Effects of David Deer Grazing on Soil Bacterial and Fungal Communities in an Eastern Coastal Wetland of China

: The grazing activity of animals has a signiﬁcant role on the environmental modiﬁcation of land. In the coastal wetlands of eastern China, long-term David deer ( Elaphurus davidianus ) grazing has caused the degradation of various ecological elements in the area. Still, few studies have been reported concerning the effects of David deer grazing on the soil microorganisms of their habitats. We analyzed the community structure of soil bacteria and fungi in an area of continuous annual grazing and another area without traces of David deer grazing so as to learn about the effects of deer grazing on the soil microbial community structure in a spatial instead of temporal way, in preparation for improving the environment for deer survival. David deer grazing drastically changed the physicochemical characteristics of the soil, accelerating the alkalinization process and inhibiting the buildup of nutrients. There were differences in the microbial community structure between the grazed and the control areas, with bacteria predominating. The control had a higher level of bacterial and fungal alpha-diversity than the grazed area. The makeup of the soil’s microbial community was also inﬂuenced, except for the dominant microbial at the phylum level. In addition to the establishment of numerous complex fungal functional types, David deer grazing increased the number of bacterial functional types linked to the carbon cycle. The impacts of soil pH and urease activity on bacterial and fungi populations were highlighted using the redundancy analysis. This study demonstrates that David deer grazing changes and complicates microbial functional kinds of composition, as well as modiﬁes the composition of the soil’s microbial community, improving the soil nutrient cycling process, mainly the carbon element.


Introduction
Grazing is a form of human land use, and ungulates' feeding and trampling behavior affect plants and soils. Most studies on the effects of grazing animals on grounds have focused on soil physical and chemical properties [1,2], with few data on biological features such as soil microbes, which are related to soil quality [3]. With the development of second-generation sequencing technologies such as high-throughput sequencing, soil microbial communities are gradually gaining attention in studies on grazing [4,5]. Bacteria have an essential role in soil nutrient and material cycling, and fungi can decompose plant and animal residues, exist as plant and animal pathogens, and form symbioses with plants in the soil [6]. The size of soil microbial communities has shrunk due to changes in the soil environment caused by grazing [7]. Overgrazing induces changes in the structure and diversity of soil microbial communities, but the patterns are inconsistent. Wang et al. [8] found that the diversity of soil microbial communities increased with longterm grazing. In a summer pasture, Zhang et al. [9] found a decline in soil fungal community diversity. Grazing alters the relative abundance of certain microorganisms in the soil microbial community [10] and the functional composition of microorganisms. For example,

Experimental Design
The cohort was divided into grazed zones and control areas based on the amount of faucal residual and the density of the hoofprints [19]. The grazed area had no vegetative cover and had been grazed continuously since 2018. S. alterniflora was present in the control areas and there were no feces or hoofprints. Details about the regions are presented in Table 1. The sampling was finished in October 2020. Three large 10 m x 10 m squares, each 150-200 m apart, were set up inside both the grazed and control areas. There were five 1 m x 1 m courts set in the corners and center of each large sample square to gather samples from the top 0-20 cm of soil. A total of 30 samples were collected. Two bags were created out of each component, for lab analysis of the soil environmental factor and microbial community sequencing by Shanghai Majorbio Bio-pharm Technology Co., Ltd., respectively.

Determination of Soil Environmental Factors
Soil samples for soil environmental factor testing were air-dried and sieved in a cool location after being cleared of visible stones and plant residues, then the following parameters were measured: soil bulk density (BD), water content (SWC), pH, Na + concentration (Na + ), soil organic carbon (SOC), total nitrogen (TN), total phosphorus (TP), and total potassium (TK), as well as the catalase (CAT), urease (URE), and sucrase (SUC) enzyme ac-

Experimental Design
The cohort was divided into grazed zones and control areas based on the amount of faucal residual and the density of the hoofprints [19]. The grazed area had no vegetative cover and had been grazed continuously since 2018. S. alterniflora was present in the control areas and there were no feces or hoofprints. Details about the regions are presented in Table 1. The sampling was finished in October 2020. Three large 10 m × 10 m squares, each 150-200 m apart, were set up inside both the grazed and control areas. There were five 1 m × 1 m courts set in the corners and center of each large sample square to gather samples from the top 0-20 cm of soil. A total of 30 samples were collected. Two bags were created out of each component, for lab analysis of the soil environmental factor and microbial community sequencing by Shanghai Majorbio Bio-pharm Technology Co., Ltd., respectively.

