Higher resistance of a microcystin (MC)-producing cyanobacterium, Microcystis, to the submerged macrophyte Myriophyllum spicatum

Outbreaks of Microcystis blooms can affect growth of submerged plants, which in turn can inhibit cyanobacterial growth. Microcystin (MC)-producing and non-MC-producing Microcystis strains typically coexist in Microcystis-dominated blooms. However, the interaction between submerged plants and Microcystis at strain level is not clear. This study was aimed at assessing the effects of a submerged macrophyte Myriophyllum spicatum on one MC-producing versus one non-MC-producing strains of the cyanobacterium Microcystis using plant-Microcystis co-culture experiments. The impacts of Microcystis on M. spicatum were also examined. It showed that the MC-producing Microcystis strain had a higher resistance to negative impacts by the cocultured submerged plant M. spicatum than the non-MC-producing strain. By contrast, the plant M. spicatum was impacted more by the MC-producing Microcystis than the non-MC-producer. The associated bacterioplankton community was affected more by the MC-producing Microcystis than the cocultured M. spicatum. The MC cell quotas were significantly higher in the coculture treatment (the PM + treatment, p < 0.05), indicating that the production and release of MCs might be a key factor responsible for the reduced impact of M. spicatum. The higher concentrations of dissolved organic and reducing inorganic compounds might eventually exacerbate the recovering capacity of coexisting submerged plants. Overall, this study indicated that the capacity to produce MCs, as well as the density of Microcystis, should be taken into account when attempting to reestablish submerged vegetation to undertake remediation works.


Introduction
Submerged macrophytes and phytoplankton are key primary producers in lakes. Increased nutrient loading and global climate change shift many lakes from a clear state dominated by submerged plants into a turbid state dominated by phytoplankton. Then, phytoplankton blooms, including cyanobacteria blooms, may occur frequently, leading to severe ecological and economic damage (Scheffer et al. 2001;Huisman et al. 2018). The reestablishment of the submerged vegetation community in lakes has been proposed to be an important strategy for controlling cyanobacterial blooms and managing eutrophicated shallow waterbodies broadly (Pan et al. 2011;Bai et al. 2020). Macrophytes have been shown to selectively inhibit cyanobacteria on multiple scales (Zhu et al. 2010;Švanys et al. 2014). The bloom-forming cyanobacterial genus Microcystis is more sensitive to the coexisting submerged plant M. spicatum than green algae (Körner and Nicklisch 2002;Jeong et al. 2021). However, dense cyanobacterial blooms have been shown to impair growth of submerged macrophytes by releasing various primary and secondary metabolites (Zhang et al. 2021a, b). The interaction between submerged macrophytes and cyanobacteria varied with biotic environmental factors including bacteria and herbivory and abiotic factors including nitrogen, phosphorus, and light (Mohamed 2017). Whether submerged macrophytes or cyanobacteria are the winners when grown together may also depend on species/strains variation and their metabolite properties, which have not been thoroughly elucidated. This situation hinders the effective implementation of submerged vegetation restoration practices.
The cyanobacterial genus Microcystis is an ubiquitous non-nitrogen-fixing bloom-former found throughout freshwater waterbodies in tropical, subtropical, and temperate zones (Xiao et al. 2018). It can produce monocyclic heptapeptide toxins, namely, microcystins (MCs), which have been consistently reported to inhibit the photosynthesis of aquatic plants over a range of exposure concentrations (Rastogi et al. 2014;Zhang et al. 2022). Strains vary in their MCproducing capacity, and non-MC-producing strains typically coexist in Microcystis-dominated blooms (Bormans et al. 2020;Sidelev et al. 2020;Liu et al. 2021). Different strains may dominate in different seasons or at different stages of bloom formation in the same or different waterbodies (Kardinaal et al. 2007;Alexova et al. 2016). In comparison with non-MC-producing strains, the selective advantage of MC-producing Microcystis generally accompanied with the increasement of temperature and nutrients and the reduction of CO 2 levels (Davis et al. 2009;Van De Waal et al. 2011;Lei et al. 2015;Alexova et al. 2016;Suominen et al. 2017). The composition of Microcystis-dominated blooms is therefore an important factor determining the population toxicity and impact on other aquatic organisms, including submerged macrophytes.
A previous study found that several MC-producing Microcystis strains are more sensitive to exposed isolated macrophyte allelochemicals than non-MC-producing strains (Gao et al. 2020). However, understanding the response of MC-producing and non-MC-producing strains to submerged macrophytes remains a puzzle. We hypothesized that MCproducing Microcystis may have a higher resistance to coexisting macrophytes rather than non-MC-producing Microcystis strains.
Bacterioplankton, a fundamental component of aquatic ecosystems, plays a key role in the transformation of nutrients and may be closely linked to primary producers Seymour et al. 2017). In the large, shallow subtropical Taihu Lake of China, the bacterioplankton community composition has been shown to vary between the macrophyte-and phytoplankton-dominated states (Wu et al. 2007). Submerged macrophytes or phytoplankton community succession have been reported separately to participate in structuring bacterioplankton community composition (Niu et al. 2011;Zeng et al. 2012). The substances released or extracted from Microcystis cells have been shown to alter the epiphytic bacterial composition on the surfaces of the submerged macrophytes (Jiang et al. 2019;Zhang et al. 2021a, b). Our recent work find planktonic and epiphytic bacteria community responded differently to the Microcystis cell extracts (Gao et al. 2022). However, the response of the bacterioplankton community to the interaction between submerged macrophytes and cyanobacteria is unclear. We hypothesized that the winner of the competition between submerged macrophytes and cyanobacteria plays the major role in affecting the bacterioplankton community structure and function.
The submerged macrophyte M. spicatum is widespread in freshwater and low-salinity coastal areas around the world (Liu et al. 2018). It has been reported to be a plant species with strong allelopathic effects against cyanobacterium Microcystis (Körner and Nicklisch 2002;Jeong et al. 2021). Hence, in the present study, we selected a MC-producing and a non-MC-producing Microcystis strain for one-on-one coexistence experiments with the submerged macrophyte M. spicatum to test the above two hypotheses. The responses of both Microcystis strains under coculture conditions were observed to determine whether their interaction outcome is related to strains with or without MC-production capacity. The responses of the bacterioplankton community and the changes of dissolved carbon, nitrogen, and phosphorus were also examined during the interaction between Microcystis and M. spicatum.

