Introduction

Submerged plants are the important primary producers in lake ecosystems, which play an important role in water purification, structure restoration, function transformation, and health maintenance of water ecosystems1,2,3. Due to the influence of human activities, such as the continuous output and accumulation of a large number of nitrogen and phosphorus caused by point and area-source pollution, lakes as the receiving water bodies are facing the issue of eutrophication and even hyper eutrophication4,5,6. The use of submerged plants to repair polluted rivers and eutrophic lakes has received widespread attention7. As an important part of the aquatic ecosystem, submerged plants can absorb nitrogen, phosphorus, and other nutrients in the water during their growth process, thereby reducing the pollution load of nitrogen and phosphorus in water bodies. In addition, the surface of submerged plants can attach suspended solids to reduce water turbidity8, inhibit the release of nutrients in sediments through their roots, and also resist the resuspension of sediments caused by wind and wave disturbances9. Therefore, submerged plants play an important role in ecological restoration of water bodies.

The growth of submerged plants and the ecological restoration process are influenced by many factors. Suspended solids is one of the important factors, which contains organic and inorganic components and is also a significant indicator of water quality10,11,12. The increase of suspended solids in water bodies will change the optical properties of water bodies, such as transparency, optical attenuation coefficient, color, and turbidity of water, as well as decrease the landscape of water bodies13. Suspended solids can release nitrogen and phosphorus through desorption processes, which is one of the most important factors affecting lake eutrophication14. The adsorption, flocculation, deposition, and resuspension of suspended solids can also cause local pollution of water bodies. In the water bodies with high content of suspended solids, the suspended particles reduce the underwater light intensity and light availability by the effects of refraction, reflection, and scattering. Thus, photosynthesis of submerged plants is inhibited15, even leading to death of plants, which directly affects its ecological restoration process.

Doyle et al. found that water turbidity could significantly affect the survival and growth of Vallisneria americana winterbuds and seedlings, relative growth of the plants, total number of rosettes and leaves16. Submerged plants obtain lower light energy due to water turbidity or sediment resuspension. Cao et al. observed that light stress gradually reduced the electron transfer rate of PS II, and concluded that turbid water mainly changed the growth of Vallisneria natans by affecting the photosynthesis17. Chen et al. studied the effects of water flow and suspended solids on the surface biofilm of two submerged plants (Myriophyllum verticillatum and Vallisneria natans)8. They found that the main peak particle size of suspended solids (8.71–13.18 μm) in the treatment group with submerged plants was lower than that (15.14–19.95 μm) without submerged plants. The addition of water flow and suspended solids significantly increased the biofilm thickness attached to Myriophyllum verticillatum, but significantly reduced the biofilm thickness of Vallisneria natans.

In summary, it is necessary to reduce the concentration of suspended solids in water bodies for submerged plants to get sufficient light, which is beneficial for improving the growth of submerged plants and their water restoration efficiency. At present, the common method of constructing submerged plant community is cultivating submerged plants after dredging or lowering the water level, or planting them in enclosure to reduce the influence of suspended solids18. This study attempted to meet the light conditions required for the growth of submerged plants in the water with high turbidity by directly adding underwater light source around the plants.

Vallisneria plants, as one kind of typical submerged plants, are widely distributed in surface water bodies around the world. Compared with other kinds of submerged plants, they have the lowest light compensation point and the lowest demand for light19,20, and can be used as a pioneer species in the water restoration process by submerged plants. In addition, the turbidity in natural water bodies is mainly caused by suspended inorganic and organic substances. Different kinds of suspended solids under similar lighting conditions may generate different influences on submerged plants and its restoration process for water bodies, and the corresponding influencing mechanism is still unclear. Therefore, this study selected the common submerged plants (V.spinulosa Yan) in surface water bodies as the research object. The effects of underwater light source on the ecosystem constructed by V.spinulosa Yan under different water turbidity conditions were explored, focusing on the physiological and biochemical characteristics of V.spinulosa Yan, water purification ability and microbial community composition of the ecosystem. We assume that the supplementation of underwater light source will affect the growth of V.spinulosa Yan and change the microbial community structure attached to the surface of V.spinulosa Yan. In this study, V.spinulosa Yan was exposed to eutrophic water containing different kinds of suspended solids under the supplementation of underwater light source. The aims were to demonstrate whether the underwater light source can improve the water purification capacity of the ecosystem, promote the growth of V.spinulosa Yan, and change the composition of microbial communities in the system. This study will provide a theoretical basis for the construction or restoration of submerged plants in natural eutrophic lakes.

Results

Efficiency of water purification

It can be seen from Supplementary Fig. 1 that there was no significant difference in water temperature between different experimental groups, but there were significant differences in pH, DO, and EC (P < 0.05). DO value in the four experimental groups with underwater light source (4.94–5.77 mg/L) was significantly higher than that without underwater light source (2.35–3.15 mg/L) (P < 0.05), pH had the similar trend. pH (7.23–9.03) and DO (4.06–7.90 mg/L) of the NWL were the highest, while the EC value (133.27–255.03 μS/cm) of the NW was the highest.

