Biogeographical distribution of bacterial communities in Changbai Mountain, Northeast China

Abstract The broad‐leaved and Korean pine mixed forest in Changbai Mountain, China is an important component of boreal forest; the area is sensitive to global climate change. To understand spatial distribution patterns of soil bacterial community along elevation, we analyzed the soil bacterial community diversity and composition along an elevational gradient of 699–1177 m in a primitive Korean pine forest in Changbai Mountain using the high‐throughput sequencing. In total, 149,519 optimized sequences were obtained. Bacterial Shannon index increased along elevation from 699 m to 937 m and started to decrease at the elevation of 1,044 m, showing a humpback curve along elevation. Evenness (ACE index) and richness (Chao index) of the soil bacterial community both decreased with elevation (the highest values of 770 and 762 at 699 m and the lowest values of 548 and 539 at 1,177 m, respectively), all the indices are significantly different between elevations. Bacterial composition at phylum and genus levels had some differences between elevations, but the dominant bacterial populations were generally consistent. Beta‐diversity analysis showed a distance‐decay pattern of bacterial community similarity at different samples. Soil physical and chemical properties explained 70.78% of the variation in bacterial community structure (soil pH explained 19.95%), and elevational distance only explained 8.42%. In conclusion, the contemporary environmental disturbances are the critical factors in maintaining the bacterial spatial distribution compared with historical contingencies.

Distribution pattern and functional characteristics of soil microbes have substantial impacts on growth of aboveground parts of plants (Chabrerie, Laval, Puget, Desaire, & Alard, 2003;Gömöryoá, Hrivnák, Janišová, Ujházy, & Gömöry, 2009). Therefore, studying distribution pattern of soil microbial community is very necessary to better preserve and utilize the old-growth forests and resolve forest eco-environment imbalances .
Research on elevation-dependent microbial diversity is indispensable for basic ecological research and extremely crucial in forecasting responses of terrestrial ecosystems to global climate changes (Shen et al., 2014). In the past few years, modern biological techniques, particularly high-throughput sequencing, supplied a powerful technical support for studying spatial distribution patterns of soil microbes.
The Changbai Mountain is located at mid-to high-latitudes, with a significant regional characteristics (Wang et al., 2015). It is very sensitive to global climate change; so has been considered as an ideal zone to study positive and negative feedback mechanisms of temperate forest to global climate change. Broad-leaved and Korean pine mixed forest (Pinus koraiensis Sieb. et Zucc as an edificator) is the most representative and diverse forest ecosystem in Northeastern China (Yu et al., 2013). Recently, the broad-leaved and primitive Korean pine forests disappeared in most of its distribution areas, whereas Changbai Mountain, as one of the core distribution areas of Korean pine forest, is still well covered by the most representative vertical gradientdistributed old-growth Korean pine forest (Shen et al., 2014;Yu et al., 2011Yu et al., , 2013. At elevations from 700 to 900 m, the main forest type is a broad-leaved and Korean pine forest in Changbai Mountain; it turns to spruce-fir and Korean pine mixed forests at higher elevations (Li, Bai, & Sang, 2011;Zhao, Fang, Zong, Zhu, & Shen, 2004). Here, we investigated diversity and composition of soil bacterial community and discussed underlying factors affecting vertical distribution patterns of the Korean pine forests by using Illumina High-throughput sequencing. We hypothesized that the variation in soil physical and chemical factors and vegetation types along elevations would have important impacts on bacterial community structure. Our results will provide a reference for a better understanding of the relative contribution of soil bacterial community to the forest response to changing environmental conditions.
We totally collected 30 soil samples at five elevational gradients in September, 2015. The selected plots characteristics can be seen in our previous study (Ping et al., 2017). We first chose a typical primitive Korean pine forest (200 m × 400 m in area) at each elevation (699, 818, 937, 1044, 1,177 m), soil samples were collected from three plots (20 m × 20 m; about 200 m apart) as three independent replicates. In each plot, soil samples (5-10 cm and 0-5 cm depth) were collected at 10 random points (15 cm × 15 cm, composited together as a single sample), then taken back to the laboratory in bubble boxes with ice bags. The fresh soil samples were passed through a 2-mm sieve to remove plant roots and residues, then subdivided into two subsamples: one was stored at −80°C prior to DNA extraction, another was air-dried to measure the physical and chemical characteristics.

| Physical and chemical properties analyses
Soil total nitrogen (TN), total organic carbon (TOC), moisture, and pH were measured according to the methods described by Ping et al. (2017).

| DNA extraction, PCR, and sequencing
DNA was extracted from soil sample (0.25 g wet weight) using the Power Soil DNA isolation kit (MoBio Laboratories, Inc., Carlsbad, CA).

| Processing and analyzing of sequencing data
Quality-filtered and operational taxonomic units (OTUs) were clustered at 97% sequence similarity using QIIME. Chimera was detected by UCHIME. Shannon diversity index, Chao index (richness) and ACE index (evenness) were estimated at 97% sequence similarity using Mothur (Sáenz de Miera et al., 2014;Xu, Ravnskov, Larsen, Nilsson, & Nicolaisen, 2012). Beta diversity was calculated using Weighted UniFrac metric. The sequences obtained in the study have been deposited at the NCBI under BioProject ID PRJNA323515.