Determination of Soil Environmental Factors
Soil samples for soil environmental factor testing were air-dried and sieved in a cool location after being cleared of visible stones and plant residues, then the following parameters were measured: soil bulk density (BD), water content (SWC), pH, Na + concentration (Na + ), soil organic carbon (SOC), total nitrogen (TN), total phosphorus (TP), and total potassium (TK), as well as the catalase (CAT), urease (URE), and sucrase (SUC) enzyme activity. BD was determined using the cutting ring method; the soil water content was determined by oven drying (oven 105 • C to constant weight); pH was determined using a pH meter (OHAUS STARTER 3100) with a water/soil ratio of 5:1; the Na + concentration was determined using the flame photometric method. The H 2 SO 4 -K 2 Cr 2 O 7 oxidation-external heating method was used to calculate the SOC. Elemental analysis was used to calculate the TN (PE-2400, USA). TP content was determined by NaOH melting-molybdenum antimony anti spectrophotometry. The NaOH melting-flame photometry was used to calculate the TK. CAT was calculated via titration with potassium permanganate. URE was calculated using colorimetry with indophenol blue. A 3,5-dinitrosalicylic acid-based colorimetric technique was used to determine the SUC.

Data Processing and Analysis
Based on whether the data conformed to the normal distribution and the homogeneity of variance, we decided to test the significant difference of the data by one-way ANOVA or nonparametric tests. The above analysis was completed in SPSS 25.0.
Mothur was used to analyze the alpha diversity indices of the bacterial and fungal communities at the OUT level (97% similarity), including the Shannon, Simpson, Chao, and Ace indexes. Qiime calculated the beta diversity distance matrix for bacterial and fungal communities between different treatments, and the R vegan package performed the NMDS analysis and mapping. The bacterial and fungal community composition was mapped using R. Bacterial community function predictions depended on the FAPROTAX database, while fungal community were carried out using FUNGuild. LEfSe analysis was used to identify communities that had a significant differential impact on treatment delineation, based on http://huttenhower.sph.harvard.edu/galaxy/root?tool_id=lefse_upload (accessed on 13 May 2022). The relationship between soil physicochemical properties and microbial communities was analyzed using RDA, and the analysis and mapping were done using the R language vegan package. The above analysis process involving microbial community data was completed on the online analysis platform provided by Shanghai Majorbio Bio-pharm Technology Co., Ltd.

Effect of Grazing on Soil Properties
Significant differences existed between the grazed and control areas in terms of soil properties, indicating changes in soil properties due to grazing ( Table 2). The pH was significantly higher in the grazed than the control area (p < 0.01). SOC, TN, and TP were substantially lower in the grazed area (p < 0.01), and TK was considerably lower in the grazed compared with the control area (p < 0.05). Grazing also affected the soil enzyme activities, with URE and SUC varying significantly between treatments, and caused a significant reduction in URE and SUC (p < 0.01).

Soil Bacterial and Fungal Communities Alpha Diversity
The bacterial community alpha diversity indices of the OTUs are displayed in Table 3. The sequencing of the bacterial communities had a high coverage of 99% or higher. The Shannon, Simpson, Ace, and Chao datasets were chosen to represent the quantity and variety of soil microbial communities in the grazed and control areas. The Shannon was substantially lower (p < 0.05) and the Simpson was significantly greater (p < 0.01) in the bacterial community diversity indices in the grazed, indicating that grazing decreased the variety of bacterial community in the soil. The Shannon and Simpson indexes were not significantly different in the fungal community diversity indexes, while the Ace and Chao in the grazed were significantly lower (p < 0.01), showing that grazing reduced the quantity of the soil fungal communities.

Soil Bacterial and Fungal Communities Beta Diversity
Differences in bacterial and fungal communities were examined at the OTU level for grazing using Bray-Curtis distances ( Figure 2). The NMDS analysis ranking had some relevance, as can be seen from the figure, where the stress coefficients for the bacterial and fungal communities were both less than 0.2, 0.05, and 0.156, respectively. Between the grazed and control areas, the bacterial and fungal communities of the soil were consider- ably different, indicating that grazing had a major impact on soil microbial communities (p < 0.01).
(p < 0.01), showing that grazing reduced the quantity of the soil fungal communities.

