Edited by: Hongyue Dang, Xiamen University, China
Reviewed by: Wei Xie, Tongji University, China; Simone Jaqueline Cardoso, Universidade Federal de Juiz de Fora, Brazil
This article was submitted to Aquatic Microbiology, a section of the journal Frontiers in Microbiology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
In this study, Illumina MiSeq sequencing technique was employed to explore the characteristics and dynamics of cyanobacteria–heterotrophic bacteria between two estuarine reservoirs in sub-tropical (reservoir A in Shanghai) and tropical (reservoir B in Singapore) regions. The results indicated that significant differences in bacterial community composition were found between two estuarine reservoirs, which influenced by varied environmental variables. The environmental heterogeneity in reservoir A was much higher, which indicated that the composition of bacterial community in reservoir A was more complex. In contrast, reservoir B provided a suitable and temperate water environment conditions for bacterial growth, which resulted in higher community diversity and less co-exclusion correlations. The molecular ecological network indicated that the presence of dominant bacterial community in each of the reservoir were significant different. These differences mainly reflected the responses of bacterial community to the variations of environmental variables. Although
In aquatic ecosystems, the balanced relationships of cyanobacteria–heterotrophic bacteria play important roles in maintaining aquatic ecologic stability (
Estuaries are important water sources for numerous cities such as Shanghai, Hong Kong, and Singapore, which located in estuarine areas around the world (
In our study, Shanghai and Singapore, as metropolitan cities located in estuary areas with different geographical characteristics, were chosen as our study sites. High-throughput sequencing (HTS) techniques were used to evaluate the diversity and variations of bacterial community composition in estuaries between different ecological regions. Especially, the molecular ecological network was implemented to explore the composing characteristics of cyanobacteria–heterotrophic bacteria between these estuarine reservoirs. We hope this research could promote the understanding of interactions within cyanobacteria–heterotrophic bacteria and how environmental factors affect their compositions in estuary reservoirs.
Two typical estuarine reservoirs in Shanghai (reservoir A) and Singapore (reservoir B) were selected in our studies, which represented estuaries ecosystems in sub-tropical and tropical regions, respectively (Supplementary Figure
Two sampling sites in each reservoir were selected to represent different hydrological conditions: QI and QE sites represented the midstream and downstream location of reservoir A, respectively. While in reservoir B, MA and MB represented the meeting points of different tributaries and main reservoir. All water samples were collected monthly for a year (May 2014 to April 2015) at a depth of 0.5 m below the surface water from these four sites.
Water temperature and pH were measured
Water quality parameters including turbidity (NTU), chlorophyll-a, total phosphorous (TP) and total nitrogen (TN) were analyzed immediately according to water and wastewater monitoring analysis methods (
The genomic DNA was extracted from 500 mL of water sample from each site by using E.Z.N.A. Water DNA Kit (Omega, Inc.) according to the manufacturer’s specifications. The extracted DNA was detected using Qubit 3.0 fluorometer (Invitrogen, Inc.) and frozen at -20C for further analysis. The PCR amplification was performed according to the two-step target amplicon sequencing protocol (
Raw sequencing data (NCBI BioProject:
The quantification of cyanobacterial 16S rRNA gene was carried out in a StepOnePlus real-time PCR (qPCR) system using primers and probes designed in a previous study (
Before statistical analysis, both biological and environmental data were first preprocessed, and standardized. The relative abundance of bacterial OTUs were square-root-transformed and environmental data were normalized. The similarity matrices of biological data and environmental characteristics were then calculated based on Bray–Curtis similarity and Euclidean distance, respectively. The contribution of each measured environmental variable to the variation within the bacterial community composition in each reservoir were evaluated based on distance-based linear model (DistLM) implemented in PRIMER v7 (PRIMER-E Ltd., Unite Kingdom) (