M. spicatum plants and Microcystis
Fresh M. spicatum plants were obtained from Honghu Lake (N29.827°, E113.476°) in Hubei Province, China, and cultivated in aquaria with 10 cm-deep sediments in a greenhouse until they grew new roots. Uniform plant shoots were selected and cleaned for subsequent preculture.
One MC-producing (FACHB 915) and one non-MC-producing (FACHB 1005) Microcystis strains were obtained from the Freshwater Algae Culture Collection of the Institute of Hydrobiology, the Chinese Academy of Sciences. The prepared plant shoots and Microcystis cells were precultured separately in 1/10 diluted BG11 medium (Gao et al. 2022) under climate-controlled conditions at 22 ± 3 °C and light exposure at 25 μmol photon (PAR) m −2 s −1 with 12 h light/12 h dark cycles in the laboratory before the experiments. The Microcystis cells in exponential growth phase were used for experiments.

Experimental design
Five experimental treatments were prepared, namely: the P, PM + , PM − , M + , and M − treatments. The monoculture control of the plant M. spicatum without the addition of Microcystis cells was the P treatment, while the monoculture controls of the MC-producing and non-MC-producing Microcystis strains without plants were set as the M + and M − treatments, respectively. For the coculture treatments with M. spicatum (the PM + and PM − treatments), plant apical tissues (12 cm long) were secured with sterile glass beads in beakers at the density of 2.0 ± 0.05 g fresh weight per liter. Then, the prepared MC-producing or non-MC-producing Microcystis cells were added to the corresponding beakers at the final concentration of 5.0 ± 0.2 × 10 9 cells L −1 . The 1/10 BG11 medium provided nutrients for plants and Microcystis cells in the beakers. Six replicates were conducted for the P, PM − , and PM + treatments, and three replicates were conducted for the M-and M + treatments. All the beakers were covered with breathable sealing membranes and cultured under the same conditions mentioned above.
Subsamples were collected on days 0, 3, 6, 9, 12, 15, and 18 to measure changes in growth and physiology of both Microcystis strains and extracellular and intracellular MC concentrations of the MC-producing strain. At the end of the experiment, the culture media in the upper-middle water column were collected for bacterioplankton community analysis. Chemical parameters were measured on day 0 and day 18. Dissolved carbon and ammonia nitrogen concentrations were measured every two days. The morphological changes in the plants were recorded by using photographs.

