which is a temperate semi-arid continental climate. The average annual temperature is -5 ~ -2℃, with the highest and lowest temperatures of 21.1℃ and − 27.2℃, respectively. The average annual precipitation is 350 ~ 400 mm, and the highest precipitation periods are from July to September (Fig. 1). The soil types of the study sites were or chestnut soil, which corresponds to Castanozems in the soil taxonomic system of FAO and the calcic-orthic aridosol in the US soil classi cation system. The study regionalization was divided into 6 grazing areas with different grazing intensities, 0.00, 0.23, 0.34, 0.46, 0.69, and 0.92 Au ha − 1 , named G0.00, G0.23, G0.34, G0.46, G0.69 and G0.92, respectively. For grazing, 500-kg beef cattle was used as the standard cattle beef unit (Fig. 2). Each grazing intensity was divided into three repetitions, and each plot area was 5 ha − 1 . Since 2009, grazing began on June 1st and stopped on October 1st each year for 120 days. Samples were collected in the 10th year of grazing treatment. Soil samples were collected in August 2018 to determine the microbial community structure and soil physical and chemical properties and investigate the characteristics of the plant community.
Five quadrats (1 × 1 m) in each test area were randomly selected to measure the height, coverage and density of plants in each quadrat, and the aboveground parts of plants in the quadrat were divided and cut using the method of cutting the whole ground. Plants were put in a sample bag, numbered and transported to the laboratory. The fresh weight was measured, and plants were dried in a paper bag under a constant temperature (85℃) for 12 h. The dry weight was measured, and the aboveground biomass of the community was calculated. The underground biomass of community was determined using the excavation method of 5 topsoil samples (0-10 cm depth) that were randomly selected from each quadrat and immediately mixed into a single soil sample. Rocks and animal and plant residues were removed from the soil for the determination of soil chemical properties, soil physical properties, bacteria and fungi, microbial carbon and nitrogen.

Determination of Physical and Chemical Properties of the Soil
A speci c method was used to determine the soil chemical properties that was different from the instrument instructions. Soil temperatures (ST) of the 0-10 cm soil layers were measured by portable thermometers that can be inserted down to different soil depths when collecting soil samples that need to determine the soil microbial community,and three replicate soil samples of each test area (0-10 cm soil layers) were collected Simultaneously. Soil moisture (SM) was measured by the oven-drying method, and the gravimetric soil water content was determined by drying the sampled soil at 105℃ for 24 h. Soil pH was measured in a 1:2.5 soil/water mixture by a multi-parameter water quality analyser. Soil total nitrogen (TN) concentration was measured the semi-micro Kjeldahl determination. The dichromate oxidation and sodium hydroxide alkali-molybdenum-antimony colorimetric methods were used to determine the soil organic carbon (SOC) concentration, and soil total phosphorus (TP) concentration. A spectrophotometer was used to measure soil total potassium (TK) Based previous reports, the primers (Biosciences, Union City, CA, USA) were washed with Tris-HCl and veri ed using 2% agarose gel electrophoresis. PCR products were quanti ed using the Quanti Fluor TMs 338F-806R [27] and 817F-1196R [28] for the 16S rRNA and 18SrRNA genes, respectively. Ampli ed products were detected using 2% agarose gel electrophoresis and recovered from the gel using the Axy Prep DNA gel extraction kit (

Statistical analysis
One-way ANOVA of soil physical and chemical properties was performed using SPSS 21.0. Signi cance was calculated using Tukey's test (P < 0.05). Origin 2017 was used for gures of MBC and MBN. The relationship between soil microbial community structure and each affecting factor was analysed using RDA and variation partitioning. RDA eliminates redundant variables depending on the other measured variables (variables with large effects are and automatically selected) and on the variance in ation factor values to gradually remove redundant parameters, and the signi cance levels are based on 999 Monte Carlo permutations. Linear discriminant analysis (LDA) coupled with effect size measurements (LEfSe) analysis was performed to search for signi cantly different biomarkers between groups [29].

