Chinese Journal of Oceanology and Limnology   2017, Vol. 35 issue(2): 336-349     PDF       
http://dx.doi.org/
Institute of Oceanology, Chinese Academy of Sciences
0

Article Information

ANG Yujing(王毓菁), LI Huabing(李化炳), XING Peng(邢鹏), WU Qinglong(吴庆龙)
Contrasting patterns of free-living bacterioplankton diversity in macrophyte-dominated versus phytoplankton blooming regimes in Dianchi Lake, a shallow lake in China
Chinese Journal of Oceanology and Limnology, 35(2): 336-349
http://dx.doi.org/

Article History

Received Oct. 13, 2015
accepted in principle Dec. 7, 2015
accepted for publication Jan. 12, 2016
Contrasting patterns of free-living bacterioplankton diversity in macrophyte-dominated versus phytoplankton blooming regimes in Dianchi Lake, a shallow lake in China
ANG Yujing(王毓菁)1,2, LI Huabing(李化炳)1, XING Peng(邢鹏)1, WU Qinglong(吴庆龙)1,3        
1 State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing 210008, China;
2 University of Chinese Academy of Sciences, Beijing 100039, China;
3 Sino-Danish Center for Science and Education, Beijing 100039, China
ABSTRACT: Freshwater shallow lakes typically exhibit two alternative stable states under certain nutrient loadings:macrophyte-dominated and phytoplankton-dominated water regimes. An ecosystem regime shift from macrophytes to phytoplankton blooming typically reduces the number of species of invertebrates and fishes and results in the homogenization of communities in freshwater lakes. We investigated how microbial biodiversity has responded to a shift of the ecosystem regime in Dianchi Lake, which was previously fully covered with submerged macrophytes but currently harbors both ecological states. We observed marked divergence in the diversity and community composition of bacterioplankton between the two regimes. Although species richness, estimated as the number of operational taxonomic units and phylogenetic diversity (PD), was higher in the phytoplankton dominated ecosystem after this shift, the dissimilarity of bacterioplankton community across space decreased. This decrease in beta diversity was accompanied by loss of planktonic bacteria unique to the macrophyte-dominated ecosystem. Mantel tests between bacterioplankton community distances and Euclidian distance of environmental parameters indicated that this reduced bacterial community differentiation primarily reflected the loss of environmental niches, particularly in the macrophyte regime. The loss of this small-scale heterogeneity in bacterial communities should be considered when assessing long-term biodiversity changes in response to ecosystem regime conversions in freshwater lakes.
Key words: bacterioplankton biodiversity     regime shift     macrophyte     phytoplankton    
1 INTRODUCTION

Freshwater ecosystems, particularly freshwater lakes, account for less than 0.01% of water worldwide and cover approximately 0.8% of the surface area of the Earth. However, this tiny fraction of global water supports approximately 40% of global fish species, 25% of global invertebrate diversity, and 6% of all described species on Earth (Dudgeon et al., 2006). Freshwater lake ecosystems are experiencing far greater declines in biodiversity compared with most affected terrestrial ecosystems. It has been suggested that deterioration of the water environment is one of the most important threats to freshwater biodiversity (Dudgeon et al., 2006). Eutrophication is a key emergent environmental issue for many freshwater lakes, reflecting the overloading of phosphorus and nitrogen in these waters. Under particular levels of nutrient loading, eutrophicated shallow freshwater lakes typically exhibit two alternative ecologically stable states (Scheffer et al., 1993). One stable state is characterized by high coverage of macrophytes, complex food webs, and high biodiversity of plants and animals, whereas large phytoplankton blooms, simple food web structures, and lower biodiversity of plants and animals are hallmarks of the other stable state (Wetzel and Søndergaard, 1998). An ecosystem regime shift from a macrophyte-dominated state to a phytoplankton blooming state profoundly affects plant and animal biodiversity; however, the response of bacterioplankton biodiversity is poorly understood. Given that bacterioplankton account for major biodiversity in freshwater lakes and play key roles in ecosystem functioning, it is urgent to determine the impact of ecosystem regime conversion to a phytoplankton blooming state on bacterioplankton biodiversity in freshwater lakes.

Previous studies have shown that the bacterioplankton community composition (BCC) differs in the water column in natural habitats (Van der Gucht et al., 2001; Wu et al., 2007) or manipulated mesocosms (Haukka et al., 2006) under two different ecological states or regimes. The biogeochemical cycling of nitrogen, phosphorus, and methane, mediated through bacteria and archaea, is also potentially altered through such ecosystem regime conversion (Wetzel and Søndergaard, 1998; Zhao et al., 2013; Zhu et al., 2013). However, there are no available studies evaluating the response of bacterioplankton biodiversity to ecosystem regime conversion in freshwater lakes. Previous studies have applied denaturing gradient gel electrophoresis (DGGE) for the assessment of BCC through limited cloning and sample sequencing (Van der Gucht et al., 2001; Wu et al., 2007), which only demonstrates the low sampling of bacterial diversity in the water columns. High-throughput next-generation sequencing technology should allow comprehensive sampling and achieve a better evaluation and comparison of the bacterial diversity between the two ecological states or regimes, though it is still PCRbased and thus has its own problems concerning diversity estimation.