The removal loads of TN in the experimental groups with underwater light source in Supplementary Fig. 2 were 1.43–3.60 × 10−2 g m−2 d−1 (NWL), 4.24–8.48 × 10−3 g m−2 d−1 (AWL), 3.03–9.37 × 10−2 g m−2 d−1 (CKL), and 4.10–11.10 × 10−2 g m−2 d-1 (IWL), which were significantly higher than that without underwater light source (P < 0.05). There was a significant difference in TN removal load between natural water and AW, IW, and CK, but there was no significant difference among AW, IW, and CK.

Single-factor analysis showed that the presence of underwater light source and the type of suspended solids had no significant effect on the removal load of NO-3-N in water. However, it can be seen from Supplementary Fig. 3a, c, d that in addition to the experimental group with organic suspended solids, the addition of underwater light source could enhance the removal loads of NO-3-N in other experimental groups. In the later stage of the experiment, the difference in the removal loads of NO-3-N between the experimental groups with and without underwater light source gradually decreased (Supplementary Fig. 3). In addition, there was a significant positive correlation between WT and the removal loads of \({\mathrm{NO}}^{-}_{3}{\mbox{-}}{\mathrm{N}}\) (P < 0.05).

Supplementary Fig. 4 illustrated that the removal loads of NH4+-N in water were not significantly affected by the presence of underwater light source, but there was significant difference among the water bodies with different kinds of suspended solids (NW, AW, and IW, P < 0.05). The removal loads of NH4+-N in AWL (3.53 × 10−2 g m−2 d−1), IWL (4.27 × 10−2 g m−2 d−1) and CKL (3.99 × 10−2 g m−2 d−1) were 2.5–3 times higher than that of NWL. Different types of suspended solids had little effect on the TP removal in water, but the presence of underwater light source had a significant effect (Supplementary Fig. 5). The removal loads of TP in the experimental groups with underwater light source were significantly higher than that without underwater light source (P < 0.05). The average removal loads of TP by AWL, IWL, and CKL was 6.80 ± 0.80 × 10−3 g m−2 d−1, but that without underwater light source was 3.94 ± 0.17 × 10−3 g m−2 d−1. In addition, the presence of underwater light source had no significant effect on the \({\mathrm{PO}}^{3-}_{4}{\mbox{-}}{\mathrm{P}}\) removal loads of the experimental groups (Supplementary Fig. 6). The \({\mathrm{PO}}^{3-}_{4}{\mbox{-}}{\mathrm{P}}\) removal loads of CKL was the highest, with an average of 6.57 × 10−3 g m−2 d−1.

Physiological and biochemical characteristics of V.spinulosa Yan

Chla is the main photosynthetic pigment for energy conversion in plants. Depending on the different protein environment, chla can both harvest light energy and act as a redox participant in the main charge separation of PSII and PSI21,22. Chlb is the main component of the light-harvesting protein complex, and its main function is to absorb and transfer solar energy23,24. The chlorophyll content of the experimental groups with underwater light source was significantly lower than that without underwater light source (P < 0.01), while there was no significant difference in chlorophyll content between the leaves of V.spinulosa Yan with different types of suspended solids. The average chlorophyll content of the experimental groups was NW (1.12 mg/g) > IW (1.09 mg/g) > AW (1.08 mg/g) > CK (1.08 mg/g) > NWL (1.03 mg/g) > CKL (0.97 mg/g) > AWL (0.91 mg/g) > IWL (0.88 mg/g).

Carotenoid is a non-enzymatic antioxidant. It can absorb different light and play the role of auxiliary pigment under light restriction, and also protect the formation of chla and chlb25. The carotenoid content of the experimental groups with underwater light source was significantly higher than that without underwater light source (Fig. 1b), while different types of suspended solids had no significant effect on it. The average carotenoid content of each experimental group was NWL (0.154 mg/g) > CKL (0.152 mg/g) > AWL (0.143 mg/g) > IWL (0.138 mg/g) > NW (0.126 mg/g) > CK (0.125 mg/g) > AW (0.124 mg/g) > IW (0.105 mg/g). Among them, the carotenoid content in the leaves of V.spinulosa Yan was the lowest in the experimental group containing inorganic suspended solids.

Fig. 1: Physiological characteristics of V.spinulosa Yan leaves.
figure 1

a Chlorophyll a + b content; b carotenoid content. ChlU+b represents chlorophyll a + b content. NW, AW, IW, and CK represent the natural lake water group, simulated lake water + Microcystis aeruginosa group, simulated lake water + kaolin group, and simulated lake water group, respectively. NWL, AWL, IWL, and CKL are the corresponding groups with underwater light source.

