Micro-electric field-associated restructuring of bacterial communities enhances wastewater purification in constructed wetlands

Summary

To improve microbial activity and shock resistance in constructed wetlands (CWs), we established micro-electric field (MEF) enhanced CW systems to investigate their regulatory effects on domestic sewage treatment performance and bacterial communities. The results showed that MEF significantly improved the removal efficiencies of TP (16.97%–17.13%), NH₄⁺-N (16.32%–28.93%), and COD (28.55%–53.81%). MEF also enhanced bacterial α diversity (Chao1, Shannon, Simpson) and altered community structure, with significant enrichment of Pseudomonadota and Bacteroidota that positively correlated with pollutant removal. Network analysis indicated that the MEF increased microbial network complexity and connectivity, and promoted competitive interactions within the community. Functional prediction revealed that MEF upregulated key pathways involved in carbon metabolism and nitrogen metabolism. We hypothesize that the MEF improves pollutant removal performance and stability in CWs by optimizing micro-environmental conditions and promoting a more diverse and functional microbial community. This work provides a microbiological foundation for developing energy-efficient, low-carbon wetland technologies.

Graphical abstract

Highlights

• •Micro-electric field enhanced CWs with high NH₄⁺-N, TP, and COD removal
• •Micro-electric field significantly altered bacterial diversity and composition
• •Micro-electric field enhanced microbial network complexity and connectivity
• •Micro-electric field strengthened carbon and nitrogen metabolism

Environmental science; Pollution; Microbiology

Introduction

The rapidly growing global demand for water resources globally is posing a series of severe challenges, including regional water scarcity, continuous deterioration of water quality, and eutrophication of water bodies.^1,2 Against this backdrop, constructed wetlands (CWs) have gained increasing attention as a green and sustainable wastewater treatment technology, owing to their favorable ecological compatibility, low energy consumption and low construction and operational costs.³ Unlike traditional treatment systems that rely on a single process, CWs remove various pollutants, such as organic matter, nitrogen and phosphorus, by mimicking the complex physical, chemical and biological interactions that occur naturally in wetlands between plants, substrates and microorganisms.⁴ They have been successfully used to treat domestic wastewater, agricultural runoff, and certain industrial effluents.³ Of the various CW systems, vertical flow CWs (VFCWs) are the most widely used. Operating under intermittent loading conditions, VFCWs effectively enhance the transfer of oxygen into the substrate layer. This promotes the aerobic degradation of organic matter and the efficient transformation of nitrogen, while also helping to reduce odors and the risk of pathogens.⁵ Additionally, VFCWs have a compact structure and small footprint, making them ideal for small-scale, decentralized wastewater treatment in areas where land is scarce.²

However, the performance of CWs in removing pollutants, particularly nitrogen and phosphorus, is often limited by factors such as insufficient dissolved oxygen (DO), carbon sources and microbial activity.⁶ These limitations result in inconsistent treatment efficiency and make it difficult to meet increasingly stringent discharge standards.⁴ To address these bottlenecks, various substrate modification strategies have been explored, including the use of iron-based materials (e.g., iron sulfide based artificial wetlands) and sulfur-based materials, which can enhance nutrient removal via chemical adsorption, redox reactions, and autotrophic denitrification.^7,8 Although these methods can improve performance in the short term, they often come with issues such as increased energy consumption, operational complexity, and risks of secondary pollution.⁴

In recent years, micro-electric field technology has shown great promise in environmental biological treatment. For example, previous studies have shown that the application of an electric field can regulate microenvironments, such as pH, DO, and redox potential, in situ.⁹ This creates more favorable habitats for functional microorganisms. Furthermore, electric fields can alter the surface properties of substrates, thereby optimizing mass transfer and adsorption performance.⁵ These effects are expected to enhance the biological and physicochemical processes involved in pollutant migration and transformation within CWs at a fundamental level. However, most existing research has concentrated on microbial fuel cell-coupled CWs (CW-MFCs), where the electric field is passively generated by microorganisms and is often unstable. Therefore, a sustainable and endogenous enhancement technology is needed to improve pollutant removal efficiency and operational stability. Moreover, previous studies have largely focused on improving pollutant removal efficiencies or providing one-dimensional descriptions of shifts in microbial community structure, while the intrinsic mechanisms by which electric fields drive system performance enhancement through the restructuring of microbial ecological networks remain unclear. Specifically, previous studies on micro-electric field-enhanced CWs have not systematically elucidated how micro-electric fields reshape microbial ecological network topology, nor have they quantitatively determined how an externally applied micro-electric field regulates microbial community functions. To advance beyond these gaps, our study integrates co-occurrence network analysis with partial least squares path modeling (PLS-PM) to provide a quantitative and topological understanding of micro-electric-field-driven bacterial community restructuring.