Growth and physiology measurement of Microcystis
A 5 mL sample of well-mixed Microcystis was collected to determine optical density at a wavelength of 680 nm by using a spectrophotometer (SPECORD210 Plus, Jena, Germany), and Fv/Fm photosynthetic yield values were measured with an AquaPen-C 100 fluorometer (Photon Systems Instruments, Drasov, Czech Republic) in accordance with the manufacturer's instructions. Then the samples were recollected and preserved with 1% Lugo fluid before cell enumeration using a biological microscope (E100, Nikon Eclipse).
Another fresh homogeneous subsample was collected for malondialdehyde (MDA) measurement by using assay kits from Nanjing Jiancheng Company, China. The increase in MDA concentrations reflects membrane lipid peroxidation due to the overproduction of reactive oxygen species and the damage to antioxidant systems in Microcystis cells (Lu et al. 2016).

Quantification of chemical parameters and MCs
The Microcystis culture solution was filtered through a 0.45 μm filter membrane, and the filtrate was collected for subsequent nutrient analyses. The dissolved total nitrogen (TDN), dissolved total phosphorus (TDP), phosphate (PO 4 -P), nitrate (NO 3 -N), nitrite (NO 2 -N), and ammonia (NH 4 -N) were measured by using colorimetric methods in accordance with the Chinese National Standards for Water Quality (The National Environmental Protection Agency of China 2002). Dissolved total carbon (TDC), dissolved organic carbon (DOC), and dissolved inorganic carbon (DIC) were determined by using a total organic carbon analyser (Multi N/C 3100, Jena, Germany) (The National Environmental Protection Agency of China 2002).
The Microcystis culture solution was centrifuged at 10 000 rpm for 15 min. The supernatant was collected and diluted 20 times for the determination of extracellular MC concentrations. The cell pellets were disrupted by freezing and thawing five times in liquid nitrogen and centrifuged at 10 000 rpm for 15 min. The supernatant was collected and diluted 20 times for the determination of intracellular MC concentrations. The concentration of MC in solutions was determined with a MC-LR ELISA kit (Institute of Hydrobiology, CAS). It has sensitivity of 0.1 μg L −1 .

Bacterioplankton community composition
A total of 200 mL of Microcystis culture solution was filtered through 0.22 μm acetate fiber filters and frozen at − 80 ℃ for further DNA extraction and analysis. The DNA extraction with the Fast DNA™ SPIN Kit and highthroughput sequencing were carried out by Majorbio BioPharm Technology Co. Ltd. (Shanghai, China). The V3-V4 hypervariable region of 16S rRNA was amplified by using a barcoded universal primer set, which included 338F (5′-ACT CCT ACG GGA GGC AGC AG-3′) and 806R (5′-GGA CTA CHVGGGTW TCT AAT -3′) (Gao et al. 2022). Highthroughput sequencing was conducted by using the Illumina MiSeq platform.

Data analysis
Microsoft Excel 2019 and IBM SPSS Statistics 26.0 were used for data analysis. Data is presented as means ± standard deviations. One-way ANOVA was performed to compare the differences in Microcystis performance and chemical parameters among experimental treatments at the same sampling point at the p < 0.05 level. The average growth rate of Microcystis cells in the monoculture treatments was calculated according to the following equation: Here, C t1 and C t0 are the cell concentrations of Microcystis at the end (t1) and the beginning (t0) of the experiment.
Bioinformatics analysis was conducted by using a Majorbio Cloud Platform (www. major bio. com) to analyse the bacterioplankton community composition and to reveal correlations between physiochemical parameters and bacterioplankton community structure, including the calculation of Shannon and Chao1 indexes. PICRUSt2 was used to predict the microbial function. Nonmetric multidimensional scaling (NMDS) was used to illustrate the differences among different treatments.