Full size image

SOD can scavenge superoxide anion radicals and plays an important role in the balance of oxidation and antioxidation in plants26,27. In Fig. 2a, there was no significant difference in SOD enzyme activity between the leaves of V.spinulosa Yan in different experimental groups. The SOD enzyme activity in the most experimental groups was slightly increased and then decreased, and presented as a decrease trend from 30 d to 60 d. MDA is one of the products of membrane lipid peroxidation. It acts as an indicator of lipid peroxidation, indicating the degree of membrane lipid peroxidation and the response of plants to stress conditions28,29. From Fig. 2b, we can see that the content of MDA in leaves of V.spinulosa Yan in each group decreased first and then increased. In the later stage of the experiment, the leaves of V.spinulosa Yan began to senescence, resulting in an increase in MDA content. Compared with the experimental group without underwater light source, the changing trend of MDA content in the experimental group with underwater light source was more obvious. The changing trend of AWL and CKL was the most obvious, and the MDA content of CKL increased rapidly as the experiment processed.

Fig. 2: Biochemical characteristics of V.spinulosa Yan leaves.
figure 2

a SOD enzyme activity b MDA enzyme content. SOD and MDA represent activity of superoxide dismutase and content of malondialdehyde, respectively. NW, AW, IW, and CK represent the natural lake water group, simulated lake water + Microcystis aeruginosa group, simulated lake water + kaolin group, and simulated lake water group, respectively. NWL, AWL, IWL, and CKL are the corresponding groups with underwater light source.

Full size image

Microbial community composition

Ace and Chao are indicators for evaluating the richness of microbial communities, and the larger Chao and Ace values indicate the more total number of microbial species30,31. Compared with other regions, Ace and Chao in water of the device had the lowest values, and the leaf area had the second lowest value (Fig. 3a, b). The microbial abundance of the leaf in the experimental groups with underwater light source was lower than that without underwater light source (Fig. 3). The microbial abundance on leaf surface of the CKL group was almost the same as that in the water. The abundance of microbial communities in the sediments and the root surface of V.spinulosa Yan was the highest, while it within the roots of V.spinulosa Yan was lower.

Fig. 3: Differences of microbial alpha diversity indexes at different locations.
figure 3

a Chao, (b) Ace, (c) Shannon, (d) Simpson.

Full size image

Shannon index can evaluate the diversity of microorganisms in samples, and both Shannon and the Simpson index are often used to reflect alpha diversity32,33. The larger Shannon value and the smaller Simpson value indicate the higher community diversity. Figure 3c, d showed lower microbial diversity in water. Compared with the water area, the leave surface of V.spinulosa Yan had the higher microbial diversity. The Ace, Chao, Shannon, and Simpson indexs in the leaf area indicate that the biofilm is attached to the leave surface of V.spinulosa Yan, forming a suitable habitat for bacterial growth. In addition, the microorganisms on the root surface of V.spinulosa Yan and the sediment surface had the highest microbial diversity.

The bacteria of all samples were classified and screened at the phylum level, and the OTUs with the relative abundance of less than 1% were assigned to others. The relative abundance map of free-living bacteria at the phylum level was shown in Fig. 4. A total of 14 dominant bacterial phyla were obtained in the five regions, including Proteobacteria, Actinobacteriota, Cyanobacteria, Acidobacteriota, Bacteroidota, Chloroflexi, Myxococcota, Firmicutes, Patescibacteria, Gemmatimonadota, Desulfobacterota, Nitrospirota, Verrucomicrobiota, and Methylomirabilota.

Fig. 4: The percentage of bacterial flora in different positions of each experimental group at the level of phylum.
figure 4

The letters L, R, Ri, S, and W represent the microorganisms taken from the leaves of V.spinulosa Yan, the area around the root system, the internal area of the root, the sediment area and the water area respectively. The numbers represent different experimental groups, 1: NW 2: AW 3: IW 4: CK 5: NWL 6: AWL 7: IWL 8: CKL.

Full size image

NMDS analysis of bacteria at the OTU level showed that there were significant differences in bacterial communities between different regions (Fig. 5a). The Stress of NMDS analysis results is 0.1 (<0.2), indicating that the analysis results are reliable. There was almost no difference in the distribution of microorganisms in the sediments of different experimental groups, while the microorganisms inside and around the roots of V.spinulosa Yan were similar. In addition, the relative distance of different points in the leave surface area of V.spinulosa Yan is far. Therefore, the microbial community on the surface of V.spinulosa Yan leaves needs further analysis. Figure 5b shows that the Stress of NMDS analysis results is 0.004 (<0.05), indicating the highly representative results. The microbial community on the leaf surface of V.spinulosa Yan in eight different experimental groups was significantly different. Underwater light source had a great impact on the microbial community of the leave surface of V.spinulosa Yan in simulated water bodies.

