Soil Microbial Community Variation With Time and Soil Depth in Eurasian Steppe (Inner Mongolia, China)

Background Soil microorganisms play an indispensable role in the material and energy cycle of grassland ecosystem, and were affected by many environmental factors, such as time and space changes. However, there are few studies on the temporal and spatial transformation of soil microbial community in typical degraded steppe. We analyzed the community structure and diversity of soil bacteria and fungi and the effects of environmental factors on the community structure in Xilingol degraded steppe. Results The abundance and diversity of bacteria and fungi were signicantly affected by depth. Bacteria and fungi diversity of 10 cm was higher than that of 20 cm and 30 cm. The abundance of Acidobacteria, Proteobacteria, Actinomycetes, Ascomycetes and Basidiomycetes varies signicantly with depth. What’s more, soil pH increased signicantly with depth increasing, while SOM, AN, VWC and ST decreased signicantly with increasing depth. In addition, Depth, TOC and AN had signicant impact on the bacterial and fungi communities (p < 0.05). and


Introduction
The grassland covering approximately 25% of the earth's terrestrial area, and plays an important role in the global material cycle and energy exchange (Foley et al. 2011,). Inner Mongolia grassland is the typical arid and semi-arid steppe which is located in the eastern of the Eurasian steppe, and is the representative of Eurasian steppe in terms of climate, terrain, soil properties, and vegetation composition. Meanwhile, Inner Mongolia grassland also has the unique history of land use and management policy (Wu et al. 2015). In grassland ecosystem, soil is the important place for material and energy exchange, and has a strong in uence on the diversity of microorganisms as the medium for microorganisms' survival.
Microorganisms in soil stay in an indispensable position during the process of organics decomposing and the regulation of C , N (Bahram et al. 2018), P (Handa et al. 2014), S (Kowalchuk et al. 2001). Soil fungal communities are highly sensitive to soil water content, nitrate, and organic matter content (Wang et al. 2018), while bacterial diversity and community differences are strongly correlated to soil pH (Gri ths et al. 2011). Microbes have the general symbiotic relationship with soil and plants. They help plants to absorb nutrients (such as C, N, P and other) that restrict the growth of plants (Der Heijden et al. 2008), and ingest nutrients from plant secretions and litters (Zhalnina et al. 2018).
Due to the nonuniform distribution of effective nutrients and plant roots in soil, soil microbial communities have different biogeographic distributions (Eilers et al. 2012). Soil microorganisms respond signi cant differently to environmental factors (soil physical & chemical properties and plant community changes) at different depths (0-10 cm and 10-20 cm). Also, from the research of a multi-scale spatial assessment of soil bacterial community across the UK, a signi cant correlation between bacterial community and spatial distance was found out (Gri ths et al. 2011). While, in wetlands (Wang et al. 2010) and fallow farmlands (Ko et al. 2017), the size, activity, and diversity of microbes will decrease when soil depth increases. But the researches on grassland soil microorganisms mostly focused on the xed depth of the surface layer (0-10/20 cm) (Leff et al. 2015;Na et al. 2019), and a few studies on the depths of the surface layer. Soil microorganisms will be different when plant phenology changed, which is in uenced by the climate and ecosystem. In temperate forest soils, the relative abundance of Actinomycetes will increase obviously in winter, but the relative abundance of Acinetobacter and Proteobacteria will decrease (Santalahti et al. 2016). It is widely concerned in recent years about the changes of soil microbial communities during the vegetation growing season. In crops root microbial communities have been varied throughout the whole life cycle (Shi et al. 2015;Zhang et al. 2018), and they tend to deviate gradually from the soil microbiome and enrich microbial speci c groups. Oppositely, the microbial community in desert steppe has remained relatively stable over the course of the year. However, in Eurasian grassland, the soil temperature and water content are very variable between seasons, which would lead to the instability of plant litter and secretions, and affects the soil carbon input ultimately (Bardgett et al. 2005). Also soil organic carbon is the most important factor in driving the spatial distribution of microbial communities. When roots grow in summer, the effectiveness of soil carbon sources will also increase, while the effectiveness of carbon sources drop when roots activities stop in autumn. As a result, the rapidly growing microbial populations (which prefer to use direct carbon sources) have to slowed down their growth rate, which ultimately affect the structure of microbial population (Barboza et al. 2018).
Our research was conducted on the representative area of Eurasian Steppes in Inner Mongolia, China.
Contrastively analyzed were performed on soil bacterial and fungal community structure, diversity, and the impact of environmental factors on the community structure. Those analyses were to explore and answer the questions as follows: (1) what are the composition of soil bacterial and fungal communities in this area? (2) what or how did time(months) and soil depth affect the structure of soil bacterial and fungal communities? (3) How is the relationship between other environmental factors and soil bacterial / fungal communities?