In this study, micro-electric field-enhanced CW treatment systems were constructed. The objectives are to systematically investigate (1) the removal efficiency of typical pollutants (total phosphorus [TP], NH₄⁺-N, and chemical oxygen demand [COD]); (2) the evolution patterns of microbial community diversity, compositional structure, co-occurrence network topology, and potential metabolic functions; and (3) the underlying mechanisms governing the removal of these pollutants under micro-electric field exposure. This research aims to provide novel theoretical insights for the development of highly efficient and stable wetland enhancement technologies.

Results

Micro-electric field enhances simultaneous removal of TP, NH₄⁺-N, and COD

To evaluate the treatment performance, influent and effluent water quality across four systems was systematically monitored during stable operation (Figure 1; Figure S2). The results revealed that the micro-electric field-enhanced groups (EF-substrate and EF-CW) exhibited efficient and stable removal of TP (Figures 1A and 1B), NH₄⁺-N (Figures 1C and 1D), and COD (Figures 1E and 1F). The EF-CW group exhibited the lowest average effluent concentration (0.61 mg L⁻¹, Figure 1A), which was below the class I (B) of the wastewater discharge standard (GB18918-2002) in China (1 mg L⁻¹). The average TP removal rates were significantly higher in the EF-substrate and EF-CW systems than in the CK-substrate and CK-CW systems (Figures 1B; S2A, p < 0.05). Specifically, the mean removal efficiencies reached 72.37% in EF-substrate and 84.16% in EF-CW, markedly exceeding those of CK-substrate (18.56%) and CK-CW (55.61%). In terms of NH₄⁺-N removal, the micro-electric field enhancement also demonstrated significant effects (p < 0.05). The average effluent concentrations of EF-CW and EF-substrate were 18.86 mg L⁻¹ and 26.21 mg L⁻¹, respectively (Figure 1C). These values were below the class II standard for wastewater discharge in China (GB18918-2002) (30 mg L⁻¹). Among them, EF-CW achieved the highest removal rate (62.05%), significantly outperforming all other groups (p < 0.05, Figures 1D and S2B). Regarding COD removal, the micro-electric field also demonstrated a significant enhancement effect (p < 0.05). The average COD removal rates in the EF-CW and EF-substrate systems were 63.79% and 61.76%, respectively. These were both significantly higher than the rates in their corresponding control groups (CK-CW: 46.66%; CK-substrate: 44.79%; p < 0.05, Figures 1F and S2C).

Figure 1.

Micro-electric field enhances removal of TP, NH₄⁺-N, and COD in CWs

(A) Influent and effluent TP concentrations.

(B) TP removal rates.

(C) Influent and effluent NH₄⁺-N concentrations.

(D) NH₄⁺-N removal rates.

(E) Influent and effluent COD concentrations.

(F) COD removal rates.

See also Figure S2.

Micro-electric field increases α diversity and alters β diversity of bacterial community