Fig. 5: NMDS analysis of bacteria in leaf area.
figure 5

a Different positions; b different light conditions (L is underwater light source, NL is no underwater light source). Stress: Test the results of NMDS analysis. It is generally believed that when stress <0.2, it can be expressed by the two-dimensional point diagram of NMDS, and the diagram has certain explanatory significance; When stress <0.1, it can be considered as a good sort; when stress <0.05, it has a good representativeness.

Full size image

Discussion

Addition of underwater light source promoted the photosynthesis of V.spinulosa Yan, which can convert more CO2 to O2, resulting in the increase of pH in the water (Supplementary Fig. 1a). Pearson correlation analysis shows that there is a significant negative correlation between water temperature (WT) and dissolved oxygen (DO) (P < 0.01). When the ambient temperature is low, the DO content in the water increases, which may be due to a decrease in the activity of aerobic microorganisms in the water and a decrease in DO consumption in the water. Kim et al. observed that nitrification was slowly processed at low temperatures in winter (the reaction rate was about half of that in summer), and nitrification activity could be controlled by adjusting DO to obtain better water quality in effluent34.

Further analysis shows that there is a significant positive correlation between WT and NO3−-N removal load (P < 0.05). When WT decreased, the NO−3-N removal loads of the experimental group decreased significantly, especially in the experimental groups with underwater light source. In addition, when WT decreases, the difference between the NO−3-N removal loads of the experimental groups with the underwater light source and without the underwater light source becomes smaller (Supplementary Fig. 3).

It is worth noting that there is a significant positive correlation between NH4+-N removal loads and WT, TN removal loads (P < 0.05) (Fig. 6). When WT decreased, the NH4+-N removal loads of the experimental groups decreased significantly. The activity of aerobic microorganisms may be reduced with the decrease of WT, as well as the activity of aerobic nitrifying bacteria, which may be a major reason for the reduction of NH4+-N removal loads in water. There is a significant positive correlation between the removal loads of TN and the removal load of NH4+-N, but there is no significant difference in the removal loads of NH4+-N between each experimental group, and the removal efficiencies achieved approximately 100%. Therefore, the difference in TN removal is mainly due to the removal of \({\mathrm{NO}}^{-}_{3}{\mbox{-}}{\mathrm{N}}\) rather than the removal of NH4+-N.

Fig. 6: Correlation heat map between environmental factors.
figure 6

Turb, WT, EC, DO, TN, NH4+-N, NO3–-N, TP, PO43−-P, and NO2−-N represent turbidity, water temperature, conductivity, dissolved oxygen, total nitrogen, ammonia nitrogen, nitrate nitrogen, total phosphorus, phosphate, and nitrite nitrogen, respectively.

Full size image

The results of chlorophyll content reveal that V.spinulosa Yan can adapt to the water environment with low light intensity by enhancing its own chlorophyll content and strengthening its own photosynthetic organs. While in the water environment with high light intensity after adding underwater light source, V.spinulosa Yan will choose priority cloning and reproduction to increase its biomass. Previous studies demonstrated that higher content of carotenoids in plants was obtained when they were exposed to higher light intensity35. This is consistent with the results of this study (Fig. 1b), revealing that the content of carotenoids in the experimental groups with underwater light source is higher than that without underwater light source. In addition, the content of carotenoids in the leaves of V.spinulosa Yan was the lowest in the water containing inorganic suspended solids. It may be due to that inorganic suspended solids have stronger scattering of light and are more easily adsorbed on the leaves of V.spinulosa Yan, leading to the decrease of light intensity and the inhibition on the photosynthesis of V.spinulosa Yan.

The SOD values in the leaves of most experimental groups were slightly increased and then decreased. The enzyme activities of V.spinulosa Yan in almost all the experimental groups exhibited a downward trend during 30–60 d. It indicates that V.spinulosa Yan is subjected to a certain degree of stress during this period, which exceeds its tolerance range and destroys its protective system, resulting in a high accumulation of reactive oxygen species and peroxidation of membrane lipid in the plants, and an increase in the MDA content (Fig. 2b). More obvious changing trend was obtained in AWL and CKL. The MDA content of CKL increased rapidly with the time, indicating that the addition of underwater light source accelerates the stress of V.spinulosa Yan. The reason may be that the algae adsorbed on the leaf surface of V.spinulosa Yan compete with V.spinulosa Yan for a favorable growth environment, which is not conducive to V.spinulosa Yan. This agrees with Song et al.‘s conclusion36.

