Inconsistent response of bacterial phyla diversity and abundance to soil salinity in a Chinese delta

Soil salinization is an increasingly serious problem and decreases crop yields in the Yellow River Delta (YRD), but its effects on bacterial community and diversity at the phylum level are not well known. We used high-throughput sequencing of soil bacterial 16S rRNA to identify soil bacterial communities and diversity across a gradient of soil salinity (electrical conductivity), namely, S1: low salinity level (1.78 ds/m), S2: medium salinity level (3.16 ds/m), S3: high salinity level (17.26 ds/m), S4: extreme salinity level (34.41 ds/m), and a non-salted site as the control (CK, 0.92 ds/m). Our results indicated the significantly higher values of soil C/N ratio in S2, S3, and S4 compared with that in CK. Significantly lower values of the Shannon and Chao 1 indexes were observed in S4 compared with the CK (p < 0.05). High salinity decreased the relative abundance of Actinobacteria and Acidobacteria, but increased that of Gemmatimonadetes and Bacteroidetes. Additionally, the Shannon diversity of Bacteroidetes increased by 15.5% in S4 compared with that in the CK. Our results indicate that soil salt is a main factor regulating bacterial phyla diversity and community in the extremely saline-alkaline soils of YRD. The high abundance and diversity of Bacteroidetes can be used for saline-alkali land restoration.

. The α-diversity and species richness of soil bacteria in the control (CK), low salinity level (S1), medium salinity level (S2), high salinity level (S3), and extreme salinity level (S4) sites. Values are the mean ± standard error. Sobs: the actual observed richness; Shannon: the Shannon diversity index; ACE: Ace index of species richness; Chao1: Chao1 index of species richness. In the table, the significant relationships at p < 0.05 are indicated by different letters based on the DUNCAN test.  www.nature.com/scientificreports/ abundances of Gemmatimonadetes, and S3 plot have the highest abundances of Bacteroidetes (Fig. 2a). However, the relative abundances of Actinobacteria and Acidobacteria under high salinity were lower than those under low salinity (Fig. 2a). Furthermore, the class Alphaproteobacteria and Betaproteobacteria under high salinity were lower than those under low salinity, and soil salinization significantly increased the relative abundances of Gammaproteobacteria (Fig. S2b).
The Shannon diversity at the phylum level is shown in Fig. 2b. Soil salinization significantly increased the Shannon diversity of Bacteroidetes. However, the Shannon diversity of Actinobacteria and Chloroflexi decreased dramatically at extremely high salinity (S4). Interestingly, the Shannon diversity of Proteobacteria, Acidobacteria, Gemmatimonadetes and Cyanobacteria showed an increase from CK to S3 and then a decrease as the salt increased from S3 to S4.
Spearman correlation analyses showed that soil pH was not correlated with the bacterial abundance and diversity (Fig. 4). The relative abundances of Acidobacteria, Actinobacteria, Nitrospirae, and Cyanobacteria were negatively correlated with soil salt, EC, and the C/N ratio and positively correlated with soil C and N. In addition, The relative abundances of Alphaproteobacteria and Betaproteobacteria were negatively correlated with soil salt, EC, and the C/N ratio and positively correlated with soil C (Fig. S2d). However, the relative abundances of Firmicutes, Gemmatimonadetes, and Bacteroidetes were positively correlated with soil salt, EC, and the C/N ratio and negatively correlated with soil C and N. Interestingly, only the Bacteroidetes Shannon diversity showed a positive correlation with soil salt, EC, and the C/N ratio compared with the other phyla (Fig. 4).