There were significant differences in the α-diversity indices (Chao1 richness index, Shannon index, and Simpson index) of the microbial communities among the different treatment groups (p < 0.05, Figure 2A). Specifically, the micro-electric field treatment significantly enhanced the species richness of bacteria (p < 0.05, Figure 2A). The Chao1 richness index of EF-substrate was approximately 37.58% higher than that of CK-substrate, and the Chao1 richness index of EF-CW was also approximately 26.22% higher than that of CK-CW (Figure 2A). Correlation analysis revealed that DO (r = 0.87), TP removal (r = 0.70), and NH₄⁺-N removal (r = 0.75) was significantly positively correlated with the Chao1 richness index (Table S4, p < 0.001). These correlations suggest that higher DO levels are associated with enhanced microbial growth. In terms of species diversity, compared to CK-substrate, the Shannon and Simpson indices of EF-substrate increased by approximately 118.53% and 56.83%, respectively (Figures 2B and 2C). Similarly, these two indices for EF-CW were approximately 21.67% and 5.60% higher than those for CK-CW, respectively (Figures 2B and 2C). Correlation analysis revealed that the Shannon and Simpson indices of the microbial community were significantly positively correlated with the removal efficiencies of TP (r = 0.87, p < 0.001; r = 0.91, p < 0.001, respectively), NH₄⁺-N (r = 0.78, p < 0.001; r = 0.74, p < 0.001, respectively), and COD removal (r = 0.46, p < 0.05; r = 0.83, p < 0.001, respectively, Table S4). For instance, the Simpson index showed significant positive correlations with both pH (r = 0.66, p < 0.01) and DO (r = 0.49, p < 0.05, Table S4). This suggested that micro-electric field treatment may enhance the richness and diversity of the microbial community by modulating micro-environmental conditions such as water pH and DO, thereby strengthening the overall pollutant purification potential of the system. There was no significant difference in Simpson index between EF-substrate and EF-CW (p > 0.05, Figure 2C), further indicating that the enhancing effect of the micro-electric field on microbial diversity exhibits a stable performance across different systems. The β diversity of the bacterial community is shown in Figure 2D. The bacterial communities from CK-substrate, CK-CW, EF-substrate, and EF-CW sites formed distinct clusters, respectively. Principal component 1 (PC1) and principal component 2 (PC2) explained 50% and 44.8% of the total variance, respectively. These results indicate that the application of the micro-electric field significantly influenced the bacterial community structure.

Figure 2.

Micro-electric field increases α diversity and alters β diversity of the bacterial community

(A) Chao1 richness index.

(B) Shannon index.

(C) Simpson index.

(D) Principal coordinate analysis.

Data are shown as boxplots with median and interquartile range (n = 5 per group). Different lowercase letters above the boxes indicate significant differences among groups (p < 0.05).

See also Table S4.

Micro-electric field selectively enriches functional bacterial phyla linked to pollutant removal

As shown in Figure 3 the micro-electric field treatment significantly altered the community structure of substrate-attached microorganisms (p < 0.05). The dominant bacterial phyla (with an average relative abundance >1% across all samples) included Pseudomonadota, Chlorobiota, Bacteroidota, Bacillota, Patescibacteria, and Halobacteriota. Their cumulative relative abundance accounted for 95.04%–98.47% of the total bacterial community (Figure 3A). Pseudomonadota was the most abundant bacterial phylum in the system (Figure 3A). Compared to their respective control groups, the micro-electric field treatment significantly increased its relative abundance (p < 0.05). Specifically, the average relative abundance of Pseudomonadota in EF-substrate showed a 78.11% significant increase over CK-substrate, while that in EF-CW also exhibited an 8.21% significant increase compared to CK-CW. Correlation analysis revealed that the relative abundance of Pseudomonadota was significantly positively correlated with the removal efficiencies of TP (r = 0.74, p < 0.001), NH₄⁺-N (r = 0.58, p < 0.01), and COD (r = 0.95, p < 0.001, Figure 3B). Consequently, the significant enrichment of Pseudomonadota in the micro-electric-field-treated groups is strongly correlated with the enhanced removal of NH₄⁺-N, TP, and COD observed in the micro-electric field systems, suggesting a potential contribution of this phylum to these removal processes. Furthermore, the relative abundance of Pseudomonadota showed a significant positive correlation with temperature (r = 0.46, p < 0.05, Figure 3B). The relative abundance of Bacteroidota ranged from 4.64% to 16.11% across all samples. Compared to CK-CW, its relative abundance in EF-CW was significantly higher by 1.74%. Bacteroidota abundance was significantly positively correlated with NH₄⁺-N removal (r = 0.57, p < 0.01) and TP removal (r = 0.45, p < 0.05, Figure 3B). The micro-electric field treatment also significantly altered the relative abundances of Bacillota, Patescibacteria, and Halobacteriota. The micro-electric field also significantly altered the abundances of Bacillota, Patescibacteria, and Halobacteriota. The average relative abundance of Bacillota in EF-substrate was significantly higher than that in CK-substrate. Correlation analysis indicated that the abundance of Bacillota showed significantly positive correlations with both temperature (r = 0.47 p < 0.05) and COD (r = 0.65, p < 0.01) removal efficiency (Figure 3B). The subsequently enhanced proliferation and metabolic activity of this phylum may have directly contributed to the improvement in COD removal performance within the system. Furthermore, the abundance of Patescibacteria was also significantly positively correlated with NH₄⁺-N removal efficiency (r = 0.54, p < 0.05, Figure 3B).