The microorganisms on the leave surface of V.spinulosa Yan were analyzed by LEfSe, aiming to find out the microorganisms with significant differences between the experimental groups with and without underwater light source, and the microorganisms with >4 of LDA values were listed (Fig. 7). The branch bacteria with a higher abundance on the leaf surface of V.spinulosa Yan without underwater light source included c_Actinobacteria, f_Saccharimonadales, c_Sphingomonadales, g_Nitrospira, f_Methylomonadace. Interestingly, both anoxic (f_Methylomonadace) and aerobic (g_Nitrospira) microorganisms existed on the leaf surface of V.spinulosa Yan under the insufficient light conditions. It indicates that biofilm with a perfect structure has been formed on the leaf surface, and the biofilm has the dominant bacteria related to water purification. f_Methylomonadace belongs to o_Methylococcales, which is a methane-oxidizing bacteria that can remove greenhouse gas (CH4). g_Nitrospira plays an important role in biological nitrification process, which can completely convert NH4+-N into \({\mathrm{NO}}^{-}_{3}{\mbox{-}}{\mathrm{N}}\). In addition, there were also some genus of c_Sphingomonadales with significant special abundance, which have the potential to improve plant growth by producing hormones for plant growth under stress conditions such as drought, salinity, and heavy metals in agricultural soils37.

Fig. 7: Microbial analysis on the leave surface of V.spinulosa Yan under different conditions.
figure 7

a Branching diagram, LEfSe multi-level species analysis of microbial abundance on the leaf surface of V.spinulosa Yan under different light conditions; b Highly differential taxonomy and bacteria with linear discriminant analysis (LDA) values greater than 4. NL and L represent the groups without and with underwater light source.

Full size image

The branching bacteria with higher abundance on the leaf surface of V.spinulosa Yan with underwater light source included p_Cyanobacteria, o_Chloroplast, f_Leptolyngbyaceae, g_Limnothrix, g_Rhodobacter, f_Sutterellaceae, and g_AAP99. The results showed that the addition of underwater light source promoted the attachment and growth of p_Cyanobacteria on the leaf surface of V.spinulosa Yan, and there is also a close relationship between f_Sutterellaceae and bloom of Microcystis aeruginosa. Although the attachment of algae may have a negative impact on the growth of V.spinulosa Yan, previous studies also found that bacteria such as f_Leptolyngbya and g_Limnothrix had a high removal rate for organic matter, N, and P. Especially in treating the wastewater from distilleries, they could achieve a high removal rate of dissolved chemical oxygen demand (up to 97.4%). Interestingly, underwater light source promoted the growth of g_Rhodobacter on the leaf surface, which belongs to nitrogen-fixing red bacteria (Photosynthetic bacteria). Photosynthetic bacteria can use organic acids with low molecular weight (e.g., butyric acid and acetic acid) and carbohydrates (e.g., glucose and other easily hydrolyzed organic matters) to produce biohydrogen through photofermentation.

It was found that the abundance of f_Methyloligellaceae in the sediment increased significantly in the system with underwater light source (Fig. 8). f_Methyloligellaceae is one kind of methanotrophs, which can use monocarbon organic matter as source of carbon and energy. It indicates that the system with underwater light source may have stronger degradation ability of monocarbon organic matter such as methane. We also found that the abundance of f_Methylophilaceae increased significantly in the water with underwater light source (Fig. 9a). f_Methylophilaceae belongs to c_Betaproteobacteria, they can metabolize methylated compounds, especially methanol.

Fig. 8: Microbial analysis in the sediment of the systems under different conditions.
figure 8

a Branching diagram, LEfSe multi-level species analysis of microbial abundance on sediments under different light conditions; b Highly differential taxonomy and bacteria with linear discriminant analysis (LDA) values greater than 3. NL and L represent the groups without and with underwater light source.

Full size image
Fig. 9: Microbial analysis in the water of the systems under different conditions.
figure 9

a Branching diagram, LEfSe multi-level species analysis of microbial abundance in water under different light conditions; b Highly differential taxonomy and bacteria with linear discriminant analysis (LDA) values greater than 3. NL and L represent the groups without and with underwater light source.

Full size image

More ammonia-oxidizing bacteria (AOB) were observed on the leave surface of V.spinulosa Yan without underwater light source (Fig. 10a). For g_Nitrosomonas, its abundance achieved 0.23–0.52% in the simulated lake water and 0.05% in the natural lake water without underwater light source, while its abundance was less than 0.1% on the leave surface of V.spinulosa Yan with underwater light source. In the simulated lake water with low content of suspended solids, the abundance of g_Nitrosomonas reached 0.42% (without underwater light source) and 0.29% (with underwater light source) in the root interior of V.spinulosa Yan. The abundance of f_Nitrosomonadaceae in the water without underwater light source was 2.1–1.02%, achieving the highest abundance in AW and the lowest abundance in NW. Its abundance in the water with underwater light source reached 0.15–0.7%, with the highest abundance in NWL and the lowest abundance in CKL. In the sediment, the abundance of f_Nitrosomonadaceae in each group was AWL (2.59%) > NWL (2.37%) > CKL (2.10%) > IWL (1.91%), NW (2.28%) > CK (2.02%) > IW (1.91%) > AW (1.81%). Therefore, the addition of underwater light source promoted its abundance in the sediment. In addition, the abundance of nitrite-oxidizing bacteria (NOB) on the leave surface of V.spinulosa Yan without underwater light source was higher than that with underwater light source (Fig. 10b). The abundance of NOB such as g_Nitrospira on the leaf surface of each group was NW (4.22%) > AW (3.80%) > IW (2.30%) > CK (2.90%) > AWL (0.33%) > NWL group (0.18%) > CKL (0.10%) > IWL (0.09%).