Study site and soil sampling
The experiment site is located in Erlitu Ranch of Zhengxiangbai Banner, XilinGol, Inner Mongolia, China (42°9' 14" N, 115°14' 39" E). Soil samples were collected every month from June to October in 2018 (represented by May, Jun, Jul, Aug, Sep ). Each sample of 0-10 cm, 10-20 cm, and 20-30 cm depth (represented by 10, 20, and 30) was collected in the shape of "S" with a soil drill of 8cm in diameter. A total of six points were selected for soil collecting, and mixed into composite samples. After removing the animal / plant residues and other impurities like gravels by sieving (2 mm). Each composite sample was then divided into 3 subsamples as repeats for further DNA extraction (store in -80℃) and other soil properties analysis (store in -20℃).

DNA extraction, PCR ampli cation and sequencing
The total DNA of soil microorganisms was extracted from 0.5 g soil samples using DNeasy PowerSoil Kit (Qiagen Benelux B.V., Venlo, The Netherlands), and each soil sample has three parallels. DNA concentration and quality were assessed by ultramicro nucleic acid quanti er, and DNA integrity was checked using 1% agarose gel electrophoresis.
After mixing homogeneously the reaction system as above, the PCR program was carried out according to the following program. Pre-denaturation at 95 °C for 5 min, 15 cycles of denaturation at 95 °C for 1 min, annealing at 50 °C for 1 min, extension at 72 °C for 1 min, and a nal extension at 72 °C for 7 min. PCR products were detected by electrophoresis with agarose 1% gels, followed by high-throughput sequencing and analysis, which were based on the Illumina HiSeq 2500 platform (Illumina Int., San Diego, CA, USA). FLASH v1.2.7 (Magoč and Salzberg. 2011), Trimmomatic v0.33 (Bolger et al. 2014), and UCHIME v4.2 (Edgar et al. 2011) were used to control the quality of data and the effect of splicing to obtain Effective Tags. QIIME v1.8.0 (Caporaso et al. 2010) was used to cluster the obtained Effective Tags at a similarity level of 97%, and obtain OUT (Operational Taxonomic Units). The representative sequences of OTU are compared with the microbial reference database to obtain the classi cation information for each species corresponding to each OTU, and then the composition of each sample community was counted at each level (phylum, class, order, family, genus, species). And the taxonomic annotation was taken for OTU basing on Silva (for bacteria) and UNITE (for fungal) taxonomy database.

Statistical analysis
R (Mass Package) was used to carry out a general linear mixed model (GLMM). With month and depth as xed factors and sampling point as random factors, the variation rules of soil physical and chemical properties, microbial abundance, uniformity and diversity with month and depth were analyzed. At the same time, we conducted multivariate analysis of variance with SPSS software to explore the effects of month, depth and their interaction on soil and microorganisms and their signi cance.
Alpha diversity was analyzed after the taxonomic annotation of OTU using Mothur (Schloss et al. 2009), and the Shannon-Wiener Index S is the total number of species, Pi is the proportion of individuals of this species to the total number of individuals) was used to measure species. Beta diversity analysis was used to compare the similarity of species diversity among different groups. Among them, Principal coordinates analysis (PCoA) used dimensionality reduction thinking to observe the differences in microbial community composition between groups (Anderson and Willis. 2003). And analysis of similarities (Anosim) can test the signi cance of difference of beta diversity between samples of different groups. All these were analyzed and drew by the vegan package of R language (Anderson and Walsh. 2013;Team RC. 2014).
Another analysis used in this study is signi cance analysis of differences between groups. The LEfSe