Dianchi Lake is a large shallow lake located in western China, and this lake was fully covered with submerged macrophytes approximately 50 years ago (Wang and Dou, 1998). Due to overloading of phosphorus and nitrogen, Dianchi Lake has become eutrophicated and has diverged into two different ecological states: macrophyte-dominated in the Caohai region and exclusively phytoplankton blooming in the pelagic Waihai region. The macrophyte-dominated Caohai is covered with macrophytes, including Hydrilla verticillata and Potamogeton pectinatus, Eichhornia crassipes, and Nelumbo nucifera. The exclusively phytoplanktondominated Waihai is dominated by Microcystis spp. and Ceratium hirundinella blooms from April to November. Using this ideal lake ecosystem, we assessed the impact of ecosystem conversion from a macrophyte-dominated to an exclusive phytoplankton blooming regime on bacterioplankton biodiversity using high-throughput next-generation sequencing for deep sampling of bacterial diversity. In particular, we determined whether this regime conversion has resulted in changes in local bacterial diversity and bacterial community differentiations.

2 MATERIAL AND METHOD 2.1 Study site and sampling design

Dianchi Lake (24 ° 40′-25°02′N, 102° 36′-102°47′E) is the largest freshwater eutrophic lake on the Yunnan Plateau (Fig. 1a). The lake covers a total area of 306 km2, with a mean depth of 4.4 m, a maximum water depth of 6.7 m and a watershed area of 2 920 km2. The length is 41.2 km from south to north, and the width from east to west is 7.2 km on average. According to an investigation conducted in the early 1960s, the vegetation coverage of Dianchi Lake was 90%, and the lake contained more than 100 different types of aquatic plants, with submerged macrophytes covering the lake from the littoral zone to the center of the lake (Wang and Dou, 1998). However, with increasing human disturbance, macrophytes sharply declined, and the vegetation coverage was reduced to 12.6% in the 1990s, with the remaining macrophytes primarily being distributed in the Caohai region and lakeside areas (Wang and Dou, 1998). Currently, the northern region of the lake (Caohai) is dominated by macrophytes, and the southern region of the main water body (Waihai) is dominated by phytoplankton exclusively.

Figure 1 a. sampling locations in Dianchi Lake, China. Three sampling plots, M1, M2, and M3, are located in the macrophytedominated ecosystem, whereas plots P1, P2, and P3 are distributed in the phytoplankton bloom-dominated ecosystem; b. nested sampling scheme for each sampling plot, where 10 sampling points (as indicated by grey stars) were designed for investigation of the bacterioplankton community composition

Six sampling plots were selected across Dianchi Lake (Fig. 1b): 3 plots (M1, M2, and M3) in the macrophyte-dominated area (Caohai) and 3 plots (P1, P2, and P3) in the phytoplankton bloom-dominated area (Waihai) (Fig. 1a). Each plot was sampled using a nested sampling strategy. The samples were centered within a 55 m×55 m square, with a shared vertex which 150 m×150 m and 400 m×400 m squares were nested, forming a total of 10 sampling points per site. The distance was 1 km from sites M1 (P1) to M2 (P2) and 8 km from sites M1 (P1) to M3 (P3). Because it was difficult to perform sampling at precise sampling points in the pelagic area of the lake, we used a highprecision global positioning system (GPS) to record the actual positions of the sampling points. The recorded coordinates were further used to compute pairwise distances between sampling points. The spatial sampling campaign facilitated a comparison of bacterial alpha and beta diversities between the two different regimes.

2.2 Sampling and determination of environmental parameters

All samples were collected at the end of March 2013. Using a 5-L Schindler sampler, water samples of approximately 2.5 L were collected at a 0.5-m depth in the water column at each sampling point and stored in pre-sterilized polypropylene bottles. Water samples of approximately 200 mL were filtered through 5-μm-pore-size polycarbonate filters (Millipore, Billerica, MA, USA) to remove large plankton, epibiotic bacteria and other particles. The filtered water was passed through 0.22-μm-pore-size polycarbonate membrane filters (Millipore, Billerica, MA, USA) to collect the bacterioplankton biomass. The samples were stored at -80℃ until further processing.

The physical and chemical properties of the water samples, including the salinity, temperature, pH, oxidation reduction potential (ORP), dissolved oxygen (DO), turbidity and conductivity, were measured in situ using a calibrated multifunction water quality sonde (YSI 6600, Yellow Springs, OH, USA) at a depth of 0.5 m. Unfiltered water samples were used to measure the concentrations of total nitrogen (TN) and total phosphorous (TP) (Rice et al., 2012). The concentrations of ammonium (NH+ 4), nitrate (NO3-), nitrite (NO2-), and orthophosphate (PO43-) were measured through continuous colorimetric flow analysis (Skalar SAN PLUS system; Skalar Analytical BV, Breda, the Netherlands) after the water samples were filtered using glass microfiber filters (GF/C, Whatman, Maidstone, UK). The chlorophyll a concentration was spectrophoto-metrically measured using hot ethanol as an extraction solvent (Jespersen and Christoffersen, 1987).

2.3 DNA extraction, PCR, and sequencing

The extraction and purification of bacterioplankton DNA from a 0.22-μm-pore-size polycarbonate membrane filter was performed as previously described (Wu et al., 2006). The DNA concentration and purity were spectrophotometrically determined with a NanoDrop2000 (ThermoScientific, Wilmington, DE, USA). Bacterial 16S rRNA genes were amplified through PCR and assessed using the Illumina MiSeq platform as previously described (Yuan et al., 2014). Briefly, PCR amplification was implemented in the Gene Amp PCR-System® 9700 (Applied Biosystems, Foster City, CA, USA) in a total volume of 25 μL, containing a 0.4 μmol/L concentration of each of the universal 515F (5′-GTGCCAGCMGCCGCGG-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′) primers, 10 ng template of DNA, 2.5 μL of 10X AccuPrimeTM PCR buffer Ⅱ (for genomic DNA) and 0.5 units of AccuPrimeTM Taq DNA Polymerase High Fidelity (Invitrogen, Carlsbad, CA, USA). The applied primer pairs targeting the V4 hypervariable regions of bacterial/Archaeal 16S rRNA genes are found well suited to the phylogenetic analysis of pyrosequencing reads (Caporaso et al., 2012). The amplification program was as follows: an initial denaturation step at 94℃ for 1 min, followed by 30 cycles at 94℃ (20 s, denaturation), 53℃ (25 s, annealing), and 68℃ (45 s, elongation), with a final extension at 68℃ for 10 min. All samples were amplified three times. The PCR products were then pooled together, analyzed by agarose gel electrophoresis, purified with the QIAquick Gel Extraction Kit (Qiagen, Germantown, MD, USA) and further quantified using PicoGreen (BMG Labtech, Jena, Germany). A sample DNA library was constructed using the purified PCRamplified products (amplicons), mixed with PhiX (Illumina, San Diego, CA, USA), and subsequently run on the Illumina MiSeq platform.