Soil Bacterial and Fungal Communities Beta Diversity
Differences in bacterial and fungal communities were examined at the OTU level f grazing using Bray-Curtis distances ( Figure 2). The NMDS analysis ranking had som relevance, as can be seen from the figure, where the stress coefficients for the bacterial a fungal communities were both less than 0.2, 0.05, and 0.156, respectively. Between t grazed and control areas, the bacterial and fungal communities of the soil were consid ably different, indicating that grazing had a major impact on soil microbial communit (p < 0.01).

Soil Microbial Community Composition
In Figure 3a, the dominant phylum of the bacterial community composition was Pr teobacteria. A total of 18 bacterial phyla were detected, and a relative abundance of le than 1% was combined. The top five bacterial phyla, in relative abundance, were Prote bacteria, Chloroflexi, Acidobacteriota, Actinobacteriota, and Desulfobacterota, with t sum of the relative abundance being more significant than 50%. The relative abundan of Proteobacteria was 31.86% in the grazed area, which was significantly better than th in the control (p < 0.01). The relative abundance of Acidobacteriota was 11.41% in t grazed and 6.64% in the control area, with a highly significant difference between the tw areas (p < 0.01). The relative abundance of Chloroflexi, Actinobacteriota, and Desulfoba terota was lower in the grazed than control area, with the relative abundance of Chlo oflexi being significantly lower (p < 0.01) and Desulfobacterota being significantly low (p < 0.05).

Soil Microbial Community Composition
In Figure 3a, the dominant phylum of the bacterial community composition was Proteobacteria. A total of 18 bacterial phyla were detected, and a relative abundance of less than 1% was combined. The top five bacterial phyla, in relative abundance, were Proteobacteria, Chloroflexi, Acidobacteriota, Actinobacteriota, and Desulfobacterota, with the sum of the relative abundance being more significant than 50%. The relative abundance of Proteobacteria was 31.86% in the grazed area, which was significantly better than that in the control (p < 0.01). The relative abundance of Acidobacteriota was 11.41% in the grazed and 6.64% in the control area, with a highly significant difference between the two areas (p < 0.01). The relative abundance of Chloroflexi, Actinobacteriota, and Desulfobacterota was lower in the grazed than control area, with the relative abundance of Chloroflexi being significantly lower (p < 0.01) and Desulfobacterota being significantly lower (p < 0.05).
As shown in Figure 3b, Ascomycota was the dominant phylum in the fungal community. Except for unclassified fungi, the fungal community at the phylum level consisted of Ascomycota, Basidiomycota, Mortierellomycota, and Rozellomycota. The relative abundance of Ascomycota in the grazed area was 61.47%, 12.17% lower than that in the control area. In comparison, the relative abundance of Basidiomycota in the grazed area was 28.82%, 8.91% higher than the control area. The relative abundance of Mortierellomycota in the grazed area was more elevated than the control, with a difference of 2.65%. The relative abundance of Rozellomycota in the grazed area was only 0.02%, which was significantly lower than the control area (p < 0.01), with a difference of 1.69% between areas.

Predicting the Function of Soil Bacteria and Fungi
In Figure 4a, the top 20 functional taxa annotated for relative abundance were selected and analyzed for differences in bacterial function between the grazed and control areas. In the set of bacteria with functional predictions, chemoheterotrophy and aerobic chemoheterotrophy were dominant, and were superior in the grazed area. The phototrophy, nitrate reduction, and photoheterotrophy varied significantly between the two areas. Phototrophy and photoheterotrophy were considerably higher in the grazed than control (p < 0.01) areas, and nitrate reduction was significantly higher in the grazed area (p < 0.05). As shown in Figure 3b, Ascomycota was the dominant phylum in the fungal community. Except for unclassified fungi, the fungal community at the phylum level consisted of Ascomycota, Basidiomycota, Mortierellomycota, and Rozellomycota. The relative abundance of Ascomycota in the grazed area was 61.47%, 12.17% lower than that in the control area. In comparison, the relative abundance of Basidiomycota in the grazed area was 28.82%, 8.91% higher than the control area. The relative abundance of Mortierellomycota in the grazed area was more elevated than the control, with a difference of 2.65%. The relative abundance of Rozellomycota in the grazed area was only 0.02%, which was significantly lower than the control area (p < 0.01), with a difference of 1.69% between areas.