Variation of soil physicochemical properties
Our results showed that soil pH and NH 4 + increased signi cantly with month, while VWC reached the maximum in August and September, and ST reached the maximum in July (p < 0.05, Table S1). Soil pH and NO 3 increased signi cantly with soil depth, while the other soil properties decreased signi cantly with soil depth (p < 0.05, Table S1). Moreover, month had signi cant effects on soil characteristics (except SOM) and bacterial abundance and evenness, but had insigni cant effects on bacterial diversity and fungal abundance, evenness and diversity. In addition, all indicators are signi cantly affected by depth (Table S2). The results of multivariate analysis of variance showed that the soil physical and chemical indexes (except NO 3 -) were also signi cantly affected by the interaction of month and depth (p < 0.05, Table 1).

Variation in bacterial and fungal community richness
The bacterial communities of all soil samples were mainly Actinomycetes, Proteobacteria, Acidobacteria, Chloro exi, Verrucobacteria, and Bacillus (95%). The fungal community was dominated by Basidiomycota, Ascomycota and Mortierella (85%)( Table S3). According to GLMM analysis results, bacterial abundance, diversity, and fungal abundance, evenness, and diversity all decreased signi cantly with the increase of depth, while bacterial evenness and fungal diversity increased signi cantly with month (Table S1). Our results showed that the changes in the abundance of phyla of bacteria and fungi were signi cantly affected by depth and month ( Table 2). The abundance of Actinobacteri, Verrucomicrobia, Gmatimonadetes, Ascomycota, Basidiomycota and Mortierellomycota increased signi cantly with the increase of depth. While the abundance of Proteobacteria, Acidobacteria and Chloro exi decreased signi cantly with the increase of depth (Table S3). In addition, the abundance of Ascomycota increased signi cantly with the increase of month, while the abundance of Verrucomicrobia, Gmatimonadetes, Basidiomycota and Mortierellomycota decreased signi cantly with the increase of month (Table S4).

Bate diversity of bacterial and fungal community
According to PCoA results, the bacterial and fungal community compositions of 10 cm were clearly different from that of 20 cm and 30 cm soil layers ( Figure 1A-B). As shown in Figure 1C-D, the bacterial community composition in August was quite different from other months, and the community composition in August and September was more specially. Moreover, according to ANOSIM 's analysis, soil fungi were not well grouped by depth and month, and soil bacteria were not well grouped by month either ( Figure 1B-D, Supplementary Figure 1).

Taxonomic composition
With the increase of soil depth (without considering the month factor), the abundance of Actinomycetes, Thermoleophilia, MB_A2_108, and some other bacteria were signi cantly increased, while the abundance of Acidobacteria, Proteobacteria, Rhizobiales, and Alphaproteobacteria were decreased (Figure 2A). Similarly, the abundance of Ascomycetes, Basidiomycetes, Agaricomycetes and some other fungi increased signi cantly with the increase of soil depth, while the abundance of Dothideomycetes, Hypocreales, Pleosporales and other fungi was signi cantly reduced ( Figure 2B).
If soil depth wasn't considered when the abundance of soil bacteria and fungi in different months were analyzed contrastively, the month did not have a signi cant effect on bacteria's abundance ( Figure 2C).
As compared with bacteria, the abundance of soil fungi was impacted more greatly by month. As shown in Figure 2D, the abundances of Mortierellomycetes and Phaeosphaeriaceae (belonging to Ascomycota) increased signi cantly in September; the abundances of Cantharellus and Ceratobasidiaceac (all belonging to Basidiomycetes) were highest in May; the abundance of Hypocreales (belonging to Ascomycota) increased in June, and the abundance of Fusarium (also belonging to Ascomycota) increased in August.