Responses of the Microcystis strains and M. spicatum
Under the P treatment, M. spicatum plant shoots grew well during the experiment in the absence of Microcystis, with the plant biomass increasing by 11% for 18 days. Under the monoculture treatments, the cell concentrations of MCproducing (the M + treatment) and non-MC-producing Microcystis strains (the M − treatment) increased during the experiment (Fig. S1). The average growth rates of the Average growth rate (%) = (ln(C t1 ) − ln(C t0 ))∕(t1 − t0) monocultured MC-producing and non-MC-producing strains from day 0 to day 18 were 0.011 and 0.036 day −1 , respectively. Under the PM + treatment, the leaves of M. spicatum fell off on day 9, and almost all plant shoots had fractured and decayed on day 15. Under the PM − treatment, some of the plant shoots continued to grow even on day 15 (Fig. S1).
The cell concentrations of Microcystis expressed as OD 680 values under the PM + and PM − treatments were only 62.7% and 42.7% of those under the M + and M − treatments on day 9 and continued to decrease to 16.6% and 5.0%, respectively, on day 18 (Fig. 1A). The photosynthetic activity, expressed as the Fv/Fm ratios, of the non-MC-producing Microcystis strain reduced more significantly under the PM − treatment than those of the MC-producing strain under the PM + treatment from day 9 (p < 0.01). The percentage of the control was 56.7% on day 9 and decreased to 48.3% on day 18 under the PM − treatment. The Fv/Fm values of the MC-producing Microcystis strain under the PM + treatment reduced to 74.6% and 73.0% of those of the control on days 9 and 18, respectively (Fig. 1B).
The MDA concentrations of both Microcystis strains cocultured with M. spicatum were significantly higher than those of the monoculture controls from day 12 (p < 0.05). The MDA concentrations under the PM − and PM + treatments were 162.2% and 156.5% of the corresponding controls, respectively, on day 12. At most sampling points, the

Changes in intracellular and extracellular microcystin concentrations
The cell quotas of intracellular and extracellular MC concentrations were significantly higher under the PM + treatment than under the M + treatment from day 12 (p < 0.05). Under the M + and PM + treatments, the average cell quotas of intracellular MC concentrations were 17.84 and 36.82 fg cell −1 , respectively, and those of extracellular MC concentrations were 14.13 and 40.54 fg cell −1 , respectively, at the end of the experiment (Fig. 2).

Responses of bacterioplankton
A total of 482 operational taxonomic units (OTUs) were classified at a 97% similarity threshold from 656 915 raw sequence reads of the 15 bacterioplankton samples. The diversity indices of the bacterioplankton community (Shannon index) ranged between 1.75 and 3.59 (Fig. 3). The Shannon value was the highest under the PM − treatment but was the lowest under the M − treatment, indicating that the coexistence of the plant M. spicatum markedly increased the bacterioplankton diversity under the treatments. The richness estimator of the bacterioplankton community (Chao1 index) ranged from 117.17 to 175.62. The highest value was observed under the P treatment, and the lowest value was found under the PM + treatment, indicating that the coexistence of the MC-producing Microcystis decreased the bacterioplankton richness.
Proteobacteria was the most dominant phylum in all samples with relative abundance ranging from 73% under the PM + treatment to 86% under the P treatment (Fig. S2). The highest relative abundance of Bacteroidetes was 12.1% under the PM − treatment, but the lowest was under the M − treatment. The relative abundance of Actinobacteriota under the M − treatment and the relative abundance of Firmicutes under the M + treatment was the highest. Gemmatimonadetes was found only under the M + and PM + treatments with the relative abundances of 5% and 14%, respectively.
The distribution of the 30 most dominant bacterial genera under the five experimental treatments is illustrated in Fig. 4. In the P treatment, Acinetobacter (Proteobacteria) was the most dominant genus, accounting for 63.2% of the bacterial abundance. In the M − treatment, Phreatobacter (Proteobacteria) accounted for 65.5% of the bacterial abundance. They were less dominant under the PM − treatment, where the relative abundances of Acinetobacter and Phreatobacter were 0.3% and 3.0%, respectively. More than 13 genera, including Chryseobacterium (Bacteroidetes) and Novosphingobium (Proteobacteria), had higher relative abundances, and at least six genera, including Phreatobacter and Rhodococcus (Actinobacteriota), had considerably lower relative abundances under the P and PM − treatments than under the M − treatment. More than 13 genera showed higher relative abundances, and 8 genera showed lower relative abundances under the M + and PM + treatments than under the P treatment.
The differences among the treatments based on NMDS analysis by using unweighted UniFrac distance are shown in Fig. 5. The first axis revealed that the bacterioplankton communities under the P and PM − treatments were different from those under the M − treatment. The second axis demonstrated the similarity between bacterioplankton compositions under the M + and PM + treatments, which were significantly different from those under the P treatment (p < 0.05).
The PICRUSt2 prediction revealed a significantly higher abundance of clusters of orthologous groups associated with energy production and conversion, lipid transport and metabolism, inorganic ion transport and metabolism, secondary metabolite biosynthesis, transport and catabolism, and other unknown functions under the M − treatment than under the PM − and P treatments (Fig. 6, P < 0.05). The abundances of clusters of orthologous groups associated with amino acid transport and metabolism; carbohydrate