2.4 Sequence analysis

>The raw MiSeq reads for the 16S rRNA genes were managed to control the quality in the Galaxy Pipeline (http://zhoulab5.rccc.ou.edu:8080), and the control steps are listed in detail on the website. The treated 16S rRNA gene sequences were considered to be the same operational taxonomic units (OTUs) at a 97% similarity level (Roesch et al., 2007). A phylogenetic tree (Price et al., 2010) was constructed based on the aligned sequences using the sequence analysis tools of the Galaxy Pipeline (http://zhoulab5.rccc.ou.edu:8080). Taxonomic information for the top 10 OTUs with the greatest niche breadth (Logares et al., 2013) and for OTUs unique to the macrophytecovered ecosystem and OTUs unique to the phytoplankton blooming ecosystem was obtained from the Ribosomal Database Project (RDP) classifier (Wang et al., 2007b). The sequences of these 30 OTUs were analyzed using ARB software (Ludwig et al., 2004) for comparison with typical freshwater bacterial clusters deposited in the ARB database (Newton et al., 2011).

2.5 Data analysis

Because of unequal numbers of sequences among sampling points, the OTUs from each sample with the poorest sequencing effort were randomly extracted (13 368 sequences). Phylogenetic analyses were based on a tree built with representative OTUs using the program FastTree2 (Price et al., 2010).

Taxonomic alpha diversity was calculated as total OTU numbers (Edgar, 2013); phylogenetic diversity was calculated using Faith’s phylogenetic diversity (PD) (Faith, 1992); and all diversity indices were compared between the macrophyte and phytoplankton sites using Student’s t-test. After all BCC data “Hellinger” transformed, a distance matrix was calculated based on the Bray-Curtis dissimilarity index matrices, and the distance matrix was used to construct a nonmetric multidimensional scaling (NMDS) plot using the program R (http://www.rproject.org) with the Vegan package (Oksanen et al., 2013). The taxonomic community composition was compared between the two different types of sites using permutational multivariate analysis of variance (Adonis) (Zapala and Schork, 2006). The environmental parameters were assessed via ANOVA to determine the difference between macrophytedominated and phytoplankton-dominated areas. The pairwise geographic distances between sampling points were calculated based on geographic coordinates and physical measurements in situ. All environmental parameters were standardized transformation (scaling the values to obtain a zero mean and unit variance), and environmental differences were calculated as Euclidean distances using the program R (http://www.r-project.org) with “vegdist” function in the Vegan package (Oksanen et al., 2013). BCC dissimilarity with geographic distance decay (community turnover) and environmental Euclidean distance with geographic distance decay were determined using the linear regression analysis method. Analyses of covariance (ANCOVA) were performed with SPSS version 16.0 to test for significant differences among the liner regression (between taxonomic dissimilarity and geographical distance, between environmental differences and geographical distance) slopes with two different ecosystem regimes.

The relationships between the BCC and the environmental parameters were evaluated using redundancy analysis (RDA) with the vegan package (Oksanen et al., 2013) in the program R. The RDA model was subjected to significance testing; the environmental parameters were subjected to standardized transformation, and the collinear variables, such as salinity, temperature, dissolved oxygen, oxidation-reduction potential, orthophosphate, nitrate nitrogen, chlorophyll a, were removed to ensure a variance inflation factor (VIF) of less than 20 to resolve the multicollinearity issue.

A permutation-based Mantel (partial Mantel, simple Mantel) test (Legendre and Legendre, 2012) was conducted to examine the significance of the correlations between bacterial communities and geographic distances or environmental parameters. All BCC data were “Hellinger” transformed, and the environmental parameters were subjected to “standardize” transformed.

We also applied null model test to analyze mechanisms determining BCCs between the two different ecosystem regimes (Chase et al., 2011). The values afford the quantification of the metric to which pairwise (site-to-site) community dissimilarity differs from that which would be simulated by random using the null model with the Raup-Crick metric (Chase et al., 2011), when the value will approach 0, it means local communities exhibit high dispersal and are controlled through stochastic rather than deterministic processes; whereas a value of 1 will be obtained when the observed communities in different habitats are more dissimilar than random expectations; and a value of -1 will be obtained when the observed communities are less dissimilar than expected at random due to niche-related processes (Chase et al., 2011).

2.6 Sequence data submission

The sequence data were submitted to the Sequence Read Archive (SRA) database (http://www.ncbi.nlm.nih.gov/sra) of the National Center for Biotechnology Information (NCBI) under accession No. SRP056718.

3 RESULT 3.1 Description of environmental conditions between the two ecosystem regimes

The lake is divided into two areas by an artificial dyke. Although the dyke is opened occasionally, the macrophyte-dominated and phytoplankton-dominated areas differed significantly in many environmental parameters (Table 1), including pH, electrical conductivity, turbidity, and the concentrations of TP, ammonium nitrogen (NH4+), and nitrite (NO2-).