Figure 3.

Micro-electric field enriches functional bacterial phyla correlated with pollutant removal

(A) Community composition of bacterial taxa at phylum level across the four treatment groups (CK-substrate, CK-CW, EF-substrate, and EF-CW).

(B) Spearman’s rank correlation heatmap between dominant bacterial phyla and key water parameters.

∗ indicates p < 0.05, ∗∗ indicates p < 0.01, ∗∗∗ indicates p < 0.001.

Micro-electric field increases network complexity and modularity

Analysis of microbial co-occurrence networks indicated that micro-electric field treatment significantly altered the topological structure of microbial interaction networks (Figure 4; Table S5). Compared to CK-substrate, the node count, edge count, and average degree in the EF-substrate network increased by 74.1%, 201.8%, and 73.3%, respectively. Relative to CK-CW, the EF-CW network showed further significant increases in node count, edge count, and average degree, with rises of 45.6%, 110.9%, and 44.9%, respectively. In the microbial co-occurrence network analysis, it is noteworthy that the EF-CW treatment group achieved the highest modularity index (0.589, Figure 4D; Table S5), indicating that its microbial community developed the clearest and most organized functional modular structure. The control groups CK-substrate and CK-CW comprised seven and nine modules, respectively, while the application of the micro-electric field reduced the number of modules to six in EF-substrate and to seven in EF-CW. Significantly, alongside this reduction in module count, EF-CW attained the highest modularity index of 0.589 among all groups. In our study, positive OTU correlations outnumbered negative ones (Figure 4). EF-substrate (46.13%, Figure 4C; Table S5) and EF-CW (47.96%, Figure 4D; Table S5) exhibited a higher proportion of negative interactions compared to CK-substrate (22.44%, Figure 4A; Table S5) and CK-CW (49.49%, Figure 4B; Table S5), which may indicate that micro-electric field enhances interspecies competition or promotes niche differentiation. The average clustering coefficient reflects the tightness of local cooperation in the microbial network. All four groups maintained a high level (>0.55), with CK-substrate showing the highest value (0.623, Figure 4C; Table S5). The average path length measures the global transmission efficiency of the network, and EF-CW exhibited the shortest average path length (3.286).

Figure 4.

Micro-electric field increases network complexity and modularity of bacterial co-occurrence networks

(A) CK-substrate.

(B) CK-CW.

(C) EF-substrate.

(D) EF-CW.

Edge colors represent positive (red) and negative (green) correlations with Spearman’s correlation coefficient |ρ| > 0.8 and p < 0.05. Topological properties of each network are summarized in Table S5. See also Table S5.

Micro-electric field enhances carbon and nitrogen metabolism-related functions

Functional prediction based on FAPROTAX analysis of the top ten metabolic functional groups revealed that micro-electric field treatment significantly altered the potential functional structure of the microbial community (Figure 5A). Specifically, compared to CK-CW, microbial functions related to carbon metabolism, including chemoheterotrophy, aerobic chemoheterotrophy, and hydrocarbon degradation, were significantly enhanced in EF-CW (Figures 5A; Table S6). Micro-electric field treatment significantly enhanced nitrogen metabolism-related functions alongside carbon metabolism (Figure 5B). Nitrate reduction and nitrogen respiration were significantly higher in the treatment groups (EF-CW and EF-substrate) than in the controls (CK-CW and CK-substrate), with denitrification showing the greatest increase in EF-substrate compared to CK-substrate (Table S6).

Figure 5.

Micro-electric field enhances carbon and nitrogen metabolism-related functions of the microbial community

(A) Relative abundance of the top ten metabolic functional groups predicted by FAPROTAX across the four treatment groups (CK-substrate, CK-CW, EF-substrate, and EF-CW).

(B) The relative abundance of carbon and nitrogen metabolism across the four groups.