Changes in the dissolved carbon, nitrogen, and phosphorus
The DOC concentrations ranged from 5.64 mg L −1 to 65.41 mg L −1 , and the DIC concentrations were between  (Table 1, Fig. S3). At the end of the experiment, the P treatment had the lowest DOC concentrations, whereas the PM + and M + treatments had the highest DOC concentrations. The DIC concentrations under the coculture treatments were higher than those under the corresponding monoculture treatments at most sampling times. The increment in DIC concentrations under the PM + and PM − treatments was larger than that under the M + , M − , and P treatments.
The DON and NO 3 -N concentrations ranged from 29.97 mg L −1 to 32.68 mg L −1 and from 12.25 mg L −1 to 21.89 mg L −1 during the experiment. The DON concentrations decreased by a range of 19.4%-90.7% under all treatments during the experiment (Table 1). At the end of the experiment, the DON concentration under the P treatment was the lowest, whereas those under the PM − and PM + treatments were the highest. The NO 3 -N concentrations increased by 78.5% under the P treatment, but decreased by 32.5%-68.5% under the other four treatments during the experiment.
The NO 2 -N and NH 4 -N concentrations under the P treatment kept a level less than 0.03 mg L −1 for NO 2 -N and 0.3 mg L −1 for NH 4 -N during the experiment (Table 1, Fig. S3). However, their concentrations increased obviously especially under the PM + and PM − treatments. At the end of the experiment, the NO 2 -N concentrations increased to 3.16 and 1.14 mg L −1 under the PM + and PM − treatments. The NH 4 -N concentrations under the PM + and PM − treatments increased to 2.45 mg L −1 and 1.68 mg L −1 , respectively, on day 18.
The DOP and PO 4 -P concentrations decreased by varying degrees under all the treatments during the experiment (Table 1). On day 18, the DOP concentrations ranged from 0.01 mg L −1 for the M + treatment to 0.78 mg L −1 for the PM + treatment. The PO 4 -P concentrations ranged from 0.05 mg L −1 for the M-treatment to 0.78 mg L −1 for the PM-treatment.