Discussion
The bacterial community under the five salinity levels was mainly formed by Proteobacteria, Actinobacteria, Bacteroidetes, Gemmatimonadetes, Acidobacteria, and Chloroflexi in this study. The phylum Proteobacteria occupied the largest proportion across the five salinity levels due to its high growth rate and strong metabolic capacity 16 . The relative abundance of Proteobacteria slightly increased as salinity increased. Similar findings reported that the Shannon diversity of Proteobacteria increased from CK to S3 and then declined significantly as salinity increased from S3 to S4 4 . Moreover, previous studies have also reported that the relative abundances www.nature.com/scientificreports/ of Alphaproteobacteria and Gammaproteobacteria were enriched in high-salinity soils 17 , and Gammaproteobacteria was also the dominant class in the saline soils in the salt lake system 18 , while Betaproteobacteria were more abundant in low-salinity soils 4 . Similar results were obtained also in the present study. Those results indicated that different bacterial species show different levels of tolerance to salt, even under the same phylum. The abundance of Betaproteobacteria decline in high-salinity soils due to the increase of the osmolarity outside the microbial cell 1 ,  . Spearman correlation analyses between soil physiochemical properties and the relative abundances and Shannon diversity of soil bacterial phyla. EC: electrical conductivity, TC: total carbon, TN: total nitrogen, and C/N ratio: soil total carbon/nitrogen. *, **, and *** indicate significance along the paths at p < 0.05, p < 0.01, and p < 0.001 levels, respectively. www.nature.com/scientificreports/ while Gammaproteobacteria was reported to be halotolerant 19 , and the salt tolerance mechanism needs further study. Furthermore, the relative abundances of Actinobacteria, Nitrospirae, Acidobacteria, and Chloroflexi were negatively correlated with soil salinity. In particular, both Actinobacteria and Acidobacteria sharply declined with increasing soil salinity in this study. These bacteria generally prefer to live in acidic soils 4,20 . Importantly, the phyla Gemmatimonadetes and Bacteroidetes are the two dominant populations in alkalinesaline soils and increased significantly as salinity increased in this study. In our previous study with higher salinity between pH 8.27 and 10.37 in semi-arid grassland, soil salinization dramatically increased the relative abundance of Gemmatimonadetes and Bacteroidetes 21 . Similarly, the relative abundance of the phylum Gemmatimonadetes increased following irrigation with salt water in irrigation system in arid desert 22 . However, some recent studies found that the abundance of Bacteroidetes was slightly affected by soil salinity in the YRD, which might be because this phylum can tolerate a wide salinity range 23,24 . Bacteroidetes showed no significant changes across a range of salinity levels from 0.34 to 6.73 ds m −14 . Interestingly, as the salinity levels increased from 0.92 to 34.41 ds m −1 in this study, the phylum Bacteroidetes was found to be significantly enriched in the extremely saline-alkaline soils. Additionally, the Shannon diversity of Bacteroidetes significantly increased in highly saline-alkaline soils. The salt tolerance mechanism and biological function of these microorganisms need further study in order to the ecological restoration in the extremely saline-alkaline soils of the YRD.
Salinity is commonly found to negatively influence the community structure and diversity of microbes in soils from reclaimed coastal land 25 and coastal wetlands 16 . Salt stress increases the extracellular osmotic pressure of the microbial cell wall, decreasing microbial activities 1 . Indeed, RDA and the Mantel test revealed that soil salinity significantly affected the bacterial community structure and Shannon diversity.
A large part of the variation in soil bacterial composition is commonly related to differences in soil properties, such as the pH value 21 , salinity value 3 , ion concentrations 4 , and soil nutrient levels 26 . Generally, nutrient availability, serving as energy sources for microbes, is known to positively determine bacterial activities 27 . In the present study, some bacteria (Firmicutes, Gemmatimonadetes, and Bacteroidetes) were negatively correlated with soil C and N, and the relative abundances of Acidobacteria, Actinobacteria, Nitrospirae, and Cyanobacteria were positively correlated with soil C and N. As a recently emerged coastal wetland, nutrient contents (such as TN or TC) in the YRD were proved lower than in other coastal wetland sediments 28 . Along with soil salinity, TN and TC were significantly reduced, which affected distribution patterns of bacterial communities in this area. A study that conducted large-scale sampling of French soils indicated that pH was the dominant environmental factor controlling the bacterial phyla distribution 26 . A meta-analysis showed that the global microbial diversity and composition in saline soil are more affected by salinity than other extreme soil factors, such as pH 15 . Our recent findings revealed that soil salinity and pH have clearly defined the microbial communities in the highsalinity soils of the grasslands in northern China 21 , and the results of another study implied that soil pH is an equally important factor as soil salinity in shaping the soil bacterial community structure 18 . The RDA, Monte Carlo permutation test, and Spearman correlations in the current study suggested that soil pH had no effect on the bacterial abundance and diversity. Conversely, the soil salt content or EC was found to significantly influence the community structure and diversity of microbes. This is in accordance with the finding that soil salt is more important than soil pH in driving the bacterial phyla distribution in saline-alkaline soils of the YRD 4 . The soil salinity and pH commonly have a collinear relationship in saline-alkaline soils 18 ; however, the soil salinity did not differ in response to soil pH based on Spearman correlation analyses in the current study. We speculate that soil bacteria in northeastern China are mainly affected by soil pH 21 , compared with salinity (soil EC and salt content) in the YRD. Hence, it is also important to study the relationship between soil saline and alkali conditions, especially in the extremely saline-alkaline soils of the YRD.

Conclusions
This study explored the distribution patterns of soil bacterial communities and diversity in the extremely salinealkaline soils of YRD. The α-diversity of soil bacteria showed a consistent response to soil salinity, and significantly lower values of the Shannon and Chao 1 indexes were observed under high salinity conditions. Bacterial diversity at the phylum level showed an inconsistent response to soil salinity compared with abundance. Compared with soil pH, the soil salt content is extremely negatively correlated with bacterial community and Shannon diversity, implying that soil salt was the main factor that shaped the soil bacterial communities and diversity in the extremely saline-alkaline soils of YRD. Considering its high salt tolerance, it is urgent to study the function of Bacteroidetes in order to the improvement and restoration of saline-alkali land. Our results provides a framework for future research to deeply analyze the mechanism and function of salt tolerance of soil bacteria in saline-alkaline environments.

