Influence of irrigation regime and microbial consortium application on pharmaceutical uptake in durum wheat irrigated with treated wastewater

1 Introduction

Water scarcity, exacerbated by climate change and the increasing demand for water for agricultural purposes, represents one of the major challenges to the sustainability of agri-food systems (1). In this context, the use of alternative water resources, such as treated wastewater (TWW), is increasingly recognized as an effective strategy to alleviate pressure on conventional freshwater resources, promote circular water management practices, and sustain crop yield and quality.

Despite growing scientific interest and the acknowledged potential of TWW reuse, its implementation remains limited, particularly in Europe, where this practice is still largely underexploited relative to the available water resources (2). This limited adoption is mainly driven by concerns regarding TWW quality, as treated effluents may still contain contaminants such as pharmaceutical active compounds (PhACs), which have received considerable attention due to their potential environmental impacts (3–5).

Many PhACs are only partially metabolized and excreted, and their widespread use, coupled with incomplete removal during conventional wastewater treatment, leads to the continuous occurrence of both parent compounds and their transformation products in TWW (4, 6). Current wastewater treatment plants, although achieving high removal efficiencies for conventional pollutants, are not designed for the complete removal of micropollutants such as PhACs. Consequently, the implementation of advanced treatment technologies requires substantial capital investment and higher operational costs, limiting, at least in the short term, their widespread adoption (4). At the same time, European water policy is progressively moving toward stricter requirements for micropollutant monitoring and control. The revision of the Urban Wastewater Treatment Directive (Directive (EU) 2024/3019) and the periodic updates of the European Watch List (Commission Implementing Decision (EU) 2022/1307) (7) highlight the increasing regulatory attention devoted to contaminants of emerging concern and are expected to foster the implementation of more advanced wastewater treatment solutions.

Although PhAC concentrations in TWW generally range from a few ng L^-1 to several μg L^-1, their continuous input into agroecosystems through irrigation has raised increasing concern regarding their environmental fate and potential transfer along the food chain (5). Once introduced into the soil–plant system, their behavior is governed by adsorption–desorption, transport within the soil solution, degradation, and root uptake which are influenced by compounds physicochemical properties, soil characteristics, crop species, and environmental conditions (4, 6, 8–10). Antibiotics, anti-inflammatory drugs, β-blockers, and antiepileptics are among the classes most frequently detected in TWW. Among them, antibiotics are of particular concern because their dissemination in environmental compartments may promote the selection and spread of antimicrobial-resistant bacteria and resistance genes, posing an increasing threat to ecosystem integrity and public health (11).

Irrigation management further influences contaminant dynamics, as the applied water volume determines the overall contaminant load entering the system and modifies soil water conditions, potentially affecting persistence, transformation, and plant uptake. As a result, predicting their behavior and transfer to plant tissues remains a major challenge.

Among the factors influencing the persistence and transformation of PhACs, the soil microbial community plays a pivotal role. Several studies have shown that exposure to pharmaceutical contaminants alters the composition, abundance, and metabolic activity of the soil microbiome, affecting key processes such as carbon mineralization, nitrogen cycling, and the degradation of organic contaminants (12–15). At the same time, microbial communities may progressively adapt to PhACs exposure, promoting enhanced biodegradation during prolonged contamination and thereby reducing the persistence of certain compounds (16–18). This role becomes even more relevant under deficit irrigation with TWW, a widely adopted water-saving strategy that exposes plants to osmotic stress while altering soil water dynamics. In this context, the application of beneficial microbial consortia and arbuscular mycorrhizal fungi (AMF) represents a promising approach to improve plant drought tolerance while simultaneously modifying rhizosphere processes, enhancing microbial activity, and potentially stimulating their attenuation and removal of PhACs contaminants (19, 20).