Fig. 10: Relative abundance of nitrogen removal bacteria in different locations.
figure 10

a AOB; b NOB; c DB. AOB, NOB, and DB represent ammonia-oxidizing bacteria, nitrite-oxidizing bacteria, and denitrifying bacteria. NW, AW, IW, and CK represent the natural lake water group, simulated lake water + Microcystis aeruginosa group, simulated lake water + kaolin group, and simulated lake water group, respectively. NWL, AWL, IWL, and CKL are the corresponding groups with underwater light source.

Full size image

The abundance of denitrifying bacteria (DB) in the sediment and the root surface of V.spinulosa Yan exhibits a significant correlation with the presence or absence of underwater light source (Fig. 10c). The main DB in the system included g_Bacillus, g_Paracoccus, g_Methylobacterium, f_Alcaligenaceae, g_Pseudomonas, the dominant DB with acetate as carbon source (f_Comamonadaceae, f_Rhodocyclaceae), and with methanol as carbon source (g_Methylophilus, g_Methylobacillus, g_Hyphomicrobium). We found that the abundance of f_Comamonadaceae was significantly reduced on the leave surface of V.spinulosa Yan with underwater light source. It may be due to the increase of DO around V.spinulosa Yan by supplying sufficient light. Interestingly, except for the root interior of V.spinulosa Yan, the abundance of f_Rhodocyclaceae in other four regions is significantly correlated with the presence or absence of underwater light source. The addition of underwater light source decreased the abundance of f_Rhodocyclaceae in the leaves and root surface of V.spinulosa Yan, sediment surface, and water. The abundance of g_Paracoccus in the water (1.42%) was higher than that in root surface (0.08%) and sediment (0.02%) with underwater light source, and the highest content (2.24%) was observed on the leaf surface of V.spinulosa Yan in NWL. But it had the highest abundance (2.38%) on the leaf surface of V.spinulosa Yan without underwater light source. However, the addition of underwater light source promoted the growth of g_Bacillus in the sediment and on the leaf surface of V.spinulosa Yan.

The above analysis showed that the experimental groups without underwater light source had a higher abundance of AOB, especially in the leave surface of V.spinulosa Yan (Fig. 10a). The abundance of NOB in the leaf surface, root interior and surface, and water (except for the water in CK) without underwater light source was higher than that with underwater light source (Fig. 10b). In addition, the abundance of NOB and AOB is significantly negatively correlated with the content of NO2−-N and NH4+-N in the water (Fig. 11). But there is no significant difference in the removal loads of NH4+-N in the water between with and without underwater light source (Supplementary Figs. 3 and 4), indicating that the removal of NH4+-N in the groups without underwater light source is closely related to nitrifying bacteria, while the removal of NH4+-N in the groups with underwater light source may not be mainly dependent on nitrifying bacteria. Similar phenomena were found between DB and \({\mathrm{NO}}^{-}_{3}{\mbox{-}}{\mathrm{N}}\) (Figs. 10c, 11), indicating that the removal of \({\mathrm{NO}}^{-}_{3}{\mbox{-}}{\mathrm{N}}\) in the groups without underwater light source is closely related to DB, while the removal of \({\mathrm{NO}}^{-}_{3}{\mbox{-}}{\mathrm{N}}\) in the groups with underwater light source may not be mainly dependent on DB. Previous studies pointed out that there is a relationship between microalgae and nitrification-related bacteria that promotes the removal of N and P in water or competes with each other. Under insufficient NH4+-N, there is a competitive relationship between microalgae and nitrification-related bacteria34, and sufficient light source promotes the growth of microalgae and inhibit the growth of nitrification-related bacteria38. Figure 7 showed that the addition of underwater light source promoted the attachment and growth of algae bacteria on the leave surface of V.spinulosa Yan. Therefore, in the groups with underwater light source, the removal of N may mainly depend on the assimilation of NH4+-N and \({\mathrm{NO}}^{-}_{3}{\mbox{-}}{\mathrm{N}}\) by the growth of algae and V.spinulosa Yan, while in the groups without underwater light source, the removal of NH4+-N and \({\mathrm{NO}}^{-}_{3}{\mbox{-}}{\mathrm{N}}\) mainly depends on the nitrifying bacteria (AOB and NOB) and DB.