Higher resistance of the MC-producing Microcystis strain
The present study found a considerable difference in the responses of MC-producing versus non-MC-producing Microcystis strains by using one-on-one coexistence experiments with the submerged macrophyte M. spicatum under controlled laboratory conditions. The MC-producing Microcystis strain showed a higher resistance to the cocultured submerged plant M. spicatum than the non-MC-producing strain. By contrast, M. spicatum was damaged seriously and rapidly in the presence of MC-producing Microcystis cells instead of non-MC-producing cells. Consistent with our study, the morphology and physiological status of the aquatic macrophyte Egeria densa showed greater impact to exposure to MC-producing Microcystis cells than non-MC-producing cells (Amorim et al. 2017). Although more Microcystis strains are called to provide stronger evidence, Table 1 Concentrations of dissolved nitrogen and phosphorus in the culture solution of the treatments at the end of the experiment and the increasing ( +) or decreasing ( −) ratio compared to the concentra-tions in the beginning. Different lowercase letters indicate significant differences among groups (one-way ANOVA, p < 0.05) the higher resistance of MC-producing strain might be related to the ecotoxicity of MCs to submerged macrophytes (Jiang et al. 2011;Rastogi et al. 2014;Li et al. 2020a). Pure MCs at low concentrations have been shown to significantly affect the physiological structure and function of the submerged macrophyte Vallisneria natans (Jiang et al. 2011). We observed significant stimulation of MC production and MC release by MC-producing Microcystis during coculture with the plant M. spicatum in our study. Additionally, the elevated production of toxic compounds by Microcystis has been shown to defend this species against the grazer Daphnia and Planktothrix, another cyanobacterium (Sadler and von Elert 2014;Briand et al. 2017). Herein, MC production and release might be a defensive strategy for the MC-producing strain to enhance its resistance against M. spicatum. Therefore, higher MC cell quotas combined with a high abundance of MC-producing Microcystis subpopulations may play an important role in affecting the survival and restoration efficiency of submerged macrophytes. This situation should be taken into account when undertaking aquatic vegetation restoration projects in eutrophic waterbodies.
This study had higher initial cell concentrations of Microcystis in coexistence experiments compared with a previous mesocosm study (Švanys et al. 2014), in which coexisting MC-and non-MC-producing M. aeruginosa subpopulations were both inhibited by M. spicatum (Švanys et al. 2014). It indicates that the cell concentration of Microcystis might be another important factor to influence the interaction outcome of submerged macrophytes and Microcystis cells. Dense Microcystis cells may easily cause light shading on M. spicatum from the beginning of the experiment. However, the rapid disappearance of non-MC-producing Microcystis cells starting on day 9 was observed in the PM-treatment. The concentrations of dissolved carbon, nitrogen and phosphorus during the experiment were considerably higher than those in the typical eutrophic lakes (Ye et al. 2011;Xu et al. 2017) and were sufficient to avoid imposing nutritional restrictions on plants and Microcystis in the coculture systems. Hence, we deduced that the stronger allelopathic effects from M. spicatum must be the main reason for the out-competition of the non-MC-producing Microcystis by M. spicatum.
When exposed to pure allelochemicals or algicides, MCproducing Microcystis strains have been shown to be more sensitive than non-MC-producing strains in some studies. In mono-and coculture systems, the toxigenic M. aeruginosa strain was more sensitive to pyrogallol, a typical allelochemical derived from M. spicatum, than the nontoxigenic strains (Gao et al. 2020). The MC-producing strain was more sensitive to hydrogen peroxide treatment than non-MC-producing strains (Latour et al. 2022). This contrasts to the present study where plants and Microcystis cells are in contact and influence each other directly in plant-Microcystis co-existing conditions. Therefore, there is value in future coculture experiments simulating natural conditions in order to understand the interactions between submerged macrophytes and cyanobacteria.