Soil and vegetation characteristics change under different grazing intensities
As can be seen from Table 1, the changes of soil physical and chemical properties under different grazing intensities were as follows: SM was signi cantly reduced with the increase of grazing intensities (P < 0.05), and the content of non-grazing plotd was the highest. NO3-N was signi cantly higher than other grazing intensity in the G0.69 plot (P < 0.05), while other factors showed no signi cant difference under different grazing intensity. Vegetation characteristics showed signi cant differences under different grazing intensities (P < 0.05)(

Changes in the soil microbial biomass carbon and nitrogen under different grazing intensities
Our results showed that the variation trends of microbial biomass carbon (MBC) and microbial biomass nitrogen (MBN) downstream of grazing treatment were the same (Fig. 3). The content in the G0.34 plots was the highest in the 0-10 cm soil layer, and the content in the G0.92 test plots was smaller than other grazing treatments. Except for the area with a grazing intensity of 0.34, the greater the grazing intensity, the lower the MBC and MBN content. The G0.00 soil in soil layers 10-20 cm and 20-30 cm was higher than the other grazing intensities. The microbial carbon and nitrogen contents decreased with increasing depth. These results are consistent with the parallel and vertical distribution of microorganisms.

Composition of microbial community under different grazing intensities Soil microbial diversity changes under different grazing intensities
The soil microbial diversity indexes of the 6 grazing treatments are shown in Fig. 4. Coverage represents the proportion of the sequence obtained by sequencing of the whole genome. The Shannon index re ects the variation or difference degree of the microbial community, which was affected by the total number of samples and uniformity. Higher values indicate higher microbial community diversity in the soil [30]. The Chao index uses the chao1 algorithm to estimate the number of operational taxonomic units in samples. Chao1 is often used to estimate the total number of species in ecology. The coverage value of each sample was higher than 0.97, which indicates that the sequence obtained by sequencing had a high coverage degree and good representativeness. This result re ects that the clone library constructed in this study represented the diversity of soil bacteria in this area.
The α diversity indexes of G0.23 and G0.34 bacteria were higher, and the α diversity indexes of G0.00 and G0.92 grasses were lower. These results are consistent with the intermediate disturbance hypothesis. The Chao of the fungus in the heavy grazing area was lower than the other treatments, but the diversity and uniformity of the fungal community was higher than the light grazing area.

Microbial community structures in soils with different grazing intensities
The histograms of dominant gates of soil bacteria and fungi ( Fig. 5A and B) were obtained according to the rule that the proportion of microbial abundance was greater than 1%. There were 10 species abundance ratios of bacteria greater than 0.01, and the 10 dominant bacteria (abundance values greater than 1%) detected in 6 different grazing treatments were Actinobacteria, Proteobacteria, Acidobacteria, Chloro exi, Verrucomicrobia, Bacteroidetes, Gemmatimonadetes, Nitrospirae, Firmicutes, and Plancomycete. Among these bacteria, Chloro exi differed signi cantly between different grazing intensities (P < 0.05), with the lowest Chloro exi in non-grazing and heavy grazing intensities, and the highest amounts in a medium grazing area (G0.46). There were 4-5 phyla with fungal abundance values greater than 1% in different grazing intensities: Ascomycota, Zygomycota, Unclassi ed_k_Fu, Basidiomycota, and Chytridiomycota. The abundance of Chytridiomycota was greater than 1% only in non-grazed (G0.00) and moderately grazed (G0.46) grasslands. Bacteria had different abundances across the 6 grazing intensities. Actinobacteria and Proteobacteria were the absolute dominant bacterial species. Ascomycota was the fungus phylum with dominant abundance values in the 6 treatments, and G0.46 was higher than the other treatments.

Microbial communities with statistically signi cant differences
Microbial communities with statistically signi cant differences were determined using α-and β-diversities. Another primary goal of comparing microbial communities was to identify specialized communities within the samples using the LEfSe tool. This tool allows analyses of microbial community data at any clade. Because the analysis of the large number of OTUs detected in the present study would be computationally too complex, statistical analysis was performed only from the domain to the genus level. Groups are shown in cladograms, and LDA scores of 2 or greater were con rmed using LEfSe (Fig. 6,7).
Most bacterial species were enriched in G0.00 and G0.69 and less abundant in the other grazing intensities (Fig. 6). Fifteen bacteria are enriched signi cantly in G0.00, including p_Bacteroidetes, o_Sphingobacteriales and c_Sphingobacteriia, which were the most important (LDA > 3. Relationship between microbial community structure and environmental characteristics Correlation between environmental factors and the microbial community Spearman correlation analysis was used to study the mutual changes between environmental factors and species and obtain the correlation and signi cance values between pair-wise comparisons. The bacteria and environmental parameters were signi cantly correlated (Fig. 8A). Norank_c_KD4-96 signi cantly positively correlated with pH. Norank_cacidobacteria signi cantly positively correlated with SAP. Rubrobacter signi cantly negatively correlated with TP, and norank_o_Gaiellales signi cantly negatively correlated with BGB. Spearman correlation analysis showed the following signi cant correlations (Fig. 8B).