Materials and methods
Study sites. This study was located in Wudi County, which is part of the YRD in northern Shandong on the southern shore of the Bohai Sea (37°54′60″N, 117°57′33″E, elevation 1 m). This area has a semi-humid continental climate characterized by a mean annual precipitation and air temperature of 600 mm and 12 °C, respectively. We selected five different salinity levels from low to extreme salinization 29 . In brief, maize croplands with low salinity were selected as the control (CK) and are mainly affected by freshwater flooding. Land covered by Setaria viridis, low salt-tolerant vegetation, was selected as low salinity level (S1). Saline-alkali land dominated by Suaeda salsa, medium salt-tolerant vegetation, was selected as medium salinity level (S2). Saline-alkali land without vegetation growth but with salt crystallization was selected as high salinity level (S3), and extreme salinity level (S4) was the saline-alkali land with salt crystallization. The soil electrical conductivity value ranged from 0.92 ds/m (CK) to 1.78 ds/m (S1), 3.16 ds/m (S2), 17.26 ds/m (S3), and finally 34.41 ds/m (S4) according to our previous study (Fig. 5)  www.nature.com/scientificreports/ Soil collection. Four transects across a distance of approximately 3 km represented four repetitions, and in each transect, five plots (CK, S1, S2, S3, and S4), which were spaced at least 500 m part, were randomly selected. In each plot (5 × 5 m 2 ), the topsoil (0-15 cm) was collected using five-spot sampling in October 2019. The plant debris was removed, and we mix the five-point sample into one sample. Hence, a total of 20 samples (5 salinity levels × 4 repetitions) were collected, and we divided the soil samples into two subsamples. One subsample was air-dried for the analysis of basic soil properties, and the other part was placed in a − 80℃ freezer for microbiological analysis 21 . The physical and chemical properties and measurement methods of the soil are listed in the Supplementary Materials. Some basic characteristics for the soils in Table S1 were cited from our previous study 29 .
High-throughput sequencing of soil bacteria. The genomic DNA was extracted from 0.30 g of soil using the MoBio PowerSoil DNA Isolation Kit (QIAGEN, Ins., USA). The V3-V4 regions of the bacterial 16S rRNA gene were amplified using universal primers 338F (5'-ACT CCT ACG GGA GGC AGC AG-3') and 806R (5'-GGA CTA CHVGGG TWT CTAAT-3') 30 . The PCR analysis included pre-denaturation at 95 °C for 3 min, 27 cycles at 95 °C for 30 s, annealing at 55 °C for 30 s, elongation at 72 °C for 45 s, and an extension at 72 °C for 10 min. Illumina MiSeq sequencing produced double-ended sequence data (2 × 300) according to standard protocols performed by MajorBio Bio-Pharm Technology Co. Ltd. (Shanghai, China). The obtained sequences were first filtered using the quantitative insights into microbial ecology. Raw FASTQ files were de-multiplexed and quality-filtered with the following criteria: (i) 300-bp reads were truncated at any site with an average quality score < 20 over a 50-bp sliding window, and truncated reads shorter than 50 bp were discarded; (ii) exact barcode matching, less than two nucleotide mismatches in the primer, and reads containing ambiguous bases were removed; (iii) only overlapping sequences longer than 10 bp were assembled according to their overlapped sequence. Reads that could not be assembled were discarded. Then, we used UPARSE ver. 7.1 to cluster the high-quality sequences with 97% identity threshold into operational taxonomic units (OTUs) 31 . The chimeric was identified and removed using UCHIME., and the chloroplasts and mitochondria sequences were removed at the classification step using Greengenes database. Because the sequencing depth varied across samples, we used a sub-sampling procedure to normalize the number of reads to the minimum observed across all samples. In total, 1,145,774 raw read sequences were obtained from 20 soil samples with at least 32,212 sequences per sample. A total of 48,240 observed OTUs were obtained from all samples with an average of 2,412 per sample. www.nature.com/scientificreports/ Statistical analysis. The α-diversity indices, including the coverage, Sobs (the actual observed richness), ACE (Ace index of species richness), Chao1 (Chao1 index of species richness), and Shannon diversity index, were classified using MOTHUR software. The Shannon diversity at the phylum level was calculated as follows 21 : where Shannon phyla is the bacterial diversity at the phylum level, N is the total number of OTUs in each bacterial phylum, and Ni is the number of individuals in group i. The significant differences in the soil total bacterial α-diversity, bacterial Shannon diversity at the phylum level, bacterial community, and soil physicochemical properties of the five salinization levels were evaluated using one-way ANOVA in SPSS (ver. 19.0). The significance was analyzed at p < 0.05 using DUNCAN's test.
Nonmetric multidimensional scaling (NMDS) based on Bray-Curtis similarity matrices was performed to identify the response of soil bacteria to salinity. The significance was tested by analysis of similarities (ANOSIM) in PAST (ver. 3.25).
The relationships between soil physicochemical properties and the soil bacterial communities at the phylum and class level and the diversity of bacterial phyla were analyzed by redundancy analysis (RDA) using CANOCO (ver. 4.5). The significance of the effect of each property was examined using the Mantel test (permutation = 999), and the significance was evaluated by the r-and p-values. Spearman analyses were performed to identify the correlations between the soil physicochemical properties and the relative abundances and diversity of bacterial phyla.

Data availability
Sequence data supporting the findings of this study have been deposited at NCBI database under Sequence Read Archive (SRA) accession number SRP268965.