The responses of abiotic microenvironment and bacterioplankton
The forms and concentrations of dissolved carbon, nitrogen, and phosphorus changed in association with the interaction between Microcystis and M. spicatum. The significantly higher concentrations of DOC and NH 4 -N under the PM + and M + treatments than those under the P, M − , and PM − treatments at the end of the experiment indicated that the presence of the MC-producing Microcystis enhanced the accumulation of dissolved organic matter and reducing substances. NH 4 -N concentrations were reported to be positively correlated with total microcystin concentrations in many lakes and reservoirs (Wei et al. 2022). The highest NH 4 -N concentrations were accompanied by an obvious increase in intra-and extracellular MC concentrations under the PM + treatment during the experiment.
The DIC, DON, NO 2 -N, and NH 4 -N concentrations in the PM + and PM − treatments were much higher than those in the corresponding monoculture treatments. These results were indicative of the quicker release and accumulation of organic and reducing inorganic substances during the decline of Microcystis cells or plant shoots in the coculture systems than in the monoculture systems. Dead or living Microcystis cells passively or actively release organic carbon into the environment (Chen et al. 2018). Organic carbon concentrations increase during the decomposition of aquatic plants (Zhang et al. 2018). The decline in Microcystis cells might result in the generation of a high amount of NH 4 -N, whose toxicity plays key roles in the deterioration of submerged macrophytes (Li et al. 2020b). The partial loss of M. spicatum under the PM − treatment might be due to the increasing amount of NH 4 -N during the decline of non-MC-producing Microcystis. The increase in organic matter and reducing inorganic substances, including NO 2 -N and NH 4 -N, might have eventually exacerbated the lose-lose situations of coexisting plants and Microcystis. However, further studies on an expanded scale of time and space at multistrain levels are needed to find the threshold at which the submerged macrophytes could survive and dominate to guide the practice of submerged vegetation restoration in Microcystis-dominated waterbodies.
The structure and predicted function of bacterioplankton community in different treatments are closely related to the competing allelopathic activities of M. spicatum and different Microcystis strains. The effective suppression of the plant M. spicatum by MC-producing Microcystis in the PM + treatment might accompany with stronger effects of MC-producing Microcystis on the bacterioplankton community than M. spicatum. Meanwhile, the obvious inhibition of non-MC-producing Microcystis by M. spicatum in the PM − treatment was suggestive of the more rapid responses of bacterioplankton to the allelopathic activity of M. spicatum than non-MC-producing Microcystis.
Rhodobacter, one genus that was dominant only under the PM + and M + treatments, is found in free-living bacterial communities in freshwater aquatic environments (Moody and Williamson 2013) and has been reported to include strains with the abilities to degrade organic pollutants (Bai et al. 2020). Interestingly, Terrimicrobium, an anaerobic bacterium, was in the highest abundance under the PM + treatment wherein NH 4 -N concentrations were the highest (Qiu et al. 2017).
The genera with higher abundance under the P treatment than under the M + and PM + treatments indicated that these genera preferred oxidizing environments with high nitrogen concentrations. Acinetobacter is a genus of denitrifying bacteria belonging to Gammaproteobacteria (Zhang et al. 2021a, b). Its high abundance might be related to the higher NO 3 -N concentrations in the P treatment. Some strains isolated from this genus execute strong cyanolytic activity (Osman et al. 2017). Chryseobacterium, which had higher abundance under the P and PM − treatments than under the M + and PM + treatments, was also isolated from Lake Taihu; this genus has the ability to produce algicidal compounds (Guo et al. 2015).
The six bacterioplankton genera, including Phreatobacter (Proteobacteria) and Rhodococcus (Actinobacteriota), with higher relative abundances under the M − treatment than under the P and PM − treatments. Phreatobacter has been reported to dominate in drinking water distribution systems with low chlorine concentrations (Perrin et al. 2019). Rhodococcus, an aerobic bacterium, has often been reported to use hydrocarbons as its only source of carbon and energy. This genus includes strains that can enhance the growth of Microcystis and degrade cyanotoxins (Berg et al. 2009;Rohrbacher and St-Arnaud 2016;Kumar et al. 2019).

Conclusions
This study demonstrated a higher resistance of the MCproducing Microcystis to the cocultured submerged macrophyte M. spicatum than that of the non-MC-producing strain. It was accompanied by more serious damage on M. spicatum by the MC-producing strain than the non-MCproducing strain. Significant stimulation of MC production and release during coculture with the plant M. spicatum (the PM + treatment) might account for the higher resistance of the MC-producing Microcystis. Meanwhile, the structure and predicted function of bacterioplankton communities at genus level responded more rapidly to the MC-producing Microcystis under the PM + treatment than M. spicatum. The increased dissolved organic and reducing inorganic compounds released by living and decaying Microcystis cells might eventually exacerbate the recovering capacity of coexisting submerged plants.
Therefore, the results of this study suggested that substantial differences can be found between cyanobacterial strains in terms of the competition between macrophytes and cyanobacteria. MC-producing capability and population abundance of MC-producing Microcystis were important factors to affect the reestablishment of submerged vegetation. This situation should be considered when undertaking remediation works.