Fig. 11: Heatmap of correlation between environmental variables and abundance of denitrification species/functional bacteria.
figure 11

AOB, NOB, and DB represent ammonia-oxidizing bacteria, nitrite-oxidizing bacteria, and denitrifying bacteria. NTU, WT, ORP, DO, TN, NH4+-N, NO3–-N, TP, PO43−-P, NO2–-N, CHl, SOD, and MDA represent turbidity, water temperature, redox potential, dissolved oxygen, total nitrogen, ammonia nitrogen, nitrate nitrogen, total phosphorus, phosphate, nitrite nitrogen, chlorophyll, activity of superoxide dismutase, and content of malondialdehyde respectively.

Full size image

The removal of phosphorus by microorganisms mainly depends on phosphorus-accumulating organisms (PAOs) with phosphorus-accumulating ability and glycogen-accumulating organisms (GAOs) without phosphorus-accumulating ability. The main PAOs in the groups were g_Accumulibacter and g_Dechloromonas and the main GAOs were g_Defluviicoccus and g_Candidatus_Competibacter. The abundance of PAOs on the root surface of V.spinulosa Yan was IW (1.13%) > AW (0.73%) > CK (0.56%) > NW (0.55%) > IWL (0.42%) > CKL (0.30%) > AWL (0.18%) > NWL (0.15%). The abundance of PAOs in the sediment was AW (0.39%) > IW (0.23%) > NW (0.07%) > CK (0.05%), IWL (0.20%) > AWL (0.10%) > NWL (0.06%) > CKL (0.05%) (Fig. 12). The addition of underwater light source reduced the abundance of PAOs in each group, especially on the root surface. Figure 13 shows that there is a significant positive association between DO and PAOs. At the same time, in the experimental groups without underwater light source, the TP removal rate and DO increased gradually in the later stage of the experiment (Supplementary Figs. 1 and 5). It may be due to that the increase of the abundance of PAOs in the system promotes the absorption of P under aerobic conditions. The TP removal is not related to PAOs and GAOs with underwater light source (P > 0.05), but the TP removal load was significantly higher than that without underwater light source (Supplementary Fig. 5). Therefore, the experimental groups with underwater light source mainly relies on the growth of microalgae and V.spinulosa Yan in the system. Compared with the other experimental groups, the group with inorganic suspended solids had a higher abundance of PAOs in each region. It may be that inorganic suspended solids can act as good carrier for attachment and growth of PAOs, and will not compete with PAOs for growth environment and nutrients, thereby promoting the growth of PAOs. The addition of underwater light source in the water with inorganic suspended solids can promote the growth of V.spinulosa Yan, and can also maintain a high abundance of PAOs in the system.

Fig. 12: Relative abundance of phosphorus removal bacteria in different locations.
figure 12

a PAOs; b GAOs. PAQs and PAOs are phosphorus-accumulating organisms and glycogen-accumulating organisms. NW, AW, IW, and CK represent the natural lake water group, simulated lake water + Microcystis aeruginosa group, simulated lake water + kaolin group, and simulated lake water group, respectively. NWL, AWL, IWL, and CKL are the corresponding groups with underwater light source.

Full size image
Fig. 13: Heatmap of the correlation between environmental variables and the abundance of P removal species/functional bacteria.
figure 13

PAQs and PAOs are phosphorus-accumulating organisms and glycogen-accumulating organisms. NTU, WT, ORP, DO, TN, NH4+-N, NO3−-N, TP, PO43−-P, NO2−-N, CHl, SOD, and MDA represent turbidity, water temperature, redox potential, dissolved oxygen, total nitrogen, ammonia nitrogen, nitrate nitrogen, total phosphorus, phosphate, nitrite nitrogen, chlorophyll, activity of superoxide dismutase, and content of malondialdehyde respectively.

Full size image

In real turbid water bodies, growth of V.spinulosa Yan is usually inhibited due to the low lighting intensity, especially in deep water areas. As a result, its ecological restoration effect for water bodies is significantly reduced. This study suggests addition of underwater light source around V.spinulosa Yan during its growth stage, aiming to improve its growth. After about two months, V.spinulosa Yan grows to a certain height when it can receive natural light, the underwater light source can be removed. During the two-month stage, the cost for adding underwater light source mainly comes from purchasing light source and electricity consumption. Based on this study, the former is about USD 15.76/(m2 water area), and the latter is approximately USD 2.24/(m2 water area). It is worth noting that the underwater light source can be removed from water bodies and reused in other ecological restoration projects after V.spinulosa Yan can receive natural light. Moreover, solar energy is suggested to electricity supply for the underwater light source. Therefore, the cost can be significantly decreased in real projects. In addition, installation and removal of the light source from water bodies can be achieved by using some ropes, suggesting convenient operation. In conclusion, field scale application of adding underwater light source to improve the growth of V.spinulosa Yan and its ecological restoration effect is feasible.