| Statistical analysis
Variation partition analysis was conducted in Vegan packages in R and canonical correspondence analysis using Canoco 4.5 software.
Taxonomy analysis was assigned to OTUs (at the level of 97% similarity) against Silva database, using the RDP classifier Jest algorithm in QIIME (Ping et al., 2017).

| Sequence results and diversity indices
A total of 149,519 sequences were generated from the 10 soil samples.
As shown in Table 1, the length of sequence varied from 434.22 bp to 440.12 bp, and the average sequence length is 437.51 bp.
Soil bacterial diversity (Shannon index), evenness (ACE index) and richness (Chao index) (97% sequence similarity) presented inconsistent changing patterns along elevation, and there was a significant difference between 1,177 m and others elevations (Table 1)  (1.47%), accounting for 97.77% of the total relative abundance.

| Bacterial community composition
Proteobacteria, Acidobacteria, Actinobacteria, and Verrucomicrobia were the most prominent phyla. The relative abundance of Acidobacteria exhibited a significant difference along elevations only.  Group 3 consisted of surface soil at 1,177 m.

| Relationship between bacterial community and environmental factors
Canonical correspondence analysis suggested soil physicochemical factors significantly influenced the relative abundance of the 10 top genera ( Figure 5); soil pH, C/N ratio and moisture are the main factors.

| Effects of elevation on bacterial diversity
Recent studies showed inconsistent conclusions on spatial distribu-  (Table 1), showing a humpbacked curve which was consist with the result of Singh et al. (2012). A possible explanation may lie in soil moisture, an important factor affecting bacterial diversity (Brockett, Prescott, & Grayston, 2012), and it explained 12.06% of the variation in bacterial community structure ( Figure 6). In addition, aboveground vegetation composition is another factor determining soil bacterial diversity. Because vegetation types determine litter composition and soil C/N ratio, then affect microbial community structure (Djukic et al., 2010;Singh, Shi, & Adams, 2013). We discovered a gradual change in the plant composition at the elevation of 1,044 m, the presence of coniferous trees increased (Ping et al., 2017), the litter composition changed then had an impact on soil bacterial community.
In this study, we found bacterial richness and evenness declined with elevation; soil pH also declined with the increasing elevation. Soil

| Effects of elevation on bacterial community composition
At phylum level, there was no significant difference in soil bacterial community composition in the sample plots at different elevations.
Janssen (2006)  Thus it can be concluded from the above results that at the phylum level, the dominant bacterial populations are generally consistent in different ecosystems and soil types (Janssen, 2006;Wang et al., 2014).
At the genus level, there was difference in bacterial composition at different elevations; the common genera accounted for 65.69% of total genera number and 96.48% of total relative abundance. The category of the common genera whose relative abundance exceeded 1% was consistent, whereas the relative abundance of 29.30% common genera had significant differences at different elevations. Elevation had impacts on the relative abundance of the bacterial community, particularly Granulicella, WD272, Rhodanobacter, Brevundimonas, and KD4-96 ( Figure 3). Our result agreed with the conclusions reported by Shen et al. (2013), Zhang, Liang, He, and Zhang (2013) and Wang et al. (2014). Although the dominant bacterial species may be the same, there was considerable variation in the abundance of members in different soil types (Janssen, 2006). The abundance may be affected by soil environmental factors, including physical, chemical, and biological factors. In this article, soil pH, C/N ratio and moisture were the main factors affecting the relative abundance of prominent genera along elevation ( Figure 5). Variation partition analysis demonstrated soil nutritional content, moisture, and pH had high contributions to variation in soil bacterial community with elevation ( Figure 6). Furthermore, variation in plant communities was a significant factor influencing the relative abundance of bacterial communities, which was confirmed by taxonomy analysis (Figure 4).

| Spatial distribution of bacterial community
According to the spatial distribution of soil microbial communities, two models (distance-decay relationship and species-area relationship) were used to describe the nonrandom distribution of microorganisms (Green & Bohannan, 2006;. Distance-decay relationship (that is β-diversity variation) has been a general biogeographic model reflecting microbial spatial distribution, which was adopted to investigate the change in communities composition similarity with the increasing spatial distance among samples (Green & Bohannan, 2006;Green et al., 2004). Bryant et al. (2008) reported that bacterial community was nonrandom distribution, exhibiting obvious spatial distribution with elevation, and the phylogenetic similarity decreased with distance (phylogenetic distance-decay). Martiny, Eisen, Penn, Allison, and Horner-Devine (2011) compared the ammonia-oxidizing bacteria community structure in 106 sediment samples from 12 salt marshes on three continents, and found the similarity between two samples declined with larger distances. In this research, β-diversity suggested that the similarity of bacterial community among the plots declined with the increasing elevational distance, presenting a distance-decay pattern ( Figure 1); bacterial community exhibited the characteristic of nonrandom distribution.
We should simultaneously consider the relative contribution of contemporary environmental factors (soil physical and chemical properties) and historical contingencies (geographic distance) when discussing the formation and maintenance mechanisms of spatial distribution pattern of the soil microbial community (Green et al., 2004;. Variation partition analysis suggested that the impact of soil physicochemical factors on bacterial community structure was much larger F I G U R E 6 Variation partition analysis of the effects of geographic distance and soil variables on the phylogenetic structure of bacterial communities than geographic distance (mainly elevational gradients) ( Figure 6).

| CONCLUSION
Elevational gradient had a significant impact on soil bacterial commu-