Table 1 Results of ANOVA for the comparison of environmental parameters between macrophyte-dominated and phytoplankton-dominated areas
3.2 Description of the overall sequences

The sequence reads spanned a portion of the Escherichia coli 16S rRNA gene from positions 515 to approximately 806, with an average read length of 250 bp. Across all 30 macrophyte samples, we obtained 1 006 920 quality sequences. There were 143 OTUs comprising more than 1 000 sequences (maximum of 87 635 sequences). Additionally, the numbers of OTUs containing≥100, ≥10, ≥2 and≥1 sequences and singletons were 881, 4 452, 13 855 and 27 388, respectively, and the Good’s coverage (Good, 1953) value at a 97% similarity level was 93.74% on average (standard deviation (SD)=1.43%). Across all 30 phytoplankton samples, we obtained 1 133 029 quality sequences. There were 167 OTUs comprising more than 1 000 sequences (the maximum was 66 286 sequences). Additionally, the numbers of OTUs containing≥100, ≥10, ≥2 and≥1 sequences and singletons were 1 153, 6 977, 23 902 and 48 363, respectively, and the Good’s coverage value at a 97% similarity level was 89.97% on average (standard deviation (SD)=1.03%).

3.3 Comparison of BCC between the two ecosystem regimes

The BCC associated with a phytoplankton blooming state was significantly different from that in the macrophyte-dominated ecosystem at the phylum level (Fig. 2a) in terms of the taxonomic composition (Fig. 2b: Adonis results, F=40.581, P < 0.001). The relative abundance of Betaproteobacteria sharply decreased in response to conversion to the phytoplankton blooming regime, from an average of 33.72% (±2.21, 95% confidence interval (CI)) to 18.18% (±2.72, 95% CI). Similar trends were observed for Actinobacteria [from 30.26% (±3.67, 95% CI) to 23.54% (±2.11, 95% CI)] and Bacteroidetes [from 20.22% (±3.94, 95% CI) to 18.73% (±1.21, 95% CI)]. Slight increases in the relative abundances of Planctomycetes, Gammaproteobacteria, Cyanobacteria, Verrucomicrobia, and Firmicutes were observed in response to the shift to a phytoplankton blooming regime (Fig. 2a).

Figure 2 Bacterial community composition in macrophyte-dominated and phytoplankton bloom-dominated ecosystems samples (a) Distribution of 16S rRNA sequences across bacterial phyla in macrophyte-dominated Caohai and phytoplankton bloom-dominated Waihai (b) Non-parametric multidimensional scaling plot of taxonomic dissimilarity (Bray-Curtis) Blue circles: samples from phytoplankton bloom-dominated Waihai; green circles: samples from macrophyte-dominated Caohai.

The percentage of shared OTUs which appeared at least once in the two different ecosystem regimes was 30.71%, the percentages of unique OTUs which found only in the macrophyte-dominated ecosystems was 19.99%, and only in the phytoplankton blooming ecosystems was 49.29% (Fig. 3). We also compared the distribution of bacterioplankton OTUs across all 60 samples, including those detected only in either the macrophyte-dominated or phytoplankton blooming ecosystem and those found in both ecosystems. The OTUs observed in both ecosystems were found in significantly fewer samples in the macrophyte-dominated ecosystem than in the phytoplankton blooming ecosystem (t=-5.1, P < 0.001, df=19 371, Fig. 4), and the OTUs that were unique to the macrophyte-dominated ecosystem occurred in significantly fewer samples than other OTUs (t=-6.9, P < 0.001, df=17 174, Fig. 4).

Figure 3 Classification of bacterial OTUs that are unique to either phytoplankton-dominated or macrophyte dominated ecosystems, or shared by both ecosystems
Figure 4 Distribution of unique and shared bacterial OTUs across sampling points located in macrophytedominated and phytoplankton bloom-dominated ecosystems.

The overwhelming majority of the top 10 OTUs with high relative abundance unique to the macrophyte-covered ecosystem belonged to Actinobacteria; the top 10 OTUs unique to the phytoplankton blooming ecosystem were members of Cyanobacteria, Bacteroidetes, Gammaproteobacteria, Alphaproteobacteria and Actinobacteria; and the top 10 OTUs with the greatest niche breadth belonged to Actinobacteria, Betaproteobacteria, Bacteroidetes and unclassified Proteobacteria. The sequences corresponding to shared OTUs were much more abundant than those of the unique OTUs (ANOVA results, F=14.391, P=0.001), and the sequence number of OTUs that were unique to macrophytedominated habitats was lower compared with phytoplankton-dominated habitats (ANOVA results, F=7.808, P=0.012) (Table 2).

Table 2 Taxonomic identification and abundance of the top 10 OTUs with the greatest niche breadth and the top 10 OTUs with the greatest macrophyte and phytoplankton distribution
3.4 Diversity patterns of bacterioplankton communities between the two regimes

The alpha diversity of the bacterioplankton communities from the phytoplankton blooming ecosystem was significantly higher compared with the macrophyte-dominated ecosystem (Fig. 5). This result was true for both taxonomic richness (t=-16.638, df=58, P < 0.001) and phylogenetic diversity (t=-16.795, df=58, P < 0.001). By contrast, the bacterioplankton community in the macrophytedominated ecosystem exhibited a significantly higher turnover (beta diversity) than was observed in the phytoplankton blooming ecosystem (ANCOVA: F=72.5, P < 0.001, Fig. 6a).