In each reservoir, the bacterial OTUs which were observed in at least five samples (>20%) and contributed at least 1% to the sample, were selected from the raw OTU data set. The observed abundance of these bacterial OTUs was not altered within any sample. The selected OTUs were further used for assessing the degree of association with respect to environmental parameters across the entire sampling period based on Pearson’s correlation coefficient (
During the sampling period, seasonal variations in reservoir A was observed with higher temperatures recorded from May to October and lower temperature from November to April (Supplementary Figure
After quality control, a total of 2,900,208 high quality sequences were observed both from reservoir A and B, with an average of 60,421 sequences in each sample. All these sequences were aligned against the SILVA database, and finally total of 2,093 bacterial OTUs were identified. These bacterial OTUs were belonged to 30 bacterial phyla and 249 genera. Among these bacterial OTUs, 20 bacterial phyla and 127 genera were observed in reservoir A. In contrast, 29 bacterial phyla and 198 genera were found in reservoir B. In reservoir A, the most dominant bacterial phylum was Actinobacteria (33.6%), and followed by Proteobacteria (22.7%), Bacteroidetes (20.0%), and Cyanobacteria (15.0%) (Figure
During the further classification levels, most of actinobacterial OTUs were classified as Actinobacteria (class level) in reservoir A, but more actinobacterial OTUs were classified as Acidimicrobiia in reservoir B (Figure
The results of high-throughput sequence indicated that the
Relative abundance of different cyanobacterial genera over time in reservoir
Comparison between two reservoirs showed that the species richness and diversity (inverse Simpson indices) were significant higher in reservoir B than in reservoir A in most of the year (Figure
Species Richness and diversity (Inverse Simpson Index) of bacterial community composition in reservoir
Results of ANOSIM and PERMANOVA tests (Supplementary Tables
The step-wise DistLM indicated that TN, TN/TP, temperature, chloride, Ca2+, Mg2+, sulfate, total cyanobacterial 16S gene copies and rainfall (30 d) were the main environmental factors influencing the variations of bacterial community composition in reservoir A (Table
DistLM results of abundant bacterial community data against environmental variables (9,999 permutations).
Variable | Reservoir A |
Reservoir B |
||||||
---|---|---|---|---|---|---|---|---|
SS(trace) | Pseudo-F | Prop. | SS(trace) | Pseudo-F | Prop. | |||
Chl-a | 3015.5 | 1.8793 | 0.0761 | 8.21E-02 | 3083.3 | 1.4897 | 0.0546 | 6.34E-02 |
pH | 2079.7 | 1.261 | 0.2228 | 5.66E-02 | 2438 | 1.1615 | 0.2207 | 5.01E-02 |
Turbidity | 3186.6 | 1.9961 | 0.0642 | 8.68E-02 | 4233.6 | 2.0985 | 8.71E-02 | |
TN | 11809 | 9.9578 | 0.32166 | 2973.7 | 1.4333 | 0.0689 | 6.12E-02 | |
TP | 2845.6 | 1.7645 | 0.0939 | 7.75E-02 | 3132 | 1.5149 | 0.0515 | 6.44E-02 |
TN/TP | 6276.7 | 4.3309 | 0.17097 | 2371.8 | 1.1283 | 0.2496 | 4.88E-02 | |
Temperature | 12128 | 10.36 | 0.33035 | 2601.3 | 1.2437 | 0.1566 | 5.35E-02 | |
Chloride | 8729.8 | 6.5516 | 0.23779 | 3748 | 1.8377 | 7.71E-02 | ||
Ca | 10699 | 8.637 | 0.29143 | 1871.2 | 0.88063 | 0.6299 | 3.85E-02 | |
Mg | 10304 | 8.1937 | 0.28067 | 2899.4 | 1.3952 | 0.0848 | 5.96E-02 | |
Sulfate | 7216.9 | 5.1384 | 0.19658 | 2807.4 | 1.3482 | 0.1064 | 5.77E-02 | |
CYAN | 3600 | 2.2832 | 9.81E-02 | 4510.5 | 2.2498 | 9.28E-02 | ||
Rainfall (30 d) | 7893.8 | 5.7523 | 0.21502 | 2665.6 | 1.2762 | 0.1382 | 5.48E-02 |
The two-dimensional nMDS plot of reservoir A showed that most of samples presented significant seasonal distribution, where samples from May to October were correlated with high temperature, chlorophyll-a, pH, total cyanobacterial 16S gene copies and rainfall (30 d); whereas samples from November to April were more correlated with high concentrations of chloride, sulfate, Ca2+, Mg2+, TN and TN/TP (Figure
The abundance of bacterial OTUs which contributed >1% to any sample in each reservoir were selected, amounting for a total of 58 bacterial OTUs in reservoir A and 85 bacterial OTUs in reservoir B, respectively. Combined with 15 environmental variables including chl-a, pH, turbidity, TN, TP, TN/TP, temperature, chloride, sulfate, Ca2+, Mg2+, total cyanobacterial 16S gene copies and rainfall (1 d, 5 d, 30 d) in each reservoir were selected to calculate the linear pairwise correlations using the rcor.test in the ltm package. Finally, 367 tests were considered significant (