Factors driving bacterial and fungal communities composition
Through Pearson correlation analysis (p < 0.05), we found that month and depth were signi cantly correlated with bacteria and fungi respectively (Figure 3). TOC and VWC were signi cantly related to the bacterial community, while TOC and ST were signi cantly related to fungal aggregation (p < 0.05, Figure  3). We build a SEM model, which was based on the correlation analysis ( Figure 4). The result showed that month, depth, pH, TOC and AN had signi cant impact on the bacterial communities (p < 0.05, Figure 4A). As shown in Figure 4B, TOC and AN had directly positive effects on fungi, soil depth and ST in uenced fungi negatively (R 2 =0.35). The month weakened positive effect of AN on fungal communities and enhanced the negative effect of ST on fungal communities by its negative effect on AN and ST ( Figure  4B).

Discussion
This study took typical degraded grassland in Inner Mongolia as the basis for multiple sampling at xed point. Month and depth were used to represented time and space. The variation of pH, VWC and ST with were analyzed to understand the soil physicochemistry and nutrients characteristics variated as time and space changed. Meanwhile, the next generation sequencing technology was used to investigate the changes of microbial diversity and the driving factors of microbial community structure transformation with temporal and spatial variation. Our results provide strong evidence that spatial heterogeneity (depth) is more important than temporal (month) in predicting changes in microbial α-diversity and β-diversity. Variation in microbial community composition was driven by changing environmental factors in their habitat. Considering that soil microbial community composition was related to community function , temporal and spatial changes mainly interfered with the ecological function of soil microbial community, rather than the stability of grassland soil ecosystem.

Vertical spatial variation of microbial communities and soil properties
Our results indicate that all measured soil physicochemical indices are signi cantly affected by depth. Soil pH increased signi cantly with depth increasing, while ST, VWC, SOM and AN decreased signi cantly with increasing depth. This may because the top soil layer (0-10 cm) was seriously in uenced by external environment conditions. In particular, in Inner Mongolia, grasslands had been in uenced by human activities like grazing and mowing for very long time, which resulted in the decline of the productivity and diversity of grassland vegetation (Xun et al. 2018). Furthermore, human activities also resulted in the increased bare area, the aggravated erosion and coarseness of surface soil, and reduced nutrient content (Fierer et al. 2009). Relatively, the environment of the deeper layer (20-30cm) is stable. However, as increasing of soil depth, the distribution of plant roots decreased, and the plant litter and secretions decreased. It may lead to a lower soil nutrient content than the surface layer (Truongand and Marschner. 2018). As soil bulk density increases, porosity and oxygen content decrease, which is not conducive to the survival of microorganisms and inhibits the activities of enzymes involved in decomposition (Bagheri et al. 2013;Holt. 1997). And leads to the decrease of soil carbon and nitrogen availability ). According to our results, most of the soil physical and chemical indexes was signi cantly affected by the month, but each index varies without rule between months, which requires further study and discussion.
We found that both soil characteristics and microbial community structure were more signi cantly correlated with soil depth than with time variations. It was consistent with previous ndings, which con rm the importance of spatial heterogeneity (Fierer et al. 2006;Lauber et al. 2013), and can vary even on the scale of meters or even centimeters (O'Brien et al. 2016). Some studies have found that microbial community structure and abundance respond to changes in environmental factors to the same degree (Bell et al. 2014;Na et al. 2019), but some studies have shown that community composition is more sensitive than community diversity (Fierer et al. 2006). However, in our study was no found that response differences in community composition and diversity. Instead, it was found that bacterial abundance and evenness were signi cantly affected by both soil depth and month, while the abundance, evenness and diversity of fungi were only signi cantly affected by depth. We speculate that the response difference between bacteria and fungi is due to their own factors. Because bacteria are more susceptible to local changes in soil properties (Sorensen et al. 2013), while the evolutionary life history of fungi enables them to form hyphae structures and highly resistant spores that are able to withstand sudden environmental changes (Sun et al. 2017). In addition, the individual size of fungi is usually larger than that of the bacterial members of the community, which results in transmission limitations severely (Young. 2006;Schmidt et al. 2014). It is also worth noting that the interaction between month and depth weakened the effect of depth on microbial community, and only had a signi cant effect on fungal diversity. This indicates that the results of single factor and multi-factor in uence are quite different and unpredictable.
Therefore, future research on microbial ecology should set up as many control factors as possible to understand more really changes of microorganisms.