Environmental factors and the RDA of the microbial community
Grazing changes microbial community structures and environmental characteristics. The disrupting effect of grazing on microbial communities may primarily be mediated by aboveground plant and soil geochemical characteristics. Therefore, the present study investigated whether microbial community structure and environmental characteristics were related. RDA revealed that the microbial community structure was formed by primary environmental characteristics, including soil moisture (SM), soil bulk density (SBD), pH, soil organic carbon (SOC), total nitrogen (TN), TP, TK, SAP, SAK, NH 4 + -N NO 3 − -N, vegetation belowground biomass (BGB), litter, and coverage. After removal of the redundant variables, 14 environmental characteristics were chosen for RDA. As shown in Fig. 7A and B, BGB (p = 0.001) signi cantly affected the bacterial community structure, and pH (p = 0.032) TN (p = 0.011) and litter (p = 0.007) signi cantly affected the fungal community structure. The rst two axes of the RDA of the bacterial community (Fig. 9A) accounted for 71.56% of the total variation in bacterial community composition, and the rst axis accounted for 60.33% of the variation. The G0.23 samples distributed together, but the sample distribution patterns of other grazing intensities were discrete. The rst two axes of the RDA of the fungal community (Fig. 9B) accounted for 25.96% of the total variation in fungal community composition, and the rst axis accounted for 22.73% of the variation. The G0.00 and G0.92 samples distributed together, but the sample distribution patterns of the other grazing intensities were discrete.
Variance partition analysis was performed to dissect the contributions of soil and plant characteristics to the microbial community structure. These selected characteristics together explained 46.98 and 61.77% of bacterial and fungal community changes, respectively (Fig. 9C,D). The contribution of soil and plant characteristics explained 45.35 and 0.75%, respectively, of bacterial community changes and 55.26 and 4.69%, respectively, of the fungal community changes. The combined contribution of soil and plant characteristics explained 0.88 and 1.82% of the bacterial and fungal community changes, respectively, which reveals a very close interaction between soil and plant characteristics.

Discussion
Vegetation, soil and microorganisms do not change in isolation, but form a feedback mechanism and interact with each other. Grazing livestock changes the growth of community plants, and soil trampling and excretion affect the nutrient cycle of the soil. Changes in plant growth patterns and the soil environment indirectly affect the microbial environment in the soil, which leads to changes in the microbial community. Soil provides a living environment for microbial communities, and soil characteristics affect microbial communities [31] .
Soil water has a signi cant inhibitory effect on bacteria, and limits the nutrient uptake and production of bacteria, which leads to metabolism restriction due to nutrient de ciency [32]. Soil pH does not direct impact the microbial community, but it indirectly affects the aboveground vegetation, changes the community characteristics of the aboveground vegetation, and reduces soil microbial energy and organic matter content, which affects the microbial community [33].

Effects of grazing intensity on soil microbial community structure
The present study examined 18 sample plots of grassland with six grazing intensities in the Hulun Buir grassland to analyse the relationships between soil bacterial and fungal communities and grazing intensity, and the interaction between plant communities and soil physical and chemical properties resulting from grazing intensity and bacterial and fungal communities. The predominant ora in the bacterial communities of Hulun Buir grassland soil was generally consistent between the six grazing regimens, but there were differences in relative abundance, and each region had its own unique fungal populations. Bacterial and fungal diversity was highest in G0.23 and G0.00, respectively. Similarly, Lienhard et al. (2013) [34] found that maximum bacterial and fungal diversity occurred with different utilization schemes. Microbial biomass or diversity increases, decreases, or remains the same depending on the type of grassland, geographical location, and the grazing system and intensity [9-10, 13, 35-39].
Although the diversity of bacterial and fungal communities respond to grazing intensity in different ways, our study showed that the soil microbial community structure changed signi cantly along a gradient of grazing intensity, which was consistent with changes in soil and plant characteristics.