In co-occurrence pattern (positive correlations) networks of reservoir A, all nodes were clustered into eight modules after modularity classification, including module I (18.57%), module II (18.57%), module III (17.14%), module IV (14.29%), module V (11.43%,), module VI (10%), module VII (8.57%), and module VIII (1.43%) (Figure
In contrast, the co-occurrence pattern network of reservoir B exhibited distinct constitutive characteristics, which further clustered into seven different modules (Figure
The co-exclusion pattern networks between two reservoirs revealed differences in composing characteristics (Figures
Although the co-excluded pattern network in reservoir B was clustered into six different modules, the composing characteristics of reservoir B was much simple, including less bacterial OTUs and environmental variables (Figure
Although many studies have been conducted in order to gain insights into microbial community diversity and dynamic variation in freshwater and estuarine ecosystems of the world, still little is known about the composing characteristics of bacteria in estuary ecosystems between different regions. Here, we provide a comprehensive evaluation by using the Illumina MiSeq sequencing technology and co-occurrence/exclusion pattern network analysis, also examined different influencing environmental factors on correlations of cyanobacteria–heterotrophic bacteria compositions between two reservoirs.
In this study, we found significant differences of bacterial community compositions between two estuarine reservoirs, mainly reflected in taxonomic composition, community diversity and interactions within molecular ecological networks (Figures
In reservoir A, temperature and rainfall (30 d) were the most important environmental factors in the summer, which further increased the relative abundance of bacterial OTUs such as
The molecular ecological network in reservoir B reflected weak interactions between bacterial OTUs and measured environmental variables, which indicated that the reservoir B located in tropical region, the seasonal differences in reservoir B were less noticeable than those in reservoir A. Therefore, the function roles and interactions between bacterial OTUs in the network of reservoir B became more prominent than in reservoir A. Within the co-occurrence pattern network in reservoir B, bacterial OTUs such as
Through the above comparison, we found that the key bacterial compositions (high betweenness centrality) were different between two reservoirs. In reservoir A, dominant heterotrophic bacterial OTUs such as
In aquatic ecosystem, Cyanobacteria could release a large number of secondary metabolites during the process of growth, such as DOC and other micro-molecular organics to surroundings. These organic matters could attract some other heterotrophic bacteria, which could decompose and use these metabolites (
In reservoir A, the dominant
Although the water environmental characteristics were much different between two estuarine reservoirs, the
Besides, with the exception of
This study highlights the comparison of microbial community, especially the characteristics and dynamics of cyanobacteria–heterotrophic bacteria in two estuarine reservoirs between tropical and sub-tropical regions. The bacterial community compositions were significantly different between two reservoirs, and mainly affected by different local environmental factors. The environmental heterogeneity in reservoir A was much higher, which indicated that the composition of bacterial community in reservoir A was more complex. The temperature, environmental conditions and nutritional status were suitable in reservoir B. Thus, the bacterial community in reservoir B have high diversity and less co-exclusion correlations. Although the
YH designed this research, and in charged of experiment in China. ZX performed experiments, analyzed data, prepared the figures, and wrote the manuscript text. ST helped sampling in Singapore and data analysis. KG in charged of experiment in Singapore and revised our manuscript earnestly. All authors have contributed and approved the final manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The Supplementary Material for this article can be found online at: