Effects of Pollutants in Urban Wastewater on Rhizoplane Microbial Communities in Constructed Wetlands: Resistance and Resilience of Macrophyte-Associated Microbiomes

Abstract

The impact of pollutants in urban wastewater on Constructed Wetlands (CWs) rhizoplane microbial communities remains quite understudied. Our study explores how civil wastewater influences the structure and ecological stability of rhizoplane microbial communities associated with three macrophytes: Nerium oleander L., Arundo donax L., and Juncus conglomeratus L. in simulated conditions as in the case of CWs. Therefore, a pot experiment was set up, using wastewater repeated exposure of the three macrophytes, to assess the microbial (bacteria and fungi) resistance and resilience by means of next-generation sequencing. The results showed that all three macrophytes contributed to pollutant removal; however, the effects on microbial communities were taxon-specific. In general, the rhizobacterial community exhibited moderate resilience and low resistance to wastewater, indicating a partial recovery post-disturbance. The fungal community showed high resistance (ResI = 0.99), in contrast with limited resilience (RI < 1), suggesting a stable but less dynamic response to the wastewater exposure. Effluent repeated addition positively influenced the relative abundance of certain bacteria taxonomical groups, specifically Firmicutes and Actinobacteria, but also of some fungal taxa. Our findings underscore the key role of microbial communities in CWs, where complementary resistance and resilience strategies contribute to system stability, plant health, and pollutant attenuation.

1. Introduction

Biodiversity is not only a hallmark of undisturbed natural ecosystems—it is also a crucial value in engineered environments such as Constructed Wetlands (CWs), traditionally designed to treat wastewater and manage stormwater. They are artificial ecosystems recognized for their potential to support a wide array of plant and animal life. Far from being simple infrastructures, CWs can serve as vibrant ecological niches where species interact, adapt, and thrive. Considering biodiversity as an integral component of these systems enhances their resilience, improves their ecological functions, and aligns with broader goals of sustainability and environmental stewardship. In this light, biodiversity is not an optional add-on, but a key asset even in those landscapes shaped by humans [1,2]. Since the beginning, research has emphasized their importance at the ecological level, and recent studies indicate the need for deeper insights into microbial dynamics and interactions [1,3]. Central to the success of CWs is the synergy between plant cycle, microbial communities, and duration of interaction with the substrate. In fact, root secretions directly influence the soil particles that closely adhere to the root surface of a plant and the associated soil microorganisms known as the root microbiome. This interaction takes place in the thin region of the soil attached to the plant root named rhizoplane, which is an extraordinarily complex microhabitat in which plant root, soil, and microorganisms interact [4]. It has been estimated that this root nutrient-rich zone contains around 1 × 10⁵ microorganism species per gram of root, including bacteria, fungi, protists, nematodes, and invertebrates [4]. In fact, in addition to their effective role on pollutants degradation, microorganisms, dwelling in the rhizoplane, have both direct and indirect effects on the plant itself, such as improving its nutrition, modifying the root system, and stimulating the production, for instance, of indoleacetic acid (IAA), which, in turn, promotes the growth of the roots and leads to a decrease in the ethylene levels, a hormone able to induce plant stress [5]. Although, research has emphasized, since the beginning, the relevance of microbial communities, recent studies suggest the need for deeper insights into their dynamics and interactions [2,3]. The rhizoplane microbial biodiversity strongly contributes to the resistance and resilience of the plants used in the setting up of CWs, where resistance refers to the ability of an ecosystem to withstand disturbances without undergoing significant change in its structure, composition, or function, whilst resilience is the capacity of an ecosystem to recover after a disturbance or stressor, returning to its original state or functioning level. Together, resistance and resilience, enhanced by biodiversity and functional redundancy, are key attributes of ecosystem stability and, in the case of CWs, of their water reclamation efficiency and efficacy. CW is a complex ecosystem, including also the abiotic components such as sand, stones, and the other matrices employed for its bed construction, also able to counteract the antibiotic resistance spread [6,7]. In fact, the heterogeneous layers—substrate, biofilm, water flow, and plant roots—create several different aerobic and anaerobic microenvironments that sustain rich microbial communities. This diversity limits the dominance of resistant strains and reduces horizontal gene transfer as in the case of antibiotic resistance. In fact, substrates and biofilms adsorb antibiotics and trap resistant bacteria, lowering selective pressure. Rhizosphere oxygenation and root exudates stimulate microbial degradation of antibiotic drugs. Physical–chemical gradients (oxygen, redox, pH, etc.) further destabilize antibiotic residues other than antibiotic resistance in both bacteria and their respective genes. CWs are able to assist transformation of organic/inorganic pollutants by plants thanks to the activation of enzymes and/or molecule production, promoting either desorption and biodegradation or their take-up [8]. Among biotic components, only plants can be chosen for CW set up. In fact, although microbial community can be augmented, its dynamics are the results of ecological relationships and diverse conditions, including chemical–physical characteristics of the pollutants, climatic factors, and biotic interaction. Among macrophyte species as Phragmites australis Cav. (Steud), Arundo donax L. and Juncus conglomeratus L. are often used in addition to Prunus laurocerasus L. and Nerium oleander L. Several publications, focused on the study of the CW microbiome, revealed the presence of OTU belonging to different phyla such as Proteobacteria, Firmicutes, Actinobacteria, and Bacteroidetes [9]. In particular, Actinobacteria, Firmicutes, and Bacteroidetes were shown to be involved in the removal/degradation of antibiotics [10], whilst Planctomycetes were often instrumental in nitrogen uptake and removal [11]. However, knowledge about the dynamics, resistance, and resilience of the microbial communities, associated with the CWs in response to either the purification of civil wastewater or the possibility that they are species specific to some plants, is still very limited. Advancing our understanding of microbial diversity in CWs offers significant benefits for both wastewater treatment and ecosystem sustainability. A deeper insight into the structure and function of microbial communities enables the optimization of treatment processes by enhancing the removal of nutrients, organic pollutants, heavy metals, and emerging contaminants as the drugs commonly used by humans and for animals. This knowledge allows for more informed and customized system design—selecting appropriate substrates, plant species, and flow regimes that support beneficial microbial communities. It also strengthens the resilience and stability of CWs, helping them withstand environmental stressors and recover more effectively after disturbances. Moreover, improved microbial management contributes to better water reuse, reduced maintenance costs, and lower energy consumption. From a broader perspective, this progress supports innovation in nature-based solutions and aligns with global environmental goals by promoting sustainable wastewater treatment and reinforcing biodiversity conservation. In essence, microbial diversity is not only a biological asset but a functional cornerstone for advancing the efficiency, adaptability, and ecological value of the CWs.

For all these reasons, the interactions of plant–microorganisms are fundamental for the survival of the plants and the associated root ecosystems. Considering the central role of plant–microbe interactions in maintaining effective CW functionality, our study hypothesized that the composition and stability of the rhizoplane microbial communities differ among macrophyte species exposed to urban wastewater. Specifically, we expected that these communities would exhibit distinct patterns of resistance and resilience to pollutant stress, measurable through changes in biodiversity indices such as the Shannon diversity index. Moreover, an index related to the relative change of microbial community was also developed, as well as the value of resistance and resilience based on the microbial community variation in the experimental groups in respect to the control.

2. Materials and Methods

2.1. Vertical Submerged Flow System (VSFS-CW) Set Up

In our study, a pot, simulating a VSFS-CW, was set up for the civil wastewater remediation, and to evaluate effect of the wastewater on the rhizosphere microbial communities of the three different macrophytes species (Figure 1—N. oleander, A. donax, and J. conglomeratus) used in the experiment. On the basis of the urban wastewater-pollution load, pots of 3.0 L volume were prepared, stratifying different materials and matrices as follows: 500 mL crushed stone (layer 1), 500 mL of medium-grained sand (layer 2), 30.0 mL of biochar (layer 3), 16.6 mL of carrier (layer 4), and 1000 mL of medium-grained sand (layer 5).