Figure 5 Bacterial alpha diversity of the samples collected from macrophyte-dominated and phytoplankton bloom-dominated ecosystems (a) total numbers of operational taxonomic units; means (n=30) are depicted with the 95% CI (b) phylogenetic diversity Means (n=30) are depicted with the 95% CI.
Figure 6 Decay of taxonomic dissimilarity (a) and Euclidean distance of environmental parameters (b) with geographic distance in the macrophyte-dominated ecosystem (black dot, solid black regression line) and phytoplankton bloom-dominated ecosystem (gray dot, dotted gray regression line)
3.5 Linking environmental and spatial factors to the bacterioplankton community structure

In the RDA model, the BCC of the macrophytedominated ecosystem was more distantly distributed than that of the phytoplankton blooming ecosystem, showing separation based on significantly different bacterial composition and environmental factors (Fig. 7). The Mantel test suggested that both geographic distance and environmental factors significantly influenced the BCC in both ecosystem regimes (Table 3). When controlling for environmental effects, geographic distance was significantly correlated with community variations for both ecosystem regimes (Table 3). When controlling for distance effects, convincingly significant environmental effects were only observed in the macrophyte-dominated ecosystem, associated with BCC variations (Table 3, Fig. 6b). The null model test also indicated that the variation of the bacterial community observed in phytoplankton-dominated Waihai was more influenced through stochastic processes than in macrophytedominated Caohai, although BCC in both lake areas was primarily structured through environmental sorting because all values are largely deviant from the value of zero (Table 4).

Figure 7 Redundancy analysis (RDA) of the bacterial community composition in relation to the examined environmental factors and ANOVA test results *: P<0.05, **: P<0.01, ***: P<0.001, #: not significant. Three sampling plots, M1, M2, and M3, are located in the macrophyte-dominated ecosystem, whereas plots P1, P2, and P3 are distributed in the phytoplankton-dominated ecosystem (Fig. 1).
Table 3 Partial Mantel and simple Mantel analysis of community distances with the ln geographic distance and Euclidian distance of environmental parameters
Table 4 Comparison of stochastic processes between the macrophyte-dominated and phytoplanktondominated ecosystems using a null model based on the modified Raup-Crick metric
4 DISCUSSION 4.1 Low overlap of BCC between the two ecosystem regimes

Not surprisingly, the results obtained in the present study clearly revealed a significant difference in the BCC between the two different ecosystem regimes (Figs. 2 and 7), suggesting that ecosystem regime conversion from a macrophyte-dominated to a phytoplankton-dominated regime results in marked divergence of the BCC. These results are consistent with previous findings in natural habitats (Van der Gucht et al., 2001; Wu et al., 2007) or manipulated mesocosms (Haukka et al., 2006) under different ecological regimes. However, previous studies applied DGGE as a tool for comparisons of BCC and were therefore unable to detect the large differences between the BCC associated with two different ecological states. Although many bacterial communities are freshwater specific and exhibit a universal distribution (Newton et al., 2011), in the present study, using a high-throughput sequencing technique, we observed low overlap of bacterial taxa between the two examined ecological states. Only approximately 30% of the observed OTUs, among which Limnohabitans was the most abundant taxon (Šimek et al., 2010), were shared by the two ecosystem regimes, and most of the detected OTUs were unique to either the macrophyte-dominated or phytoplanktonblooming ecosystem during the sampling period. As a result, we investigated 30 sampling points in 3 plots in each ecosystem, and the observed low overlap of the BCC between the two ecosystems was surprising. We found that abundant bacterial OTUs were present in both ecosystems in most cases, whereas only extremely rare bacterial OTUs (unclassified or ungrouped into any known freshwater bacterial cluster) were specific to the different ecosystems (Table 2). Because the two examined ecosystems are located close to each other, bacteria can easily be dispersed. Thus, differences in environmental factors between the two ecological sates are likely responsible for the marked divergence in the BCC, particularly favoring the growth of rare bacteria that are unique or specific to different ecosystems. Not only the distribution of macrophytes had obvious difference, but also phytoplankton composition were distinctly different between the two ecological regimes (Table 5), the host specificity between bacteria and phytoplankton and macrophyte species has previously been demonstrated (Zeng et al., 2012; Gallego et al., 2014; Liu et al., 2014). The undetected high diversity of bacterioplankton might be sustained through as yet undiscovered environmental niches, although the ecological function of these rare bacteria should be assessed in future studies.

Table 5 Phytoplankton composition between the two different ecosystem regimes
4.2 Contrasting patterns of bacterioplankton diversity between the two regimes

An ecosystem regime shift from a macrophytedominated to a phytoplankton blooming regime typically reduces the number of species of invertebrates and fishes and results in the homogenization of communities in freshwater lakes (Jeppesen et al., 1997). Previous studies have indicated that the richness of benthic invertebrates is much higher in macrophyte-dominated Caohai than in the phytoplankton blooming Waihai region of Dianchi Lake (Wang et al., 2007a). However, we observed an increased alpha diversity of bacterioplankton in the phytoplankton blooming state, suggesting that this traditional tenet associated with freshwater ecosystems might not be applicable for microorganisms. There are several potential explanations for these results. First, the higher primary productivity and turnover rates observed in phytoplankton blooming states might favor high bacterial diversity in the water column, as a relationship between primary production and diversity has also been observed for several phyla of freshwater bacteria (Horner-Devine et al., 2003), whereas the primary production of macrophytes might support a high diversity of epiphytic bacterial communities (He et al., 2012, 2014). Second, there could be greater resuspension of sediment bacteria (either attached to sediment or soil particles or released into the water column), which possess much higher diversity (Lindström and Bergström, 2004; Crump et al., 2012), into the water column in pelagic Waihai than in macrophyte-dominated Caohai because macrophytes strongly decrease sediment resuspension (Jeppesen et al., 1997). This seems to be supported by the fact that higher relative abundances of Gammaproteobacteria, Planctomycetes, Verrucomicrobia, and Acidobacteria were found in pelagic Waihai than in macrophytedominated Caohai. The colony-forming cyanobacteria Microcystis spp. might support a small amount of attached bacterial abundance (Wetzel and Søndergaard, 1998; Brunberg, 1999) and diversity (Eiler and Bertilsson, 2004), which is however, not the case in the macrophyte-dominated system.