Predicting the Function of Soil Bacteria and Fungi
In Figure 4a, the top 20 functional taxa annotated for relative abundance were selected and analyzed for differences in bacterial function between the grazed and control areas. In the set of bacteria with functional predictions, chemoheterotrophy and aerobic chemoheterotrophy were dominant, and were superior in the grazed area. The phototro-a b Figure 3. Composition of the soil microbial communities in the grazed and control areas. The vertical and horizontal coordinates are the group names and proportion of species in the group, respectively. The columns with different colors represent different species, and the length of the columns represents the proportion of the species. * 0.01 < p ≤ 0.05, ** 0.001 < p ≤ 0.01, *** p ≤ 0.001. (a) Bacterial community and (b) fungal community. Figure 4b showed the predicted soil fungal communities functions for the grazed and control areas. In the figure, the fungal functional types were predominantly saprophytic. The proportion of annotated fungal functions was more significant than 40%, and fungal functional types with a relative abundance of less than 1% were combined. In addition to the unannotated fungal functional types, undefined saprotroph had the highest abundance. Dung saprotroph-plant saprotroph, wood saprotroph, and fungal parasite-undefined saprotroph were lower in the grazed area, while endophyte-litter saprotroph-soil saprotrophundefined saprotroph were higher. Animal pathogen was detected in the grazed area with a relative abundance of 7.23%.   Figure 4b showed the predicted soil fungal communities functions for the grazed and control areas. In the figure, the fungal functional types were predominantly saprophytic. The proportion of annotated fungal functions was more significant than 40%, and fungal functional types with a relative abundance of less than 1% were combined. In addition to b

Soil Microbial Community LEfSe Analysis
The differences in soil bacterial and fungal communities within the grazed and control areas were analyzed using an LDA threshold of 4 ( Figure 5). The indicators of micro-bes for bacteria within the grazed area were Proteobacteria (including Pseudomonadales, Nitroso-coccales and Tistrellales), Chloroflexi (including S085), Desulfobacterota (including ordernorank which belongs to Desulfuromonadia), Methylomirabilota (including Methylomirabilales), Planctomycetota and Acidobacteriota while for fungi they were Ascomycota (including Sordariales and Monosporascus) and Basidiomycota (including Trichosporonales). The indicator microbes for bacteria in the control were Proteobacteria (including B2M28 and Rhizobiales), Chloroflexi (including SBR1031), and Desulfobacterota (including _Desulfobacterales and Desulfobulbales) for fungi were Ascomycota (including order-unclassifiedwhich belongs to Sordariomycetes, Phaeosphaeriaceae and Hypocreales_fam_Incertae_sedis and Lulworthia), Basidiomycota (Polyporales), and Rozellomycota. trol areas were analyzed using an LDA threshold of 4 ( Figure 5). The indicators of microbes for bacteria within the grazed area were Proteobacteria (including Pseudomonadales, Nitrosococcales and Tistrellales), Chloroflexi (including S085), Desulfobacterota (including order-norank which belongs to Desulfuromonadia), Methylomirabilota (including Methylomirabilales), Planctomycetota and Acidobacteriota while for fungi they were Ascomycota (including Sordariales and Monosporascus) and Basidiomycota (including Trichosporonales). The indicator microbes for bacteria in the control were Proteobacteria (including B2M28 and Rhizobiales), Chloroflexi (including SBR1031), and Desulfobacterota (including _Desulfobacterales and Desulfobulbales) for fungi were Ascomycota (including order-unclassifiedwhich belongs to Sordariomycetes, Phaeosphaeriaceae and Hypocreales_fam_Incertae_sedis and Lulworthia), Basidiomycota (Polyporales), and Rozellomycota. Figure 5. LEfSe analysis of soil microbial communities in grazed and control areas. Different color nodes represented microbial groups that were significantly enriched in the corresponding groups and had a significant impact on the differences between groups; pale yellow nodes represented microbial groups that had no significant difference in different groups, or had no significant impact on differences between groups. (a) Bacterial community and (b) fungal community.

RDA Analysis of Soil Microbial Community and Soil Environmental Factors
As shown in Figure 6, the differences in the relationships between dominant soil microorganisms and soil environmental factors at the phylum level emerged. Axis 1 and axis 2 of the RDA analysis explained 64.83% and 46.79% of the total variance in the bacterial and fungal communities, respectively. Combined with the results of the biological significance analysis of the RDA analysis in Table 4, pH, SOC, TN, TP, URE, and SUC showed significant impacts on the bacterial community, and URE was the environmental factor that most affected the fungal community. These results indicated that environmental factors influence bacteria more than fungi. Among the top five bacterial phyla in relative abundance, Proteobacteria and Acidobacteriota were significantly and positively correlated with pH, while Actinobacteriota, Desulfobacterota, and Chloroflexi were the most correlated with URE and were significantly and positively correlated. In Figure 6b, the sample points of the grazed and control areas were complexly distributed in each quadrant with insignificant regional boundaries. URE was the main factor affecting the fungal community structure, and Ascomycota significantly correlated with URE. a b Figure 5. LEfSe analysis of soil microbial communities in grazed and control areas. Different color nodes represented microbial groups that were significantly enriched in the corresponding groups and had a significant impact on the differences between groups; pale yellow nodes represented microbial groups that had no significant difference in different groups, or had no significant impact on differences between groups. (a) Bacterial community and (b) fungal community.