Methods

Experimental materials

V.spinulosa Yan used in this study was taken from a natural lake. The growth substrate was the planting soil in an ecological botanical garden, the gravels are removed by sieves and mixed well for later use. The water used in the experiment contained both natural and simulated lake water. The water quality was TN 2.0 mg/L, NH4+-N 1.0 mg/L, NO3−-N 1.0 mg/L, TP 0.2 mg/L. The underwater light source was a 7 W LED lamp. NH4Cl, KNO3, K2HPO4, and kaolin used to simulate lake water with different kinds of suspended solids were purchased from the National Pharmaceutical Group Chemical Reagents Co., Ltd. (China), and their purity was analytical purity.

Experimental design

The experiment was carried out in an ecological botanical garden of Wenzhou University (Zhejiang Province, China). The growth system of V.spinulosa Yan was established in a device made of polypropylene (length × width × height = 70 cm × 75 cm × 80 cm), which was divided into three independent units (length × width × height = 70 cm × 25 cm × 80 cm) set as parallel groups. In each unit, 18 plants of V.spinulosa Yan were planted in the soil with the layer height of 10 cm. After the cultivation and stable growth of V.spinulosa Yan with 15 d, the experiment was started and lasted for 60 d. Hydraulic retention time (HRT) and water depth was set as 4 d and 40 cm. Because the photosynthetically active radiation received by V.spinulosa Yan in the bottom of the device was close to 0, the artificial light source was added and fixed on the inner wall of the device and located at 10 cm above the soil layer. The light supplement time was 6: 00 am–6: 00 pm every day. Eight experimental groups were set to stimulate different growing conditions for V.spinulosa Yan, the specific settings were shown in Table 1. Microcystis aeruginosa and kaolin were used to simulate organic and inorganic suspended solids according to previous studies, respectively39,40.

Table 1 Experimental setting for different groups
Full size table

Measurement of water quality

Dissolved oxygen (DO), water temperature (WT), pH, and conductivity (EC) in the influent and effluent of the device were measured by YSI meter (YSI ProQuatro, USA) on-site every four days. HACH turbidimeter (2100Q, HACH, USA) was used to measure the turbidity of water samples. Total nitrogen (TN), total phosphorus (TP), nitrate nitrogen (NO3−-N), nitrite nitrogen (NO2−-N), ammonia nitrogen (NH4+-N), phosphate (PO43−-P) were determined by the national standards.

Characterization of V.spinulosa Yan growth

The leaves of V.spinulosa Yan were collected every 15 days to determine the chlorophyll concentration and enzyme activity of V.spinulosa Yan. Chlorophyll concentration was detected by ethanol immersion method. The weighed leaves were first soaked in 95% ethanol for 24 h, then spectrophotometer was used to detect the soak solution at the wavelengths of 665 nm, 649 nm, and 470 nm to calculate the chlorophyll content. Activity of superoxide dismutase (SOD) and content of malondialdehyde (MDA) in the plant tissues were measured by kit method.

Analysis of microorganisms

At the end of the experiment, microbial samples were collected from the five areas in each environmental unit, including leaf surface, root surface, and root interior of V.spinulosa Yan, water, and sediment surface. The quality and quantity of DNA were examined by 1% agarose gel electrophoresis and NanoDrop2000, respectively. 16S rRNA (338 F, 5′-Actcctacgggaggcagcag-3′ and 806 R, 5′-Ggactacnvgggtwtctaat-3′) V3-V4 variable regions were amplified by PCR. Sequencing was performed using Illumina’s Miseq PE300/NovaSeq PE250 platform (Shanghai Meiji Biomedical Technology Co., Ltd, China). USEARCH11 was used for sequence analysis. Sequences with ≥97% of similarity were assigned to the same operational taxonomic unit (OTU). QIIME (v1.9.1) was used for analysis of sparse data set, including extraction and analysis of OTU, and annotation based on Bray-Curtis distance. All sequences could be obtained from NCBI SRA.

Data analysis

SPSS 26.0 software was used for statistical analysis of data. One-way analysis of variance was used to compare the difference between the experimental groups and the control groups, and the statistical significance level was 0.05. Kruskal–Wallis nonparametric test was used to analyze the data of individual parameters that did not conform to the normal distribution, determining the direct and indirect interactions of water characteristics (WT, DO, pH, and EC), water nutrients (TN, TP, \({\mathrm{PO}}^{3-}_{4}{\mbox{-}}{\mathrm{P}}\), \({\mathrm{NO}}^{-}_{3}{\mbox{-}}{\mathrm{N}}\), and NH4+-N), and large plant traits (Chlorophyll, biomass, SOD, and MDA). Microbial high-throughput data were analyzed and plotted using R 4.2.1 language. Effect size pipeline (LEfSe) program of the linear discriminant analysis (LDA) was used to test the significant discriminant groups under different stress conditions. In the LEfSe analysis, OTUs with LDA value > 3.0 were considered to be significant discriminant groups. The correlation between environmental factors, plant traits, and microbial indicators was analyzed by the R 4.2.1 method. In the correlation analysis, the first axis of the principal component analysis (PCA1) was used to represent the composition of the microbial community.