In contrast to the increase of bacterial alpha diversity observed in the blooming state, we detected a decrease in the beta diversity of bacterioplankton in the exclusive phytoplankton blooming state compared with the macrophyte-dominated state. Because the high coverage of macrophytes can readily decrease the flow speed and result in low hydraulic conductivity and mass effects, the turnover rate of BCC in this area was higher than that in the phytoplankton-dominated area. The partial Mantel tests also indicated that in both ecosystems, the variations in BCC largely reflected geographic distances, but when the distance effect was controlled, the environmental factors only significantly explained the variation of the BCC in a macrophyte-covered state (Table 3, Fig. 7). This observation is also an indication that environmental niches created by macrophyte coverage might be responsible for the changes in BCC at different sampling points in the macrophyte-covered states. Accordingly, local factors, including the coverage of aquatic plants, determine the BCC over a large geographic distance (Van der Gucht et al., 2007). The simulation results of a null model (Table 4) showed that niche-based processes govern the BCC under both ecosystem regimes, whereas the relative importance of stochastic processes is slightly higher in the phytoplankton-dominated state than in the macrophyte-dominated state. These results indicate that the occurrence of various probabilistic processes was more random in phytoplankton-dominated areas, which might help maintain a higher alpha diversity of bacterioplankton than in a macrophyte-dominated state (Chase et al., 2011).

Previous studies have shown that algae and benthic invertebrate communities respond to anthropogenic disturbance with decreases in beta diversity (Balata et al., 2007; Passy and Blanchet, 2007; Donohue et al., 2009; Maloney et al., 2011) in aquatic ecosystems. Similar patterns of bacterial alpha and beta diversity after ecosystem regime changes to those observed in our investigation have also been observed in the Amazon where a forest system was converted to pasture (Rodrigues et al., 2013). The decrease in beta diversity in response to an ecosystem regime shift is common for animal communities in freshwater habitats (Jeppesen et al., 1997), but has never previously been observed for microorganisms. Olden and Poff (2004) proposed that the decreased community differentiation occurs through loss of species within certain spatial scales, widespread introductions of cosmopolitan species, and increased distribution of local species (Olden and Poff, 2004). In our investigation we did find an increase in the distributions of existing OTUs (Fig. 4) in the exclusively phytoplankton regime and an increased loss of OTUs unique to macrophyte regime (Fig. 4). We need to care about the increased loss of bacterial OTUs unique to the macrophyte regime as they might have specific niche requirements (Zeng et al., 2012). As bacterial diversity also varies with time in freshwater lakes (Jones et al., 2012), and cyanobacterial blooming can persist for nearly 8 months in this lake, we also need to know whether this conversion would also decrease bacterial beta diversity at temporal scales in the future. Currently we do not know if the changes in bacterial diversity and functional traits are reversible. In addition, because the global loss of macrophyte habitats reflects eutrophication and cyanobacterial blooms as well as global warming, in the future we must understand whether the decrease in bacterial beta diversity on a large geographic scale will inevitably lead to a net loss of overall bacterial diversity in freshwater lake.

5 CONCLUSION

Shallow lakes are abundant and are ecologically and economically significant in the global landscape, and freshwater biodiversity is particularly vulnerable to anthropogenic changes such as eutrophication. Under eutrophic conditions, regime shift from a macrophyte-dominated state to an exclusive phytoplankton blooming state in Lake Dianchi profoundly affects bacterioplankton biodiversity and composition. Although species richness was higher in the exclusive phytoplankton-dominated ecosystem, the beta diversity of bacterioplankton community decreased. This decrease in beta diversity reflects the loss of planktonic bacteria unique to the macrophytedominated as well as the loss of environmental niches. Our study illustrates the importance of macrophytes in maintaining bacterial diversity as have been commonly found for diversity of macroorganisms in shallow freshwater lakes.

6 ACKNOWLEDGEMENT

The authors thank Min Pan from the Institute of Dianchi Lake Ecology for assistance with field sampling and the collection of original data for Lake Dianchi. The authors also thank Dr. Yu Shi and Dan He for assistance with the statistical analysis.