Data are shown as mean ± SD (n = 5). Different lowercase letters above the bars indicate significant differences among groups (p < 0.05).

See also Table S6.

PLS-PM and Mantel test reveal bacterial community as the key driver of pollutant removal

To elucidate the driving forces of pollutant removal, this study employed partial least squares path modeling (PLS-PM) to analyze the effects of various factors (Figure 6A). The results of PLS-PM showed that the global goodness-of-fit (GOF), a comprehensive index for evaluating the predictive power of a PLS-PM model, was calculated to be 0.56. The results indicated that water environmental factors, bacterial community characteristics, and key metabolic functional groups all exerted significant direct positive effects on the removal of nitrogen and phosphorus pollutants (p < 0.001). Among these, bacterial community characteristics exhibited the strongest direct effect (path coefficient = 0.66), followed by water environmental factors (0.38), while the direct effect of key metabolic functional groups was negligible (0.03). Furthermore, water environmental factors had a significant direct effect on bacterial community characteristics (0.69) and on key metabolic functional groups (0.37). The Mantel test results further supported this conclusion, showing significant positive correlations between bacterial community characteristics and the removal rates of TP and NH₄⁺-N (p < 0.01, Figure 6B). Meanwhile, water environmental factors (EC, DO) were significantly positively correlated with bacterial community characteristics (p < 0.01).

Figure 6.

PLS-PM and Mantel test reveal bacterial community as the key driver of pollutant removal

(A) Path analysis of impacts on NH₄⁺-N, TP, and COD removal.

(B) Mantel test of the relationships of bacterial community, key bacterial taxa involved in carbon and nitrogen metabolism and environmental factors.

The red line indicated a significant positive correlation. ∗ indicates p < 0.05, ∗∗ indicates p < 0.01, ∗∗∗ indicates p < 0.001.

Discussion

The present results demonstrate that micro-electric field treatment substantially enhances the simultaneous removal of TP, NH₄⁺-N, and COD in CWs. The high TP removal efficiency may be due to the enrichment of phosphorus-accumulating organisms, the substrate’s improved adsorption capacity, or the electric field’s enhanced phosphate precipitation.⁵ This improvement may be due to the electric field’s ability to optimize aerobic microenvironments, stimulate nitrifying bacterial activity, or enhance the coupled nitrification-denitrification process.^9,10 This enhancement may be attributed to the promotion of microbial metabolism, increased dehydrogenase activity or optimized electron transfer processes by the electric field, thereby accelerating the biodegradation of organic matter.¹¹ In summary, the micro-electric field enhanced the synergistic removal of COD, NH₄⁺-N, and TP in the CWs, demonstrating promising application potential and technical advantages. In addition to these direct effects on pollutant removal, the micro-electric field also significantly increased bacterial species richness. Importantly, this effect may not be a direct driver of diversity changes; rather, the electric field may indirectly influence the microbial community by altering water parameters such as DO and pH. This is consistent with previous findings that higher DO levels are associated with greater diversity of functional taxa involved in nitrogen and phosphorus removal, which in turn correlates with higher removal efficiencies of TP and NH₄⁺-N.¹² Indeed, pH and DO have been established as key environmental factors shaping microbial community structure,^13,14 supporting our observation that the micro-electric field modulates these parameters to enhance community richness and diversity. Collectively, these results suggest that the micro-electric field not only directly promotes pollutant degradation through electrochemical and biological mechanisms but also indirectly improves system performance by reshaping the microbial community via environmental regulation. This dual action underscores the promising application potential and technical advantages of micro-electric-field-enhanced CWs for wastewater treatment.

The observed dominant phyla are consistent with previous findings in domestic wastewater treatment studies, and these microbial groups are known to play important roles in water purification.^2,15 The significant enrichment of Pseudomonadota is primarily attributed to its numerous functional genera involved in denitrification, phosphorus removal, and organic matter degradation.^3,16 Its positive correlation with temperature suggests that suitable temperatures enhance the metabolic activity of these functional genera.^17,18 Bacteroidota, known for its capacity to degrade complex organic substrates such as starch and cellulose^19,20 and its involvement in electron transfer,²¹ likely contributes to enhanced COD removal. Its involvement in anaerobic ammonium oxidation²² further explains its positive correlation with NH₄⁺-N removal. Bacillota, a ubiquitous group driving organic matter decomposition in aquatic ecosystems,²³ may have been promoted by micro-electric-field-mediated temperature regulation,¹⁷ thereby improving COD removal. The enrichment of Patescibacteria may enhance nitrogen removal by augmenting nitrogen transformation pathways.^1,24 Collectively, these findings indicate that the micro-electric field reshapes the bacterial community at the phylum level, selectively enriching functional taxa that are closely linked to pollutant degradation, and thereby contributes to the overall improvement in treatment performance.