RDA Analysis of Soil Microbial Community and Soil Environmental Factors
As shown in Figure 6, the differences in the relationships between dominant soil microorganisms and soil environmental factors at the phylum level emerged. Axis 1 and axis 2 of the RDA analysis explained 64.83% and 46.79% of the total variance in the bacterial and fungal communities, respectively. Combined with the results of the biological significance analysis of the RDA analysis in Table 4, pH, SOC, TN, TP, URE, and SUC showed significant impacts on the bacterial community, and URE was the environmental factor that most affected the fungal community. These results indicated that environmental factors influence bacteria more than fungi. Among the top five bacterial phyla in relative abundance, Proteobacteria and Acidobacteriota were significantly and positively correlated with pH, while Actinobacteriota, Desulfobacterota, and Chloroflexi were the most correlated with URE and were significantly and positively correlated. In Figure 6b, the sample points of the grazed and control areas were complexly distributed in each quadrant with insignificant regional boundaries. URE was the main factor affecting the fungal community structure, and Ascomycota significantly correlated with URE.

Discussion
The present study shows that the diversity and richness of the soil microbial comm nities in the grazed areas were lower than those in the control areas (Table 3), which consistent with the results of Chen et al. [4]. Long-term disturbances in the grazed area David deer (such as trampling, foraging, and lying) increased soil bulk density and s water evaporation; decreased vegetation cover, which is unfavorable to the growth of ae obic microorganisms; and resulted in a lower α-diversity of soil bacteria and fungi in t grazed area. The β-diversity of the microbial community evolved along with ma changes in the alpha diversity of the soil microorganisms. Figure 2 shows distinct bacter and fungal community borders between regions, and a dramatic change in commun structure. Soil bacterial diversity increases when land-use patterns change [22] in the r gion, suggesting that land cover directly affects the diversity of the soil microbial comm

Discussion
The present study shows that the diversity and richness of the soil microbial communities in the grazed areas were lower than those in the control areas (Table 3), which is consistent with the results of Chen et al. [4]. Long-term disturbances in the grazed area by David deer (such as trampling, foraging, and lying) increased soil bulk density and soil water evaporation; decreased vegetation cover, which is unfavorable to the growth of aerobic microorganisms; and resulted in a lower α-diversity of soil bacteria and fungi in the grazed area. The β-diversity of the microbial community evolved along with major changes in the alpha diversity of the soil microorganisms. Figure 2 shows distinct bacterial and fungal community borders between regions, and a dramatic change in community structure. Soil bacterial diversity increases when land-use patterns change [22] in the region, suggesting that land cover directly affects the diversity of the soil microbial community. Human interference mode such as grazing is an indirect factor affecting the soil microbial community.
Different compositions of the bacterial community were found between the grazed and control sites. Similar to the findings of previous studies [23,24], Figure 3a displays the bacterial community composition and significance analysis results in the grazed and control areas. In the grazed area, Proteobacteria were more prevalent compared to with the control area. According to studies, Proteobacteria are relatively more prevalent in habitats with high nutrient levels (high rates of carbon mineralization) [25]. The addition of nutrients to the soil through manure and urine from grazing animals results in soil nutrient enrichment [26,27] and makes soil more conducive to Proteobacteria-dominated bacterial life. The multi-year colonization of S. alterniflora has led to an increase in soil carbon stocks [28]. A large number of Chloroflexi were found in riverbank sediments and have an essential role in the carbon cycle [29]. The soil microrelief in the control was suitable for the survival of Chloroflexi. Acidobacteriota are generally more likely to be detected in environments with a low pH [30]. Nevertheless, in this paper, grazing caused a significant increase in the relative abundance of Acidobacteriota. The growth of Acidobacteriota is highly influenced by the carbon content of the soil [31], the organic carbon content of the grazed area was lower than the control, which may have led to the increase in the relative abundance of Acidobacteriota. Proteobacteria are the fast-growing bacteria among the phyla of relatively abundant bacteria mentioned above, while Acidobacteriota and Chloroflexi are slow-growing bacteria [32], suggesting that grazing caused a change in the soil bacterial community towards fast-growing types.
The functional makeup of the bacteria changed along with the composition of the soil bacterial population because of grazing. Chemoheterotrophy and aerobic chemoheterotrophy were among the top two relative abundances in the bacterial community's functional annotation (Figure 4a). It is understood that chemoheterotrophy is a crucial bacterial function in carbon cycling [33], and the relative abundance of chemoheterotrophy increased in the grazed area, where the legacy of David deer droppings in the area input exogenous carbon and promotes soil carbon cycling processes.
Following the administration of synthetic sheep urine, Rooney et al. [34] discovered that the soil pH increased from 5.4 to 6.4, indicating that grazing animal urine was a contributing component to an increase in pH. Due to the proximity of the studied region to the ocean, the baseline soil pH is alkaline, and under the influence of David deer urine, the soil pH of the grazed area shifts towards alkalinity.
According to the RDA analysis, pH was the main environmental factor influencing the bacterial community, which was in line with the findings of previous studies [35]. Contrary to the findings that Proteobacteria were found in high nutrient settings, the Proteobacteria were strongly positively linked with pH instead of nutrient contents such as SOC, which calls for further research. The distribution of large amounts of S. alterniflora in the control increased the soil nitrogen storage [28], and the Actinobacteriota and Chloroflexi were relatively abundant in the control and were significantly and positively correlated with urease activity, which is primarily involved in the decomposition of nitrogen [36]. A high urease activity indicates a rapid conversion rate of the soil nitrogen element, which is beneficial to microbial growth. Figure 3b shows that the sum of the relative abundance of Ascomycota and Basidiomycota within the areas were more than 90%, with Ascomycota dominating, with 98% of terrestrial fungi now known to belong to Ascomycota and Basidiomycota, with Ascomycota being dominant in fungal community [37]. Table 2 shows that the control had higher levels of soil organic carbon, total nitrogen, total phosphorus, and total potassium, while Figure 4 suggests that the relative abundance of Ascomycota increased. Chen et al. [38] found that the relative abundance of Ascomycota was the key difference in soil fungal communities from desert to normal grassland, and considered that Ascomycota may be adapted for nutrient-poor conditions. The results of this paper contradicted the conclusion as coastal mudflats do not fall into any of the foregoing categories, and the features of invasive plants do not appear to have been considered. Ascomycota, an important phylum of saprophytic fungi, tend to grow well in moist environments compared with Basidiomycota [39], and the siltation-promoting beaching effect of S. alterniflora increases the soil water content of the invaded sites, a property reflected in the regional differences in soil water content in Table 2, while the decomposition of the S. alterniflora residues increases the nutrient return to soil. Several fungi are predominantly saprophytic and nutrient-dependent [40]. The functional annotation results of the fungal community in this paper (Figure 4b) are consistent with this. Only 50% of fungal functional types were annotated when functional predictions were made for fungal communities. Fungal functional types were divided into three trophic types, saprophytic, symbiotic, and pathotrophic, while multiple complex trophic types were also found. In combination with Figure 6b, fungi were less affected by soil environmental factors than bacteria, indicating that fungi are more resistant to external disturbances than bacteria. The presence of multiple complex trophic types of fungi in the functional annotation may be an adaptation by fungi to resist external disturbances. Similarly, URE was the environmental factor that most affects the fungal community.

Conclusions
The soil microbial community, including the functional composition and bacterial and fungal community composition, presented significant changes within the grazed region. Deer excrement led to the enrichment of functional bacteria associated with the carbon cycle as a microbiological response to grazing. The grazing evolved into a complex of fungal functions, the majority of which were saprophytic. According to the RDA analysis, soil pH and URE had the greatest impact on the bacterial and fungal communities, respectively. In short, soil characteristics are the primary direct factors affecting the structure of the soil microbial community, and grazing activities, which introduce exogenous elements into the soil, are the primary indirect factors.

Institutional Review Board Statement: Not applicable.
Data Availability Statement: Data are available from the China National Center for Bioinformation (GSA number: CRA008232 and CRA008229) that are publicly accessible at https://ngdc.cncb.ac. cn/gsa (accessed on 21 September 2022). Any further data required can be requested from the corresponding author.