References
Balata D, Piazzi L, Benedetti-Cecchi L, 2007. Sediment disturbance and loss of beta diversity on subtidal rocky reefs. Ecology, 88(10): 2 455–2 461. Doi: 10.1890/07-0053.1
Brunberg A K, 1999. Contribution of bacteria in the mucilage of Microcystis spp.(Cyanobacteria) to benthic and pelagic bacterial production in a hypereutrophic lake. FEMS Microbiol. Ecol., 29(1): 13–22. Doi: 10.1111/fem.1999.29.issue-1
Caporaso J G, Lauber C L, Walters W A, et al, 2012. Ultra-highthroughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J., 6(8): 1 621–1 624. Doi: 10.1038/ismej.2012.8
Chase J M, Kraft N J B, Smith K G, Vellend M, Inouye B D, 2011. Using null models to disentangle variation in community dissimilarity from variation in α-diversity. Ecosphere, 2(2): 1–11.
Crump B C, Amaral-Zettler L A, Kling G W, 2012. Microbial diversity in arctic freshwaters is structured by inoculation of microbes from soils. ISME J., 6(9): 1 629–1 639. Doi: 10.1038/ismej.2012.9
Donohue I, Jackson A L, Pusch M T, Irvine K, 2009. Nutrient enrichment homogenizes lake benthic assemblages at local and regional scales. Ecology, 90(12): 3 470–3 477. Doi: 10.1890/09-0415.1
Dudgeon D, Arthington A H, Gessner M O, et al, 2006. Freshwater biodiversity:importance, threats, status and conservation challenges. Biol. Rev., 81(2): 163–182. Doi: 10.1017/S1464793105006950
Edgar R C, 2013. UPARSE:highly accurate OTU sequences from microbial amplicon reads. Nat. Methods, 10(10): 996–998. Doi: 10.1038/nmeth.2604
Eiler A, Bertilsson S, 2004. Composition of freshwater bacterial communities associated with cyanobacterial blooms in four Swedish lakes. Environ. Microbiol., 6(12): 1 228–1 243. Doi: 10.1111/emi.2004.6.issue-12
Faith D P, 1992. Conservation evaluation and phylogenetic diversity. Biol. Conserv., 61(1): 1–10. Doi: 10.1016/0006-3207(92)91201-3
Gallego I, Davidson T A, Jeppesen E, Pérez-Martínez C, Fuentes-Rodríguez F, Juan M, Casas J J, 2014. Disturbance from pond management obscures local and regional drivers of assemblages of primary producers. Freshwater Biol., 59(7): 1 406–1 422. Doi: 10.1111/fwb.2014.59.issue-7
Good I J, 1953. The population frequencies of species and the estimation of population parameters. Biometrika, 40(3-4): 237–264. Doi: 10.1093/biomet/40.3-4.237
Haukka K, Kolmonen E, Hyder R, Hietala J, Vakkilainen K, Kairesalo T, Haario H, Sivonen K, 2006. Effect of nutrient loading on bacterioplankton community composition in lake mesocosms. Microb. Ecol., 51(2): 137–146. Doi: 10.1007/s00248-005-0049-7
He D, Ren L J, Wu Q L, 2012. Epiphytic bacterial communities on two common submerged macrophytes in Taihu Lake:diversity and host-specificity. Chin. J. Oceanol. Limnol., 30(2): 237–247. Doi: 10.1007/s00343-012-1084-0
He D, Ren L J, Wu Q L, 2014. Contrasting diversity of epibiotic bacteria and surrounding bacterioplankton of a common submerged macrophyte, Potamogeton crispus, in freshwater lakes. FEMS Microbiol. Ecol., 90(3): 551–562. Doi: 10.1111/fem.2014.90.issue-3
Horner-Devine M C, Leibold M A, Smith V H, Bohannan B J M, 2003. Bacterial diversity patterns along a gradient of primary productivity. Ecol. Lett., 6(7): 613–622. Doi: 10.1046/j.1461-0248.2003.00472.x
Jeppesen E, Jensen J P, Søndergaard M, Lauridsen T, Pedersen L J, Jensen L, 1997. Top-down control in freshwater lakes:the role of nutrient state, submerged macrophytes and water depth. Hydrobiologia, 342-343: 151–164. Doi: 10.1023/A:1017046130329
Jespersen A M, Christoffersen K, 1987. Measurements of chlorophyll a from phytoplankton using ethanol as extraction solvent. Archiv für Hydrobiologie, 109(3): 445–454.
Jones S E, Cadkin T A, Newton R J, McMahon K D, 2012. Spatial and temporal scales of aquatic bacterial beta diversity. Front Microbiol., 3: 318.
Legendre P, Legendre L F. 2012. Numerical Ecology. 3rd ed.Elsevier Science, Amsterdam.
Lindström E S, Bergström A K, 2004. Influence of inlet bacteria on bacterioplankton assemblage composition in lakes of different hydraulic retention time. Limnol.Oceanogr., 49(1): 125–136. Doi: 10.4319/lo.2004.49.1.0125
Liu L M, Yang J, Lv H, Yu Z, 2014. Synchronous dynamics and correlations between bacteria and phytoplankton in a subtropical drinking water reservoir. FEMS Microbiol.Ecol., 90(1): 126–138. Doi: 10.1111/fem.2014.90.issue-1
Logares R, Lindström E S, Langenheder S, Logue J B, Paterson H, Laybourn-Parry J, Rengefors K, Tranvik L, Bertilsson S, 2013. Biogeography of bacterial communities exposed to progressive long-term environmental change. ISME J., 7(5): 937–948. Doi: 10.1038/ismej.2012.168
Ludwig W, Strunk O, Westram R, et al, 2004. ARB:a software environment for sequence data. Nucl. Acids Res., 32(4): 1 363–1 371. Doi: 10.1093/nar/gkh293
Maloney K O, Munguia P, Mitchell R M, 2011. Anthropogenic disturbance and landscape patterns affect diversity patterns of aquatic benthic macroinvertebrates. Journal of the North American Benthological Society, 30(1): 284–295. Doi: 10.1899/09-112.1