The increased node/edge counts and average degree demonstrate that micro-electric field treatment enhances interactions among key microbial taxa and significantly raises the overall connectivity complexity of the community. The difference in network complexity may reflect the differential influence of oxygen availability on microbial community assembly, as aerobic conditions promote niche differentiation and metabolic diversification, whereas anoxia reduces microbial diversity.²⁵ This is consistent with our finding that EF-CW exhibited the highest DO level and the greatest microbial diversity. The highest modularity index in EF-CW, alongside the reduction in module count from 9 to 7, indicates that the micro-electric field did not simply simplify the community structure but drove a “functional reorganization” of the community. This phenomenon may be attributed to alterations in the aqueous micro-environment induced by the micro-electric field, such as increased availability of DO or reactive hydrogen through electrolysis.¹¹ Furthermore, environmental heterogeneity can promote modularity and niche differentiation in microbial communities,²⁶ providing a plausible mechanism for the observed simultaneous reduction in module number and enhancement in modularity under micro-electric field treatment. In our study, positive OTU correlations outnumbered negative ones (Figure 4), implying a general tendency toward cooperative or co-occurring relationships among the bacterial communities.²⁷ Positive associations may indicate cooperation, co-occurrence, spatial overlap, or niche sharing among microorganisms, while negative associations often suggest growth inhibition, antagonistic interactions, or niche differentiation.^28,29 The higher proportion of negative interactions in EF-substrate and EF-CW may indicate that micro-electric field enhances interspecies competition or promotes niche differentiation. Competition among bacteria for resources or niches can accelerate carbon and nitrogen conversion processes, thereby improving microbial ecological functions and system stability.³⁰ The high clustering and short path length are characteristic of networks with efficient local and global connectivity, which may facilitate rapid information and material exchange among microorganisms and could be conducive to pollutant degradation.²⁹ Collectively, these topological changes suggest that the micro-electric field reshapes microbial co-occurrence networks toward a more complex, modular, and competitive architecture, which may enhance system functional resilience and pollutant removal efficiency.

The observed functional shift aligns with the ecological roles of key bacterial phyla such as Bacteroidota and Pseudomonadota (Figure 2A), which are known to hydrolyze starch and cellulose into amino acids and organic acids, playing a central role in organic matter degradation.³¹ These predicted results are consistent with the measured performance: micro-electric field treatment significantly enhanced COD removal efficiency (Figure 1), further supporting the biological plausibility of the functional predictions. The enhanced nitrogen metabolism functions suggest that the electric field may have the potential to directionally intensify microbial denitrification processes. This functional enhancement is likely linked to the enrichment of specific microbial groups. The increased abundance of the dominant Patescibacteria, which is involved in nitrite reduction,³² could directly facilitate this step. Furthermore, recognized denitrifying phyla such as Bacteroidota and Pseudomonadota may also be enriched.³³ It is therefore suggested that the micro-electric field, by creating a conducive electrochemical environment, may have selectively promoted the synergy among these key functional bacteria, thereby driving efficient biological nitrogen removal.