Relationship between microbial community and environment
We found that the environmental changes that occurred with changes in grazing intensity contributed differently to different microbial groups in the community (Fig. 9A,B). The results showed a signi cant correlation between changes in the bacterial community and the underground biomass of the community (P < 0.05), but this correlation was not found in the fungal community. There was a signi cant relationship between litter and the fungal community (P < 0.05), but no signi cant relationship with bacteria was observed. Zhou et al. [13] found that livestock feeding affected aboveground vegetation biomass and community structure, and indirectly changed soil physical and chemical properties. These changes were a result of the interaction of microorganisms with plants [40]. The fungal community structure in our study area was not as sensitive as the bacterial community structure to vegetation biomass changes. This difference may be because fungi are more likely to degrade lignocellulose from different plants than bacteria, which allows them to rst obtain resources from many of the relevant available substances [41].
We also found signi cant direct relationships of bacterial and fungal community structures with soil total nitrogen. Soil nitrogen storage decreases with increasing grazing intensity [42][43]. Nitrogen is one of the most important nutrients for life. Therefore, plant and microbial activities may gradually reduce the content of nitrogen in soil [7]. However, correlations between microbial communities and environmental factors must be carefully explained because it is often very di cult to rmly establish the relationship between microbial communities and soil nutrient cycling [44].
We found that the soil characteristic factors evaluated explained over 45% of the shift in the microbial communities in this study, which suggests that soil characteristics factors are the primary factors in uencing microbial community structure. However, 40-55% of the factors in uencing the dynamic changes in the microbial community could not be determined. Multiple studies showed that the effect of mild or moderate grazing on grassland soil was relatively small, and these grazing intensities are bene cial to dry matter production, nutrient cycling, and carbon and nitrogen storage [45][46].

Conclusion
The present study examined the soil microbial structure and its relationship with the environment in 6 leymus chinensis meadow steppes with different grazing intensities. High-throughput sequencing results revealed that bacterial and fungal communities were extremely sensitive to grazing and varied with grazing intensity. The diversity of soil bacteria in the light grazing test area was higher than the heavy grazing area, and the diversity of fungi in the nongrazing area was higher than the speci c grazing area. Heavy grazing-induced reduction the diversity of soil bacteria and fungi. The underground biomass of the community signi cantly in uenced the bacterial community structure, and pH, total nitrogen and litter signi cantly in uenced the fungal community. In conclusion, heavy grazing reduces the diversity of the community and is not conducive to the development of Hulunbuir meadow grassland. Availability of data and materials YZ participated in the design of the study ideas,carried out the diversity of soil microbial communities, performed the statistical analysis and drafted the manuscript. XPX participated in the analyzed the data. RQL participated in the design of the study ideas. WBX participated in the design of the study ideas.
RFZ participated in the analyzed the data. LHL participated in the analyzed the data. XW participated in the analyzed the data. JQC participated in the analyzed the data.RY conceived of the study, and participated in its design and coordination and helped to draft the manuscript. Tables Table 1 Changes  Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.     According to the rule that the proportion of microbial abundance was greater than 1%, the dominant phylum histograms of soil bacteria (A) and fungi (B) were drawn. The abscissa is the group name, and the ordinate is the proportion of the species in the sample. The columns of different colours represent different species, and the length of the column represents the proportion of the species. Bacterial (C) and fungal (D) community heatmaps. The abscissa/ordinate is the sample name. In gure A and B, the vertical/horizontal coordinates are the proportion of species in this sample. Columns of different colours represent different species, and the length of columns represents the proportion of this species. In C and D, the ordinate is the species name, and the variation of abundance of different species in the sample is shown by colour gradient of the colour block. The value represented by the colour gradient is on the right of the gure. Colour gradients are used to represent the size of the data in a two-dimensional matrix or table and to present information about community species composition and species abundance. Clustering was carried out according to the similarity of the abundance between species or samples, and the results are presented in the community heatmap, which groups the species with high and low abundance clusters in blocks, and re ects the similarity and difference of community composition of different groups (or samples) at each taxonomic level through colour change and similarity degree.   Heatmap analyses of the correlations between bacterial and fungal species abundance and environmental factors. By calculating the Spearman correlation coe cient between the environmental factors of the top 10 species and the total abundance, the obtained numerical matrix is visually displayed in the heatmap. The X axis and the Y axis are the environmental factors and species, respectively. Colour changes re ect the data information in the twodimensional matrix or table. The colour depth represents the size of the data value, which is directly represented by the de ned colour depth. The legend on the right is the colour interval of different R values.* 0.01 < P ≤ 0.05, ** 0.001 < P ≤ 0.01, *** P ≤ 0.001. Figure 10