Newton R J, Jones S E, Eiler A, McMahon K D, Bertilsson S, 2011. A guide to the natural history of freshwater lake bacteria. Microbiol. Mol. Biol. Rev., 75(1): 14–49. Doi: 10.1128/MMBR.00028-10
Oksanen J, Blanchet F G, Kindt R, Legendre P, Minchin P R, O'Hara R, Simpson G L, Solymos P, Stevens M H H, Wagner H. 2013. vegan:community ecology package. R package version 2, http://CRAN.R-project.org/package=vegan. Accessed on 2015-01-01.
Olden J D, Poff N L, 2004. Ecological processes driving biotic homogenization:testing a mechanistic model using fish faunas. Ecology, 85(7): 1 867–1 875. Doi: 10.1890/03-3131
Passy S I, Blanchet F G, 2007. Algal communities in humanimpacted stream ecosystems suffer beta-diversity decline. Diversity and Distributions, 13(6): 670–679. Doi: 10.1111/j.1472-4642.2007.00361.x
Price M N, Dehal P S, Arkin A P, 2010. FastTree 2-approximately maximum-likelihood trees for large alignments. PLoS One, 5(3): e9490. Doi: 10.1371/journal.pone.0009490
Rice E W, Baird R B, Eaton A D, Clesceri L S. 2012. Standard Methods for the Examination of Water and Wastewater.American Public Health Association, Washington DC.
Rodrigues J L M, Pellizari V H, Mueller R, et al, 2013. Conversion of the Amazon rainforest to agriculture results in biotic homogenization of soil bacterial communities. Proc. Natl. Acad. Sci. U. S. A., 110(3): 988–993. Doi: 10.1073/pnas.1220608110
Roesch L F, Fulthorpe R R, Riva A, Casella G, Hadwin A K, Kent A D, Daroub S H, Camargo F A, Farmerie W G, Triplett E W, 2007. Pyrosequencing enumerates and contrasts soil microbial diversity. ISME J., 1(4): 283–290.
Scheffer M, Hosper S H, Meijer M L, Moss B, Jeppesen E, 1993. Alternative equilibria in shallow lakes. Trends Ecol.Evol., 8(8): 275–279. Doi: 10.1016/0169-5347(93)90254-M
Šimek K, Kasalický V, Horňák K, Hahn M W, Weinbauer M G, 2010. Assessing niche separation among coexisting Limnohabitans strains through interactions with a competitor, viruses, and a bacterivore. Appl. Environ.Microbiol., 76(5): 1 406–1 416. Doi: 10.1128/AEM.02517-09
Van der Gucht K, Cottenie K, Muylaert K, et al, 2007. The power of species sorting:local factors drive bacterial community composition over a wide range of spatial scales. Proc. Natl. Acad. Sci. U. S. A., 104(51): 20 404–20 409. Doi: 10.1073/pnas.0707200104
Van der Gucht K, Sabbe K, de Meester L, Vloemans N, Zwart G, Gillis M, Vyverman W, 2001. Contrasting bacterioplankton community composition and seasonal dynamics in two neighbouring hypertrophic freshwater lakes. Environ. Microbiol., 3(11): 680–690. Doi: 10.1046/j.1462-2920.2001.00242.x
Wang L Z, Liu Y D, Chen L, Xiao B D, Liu J T, Wu Q L, 2007a. Benthic macroinvertebrate communities in Dianchi Lake Yunnan and assessment of its water. Acta Hydrob. Sin., 31(4): 590–593.
Wang Q, Garrity G M, Tiedje J M, Cole J R, 2007b. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl. Environ.Microbiol., 73(16): 5 261–5 267. Doi: 10.1128/AEM.00062-07
Wang S M, Dou H S. 1998. Lakes in China. Science Press, Beijing, China. (in Chinese)
Wetzel R G, Søndergaard M. 1998. Role of submerged macrophytes for the microbial community and dynamics of dissolved organic carbon in aquatic ecosystems. In:Jeppesen E, Søndergaard M, Søndergaard M, Christoffersen K eds. The Structuring Role of Submerged Macrophytes in Lakes. Springer Press, New York. p.133-148.
Wu Q L, Zwart G, Schauer M, Kamst-van Agterveld M P, Hahn M W, 2006. Bacterioplankton community composition along a salinity gradient of sixteen highmountain lakes located on the Tibetan Plateau, China. Appl. Environ. Microbiol., 72(8): 5 478–5 485. Doi: 10.1128/AEM.00767-06
Wu Q L, Zwart G, Wu J, Kamst-van Agterveld M P, Liu S J, Hahn M W, 2007. Submersed macrophytes play a key role in structuring bacterioplankton community composition in the large, shallow, subtropical Taihu Lake, China. Environ. Microbiol., 9(11): 2 765–2 774. Doi: 10.1111/emi.2007.9.issue-11
Yuan Y, Chen C, Liang B, et al, 2014. Fine-tuning key parameters of an integrated reactor system for the simultaneous removal of COD, sulfate and ammonium and elemental sulfur reclamation. J. Hazard. Mater., 269: 56–67. Doi: 10.1016/j.jhazmat.2013.12.014
Zapala M A, Schork N J, 2006. Multivariate regression analysis of distance matrices for testing associations between gene expression patterns and related variables. Proc. Natl.Acad. Sci. U. S. A., 103(51): 19 430–19 435. Doi: 10.1073/pnas.0609333103
Zeng J, Bian Y Q, Xing P, Wu Q L, 2012. Macrophyte species drive the variation of bacterioplankton community composition in a shallow freshwater lake. Appl. Environ.Microbiol., 78(1): 177–184. Doi: 10.1128/AEM.05117-11
Zhao D Y, Liu P, Fang C, Sun Y M, Zeng J, Wang J Q, Ma T, Xiao Y H, Wu Q L, 2013. Submerged macrophytes modify bacterial community composition in sediments in a large, shallow, freshwater lake. Canadian Journal of Microbiology, 59(4): 237–244. Doi: 10.1139/cjm-2012-0554
Zhu G B, Wang S Y, Wang W D, et al, 2013. Hotspots of anaerobic ammonium oxidation at land-freshwater interfaces. Nat. Geosci., 6(2): 103–107. Doi: 10.1038/ngeo1683