PLS-PM analysis yielded a GOF of 0.56, exceeding the threshold of 0.36 that indicates a strong model fit,³⁴ thus confirming the model’s high explanatory power for the observed data. The PLS-PM results indicate that the micro-electric field primarily optimizes water environmental parameters (e.g., DO, pH) to reshape the bacterial community structure, thereby indirectly driving pollutant removal; the independent contribution of single metabolic functional groups is almost negligible, and their role mainly depends on the support of the community structure. The Mantel test results are highly consistent with the PLS-PM path coefficients, further confirming the key role of bacterial community characteristics in pollutant removal. As noted above, water environmental factors (path coefficient = 0.38) also exhibited a significant direct effect. This might be attributed to the externally applied micro-electric field, which generated a stable weak DC electric field between the electrodes and directly promoted electrochemical reactions in the aqueous phase. Specifically, the electric field can induce water electrolysis to produce trace oxygen or reactive oxygen species, thereby enhancing the indirect oxidation of organic matter. Meanwhile, cathodic reduction reactions may facilitate nitrate reduction or ferrous/iron cycling, affecting phosphorus precipitation and adsorption.^6,9 In addition, the electric field can accelerate extracellular electron transfer, increase dehydrogenase activity, and promote organic degradation and nitrogen conversion.¹¹ Furthermore, the volcanic rock substrate used in this study possesses a porous structure and good adsorption capacity. The micro-electric field can alter the charge distribution on the substrate surface, enhancing electrostatic adsorption of negatively charged phosphate ions and positively charged ammonium ions. Simultaneously, the field may promote the release of metal ions (e.g., Fe, Al) from the substrate, further removing phosphorus through precipitation or complexation. Integrating the PLS-PM and Mantel test results, the dominant mechanism of micro-electric field enhancement might be reshaping the bacterial community structure by regulating water environmental factors, thereby endowing the system with higher metabolic potential and operational stability. This provides a new quantitative perspective for understanding how micro-electric fields enhance wastewater treatment performance through ecological network regulation.

Limitations of the study

From an application perspective, graphite electrodes are relatively inexpensive and durable, the DC power supply is a general-purpose device, and no external carbon source is required, indicating good cost-effectiveness potential for this technology. When scaling up, it is advisable to maintain a similar electric field intensity (0.25 V cm⁻¹) while adjusting electrode spacing and voltage accordingly. However, it should be noted that this study was conducted at laboratory scale, and pilot-scale or full-scale validation is still needed. In addition, the sampling frequency (once every 5 days) was designed to capture the performance at the end of each batch cycle, and therefore short-term fluctuations in water quality and microbial activity within the cycle could not be assessed. Detailed characterization of the electrode materials, energy consumption analysis, and nutrient mass balance accounting were not performed. Furthermore, FAPROTAX functional predictions based on 16S rRNA are indirect and lack direct validation by enzyme activity assays or meta-transcriptomics; the formal experiment lasted only 75 days, and long-term electrode fouling, degradation, and inter-annual microbial community dynamics were not assessed. It is recommended that future pilot-scale long-term studies be carried out, focusing on electric field uniformity, electrode arrangement optimization, and hydraulic retention time effects, along with detailed techno-economic analysis.

Resource availability

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Yungen Liu (henryliu1008@163.com).

Materials availability

This study did not generate new unique reagents.

Data and code availability

• •Data reported in this study will be shared by the lead contact upon request.
• •This study does not report original code.
• •Any additional information required to reanalyze the data reported in this study is available from the lead contact upon request.

Acknowledgments

This study was supported by the Key Project of Yunnan Province Science and Technology Department (202502AE090046 and 202301AS070042) and the Open Fund of Yunnan Provincial First-Class Discipline for Soil and Water Conservation and Desertification Combating (SBK20240006).

Author contributions

Conceptualization, visualization, formal analysis, writing – original draft, and writing – review and editing, Q.W.; conceptualization, project administration, supervision, and writing – review and editing, Y.L.; writing – review and editing, J.Z.; investigation, X.L.; investigation, H.Z.; project administration, supervision, and writing – review and editing, S.Y.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Biological samples
Zizania latifolia	This paper	N/A
Software and algorithms
R version 4.4.2	R Foundation	https://www.r-project.org/

Method details

Experimental system design

In our study, we used Zizania latifolia as the test plant, selecting robust seedlings with uniform growth for the experiment. Four treatment groups were designed: constructed wetland substrate control group (CK-Substrate), micro-electric field-enhanced constructed wetland substrate group (EF-Substrate), constructed wetland treatment group (CK-CW), and micro-electric field-enhanced constructed wetland group (EF-CW, Figure S1). Each system consisted of a cylindrical container with a diameter of 30 cm and a height of 60 cm, filled with volcanic rock (3–4 mm in diameter) as the substrate to a height of 55 cm. All reactors were arranged in a completely randomized layout within the greenhouse. The culture water was sourced from domestic wastewater collected at the wastewater treatment plant of Southwest Forestry University. The influent wastewater characteristics were as follows: COD 354.75–434.75 mg/L, NH₄⁺-N 47.4–52.94 mg/L, and TP 2.72–5.77 mg/L. The hydraulic retention time was uniformly set at 120 h for all groups, with five replicates per treatment. The hydraulic retention time (HRT) of 120 h was determined based on preliminary experiments from our research group (unpublished data). In the CK-CW and EF-CW groups, five healthy Zizania latifolia seedlings were transplanted into each unit. For the EF-Substrate and EF-CW groups, two graphite electrode plates were inserted in parallel into the reactor with a spacing of 20 cm. The electrodes were connected to an external direct current (DC) power supply (wanptek GPS305D) to generate a stable weak DC electric field (the power supply was set to output a constant voltage of 5 V). The actual voltage across the electrodes was monitored daily using a digital multimeter, and remained stable at 5.0 ± 0.1 V, confirming that the electric field intensity was maintained at approximately 0.25 V cm⁻¹. The corresponding CK-Substrate and CK-CW groups were also equipped with graphite electrodes, but they were not connected to the external DC power supply (electrodes were placed only). Apart from this, all other cultivation conditions were the same as those for the EF-Substrate and EF-CW groups. When the system was running stably for 30days, we started to conduct experiments.

Water sample collection and determination

All systems were operated in batch tests. During the experimental period, influent and effluent water samples from each system were collected every 5 days, and a total of 15 batch cycles were completed. covering a total of 15 sampling cycles. Chemical oxygen demand (COD), ammonium nitrogen (NH₄⁺-N), and total phosphorus (TP) were determined according to standard methods. To ensure measurement accuracy, each water sample was analyzed in triplicate. Dissolved oxygen (DO), electrical conductivity (EC), pH, and temperature were measured on-site using a multiparameter water quality analyzer.

Determination of bacterial community

Representative substrate samples were collected using the five-point sampling method. After rinsing with deionized water, the samples were centrifuged at 4°C and 8000 × g for 10 min, and the resulting microbial pellet was collected for subsequent DNA extraction. DNA was extracted using the EZNA Soil DNA Kit, and its quality was verified via 1.0% agarose gel electrophoresis. The V3–V4 region of the bacterial 16S rRNA gene was amplified via PCR using primers 338F/806R. The resulting PCR products were purified and subsequently subjected to paired-end 300 bp sequencing on the Illumina MiSeq platform.

Quantification and statistical analysis

α-diversity indices were calculated using the “vegan” package in R. Principal coordinate analysis (PCoA) based on Bray-Curtis distance was performed to visualize β-diversity. Spearman correlation analysis, implemented via the “psych” package, was used to examine the relationships between dominant bacterial taxa (mean relative abundance >1%), α-diversity indices, and environmental parameters. Prior to performing one-way analysis of variance (ANOVA), the normality and homogeneity of variance assumptions were assessed using the Shapiro-Wilk test (Table S1) and Levene’s test (Table S2), respectively. After confirming that these assumptions were met, differences among groups were assessed by one-way analysis of variance (ANOVA) followed by Duncan’s multiple range test, with the significance level set at p < 0.05. Microbial co-occurrence networks were constructed based on Spearman correlations. Prior to network construction, OTUs with detection frequencies <80% were excluded. Edges with correlation coefficients |ρ| > 0.8 and false discovery rate (FDR)-adjusted p < 0.05 were retained.³⁵ To assess the potential impact of multicollinearity on model results, we calculated the variance inflation factor (VIF) for each observed variable (Table S3). The networks were visualized using Cytoscape. Functional annotation of bacterial taxa was performed using FAPROTAX to identify the top 10 most abundant metabolic functional groups. Mantel tests were applied to analyze the influences of dominant taxa and metabolic functional groups on pollutant-removal efficiency and environmental factors. PLS-PM with the “plspm” package was used to perform path analysis based on NH₄⁺-N and TP removal. All statistical details are provided in the figure legends, figures, and Results section. Data are presented as mean ± standard deviation (SD) for at least three biological replicates (n = 5). The center value represents the mean, and dispersion is expressed as SD. The statistical significance was defined as p < 0.05 unless otherwise specified.