The macrophytes species employed for studying microbial community variations, due to the civil wastewater exposure, are those previously mentioned in the above chapter (2.1). These species belong to the class of emerging rooted macrophytes, usually found in swamps and on the shores of lakes [12], and commonly adopted in Europe for the set up/realization of CWs. A. donax, commonly known as Giant Reed, is a perennial herbaceous plant native from the Mediterranean basin to the Middle East up to India, but currently it can be found both planted and naturalized in temperate and subtropical regions of both hemispheres. It is a non-food plant that has a high biomass yield, tolerates different kind of pollutants [13,14], and is characterized by a great adaptability to different soil and climate conditions [15,16]. N. oleander, commonly known as Oleander, is an evergreen shrub belonging to the Apocinaceous family, and it is considered a useful bioindicator of Zn and Cu, two micronutrients often present in different types of soils [17]. J. conglomeratus, commonly known as Compact Rush, is recognized for its ability to stabilize HMs. These species can also grow under conditions of water saturation; moreover, it shows high amounts of Fe, Cu, and Pb associated with its roots; nevertheless, Compact Rush managed to keep the levels of these metals below the toxicity limits in their sprouts.

2.2. Set Up of the Experimental Plan

The set up of the experimental plan derived by our previous experience and pilot experiments was established with the aim to define the CW stratification, plants, and time collection point optimized with respect to the water reclamation performances [18]. Therefore, 27 pots were prepared as described at Section 2.1 and underwent controlled greenhouse condition (16 h of light and 8 h of dark at 25 °C). The pots were subdivided into three experimental groups (Figure 2): the first group was considered as control (CNT) and treated with spring water; the second group was treated for 10 days with civil wastewater (WW), collected after the first step of sedimentation in wastewater treatment plant, located in the Avellino province (IT); the third one was treated as the previous one (WW) but followed by repeated exposure of spring water (WW + W) for additional five days. Each group included three biological replicates for each of the three macrophytes species employed in the experiment. The plant specimens used in our experiments were obtained by vegetative reproduction to limit genetic diversity, and they were characterized by a height of about 50 cm reached after one year and half of growth in a greenhouse.

Five hundred mL of spring water (for CNT; for the last 5 days of the WW + W experimental group) or wastewater (for WW; for the first 10 days of WW + W experimental group) were added to the pots. After 24 h, the effluent was collected in the Beckers and manually reloaded. After 48 h from the beginning of the experimentation, the effluents were substituted with new spring water or wastewater, respectively. This was repeated all along the experimental time. In the case of WW + W experimental group, after 10 days of wastewater repeated exposure, spring water was added and reloaded for 5 days. This time period was established based on our previous experiment [19] where the microbial communities exhibited stability after 5 days. Every 48 h, the percolated waters through the pots (a volume comprises between 250–300 mL) were collected in the flowerpot holders and then analyzed to estimate the contents of the pollutants. The rhizo-microbial communities were investigated at the end of the pot experiment.

2.3. Wastewater Physical–Chemical Characterization

The wastewater influent was sourced from the primary sedimentation stage of a biological wastewater treatment plant located in the Avellino province, Italy. Both the influent and the effluents from the CWs underwent physical–chemical characterization. To assess key parameters such as Chemical Oxygen Demand (COD), nitrates, nitrites, ammonia, phosphate, and chloride absorbance, measurements were conducted using a HACH DR3900 spectrophotometer (Hach, Milano, Italy), equipped with sample standards and appropriate accessories, including cuvettes. Meanwhile, metals by means and antibiotics concentrations were estimated by a ICP-OES (Optima 7000 DV, PerkinElmer, Milan, Italy) or HPLC-ESI MS/MS (7T SolariX XR FT-ICR-MS system, Bruker Daltonics, Bremen, Germany), respectively. Each parameter abovementioned was estimated in three technical replicates for each biological replicate (5) obtained by each experimental group.

2.3.1. Chemical Oxygen Demand (COD)

The COD of both influent and effluent samples was determined using the Hach (Milan, Italy) cuvette system. Briefly, 2.0 mL of each sample was pipetted into a manufacturer-supplied cuvette and digested at 148 °C for 15 min using the HT 200 S thermostat (Hach). The COD values were then calculated by measuring the absorbance of the treated samples.

2.3.2. Nitrate (NO₃⁻), Nitrite (NO₂⁻), and Ammonia (NH₄⁺)

Nitrate concentrations were determined using the Hach cuvette LCK 339. For each measurement, 1.0 mL of the influent or effluent sample was pipetted into the cuvette, followed by the addition of 0.2 mL of solution provided by the manufacturer. The nitrate values were then calculated based on absorbance readings.

For nitrite determination, 2.0 mL of each sample was separately pipetted into the Hach cuvette LCK 341 and processed according to the manufacturer’s instructions. The NO₂⁻ concentration was then calculated on the basis of the absorbance measurement.

Ammonia values were assessed using the Hach cuvette LCK 303. Two hundred microliters of each sample of either influent or effluent was pipetted separately into each cuvette and processed according to the manufacturer’s procedures. The ammonia concentration was determined by measuring the absorbance of the treated sample.

2.3.3. Chloride (Cl⁻)

The chloride content in both influent and effluent samples was measured using the Hach cuvette system (LCK 311), following the manufacturer’s procedures. Briefly, 1.0 mL of the sample was pipetted into the cuvette, followed by the addition of 200 μL of the reagent A provided by the manufacturer. The mixture was then allowed to react for the prescribed time. The chloride concentration was determined by measuring the absorbance of the treated sample using the HACH DR3900 spectrophotometer.

2.4. Microbial Community Composition

2.4.1. Wastewater DNA Purification

Three aliquots of the wastewater were analyzed to determine the microbial community. In particular, 450 mL of the influents were filtered through sterile polycarbonate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Milan-Italy). Then, the filters were transferred in 15 mL Falcon (Milan, Italy) tubes, containing 10 mL of saline solution (NaCl 0.9%), shaken for 10 min at room temperature to release, and then collect, both bacteria and fungi. The concentration of microorganisms present in the samples was estimated spectrophotometrically at OD₆₀₀ nm considering that one OD₆₀₀ corresponds to roughly 10⁸ cells/mL in the case of bacteria and 10⁷ cells/mL for the fungi. Subsequently, DNA extraction was carried out from 0.1 OD₆₀₀ (corresponding to roughly 10⁷ cells/mL of bacteria and roughly 10⁶ cells/mL of fungi) of each sample, in triplicate, using Extract-N-Amp™ Plant PCR kit (Merk Life Science, Milan-Italy following the manufacturer’s instructions. In particular, 10 μL of 0.1 OD₆₀₀ solution was added to 90 μL of extraction buffer and incubated at 94° C for 20 min, then, 100 μL of the dilution buffer (included in the kit) was added and mixed by pipetting.

2.4.2. Rhizo-Microbial Collection and DNA Purification

At the end of the repeated exposure, the microbial communities were collected from the rhizoplanes of the three macrophyte plant groups. First, 2.5 g of roots (separately from each plant of each treatment) were weighed and placed in 50 mL Falcon tubes containing 25 mL of sterile physiological solution (NaCl 9.0 g L⁻¹), then the tubes, containing the roots resuspended in the physiological solution, were stirred at 25 °C for 30 min at 200 rpm (rotation per minute). After stirring, the roots were removed using sterile tweezers, and the physiological solutions were centrifuged (Eppendorf, Milan, Italy) for 10 min at 600 rpm. A total of 12 mL of supernatant was carefully pipetted and transferred to new sterile 15 mL Falcon tubes and centrifuged again for 15 min at 3000× g. At the end of centrifugation, the supernatant was discarded, and the microbial pellet resuspended in 10 mL of sterile physiological solution. After a centrifugation for 15 min at 3000× g, the microbial pellet was resuspended in 2.0 mL of a 20% glycerol–water solution, separated into aliquots, frozen in liquid nitrogen, and then transferred to an ultra-freezer at −80 °C and maintained there for further experimental activities.

An aliquot of microbial pellets was thawed and then centrifugated for 15 min at 3000× g. After removing the supernatant, the pellet was resuspended in 200 µL of sterile physiological solution (NaCl 0.9%); three separated aliquots of each root macrophyte species were combined at this step and the concentration of microorganisms in the sample was estimated (OD) spectrophotometrically at 600 nm. Subsequently, 10 μL of 0.1 OD₆₀₀ solution was added to 90 μL of extraction buffer and incubated at 94° C for 20 min. Then, 100 μL of dilution buffer (included in the kit) was added and mixed by pipetting, in triplicate, using Extract-N-Amp™ Plant PCR kit (Merk Life Science, Milan, Italy) following the manufacturer’s instructions.

2.4.3. Amplicon Library Preparation and Next-Generation Sequencing (NGS)

The extracted DNA samples, obtained from the wastewater influent and those purified from the plant rhizoplanes, were used to study the microbial communities. In particular, the hypervariable region V3-V4 of the bacterial 16S rRNA gene was amplified, in triplicate, with 341F (5′-CCTACGGGRSGCAGCAG-3′)- and 909R (5′-TTTCAGYCTTGCGRCCGTAC-3′)-specific primers. In particular, 4.0 μL of DNA was added to the PCR mix composed by 10 μL of RedTaq mastermix, primer forward and reverse at the final concentration of 0.25 μM, and sterile water up to the volume of 20 μL. The thermal profile consisted of an initial step of denaturation at 95 °C for 3 min, followed by 40 cycles of denaturation at 95 °C for 1 min, annealing at 58 °C for 1 min, and elongation at 72 °C for 1 min, with an additional final elongation step at 72 °C for 5 min. The fungal ITS2 region was amplified using the ITS3 (5′-GCATCGATGAAGAACGCAGC) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′) primers, adopting the same thermal profile as described above, but with an annealing temperature of 55 °C. The amplicon libraries were set up and sequenced by means of NGS approach (Genomics4life company, Baronissi, SA, Italy).

2.5. Sequences Analysis

Illumina sequence data were sorted based on unique barcodes and quality-controlled using the Quantitative Insights into Microbial Ecology 2 (Qiime2, version 2019.4) with plugins-demux2 dada2 [20] and feature-table [21]. Alpha- and beta-diversity analyses were performed using plugins alignment diversity3 [22]. For taxonomic analysis, a pre-trained Naive Bayes classifier based on the SILVA 138 (Operational Taxonomic Units) OTUs database, in the case of 16S rDNA (trimmed to include the V3-V4 region of 16S rRNA gene bound by the 341F/909R primer pair), was used. Meanwhile, the classifier for fungi ITS2 DNA sequences was pre-trained on the UNITE database (version 7–99%) and applied to paired-end sequence reads to generate taxonomy tables. Taxonomic and compositional analyses were conducted using plugins feature-classifier5 [23], taxa, and composition [24].

2.6. Microbiome Diversity Indices

The raw data were used for preliminary α-diversity analyses, employing observed OTUs and Shannon metrics through the Qiime2 α-diversity plugins [25]. The samples were rarefied to a total frequency of 6820 for 16S rDNA and 59,788 for ITS before calculating diversity metrics. Differences between experimental groups were evaluated using the Kruskal–Wallis test. All results are presented as means ± standard deviations from the triplicate analyses. A permutation-based Analysis of Variance (PERANOVA) was conducted, with p-values determined using 999 permutations.

In order to compare the results of the rhizospheric biodiversity of the three macrophyte plant species, the Treatment Normalization Index (TNI) was calculated as the ratio between each OTU identified in the different experimental group (WW, WW + W) and the same identified in the CNT. In this perspective, the 0 value indicates that the specific OTU is present in the CNT, but absent in the WW or WW + W experimental group. Otherwise, ND represents the condition where the OTU is present in the WW or WW + W experimental group but absent in the CNT.

The microbial community resistance (referring to the ability of a system to remain unchanged when subjected to a disturbance) was evaluated following methods illustrated below:

Comparison before and after a disturbance quantified using a resistance index, such as

Resistance index (ResI) = 1 − D

where D is the Bray–Curtis (BC) dissimilarity metric.

$$BC={\displaystyle \frac{\left\{{\sum }_{\left\{\mathrm{i}=1\right\}}^{\left\{n\right\}}\left|{\mathrm{c}}_{\mathrm{i}}-{\mathrm{d}}_{\mathrm{i}}\right|\right\}}{\left\{{\sum }_{\left\{\mathrm{i}=1\right\}}^{\left\{n\right\}}\left({\mathrm{c}}_{\mathrm{i}}+{\mathrm{d}}_{\mathrm{i}}\right)\right\}}}$$

where

• c = Pre-treatment (control);
• d = Post-treatment (after disturbance);
• ResI = 1 → no change = high resistance;
• ResI = 0 → complete change = no resistance.

The microbial resilience was assessed by applying a disturbance and tracking the system’s recovery over time. A quantitative way to express recovery was the Resilience Index (RI).

$$\mathrm{R}\mathrm{e}\mathrm{s}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e} \mathrm{I}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{x}={\displaystyle \frac{\left\{{\mathrm{D}}_{\left\{\mathrm{C}\mathrm{D}\right\}}-{\mathrm{D}}_{\left\{\mathrm{C}\mathrm{R}\right\}}\right\}}{\left\{{\mathrm{D}}_{\left\{\mathrm{C}\mathrm{D}\right\}}\right\}}}$$

where

• D_$\left\{\mathrm{C}\mathrm{D}\right\}$: dissimilarity between pre- and post-treatment;
• $\mathrm{D}\_\left\{\mathrm{C}\mathrm{R}\right\}$: dissimilarity between pre- and recovered;
• C = pre-treatment (CNT);
• D (disturbed) = immediately post-treatment (WW);
• R (recovered) = after recovery period (WW + W);
• RI = 1 indicates full recovery;
• RI < 1 indicates partial recovery;
• RI > 1 indicates overcompensation (a new stable state).

3. Results

3.1. Wastewater Treatment Efficiency

The influents and effluents collected in flowerpot holders of the stratified VSFS pots, planted with the three macrophytes, were characterized at the start of the experiment (T0) and after each 48 h of repeated exposure to estimate the content of pollutants still present in the wastewater. The T0 concentrations of the heavy metals and phosphate were below the limit of detection (10 μg L⁻¹) as well as in the case of pharmaceutical compounds (e.g., antibiotics). After repeated exposure, NH₃ and COD contents were significantly reduced (up to roughly 90% in the case of COD). Ammonia was totally oxidized to NO₃⁻; in fact, it increased in all the effluents (

Table 1. Average pollutant contents before (T0) and after each 48 h of wastewater repeated exposure in the VSFS pots planted with different macrophyte species. The * indicate the statistically significant differences (p < 0.05, Kruskal–Wallis; post hoc Nemenyi) between the plant species considered and the T0 influent (the same for all the experimental groups).

Pollutants (mg L−1)	T0	After 48 h
	Influent	A. donax (n = 3)	N. oleander (n = 3)	J. conglomeratus (n = 3)
COD	1353 ± 112	186 * ± 7	190 * ± 3	177 * ± 3
N-NH4	10.5 ± 0.8	1.3 * ± 0.2	ND	ND
N-NO3	3.5 ± 1.6	18.5 *± 0.2	16.0 *± 0.1	15.3 * ± 0.1
N-NO2	1.2 ± 0.1	ND	ND	ND
Cl−	283 ± 24	211 ± 22	193 ± 15	123 * ± 9

) analyzed. When the chlorides were considered, J. conglomeratus showed the best capability to reduce them significantly (p < 0.05), and contrarily, although a considerable reduction was observed for the other macrophytes, and in particular for N. oleander, it was not statistically significant.

3.2. Microbiome Analysis

3.2.1. Wastewater Microbiome Characterization

The bacterial and fungi communities of the wastewater were investigated before their use as influents in the pots. The results highlighted the presence of 20 bacteria and 27 fungi phyla. The most represented bacterial phyla were Proteobacteria (about 58%), Firmicutes (about 18%), and Actinobacteria (about 7%), as shown in

Table 2. Bacterial composition of the wastewater at the phylum taxonomy level.

INDEX	Frequencies %
Archaea; Crenarchaeota	0.02
Archaea; Euryarchaeota	0.29
Bacteria; Acidobacteria	2.86
Bacteria; Actinobacteria	6.76
Bacteria; Bacteroidetes	1.47
Bacteria; Cyanobacteria	0.04
Bacteria; Firmicutes	18.20
Bacteria; Patescibacteria	1.99
Bacteria; Proteobacteria	58.28
Bacteria; Spirochaetes	0.06
Bacteria	9.94
Unassigned	0.04

. In the case of fungi (

Table 3. Fungal composition of the wastewater at phylum taxonomy level.

INDEX	Frequencies %
k_Fungi;_	26.43
K_fungi; p_Basidiomycota	5.88
K_fungi; p_Chytridiomycota	3.29
K_fungi; p_Glomeromycota	3.30
K_fungi; p_Mortirellomycota	2.34
K_fungi; p_Rozellomycota	3.54
K_fungi; p_Ascomycota	44.27
K_fungi; p_Blastocladiomycota	2.29
K_fungi; p_Fungi_phy_Incertae_sedis	2.64
K_fungi; p_Monoblepharmomycota	2.29
K_fungi; p_Olpidiomycoita	3.69

), 27 phyla were identified, and the most represented phyla were Ascomycota (about 44%) and Basidiomycota (about 6%), while the others were present in a similar percentage (about 3–4%).

3.2.2. Rhizoplane Microbial Community Characterizing Different Macrophyte Plant Species

The microbial communities of the three macrophytes were analyzed to assess the civil wastewater effects on root microorganism biodiversity, as well as the ability of these communities to recover when the spring water was recirculated after the wastewater treatment.

Rhizoplane Communities of Arundo donax L.

A remarkable impact of the wastewater repeated exposure was detected on the microbiome of A. donax. In total, 18 bacteria OTUs were identified from the rhizoplanes of the three experimental groups. However, the relative abundances of bacterial phyla from the roots exposed to wastewater decreased when compared to the roots of the CNT irrigated with spring water. This effect was most pronounced for Alphaproteobacteria (reduced from 16% to 9%) and Clostridiae (reduced from 10% to 3%); other phyla, such as Actinobacteria and Bacilli, increased from 4% up to 12% and from 8% up to 19%, respectively (Figure S1). To this regard, the Resistance Index (calculated as illustrated in Section 2) revealed significant effects of the wastewater and its pollutant concentrations on the microbial community. In general, for the A. donax rhizobacteria community, the ResI value was 0.18, indicating a low resistance of its rhizobacteria community to the disruptive effects of the wastewater. Notably, certain bacterial OTUs were particularly susceptible to the wastewater, such as Gammaproteobacteria, Acidobacteria, and Firmicutes (ResI close to zero), while others exhibited greater resistance, as observed in Planctomycetacia, Spirochaetia, and Halobacteria (ResI close to one). Notably, the relative abundances of bacterial phyla increased when plants, previously exposed to wastewater, were treated with spring water for the next five days. In this case, the relative abundances returned to similar levels of the CNT group or, in some cases, even exceeded those values (Gammaproteobacteria 39%, Bacilli 4%, and Actinobacteria 3%).

The high resilience level was also highlighted by the Resilience Index (RI—calculated as reported in Section 2). In fact, the RI calculated for the rhizobacterial community was close to one (0.94), indicating a full recovery of this community. In particular, some bacterial OTUs showed a strong resilience to the disturbing effects caused by the wastewater. In fact, Bacterioidia, Patescibacteria, and Proteobacteria, although affected by the wastewater treatment, were able to over grow when the wastewater was substituted with spring water as influent. On the contrary, some OTUs such as Acidobacteria and Oxyphotobacterria were not able to re-grow when the disturbance ended

As regards the fungal communities (Figure S2), the composition of the rhizoplane community associated with A. donax was slightly affected in response to the different treatments. The results showed that in the CNT, all the identified OTUs belonged to the Agaricomycetes class. Moreover, among the identified orders, Agaricales represented almost 60%. In the WW group, the fungal rhizoplane community was strongly affected; in fact, the Ascomycota phylum appeared, in respect with the CNT group. Among the OTUs belonging to the Ascomycota, the orders Capnodiales, Glomerellales, and Hypocreales were most likely related to the wastewater treatment. The repeated exposure of spring water for additional five days, after the wastewater treatment, also modified the fungal community composition of the rhizoplane; in particular, the order of Capnodiales was not anymore present, whilst the Glomerellales increased its relative abundance. The order of Cantharellales, belonging to the Ascomycota class, appeared just in the rhizoplane of WW + W plants. However, the ResI and RI indices highlighted that the fungal community showed a high resistance when exposed to the WW (ResI = 0.91).

Rhizoplane Communities of Juncus conglomeratus L.

Bacterial rhizoplane community of J. conglomeratus was strongly affected by the wastewater repeated exposure. In particular, the relative abundances of bacterial phyla Spirochaetes, Planctomycetes, and Actinobacteria increased with the wastewater repeated exposure (WW group), but, subsequently, decreased after the spring water repeated exposure (WW + W group), even if the values of relative abundance remained higher than those observed in the case of CNT group. Nevertheless, the relative abundances for Cyanobacteria, Patescibacteria, Acidobcteria, and Firmicutes increased from CNT to WW groups, and then to one of the WW + W groups (Figure S3). In the CNT group, the community was prevalently composed of Gammaproteobacteria (40%), Alphaproteobacteria (18%), Clostridia (10%), Bacilli (Firmicutes) (8%), and Actinobacteria (4%), whilst, in the case of the WW group, a decrease in the relative abundance of Gammaproteobacteria (37%), Alphaproteobacteria (11%), and Clostridia (3%) was observed; contrarily, the relative abundance of Firmicutes (17%) and Actinobacteria (13%) increased. In this context, the Resistance Index (ResI) highlighted a marked influence of wastewater exposure and its associated pollutant load on the structure of the microbial community. Specifically, the rhizobacterial community associated with J. conglomeratus exhibited a low overall resistance to disturbance (0.13 ResI). This suggests a limited ability of the community to withstand the environmental stress imposed by the wastewater treatment. Notably, several OTUs, including members of Gammaproteobacteria, Acidobacteria, and Proteobacteria, showed close to zero ResI values, indicating high sensitivity to the pollutants. In contrast, taxa such as Halobacteria displayed ResI values close to one, reflecting a comparatively greater resilience to the wastewater-induced perturbation.

In the rhizoplane of the WW + W group, for Firmicutes (Bacilli) and Actinobacteria, the relative abundances almost returned to the initial values of the CNT group. This highlighted the significant effect of effluent repeated exposure on these OTUs. The RI, calculated for the rhizobacterial community was, in fact, close to one, indicating a partial recovery of this community when the disturbance ended.

In the case of fungi, the rhizoplane community associated with J. conglomeratus exhibited the highest number of OTUs (27). The data indicated that the CNT group had a higher abundance of fungal OTUs, including Sordariomycetes (37%), Ascomycota (14%), and Dothideomycetes (8%), compared with the WW group. Notably, when considering the microbiome of the WW group, a reduction in the relative abundance of Ascomycota was observed compared to the CNT. Conversely, Dothideomycetes and Sordariomycetes significantly increased their relative abundance, reaching up to 17% and 61%, respectively. The repeated exposure of spring water for five days following wastewater treatment (WW + W group) did not alter the fungal community (Figure S4). The analysis of the ResI (0.99) and RI indicated that the fungal community exhibited a high degree of resistance to wastewater exposure. Conversely, the RI value below 1 suggests a limited resilience, implying that although the fungal community can withstand the initial disturbance, its ability to recover following perturbation remains constrained.

Rhizoplane Communities of Nerium oleander L.

The composition of the rhizoplane microbiome in N. oleander was significantly affected by wastewater treatment. In fact, in the CNT group, the microbial community was prevalently composed by Gammaproteobacteria (40%), Alphaproteobacteria (16%), Clostridia (10%), Bacilli (9%), and Actinobacteria (3%). After WW treatment, a decrease in Alphaproteobacteria (11%) and Clostridia (2%) was observed, while Bacilli (16%) and Actinobacteria (14%) increased (Figure S5). To this regard, the ResI revealed significant effects of the wastewater on the microbial community. In general, for the N. olender rhizobacteria community, the ResI value was very low (8 × 10⁻¹⁶), indicating a rather limited resistance of its rhizobacteria community to the disruptive effects of the WW treatment. Notably, certain bacteria exhibited a high resistance, as observed in the case of Halobacteria, Spirochaetia, and Firmicutes (ResI close to one). Otherwise, the large part of the OTUs were particularly sensitive to the wastewater (ResI close to zero).

In the WW + W group, an opposite trend of the microbiome composition was observed; in particular, the relative abundances of Alphaproteobacteria (2%) and Clostridia (11%) returned to a percentage quite similar to those observed in the CNT group. In contrast, the reduction in Gammaproteobacteria (37%), Actinobacteria (2%), and Bacilli (4%) highlighted the detrimental effects of wastewater on these bacterial communities. The RI value (0.96) also indicated a full recovery of this community. In fact, all the identified bacterial OTUs were able to re-grow when the disturbance ended.

The fungal community composition associated with N. oleander roots in the CNT group consisted primarily of Chytridiomycetes (49%) and Spizellomycetes (30%) at class level. A significant reduction in the relative abundance of Chytridiomycetes was observed comparing the WW treatment to the CNT, as also in the case of Basidiomycota (Figure S6). As in the case of the other macrophytes used in our experiment, N. oleander was also characterized by a very high resistance (ResI close to 1). Otherwise, the RI value indicated a partial recovery of this community when the disturbance ended.

3.3. Effects of Wastewater on Rhizospheric Biodiversity of the Microbial Community

The data obtained relatively to the microbial communities of each macrophyte species were elaborated to calculate the α-diversity; in particular, the Shannon index was estimated for each macrophyte species and for each experimental group.

3.3.1. Rhizoplane Biodiversity Index of Arundo donax L.

The 16S rRNA and ITS sequencing generated 824.447 and 596.965 gene sequences, respectively, and after quality filtering, denoising, and removal of chimeric sequences, 623.974 and 382.248 genes sequences remained. The α-biodiversity was investigated for rhizoplane bacterial and fungal communities. In the case of the bacteria, the α-diversity, calculated through the Shannon index, showed a significant decrease in WW group when it was compared to the CNT group. In fact, in this case, the index was equal to roughly 6.5, whilst for the CNT group, it was equal to 7.3. Interestingly, in the case of WW + W, the index increased, reaching the value of roughly 7.7, even greater than that observed in the CNT group (Figure 3).

Contrarily, the fungi α-diversity did not vary significantly when the three groups CNT, WW, and WW + W were compared and the higher Shannon index value was observed in the case of the WW + W group.

3.3.2. Rhizoplane Biodiversity Index in Juncus conglomeratus L.

The 16S rRNA and ITS sequencing generated 976.327 and 867248 gene sequences, respectively. After quality filtering, denoising, and removal of chimeric sequences, 766,936 and 581,143 genes sequences remained. The α-biodiversity was calculated for both microbial communities. In the case of the bacterial communities, the α-diversity, based on the Shannon index, significantly decreased in the WW group, compared with the CNT group. In fact, the index was equal to roughly 6.5 in the case of the WW group, whilst for the CNT group, it was equal to 7.6. Interestingly, when the macrophyte pots, undergoing the wastewater treatment, were recirculated with spring water for five days, the index increased and reached the value of roughly 7.8, slightly greater than that observed in the case of the CNT group (Figure 4).

In the case of fungal communities, the α-diversity did not show any significant variation when the three groups were compared (CNT, WW, and WW + W), even if a higher Shannon index was observed for the WW + W group.

3.3.3. Rhizoplane Biodiversity Index in Nerium oleander L.

The 16S rRNA and ITS sequencing generated 851282 and 757087 gene sequences, respectively. After quality filtering, denoising, and removal of chimeric sequences, 687.583 and 470.885 genes sequences remained. The α-biodiversity calculated for bacterial communities (Shannon index) significantly decreased in the WW group when compared with the CNT one. In fact, the index, in this case, was equal to roughly 6.2, whilst for the CNT group, it was equal to 7.4. Interestingly, when the macrophyte pots, undergoing the WW treatment, were recirculated with spring water for five days, the index increased and it reached the value of roughly 7.6, greater than that calculated for the CNT group (Figure 5).

Even for this macrophyte species, the α-diversity calculated for fungal communities did not show significant variation when CNT, WW, and WW + W were compared. The higher Shannon value was detected in the case of the WW + W group.

3.4. Normalization Index Definition

In order to compare the rhizoplane biodiversity of microbial communities associated with the macrophytes used for the experiment, the Treatment Normalization Index (TNI), was calculated as reported in Section 2.

The TNI represents the normalization of the relative frequency (observed in the case of the WW and WW + W groups) with respect to those of its own CNT group, either for bacteria or fungi.

3.4.1. Normalization Index for Bacteria

The results showed the differences in the microbial communities among the three different macrophytes species that underwent wastewater (WW) repeated exposure (

Table 4. Bacteria differences expressed as TNI among the three different macrophytes that underwent wastewater (WW) repeated exposure.

INDEX	WW/CNT TNI J. conglomeratus	WW/CNT TNI N. oleander	WW/CNT TNI A. donax
Thermoprotei	0	0	0
Halobacteria	0	0.05	0.73
Subgroup 6	1.57	1.83	2.32
Actinobacteria	3.32	4.04	0.62
Bacteroidia	0.77	0.99	1.72
Oxyphotobacteria	0	0	0
Bacilli	1.9	1.78	0.49
Clostridia	0.39	0.28	1.13
Firmicutes	0.65	0.3	1.12
Patescibacteria-ABY1	1.84	1.93	1.72
Planctomycetacia	1.72	0	0
Alphaproteobacteria	0.64	0.7	0.87
Deltaproteobacteria	1.53	1.14	3.62
Gammaproteobacteria	0.93	1.01	0.97
Proteobacteria	1.2	0.75	1.63
Spirochaetia	0	0	0
Bacteria	0.67	0.4	1.02

). The relative abundance of each OTUs identified in the WW experimental group of each species was normalized with respect to the relative abundance observed in its CNT group. The three rhizoplane communities differed both in terms of bacterial classes and in their TNI. For instance, Actinobacteria increased up to 3.0- and 4.0-fold in J. conglomeratus and N. oleander when the macrophyte underwent the wastewater treatment, whilst, in the case of A. donax, the wastewater caused the halving of the Actinobacteria. Moreover, the Planctomycetacia was specific to J. conglomeratus rhizoplane. Other bacterial OTUs were identified in all the three rhizoplane macrophytes and also showed a similar TNI, as in the case of Patescibacteria ABY1 and Gammaproteobacteria. Conversely, some bacterial OTUs exhibited different TNI, as in the case of Deltaproteobacteria, and similar in J. conglomeratus and N. oleander, but more abundant in A. donax (about 3-fold).

The results of this analysis show the differences in the bacterial communities among the three different macrophytes that underwent the WW + W treatment (

Table 5. Bacterial differences among the three different macrophytes that underwent the wastewater + water (WW + W) treatment.

INDEX	TNI WW + W/CNT J. Conglomeratus	TNI WW + W/CNT N. oleander	TNI WW + W/CNT A. donax
Thermoprotei	0	0	0
Halobacteria	1.06	1.27	0.73
Euryarchaeota	0	0	0
Subgroup 6	2.35	2.22	2.32
Actinobacteria	0.62	1.92	0.62
Bacteroidia	1.99	1.41	1.72
Oxyphotobacteria	0	0	0
Bacilli	0.40	0.53	0.49
Clostridia	1.15	0.95	1.13
Firmicutes	1.07	0.79	1.12
Patescibacteria-ABY1	2.05	3.25	1.72
Patescibacteria	0	0	0
Planctomycetacia	0	0	0
Alphaproteobacteria	0.76	0.75	0.87
Deltaproteobacteria	3.86	4.45	3.62
Gammaproteobacteria	0.91	0.96	0.97
Proteobacteria	2.05	1.56	1.63
Spirochaetia	0	0.02	0
Bacteria	1.34	0.83	1.02

). The relative abundance of each OTUs identified in the WW + W group, of each macrophytes species, was normalized with respect to the relative abundance of the CNT group. The three plant rhizoplanes differed in terms of bacterial OTUs identified, and also in terms of TNI. The TNI highlighted that the bacterial composition of the rhizoplanes of J. conglomeratus and A. donax were quite similar after the repeated exposure of the wastewater, whilst that of N. oleander deeply differed from both. In fact, in the case of N. oleander, the TNI of Actinobacteria was significantly greater with respect to A. donax and J. conglomeratus that underwent the same experimental conditions (WW + W).

3.4.2. Normalization Index for Fungi

The TNI which relates the WW treatment to the CNT one (TNI WW/CNT) showed a comparable number (roughly 20) of observed OTUs in the three macrophytes. However, the TNI showed some species-specific differences.

In fact, in the case of the WW group, the TNI (

Table 6. Fungi differences among the three different macrophytes species that underwent wastewater (WW) repeated exposure. ND means not defined.

INDEX	TNI WW/CNT N. oleander	TNI WW/CNT A. donax	TNI WW/CNT J. conglomeratus
k_Fungi	1.18	0.94	0.35
Ascomycota	0	0	0.03
Ascomycota_cls_Incertae_sedis	0	0	6.01
Dothideomycetes	0	0	2.06
Eurotiomycetes	0	1.02	0.72
Lecanoromycetes	0	0.68	0
Leotiomycetes	0	0.49	2.53
Saccharomycetes	0	1.30	4.31
Sordariomycetes	12.47	0	1.64
Taphrinomycetes	0.22	0	0
Basidiomycota	14.51	0	3.04
Agaricomycetes	0.92	0	1.71
Cystobasidiomycetes	0	0	4.25
Exobasidiomycetes	0	0	0
Malasseziomycetes	0	0	0.91
Microbotryomycetes	0	0	0
Tremellomycetes	0	0	0.09
Ustilaginomycetes	0	ND	0
Blastocladiomycetes	0	ND	0
Chytridiomycetes	0	ND	0
Lobulomycetes	0	0	19.80
Fungi_cls_Incertae_sedis	0	ND	0
Glomeromycota	0	ND	0
Glomeromycetes	0	0	0.19
Monoblepharidomycetes	0	ND	0
Olpidiomycetes	0	0	1.62
Rozellomycota_cls_Incertae_sedis	0	0	0.41

) revealed a species-specific OTU for N. oleander and A. donax, such as Taphirinomycetes and Lecanoromycetes, respectively. Two OTUs, Basidiomycota and Sordariomycetes, were present in both N. oleander and J. conglomeratus; however, they differed in their TNI. Finally, a single OTU, Euromycetes, was found in both J. conglomeratus and A. donax, with higher TNI in A. donax with respect to J. conglomeratus. This analysis found that the wastewater repeated exposure caused the significant increase (up to 12-fold) of Sordariomycetes in N. oleander, in the case of Basidiomycota. Cystobasidiomycetes was negatively affected by wastewater repeated exposure in the case of N. oleander and A. donax, whilst, in the case of J. conglomeratus, it quadrupled in the WW group with respect to the CNT one.

The TNI, calculated to compare the WW group with the WW + W one, among the three macrophytes, found a species-specific OTU for N. oleander (Agaricales) and one for A. donax (Nectriaceae). For instance, the WW + W treatment caused an increase in Plectosphaerellaceae in A. donax with respect to the untreated one (CNT group), whilst it disappeared in N. oleander or was halved in J. conglomeratus. Furthermore, J. conglomeratus showed several species-specific OTUs such as Glomeromycota, Glomeromycetes, Monoblepharidomycetes, and Olpidiomycetes (

Table 7. Fungi differences among the three different macrophytes that underwent the WW + W treatment. ND means not defined.

INDEX	TNI WW + W/CNT N. oleander	TNI WW + W/CNT A. donax	TNI WW + W/CNT J. conglomeratus
k_Fungi	1.04	1.14	0.85
Cladosporiaceae	0	0	0.22
Pleosporaceae	0	0	0.01
Orbiliaceae	0	0	1.33
Plectosphaerellaceae	0	1.68	0.62
Hypocreaceae	0	0.37	0
Nectriaceae	0	0.12	0
Stachybotryaceae	0	0.76	6.23
Basidiomycota	7.00	0	1.10
Agaricomycetes	0.16	ND	0
Psathyrellaceae	1.39	0	1.47
Lobulomycetales_fam_Incertae_sedis	0	0	0
Fungi_fam_Incertae_sedis	0	0	0
Malasseziomycetes	0	0	2.09
Malasseziomycetes	0	0	0
Microbotryomycetes	0	0	0.34
Tremellomycetes	0	0	0
Ustilaginomycetes	0	ND	0
Blastocladiomycetes	0	ND	0
Chytridiomycetes	0	ND	0
Lobulomycetes	0	0	0.52
Fungi_cls_Incertae_sedis	0	ND	61.95
Glomeromycota	0	ND	13.13
Glomeromycetes	0	0	0.59
Monoblepharidomycetes	0	ND	1.45
Olpidiomycetes	0	0	0.72

). Among these OTUs, Glomeromycota was increased (up to 13-fold) by the spring water repeated exposure after the wastewater treatment only in the case of J. conglomeratus, whilst it disappeared in the case of N. oleander and A. donax. Similarly, Basidiomycota increased its presence up to 7-fold in N. oleander, whilst it disappeared in A. donax, returning to a similar value observed in the CNT group as in the case of J. conglomeratus.

4. Discussions

Microorganisms play an important role for the health of Earth inhabitants and have a pivotal role in the pollutant removal, and/or their degradation when contaminating different environmental matrices (e.g., water). In view of the reduction in water availability, due to the increasing anthropogenic pressures on natural and urban environments, together climate changes, the role and the interactions among microorganisms and plants in CWs is fundamental for providing safe water reuse as recently requested by numerous national or community (e.g., European Union) regulations. Therefore, the investigation of microbial communities is crucial for understanding the mechanisms underlying remediation processes. However, very few studies showed how microbial communities are modified in their composition during the process and water repeated exposure in the CWs [2,26]. In this context, our study could be considered an exploitable contribution for understanding the rhizo-microbial community variability in those macrophytes commonly employed in the CWs implementation.

The urban wastewater used in our study was previously characterized from a physical–chemical and microbiological point of view. The lab-scale CWs differ only in their macrophyte species, but not in their substrate stratification; nevertheless, they were able to significantly reduce organic matter and oxidate the ammonium, although some of them were more effective in removing both chloride and nitrate. Although chloride is an essential micronutrient for higher plants, its accumulation indeed varies among the macrophytes; the macrophytes we employed differ in the salt level tolerance and in strategies for managing chloride and other ions. Our results highlighted that the best performances in chloride removal from wastewater were observed in the case of J. conglomeratus, whilst the less effective one was A. donax. N. oleander, which showed an intermedia capacity to accumulate chloride, as reported also by Flowers and Colmer [27]. J. conglomeratus is known to live in salt marshes, giving it superior capability to tolerate chloride and sodium [28]. Halophytic species adopt different mechanisms to cope with salinity, including accumulation, exclusion, and secretion. Accumulator plants internalize salts and compartmentalize them in vacuoles to maintain osmotic balance; excluders restrict chloride uptake at the root level, accumulators store only small quantities within their tissues—contributing to removal mainly when biomass is harvested—and secretors relocate salts to leaf surfaces, from which they may readily leach back into the system. However, salt marsh tolerance does not necessarily translate into superior chloride removal in CWs. These mechanisms primarily protect plant physiology rather than promote active salt extraction. Consequently, halophytic tolerance ensures plant survival under saline stress but does not inherently enhance chloride removal efficiency in CWs [27].

The NGS microbiome analysis revealed that the wastewater we employed for the experiment was initially rich in Proteobacteria and Ascomycota phyla as reported in the scientific literature [29,30,31,32].

Otherwise, the composition of the rhizobacterial communities of the three macrophytes significantly varied on the basis of the treatment they underwent. In fact, in the case of the CNT group, rhizobacterial community of the three macrophytes was characterized by the presence of Gammaproteobacteria, Alphaproteobacteria, Clostridia, and Actinobacteria classes, although with differences in their relative abundance. Some species-specific phyla such as Firmicutes, in the case of J. conglomeratus, or Bacilli classes, in the case of N. oleander and A. donax, were detected. The wastewater repeated exposure caused a decrease in the relative abundance of Gammaproteobacteria, Alphaproteobacteria, and Clostridia, combined with an increase in Firmicutes and Actinobacteria for all the three macrophyte species. Actinobacteria are known for breaking down complex polymers and producing bioactive molecules for bioremediation, while Firmicutes are involved in the degradation of various pollutants but are often sensitive to some compounds like phthalates (PAEs), which can inhibit their growth and lead to increased Proteobacteria and Actinobacteria abundance. Moreover, Actinobacteria play important ecological roles in the environment, such as degrading complex polymers, recycling compounds, and even restoring environments [33].

When the spring water was recirculated after WW treatment, the relative abundances of these phyla and classes returned to the levels similar to those observed in the CNT group, so highlighting the significant influence of effluent repeated exposure on these taxonomic bacterial groups. Allison and Martiny [34] highlighted that microbial communities often exhibit a lack of immediate resilience to disturbances, resulting in compositional shifts that affect ecosystem functions. In fact, the authors suggested that these changes may influence ecosystem processes due to the reduced resilience of microbial communities as a response to environmental changes. In our study, the influent repeated exposure acted as a disturbance, causing specific bacterial taxa shifts. However, the tendency of the rhizoplane microbiome to revert to the condition observed in the CNT group, when spring water was recirculated, was analogous to the model of resilience described by Allison and Martiny [34]. The return to original microbial community composition after the removal of the disturbance shows how, under reduced stress, the microbial communities can recover and probably regain functional traits relevant for the ecosystems.

Similarly, Guan, Yin [35] showed the impact of the substrates on CW microbial communities. In fact, they highlighted how their pilot-scale CW, built with different substrate materials (sand and gravel) and based on a vertical flow system, showed a well-defined and specific composition of the bacterial community after the water repeated exposure. The main phyla found were Proteobacteria and Actinobacteria in percentages quite similar to those that we observed. Therefore, it might be assumed that the recirculated wastewater and the material and matrices used for assembling the CWs might have a role in the composition of the bacterial community.

The specific fungal communities found in both WW and WW + W groups associated with the rhizoplanes of N. oleander and A. donax, highlighted the relevant role of the macrophytes species in shaping rhizoplane microbial dynamics, as in the case of the Agaricomycetes class and Psathyrellaceae family. In particular, in the case of Agaricomycetes, both macrophytes species followed a similar trend, with abundance declining as treatments shifted from the CNT group to the WW one and then to the WW + W group as also reported by Mesacasa, Cabral and He, Huang [36,37].

Interestingly, the fungal family Psathyrellaceae exhibited a species-specific response; for instance, in N. oleander, its abundance first decreased in the case of WW treatment and then increased when WW + W treatment was considered. Conversely, in A. donax, the Psathyrellaceae increased with WW treatment but then declined in the case of WW + W treatment, indicating a positive response to the spring water repeated exposure. In fact, the repeated exposure to water might either dilute some contaminants or alter the rhizoplane ecosystem. This contrast in fungal response, with respect to the bacterial one, is likely due to the fact that each macrophyte produces unique root exudates and establishes specific microbial interactions; in fact, the scientific literature taken into consideration highlights the role of macrophytes specificity in fostering distinct microbial communities related to the wastewater treatment and pollutant removal [37,38].

In the case of J. conglomeratus, the rhizoplane was characterized by a greater fungal richness already present in the CNT group, including Ascomycota phylum, and both Sordariomycetes and Dothideomycetes classes. This observation suggests that J. conglomeratus may naturally support a more diversified fungal community, potentially due to more complex and diverse root exudates, or structural root properties, which could enhance pollutant degradation processes occurring in the CWs [37,38]. However, for this macrophyte, the WW group showed a reduction in the relative abundance of Ascomycota, but an increase in both Dothideomycetes and Sordariomycetes. These findings suggest that treatments significantly influence fungal diversity and its community composition which play a pivotal role in plant health and pollution degradation [37]. In fact, some fungus species belonging to Ascomycota and Basidiomycota phyla are able to establish symbiotic associations (mycorrhiza) with about 85% of the plant species [38] and also have positive effects on pollutant tolerance and plant growth [36,39,40]. These considerations support our hypothesis that the capability of CWs to degrade pollutants is due to the synergistic effect of the biotic components. In addition, the fungal species, already present in the plant rhizoplane, have proved to be particularly resistant and resilient to the stresses related to wastewater repeated exposure, increasing their abundance. These results highlighted the importance of microbial communities, in terms of taxonomic groups and/or relative abundance present in the CWs associated with the roots of each macrophyte; therefore, some plants can be strategically employed in the CW assembling to target specific microbial communities, potentially optimizing wastewater treatment and enhancing bioremediation effectiveness. In this regard, we observed some significant shifts in OTU abundance in the case of bacterial community, also highlighted by the α-biodiversity which was statistically significantly different when the WW group was compared either with the CNT group or WW + W one. This finding highlights the low resistance of the rhizobacterial communities to the disturbance caused by the wastewater repeated exposure, but also their high resilience to recover once the disturbance ends. On the contrary, the fungal communities showed a high resistance to the wastewater repeated exposure, and this is also supported even by the α-biodiversity index that did not substantially change among the experimental groups. Fungi exhibited higher resistance but lower resilience than bacteria due, probably, to their structural robustness, slower growth, and stable symbiotic associations within the rhizoplane. Their chitin-rich cell walls and ability to form spores made them less sensitive to chemical stress, ensuring stability under disturbance [41]. However, their low turnover rates limited recovery once the stress was removed [42]. Conversely, bacteria which have rapid multiplication, metabolic flexibility, and functional redundancy, were more sensitive to wastewater pollutants but recovered quickly after stress release [43]. The spring water recirculation diluted pollutants, restored oxygen balance, and stimulated root exudation, enhancing recolonization by the sensitive bacterial taxa [44]. Overall, fungi maintained the CWs stability during stress, while bacteria drove recovery, revealing complementary resistance–resilience dynamics essential for CW ecological functioning and pollutant remediation.

The response of the rhizo-microbiome of the three macrophytes to the different treatments (WW and WW + W groups) was investigated, comparing the TNI. TNI was calculated for normalizing the data with respect to the starting microbial community; in this way, the different ratio was related mainly to the treatment that the macrophytes underwent. TNI allowed us to highlight the differences in microbial communities caused, at first, by the wastewater treatment, and then by the five additional days of spring water repeated exposure. In the case of both microbial communities, the TNI, estimated for the three macrophytes differed, suggested that wastewater modified the rhizobacterial community even on the basis of the considered macrophyte species. This is because the disturbance effect, played by the wastewater, modified the original microbial community in a plant species-specific manner. However, in the case of bacteria, when the disturbance was removed, and the spring water was recirculated in the pots (WW + W group), the TNI of the three macrophytes were still different, highlighting also, in this case, a species-specific response. However, it is noteworthy that J. conglomeratus and A. donax showed a similar TNI, when WW + W groups were considered, although they were characterized by different OTUs or different relative abundance in their microbial communities, in respect to both CNT and WW groups. This suggests that, when the disturbances ended, J. conglomeratus and A. donax rhizobacterial communities reacted similarly. Probably, the spring water might have exerted a dilution effect of the wastewater contaminants or could be a carrier of nutrients able to reshape the altered microbiome.

In the case of fungi, the TNI showed that the three macrophytes differed in the fungal rhizomicrobiome when plants were exposed to wastewater. In fact, the large part of fungal taxonomic groups was characterized by a TNI equal to zero. From a mathematical point of view, with the TNI being the ratio between the OTUs identified in the treatment (WW or WW + W) and those present in the CNT group, when it is zero, it implies that a certain OTU is present in the CNT but absent in the treatments. This result suggests that many OTUs observed in the CNT were significantly reduced in the WW or WW + W groups. Contrarily, when the specific OTU was present in the WW or WW + W groups, but absent in the CNT, the TNI was not defined (ND).

However, the three macrophytes differently respond to the disturbance caused by the wastewater treatment; in particular, a large part of the OTUs associated to the N. oleander rhizoplane was depleted by wastewater; on the contrary, Basidiomycota phyla were enriched after the wastewater repeated exposure with respect to the CNT group as well as the Sordariomycetes classes.

When A. donax was taken into consideration, several OTUs showed an ND-TNI because this macrophyte was characterized by very few fungal OTUs in the CNT group. Differently, J. conglomeratus was characterized by a greater number of OTUs in the CNT group with respect to the other two macrophytes; this is reflected by the absence of ND for the TNI value. However, the wastewater repeated exposure enriched the fungal rhizo-microbial community as highlighted in the case of Ascomycota phylum, Lobulomycetes, Saccharomycetes, and Cystobasidiomycetes, classes. These taxonomic groups are known to be associated with the plant rhizoplane [45] but also play an important ecological role. For instance, Lobulomycetes can degrade organic matter as chitin, cellulose, etc., and can act like saprotrophs [46]. Members of the Ascomycota phylum play a central role in carbon and nitrogen turnover, particularly in nutrient-limited or arid environments [47]. Saccharomycetes and Cystobasidiomycetes are of considerable relevance for the degradation of organic matter and trace organic pollutants [48], but are also able to produce antioxidants and photoprotective carotenoids [48].

In the case of the TNI relative to WW + W group, the index suggested that the three macrophytes showed a similar behavior, given that the rhizo-communities were not significantly affected by spring water repeated exposure, as also highlighted by their α-biodiversity.

The observed patterns of microbial resistance and resilience in response to wastewater exposure provide critical insights into the ecological stability of rhizoplane-associated communities undergoing environmental stress. The low ResI value (0.13) observed for the macrophyte rhizobacterial community indicates a pronounced sensitivity to wastewater and its associated pollutant load. This finding is consistent with previous studies highlighting that rhizoplane microbial assemblages, particularly those in wetland environments, are susceptible to compositional shifts when exposed to nutrient-rich or chemically complex influents [49,50]

The taxon-specific responses further reinforce the concept of phylogenetic determinism in microbial resistance. The OTUs affiliated with Gammaproteobacteria, Acidobacteria, and Proteobacteria phyla displayed near-zero ResI values, suggesting high sensitivity to environmental disturbance. In fact, these taxonomic groups are known to be metabolically versatile even if poorly buffered against sustained stressors, as previously reported in studies on disturbed soil and aquatic microbiomes [34,51]. In contrast, the Halobacteria class exhibited near-complete resistance (ResI ≈ 1), likely due to their intrinsic adaptations to high salinity and oxidative stress, such as specialized membrane lipids or osmoprotectant compunds [44].

Of particular interest is the partial recovery observed in the WW + W group, when spring water repeated exposure was applied. The relative abundances of Firmicutes (notably Bacilli) and Actinobacteria nearly returned to control (CNT) levels, suggesting a stabilizing effect of operational management on community composition. This pattern is reflected in the Resilience Index (RI), which approached 0.50, indicating moderate recovery after stressor removal. These results align with recent findings demonstrating that repeated exposure regimes can modulate redox gradients, reduce nutrient shocks, and promote microbiome re-establishment in CWs and similar engineered ecosystems [51]. In support of this, Ansola et al. [52] reported significantly lower redox potentials in Constructed Wetlands (−152 to −193 mV) than in natural wetlands (−103 to −135 mV), indicating stronger reducing conditions. The authors further linked the observed bacterial community distribution to oxygen availability along flooded-to-dry gradients, confirming the role of redox modulation in shaping microbial assemblages.

Collectively, these findings emphasize the complementary roles of resistance and resilience in understanding microbial ecosystem function in the CWs. While resistance characterizes the capacity to endure disturbance without an evident change, resilience reflects the ability to recover original structure and function. Assessing both indices provides a more nuanced understanding of microbial stability in anthropic impacted environments. Further investigations should integrate functional profiling (e.g., metagenomics and meta-transcriptomics) to disentangle the metabolic underpinnings of these community-level responses and identify key functional traits associated with adaptive success.

5. Conclusions

This study, using a next-generation sequencing (NGS) approach, provided novel insights into the structure and ecological behavior of rhizoplane microbial communities associated with three macrophytes commonly employed in Constructed Wetlands (CWs) assemblages. Nowadays, the knowledge about the role of plants, and of the microbial community shift, related to the CWs, is far from being totally understood. Our results firstly contributed to identifying the rhizo-microbial community associated with three different macrophytes, highlighting how they were characterized by different microbial communities, from qualitative (OTUs identified in their rhizo-plane) and, in particular, quantitative (relative abundances and TNI) points of view. These findings suggest the importance of macrophyte choice in CW structuring and assembling. It is noteworthy how wastewater repeated exposure differentially affects their bacterial and fungal communities. Notably, wastewater repeated exposure exerted differential effects on bacterial and fungal communities. Across all the tested macrophytes, rhizobacterial communities displayed low resistance but moderate resilience, suggesting sensitivity to disturbance but also a capacity for partial recovery. In contrast, fungal communities were characterized by high resistance but lower resilience, with community composition remaining relatively stable despite perturbation. However, the reintroduction and repeated exposure of spring water post-disturbance promoted the relative abundance of several fungal taxonomic groups, suggesting that operational strategies can modulate microbial dynamics. This was also confirmed by the shift in their α-biodiversity. In conclusion, our study suggests the hypothesis that the microbial community has a key role in the CWs and pollutants accumulation/degradation thanks to its resistance and/or resilience, supporting, in this way, the plant health and function. However, further research, integrating functional metagenomics and system-level modelling, is needed to elucidate the complex plant–microbe–pollutant interactions that govern CW efficiency.