Driving factors of soil microbial community structure
Changes of soil properties had the potential impacts on variation of soil microorganisms in the vertical section. According to our results of LEfSe analysis, Acidobacteria and Proteobacteria were enriched in the top layer (0-10 cm), while Actinomycetes, Ascomycetes and Basidiomycetes had the high abundance in deeper layer (20-30 cm) (Figure 3). Simultaneously, VWC and SOM gradually reduced as depth increasing, in the opposite pH increased gradually (Table S1). Therefore, the variation of VWC and SOM explained here the abundance changes of Acidobacteria and Proteobacteria, and the variation of pH explained the abundance changes of Actinomycetes, Ascomycetes, and Basidiomycetes. Studies had shown that changes in soil fungal communities are signi cantly correlated with soil moisture and pH (Zheng et al. 2009), and pH is the main driving force for formation of soil microbial communities ). On the other hand, different microbial communities have different utilization of nutrients. Compared with fungi and actinomycetes, bacteria use smaller organic matter molecules, while fungi and actinomycetes can decompose substrates with relatively large molecules by producing lignin degrading enzymes (Bonanomi et al. 2017;Bonanomi et al. 2017). This may also be the reason why the fungal diversity in the soil layer of 30cm was greater than that of 10cm in the analysis of α diversity.
We also found that the temperature is highest in July and August, and the soil moisture content is highest in June and July. Soil pH is alkaline, NO 3 -, NH 4 + content is limited. The contents of NH 4 + and NO 3 are the minimum in August, and the pH is the highest in August. According to the analysis of in uencing factors, month has a direct and signi cant positive effect on bacteria, and promotes its signi cant positive effect on bacteria by indirectly affecting pH and VWC. However, from the perspective of taxonomic composition, this effect did not signi cantly affect the species abundance of the community. Indicating that the composition of soil bacterial community was relatively stable during the whole plant growing season.
However, month has no direct effect on fungi, but indirectly affects fungi through AN and ST. This indirect effect causes signi cant changes in species abundance of fungal community. In other words, AN and ST affected the changes of Ascomycetes, Basidiomycetes and Mortierellomycetes. However, some studies have shown that in the arid and semi-arid grassland ecosystem in the eastern part of Inner Mongolia, soil microbial biomass (Cmic, Nmic), soil TOC, TN, NH 4 + all increase with the increase of precipitation, while pH value decreases with the increase of precipitation (Yao et al. 2017) Nevertheless, 65-70% of the variation in microbial composition was not explained by month and depth, or environmental variables in our study. The possible reason is the existence of other unmeasured environmental factors that vary in space and time (Bahram et al. 2015), including biotic interactions such as competition, mutualism, and predation between microbial taxa (Zhou et al. 2017) and ecological processes such as dormancy and persistence traits of microbial communities and their members (Averill et al. 2019).

Conclusion
The article analyzed the spatiotemporal variation and driving factors of soil microbial community structure in typical degraded steppe. The results show that both month and soil depth had signi cantly effect on microbial community structure and soil properties, but depth has a more signi cant effect. The abundance and diversity of bacteria and fungi were signi cantly affected by depth. The abundance of Acidobacteria, Proteobacteria, Actinomycetes, Ascomycetes and Basidiomycetes varies signi cantly with depth. What's more, soil pH increased signi cantly with depth increasing, while SOM, AN, VWC and ST decreased signi cantly with increasing depth. Therefore, we speculate that SOM and VWC account for the abundance variations of Acidobacteria and Proteobacteria, and pH cause the abundance changes of Actinomycetes, Ascomycetes and Basidiomycota. In addition, this study only analyzed the microbial changes in each month of the plant growing season in a year, which may underestimate the real microbial time changes. Therefore, more and longer time points should be included in the design of future similar studies, including those on microbial biogeography. In conclusion, spatial and temporal studies of soil microbial ecology provide a more comprehensive basis for understanding the key factors that regulate biodiversity in soil ecosystems.