Crop species is also a key factor governing the uptake and translocation of PhACs. Indeed, considerable interspecific variability has been reported, with leafy and root vegetables generally exhibiting higher accumulation than cereal crops. Contaminant transfer is further influenced by transpiration rate, which depends on species- or genotype-specific physiological traits as well as environmental conditions during the growing season, and by crop duration, which determines the overall period of plant exposure to contaminants (5, 21–23). However, the accumulation of contaminants in organs other than leaves suggests that translocation is not regulated solely by the transpiration stream, but also by species-specific physiological mechanisms controlling contaminant movement among different plant tissues. In cereal crops, specialized anatomical structures such as the exodermis and the Casparian strip in roots act as selective physical barriers, restricting the xylem loading of more hydrophilic compounds while promoting their vacuolar sequestration in root tissues, thereby minimizing translocation to the grain (24).

Despite the growing body of research on TWW reuse, information on the combined effects of irrigation water quality, irrigation regime, and microbial inoculation on the fate of PhACs within the soil–plant system remains limited. Furthermore, although the behavior of contaminants of emerging concern has been extensively investigated in several crops, including some cereal species, evidence for durum wheat under environmentally relevant exposure conditions remains limited. Addressing these knowledge gaps is essential for improving our understanding of the processes governing contaminant persistence, transformation, and bioaccumulation, as well as for developing agronomic strategies capable of minimizing their transfer to edible plant tissues.

In this context, we hypothesized that (i) durum wheat would exhibit a low transfer of PhACs to the grain, and (ii) the application of a beneficial microbial consortium would enhance rhizosphere functioning, promoting the attenuation of PhACs in the soil and reducing their bioavailable fraction, even under deficit irrigation conditions.

Accordingly, the present study investigated the fate of PhACs in the water–soil–durum wheat system under different irrigation regimes and microbial inoculation treatments. Specifically, the study aimed to evaluate the accumulation and translocation of PhACs within plant tissues and to assess the potential of microbial inoculation to enhance contaminant attenuation, thereby supporting the development of sustainable strategies for the agricultural reuse of TWW.

2 Materials and methods

2.1 Experimental set-up and crop cultivation system

The experiment was carried out using a pot system installed at the “Orleans” experimental farm of the Department of Agricultural, Food and Forestry Sciences, University of Palermo (Palermo, Italy; 38°06′15″ N, 13°20′58″ E; 70 m a.s.l.), as previously described in Tortorici et al. (2). The experimental set-up is here briefly reported for completeness. The study was conducted from mid-January to early June 2025. To prevent uncontrolled water inputs, pots were placed within a rainout shelter (12 m × 8 m × 2.8 m), ensuring full control of soil water content throughout the crop cycle.

Each experimental unit consisted of a circular pot (22 cm diameter × 22 cm height; volume 6.84 L) filled with 6 kg of soil collected from an agricultural field located in Collesano (Palermo, Italy; 37°55′16″ N, 13°56′17″ E; 475 m a.s.l.). The main physicochemical characteristics of the soil are reported in Table 1. Figure 1 summarizes the meteorological conditions during the growing season, including air temperature, precipitation, and vapor pressure deficit (VPD), recorded by an on-site weather station.

Table 1. Soil physico-chemical characteristics.

TN, total nitrogen.

Durum wheat (Triticum turgidum subsp. durum (Desf.) Husn.) cultivar ‘Antalis’ (Semìa), commonly grown in semi-arid environments of Southern Italy, was selected for the study. Sowing was performed on 14 January 2025, placing 20 seeds per pot at 2 cm depth. After emergence, seedlings were thinned at the one-leaf stage (BBCH 11) to a final density of 13 plants per pot. No mineral fertilization was supplied during the experiment. Weeds were removed manually, and fungal diseases were controlled through the application of a systemic fungicide containing prothioconazole and tebuconazole. Plants were harvested at commercial maturity (6 June 2025), corresponding to a grain moisture content of approximately 12%.

The experimental design was based on the combination of three factors: (i) irrigation water quality, consisting of fresh water (FW) and treated wastewater (TWW); (ii) four soil water regimes, defined as 100% of field capacity (I100), and 70%, 50%, and 30% of I100 (I70, I50, and I30, respectively); and (iii) application of a microbial consortium (presence: M+ vs absence: M−). Treatments were arranged in a completely randomized design (CRD) with three replicates. The factorial combination of the three experimental factors resulted in 16 treatment combinations, with 10 pots per treatment per replicate, for a total of 480 pots.

2.2 Water management and microbial consortium application

The experiment included two irrigation water sources: fresh water (FW) and treated wastewater (TWW). FW was obtained from groundwater extracted at the experimental site, while TWW was collected from the final effluent of the Municipal Wastewater Treatment Plant of Campofelice di Roccella (Palermo, Italy; 37°59′23″ N, 13°53′03″ E; 68 m a.s.l.), after complete treatment and prior to discharge into the sea, ensuring that it represented fully treated effluent. The treatment process includes preliminary mechanical screening, grit and grease removal, biological treatment based on pre-denitrification and nitrification–oxidation stages, secondary sedimentation, and final UV disinfection prior to discharge. Both FW and TWW physicochemical and microbiological characteristics, determined by the Regional Environmental Protection Agency (ARPA Sicilia), are reported in Table 2.

Table 2. Chemical and microbiological parameters of TWW/FW, and analytical methods used.

U.M., unit of measurement.

The TWW was used as supplied by the treatment plant without any further filtration or post-treatment.

Soil water content (SWC) was monitored daily for each treatment combination using TEROS 12 capacitive sensors (METER Group, Inc., Munich, Germany) installed at a depth of 0–0.20 m in one representative pot per treatment, while soil matric potential was measured using TEROS 21 sensors at the same depth. Sensors recorded data continuously at 15-min intervals. Field capacity (θfc = 0.319 m³ m^-3) and wilting point (θwp = 0.191 m³ m^-3) were previously determined using pressure plate methods (25). Based on a rooting depth of 0.20 m, total available water (TAW) was calculated as 25.7 mm (0.88 L per pot), while readily available water (RAW), using a depletion coefficient p = 0.55 (26), was 14.2 mm (0.48 L per pot). The irrigation threshold (θp), used to trigger irrigation events, was calculated according to Allen et al. (26):

θp=θfc−( p ×TAW )

Irrigation was applied manually when θp was reached. In the full irrigation treatment (I100), soil water content was restored to field capacity. The deficit irrigation treatments (I70, I50, and I30) were imposed as fixed fractions of the I100 regime. Irrigation was always maintained within field capacity limits in the I100 treatment.

All pots were initially saturated with fresh water after sowing and maintained under full irrigation until 3 February to ensure uniform germination and establishment. Thereafter, the irrigation water treatments (FW and TWW) and deficit irrigation regimes were applied. Irrigation was continued until BBCH 89 (approximately 21 days before harvest). In the most severe water deficit treatment (I30), irrigation was discontinued on 30 April due to plant mortality. Total irrigation volumes per treatment are reported in Table 3.

Table 3. Monthly irrigation volumes (L per plot; 1 plot = 10 pots) applied to each treatment.

FW, fresh water; TWW, treated wastewater; M+, presence of microbial consortium; M−, absence of microbial inoculum; I100 , 100% FC; I70 , 70% I100; I50 , 50% I100; I30 , 30% I100.

The microbial consortium treatment consisted of a commercial formulation (Coveron, ITALPOLLINA S.p.A., Rivoli Veronese, Verona, Italy) containing arbuscular mycorrhizal fungi (Glomus intraradices, 300 spores g^-1, and Glomus mosseae, 200 spores g^-1), rhizosphere bacteria (10⁷ CFU g^-1), and Trichoderma atroviride (3 × 10⁸ CFU g^-1). Seeds were coated the day before sowing following the manufacturer’s recommended application rate of 150 g q^-1 seed. The consortium was applied to half of the seeds, while the remaining seeds were left untreated and used as control.

2.3 Plant physiological responses and yield performance

Physiological responses (gas exchange and SPAD) were evaluated at the heading stage (BBCH 51; 3 April) on fully light-exposed flag leaves. For each treatment replicate, five plants were randomly selected.

Leaf gas exchange parameters, including net photosynthetic rate (A), stomatal conductance (gs), and transpiration rate (E), were measured using a portable photosynthesis system (CIRAS-3, PP Systems, Hitchin, UK). Measurements were taken on the central portion of the flag leaf under ambient environmental conditions. Chamber conditions were standardized as follows: photosynthetic photon flux density (PPFD) was set according to ambient light conditions measured with an external quantum sensor (LI-190SA); reference CO₂ concentration was maintained at 400 µmol mol^-1; leaf temperature and relative humidity were not actively controlled; and airflow rate was set at 200 µmol s^-1. Water vapor was partially scrubbed using fresh desiccant, resulting in chamber relative humidity of approximately 50–60% when the cuvette was fully covered by the leaf.

SPAD values, used as a proxy for leaf chlorophyll content, were recorded on the same leaves using a SPAD-502 chlorophyll meter (Konica Minolta, Japan).

Aboveground biomass and grain yield were determined at the replicate level to evaluate overall agronomic performance.

2.4 Chemicals and materials

The target analytes and their corresponding metabolites were selected based on PhACs consumption data in Italy reported by the Italian Medicines Agency OsMed Report (27) and on previous studies investigating PhACs in TWW, soils, and plants, including those by De Mastro et al. (28) and Ben Mordechay et al. (29).

All analytical standards (purity > 98%) were supplied by Lab Instruments. The investigated PhACs included the antiepileptic Carbamazepine (CBZ) and its metabolites: 3-hydroxycarbamazepine (3OH-CBZ), 10,11-dihydro-10-hydroxycarbamazepine (10,11-DH-10-OH-CBZ), acridine (ACD), carbamazepine-10,11-epoxide (CBZ-EP), and carbamazepine diol (CBZ-diol); the antibiotic Clarithromycin (CLR) and its metabolite N-desmethyl clarithromycin (NCLR); the antifungal Climbazole (CLZ); the anti-inflammatory drug Diclofenac (DCF) and its metabolites 4-hydroxydiclofenac (4-OH-DCF) and 5-hydroxydiclofenac (5-OH-DCF); the beta-blocker Metoprolol (MTP) and its metabolite metoprolol acid (MTP acid); and the antihypertensive Telmisartan (TMS). The chemical structures and physicochemical properties of all compounds are reported in Supplementary Table S1.

For the extraction procedures, acetonitrile (ACN), LC–MS grade water, anhydrous magnesium sulfate (MgSO₄), sodium citrate, and primary secondary amine (PSA) were purchased from the Sigma-Aldrich catalog. QuE-Lab^® extraction tubes were obtained from Lab Instruments.

2.5 Extraction and quantification of PhACs from water, soils and plant organs

During the experimental trial, the occurrence of PhACs and their metabolites in TWW was monitored through weekly sampling. To evaluate the initial presence of these compounds in soil, triplicate soil samples were collected at a depth of 0–0.15 m prior to the application of irrigation treatments. At the end of the experimental trial, soil samples and different wheat plant tissues, including roots, straw, and grains, were collected from three independent pots per treatment (n = 3). Each independent sample was subsequently analyzed in triplicate as analytical replicates. All samples were stored at −20 °C until laboratory analysis. Pharmaceuticals and their metabolites were extracted from air-dried soil samples sieved through a 2 mm mesh, as described by De Mastro et al. (30), and from the different wheat plant tissues using a modified QuEChERS method reported by Denora et al. (31).

Quantification of PhACs and metabolites in the extracts was performed using a UHPLC system (UltiMate 3000, Thermo Fisher Scientific) coupled to a TripleTOF 5600+ mass spectrometer (AB Sciex), equipped with a DuoSpray™ ion source operated in electrospray ionization mode. Full details of the LC–MS/MS conditions are reported in De Mastro et al. (19). Data acquisition and processing were performed using instrument-specific AB Sciex software, while PeakView 2.2 and SciexOS 1.2 were used for data interpretation and analyte determination.Carbamazepine-D10 was used as the internal standard, and quantification was performed using matrix-matched calibration curves covering the concentration range of 1–250 μg L^-1 (1, 5, 10, 25, 50, 100, 150, and 250 μg L^-1). The matrices used to prepare the calibration standards were previously analyzed to verify the absence of the target analytes, thereby excluding potential background contamination before their use for calibration. Procedural blanks were also included throughout the analytical procedure to monitor potential contamination introduced during sample preparation and analysis. Limits of detection (LOD), limits of quantification (LOQ), and recovery values for soil and plant matrices are reported in Supplementary Table S2.

2.6 Calculation of bioconcentration and transfer factors

To evaluate the accumulation and internal mobility of PhACs and their transformation products within the wheat plants, the bioconcentration factor (BCF) and transfer factor (TF) were calculated based on the total concentration of each detected PhACs and its transformation products (ΣPhACs), following the approaches described by Carter et al. (32) and Pérez et al. (33). All concentrations used for the calculation of both BCF and TF were expressed on a dry-weight basis. The BCF was calculated as the ratio between the total concentration of ΣPhACs in plant tissues and the corresponding concentration in soil:

BCF=∑PhACsplant∑PhACssoil

The TF was calculated as the ratio between the concentration of ΣPhACs in straw and roots:

TF=∑PhACsstraw∑PhACsroot

In the present study, ΣPhACs corresponded to the sum of carbamazepine and its transformation products, as these were the only compounds detected in wheat tissues. For the BCF calculation, both roots and straw were considered, with the plant concentration weighted according to the respective mass of each plant compartment. Grains were not included because carbamazepine and its transformation products were below their respective LOQs in the grains.

2.7 Data analysis

All statistical analyses were performed according to a completely randomized design (CRD), considering each treatment combination. Data were processed using Minitab Statistical Software 22 (Minitab, LLC, State College, PA, USA). A general linear model (GLM) was applied to evaluate the effects of the experimental factors. Prior to analysis, data were tested for normality using the Ryan–Joiner test (α = 0.05) and for homogeneity of variances using Levene’s test (α = 0.05). When significant effects were detected, treatment means were separated using Tukey’s test (p ≤ 0.05).

3 Results

3.1 Dynamics of the PhACs and metabolites in water-soil-plant system

3.1.1 PhACs and metabolites in FW and TWW

Table 4 shows the concentrations of PhACs and their metabolites detected in TWW used for irrigation.

Table 4. Occurrence and concentrations of PhACs and transformation products in fresh water (FW) and treated wastewater (TWW) used for irrigation.

LOQ, Limit of Quantification.

Bold values indicate the concentrations of the parent pharmaceutical active compounds (PhACs).

No PhACs were detected in FW, whereas several PhACs and related metabolites were identified in the TWW collected during the experimental campaign, with concentrations generally below 1 μg L^-1. Among the detected compounds, clarithromycin (CLR) and diclofenac (DCF) showed the highest concentrations, reaching 0.57 and 0.41 μg L⁻¹, respectively, whereas CLZ was detected only at trace levels. Intermediate concentrations were observed for carbamazepine (CBZ) and metoprolol (MTP), averaging approximately 0.20 μg L⁻¹. Notably, for CBZ, some metabolites were detected at concentrations higher than those of the parent compound.

3.1.2 PhACs and metabolites in soil

Analyses performed before the onset of the experiment confirmed the absence of PhACs residues in the soil used for the study (data not shown). At the end of the experimental period, only CBZ and CLR, together with some of their metabolites, were detected in soils irrigated with TWW, despite the broader range of compounds identified in the irrigation water (Figure 2).

Overall, CLR concentrations were slightly higher than those of CBZ across all treatments, with a mean difference of 1.16 ng g^-1. In addition, the CBZ metabolites 3OH-CBZ and ACD consistently accumulated at higher concentrations than the parent compound, with concentrations ranging from 1.22 to 12.30 ng g^-1 and from 3.86 to 7.70 ng g^-1, respectively. Conversely, no marked differences were observed between CLR and its metabolite NCLR, whose concentrations remained generally comparable among treatments, ranging from 1.18 to 4.17 ng g^-1. With respect to the irrigation regime, the concentrations of all detected PhACs and metabolites increased significantly, with the highest accumulation observed under the I100 treatment in M− soils. Conversely, microbial consortium inoculation (M+) significantly reduced contaminant concentrations under most irrigation regimes compared with the corresponding non-inoculated treatments (M−).

3.1.3 PhACs and metabolites in plants organs

The concentrations detected across the different plant organs highlighted a selective uptake of specific PhACs and matabolites, along with distinct translocation patterns among plant tissues. At the end of the experimental trial, among the compounds detected in the soil, only CBZ and two metabolites (3OH-CBZ and ACD) were translocated to the roots and straw of wheat plants (Figures 3, 4).

Specifically, in the roots of plants irrigated with the I30, CBZ concentrations were below the limit of quantification (LOQ), as was CBZ-diol, a metabolite present in the TWW used for irrigation but not detected in the soil. The concentration of CBZ increased significantly and proportionally with increasing irrigation volume, with mean values rising from 1.38 ng g^-1 under I50 to 2.15 ng g^-1 under I70 and 4.43 ng g^-1 under I100. A similar trend was observed for the other detected metabolites, although 3OH-CBZ exhibited markedly higher concentrations than the other compounds, ranging from 25.60 to 49.36 ng g^-1. Moreover, consistent with the results observed in the soil, roots treated with the microbial consortium showed significantly lower concentrations compared with the untreated control, particularly under I70 and I100 conditions.

In straw, CBZ was below the limit of quantification across all treatments, in contrast to its metabolites. Concentrations of CBZ-diol and ACD were slightly higher than those observed in roots, whereas 3OH-CBZ was significantly lower. In addition, as observed for roots, contaminant levels in straw increased proportionally with irrigation volume and were significantly reduced following application of the microbial consortium. Finally, all investigated PhACs and metabolites were below their respective LOQs in the grains.

3.1.4 Bioconcentration and translocation of CBZ-related compounds

The bioconcentration factor (BCF) and transfer factor (TF), calculated considering the sum of CBZ and its transformation products (ΣCBZ), highlighted different accumulation and translocation patterns within wheat plants (Table 5).

Table 5. Bioconcentration (BCF) and translocation (TF) factors of CBZ and its transformation products (ΣCBZ) in wheat under different irrigation and microbial treatments.

Values are given as mean ± standard deviation (SD). Water regime (I100 = 100% FC; I70 = 70% I100; I50 = 50% I100; I30 = 30% I100), microbial consortium (M− = absence; M+ = presence).

The BCF values ranged from 4.16 to 7.46, indicating a greater relative accumulation of ΣCBZ in plant tissues compared with soil concentrations. The highest BCF value was observed in the I50 treatment inoculated with the microbial consortium (7.46), whereas the lowest value was recorded in M− plants under I70 conditions (4.16). Overall, inoculated plants showed higher BCF values under I50 and I70 irrigation regimes, while only minor differences were observed under I100 conditions.

The TF values ranged from 0.24 to 0.60, confirming the limited translocation of ΣCBZ from roots to straw. The lowest TF values were observed in M+ plants under I50 (0.24) and I70 (0.33) treatments, whereas the highest translocation was recorded in M− under I70 conditions (0.60). Under I100 irrigation, inoculation resulted in a slight increase in TF values (0.56 compared with 0.45 in M−), despite the lower absolute concentrations of ΣCBZ observed in plant tissues.

3.2 Leaf physiological responses

The analysis of variance (ANOVA), performed on data collected at the heading stage (BBCH 51), showed that SPAD was not statistically significant for any main effect or interaction. Regarding leaf gas exchange parameters, including net photosynthetic rate (A), transpiration rate (E), and stomatal conductance (gs), the three-way interaction among water quality (WQ), water regime (WR), and microbial consortium (MC) was not statistically significant. However, a significant effect was observed for the two-way interaction between WQ and WR (Table 6).

Table 6. Analysis of variance (ANOVA) of main effects and interactions on SPAD and leaf gas exchange parameters.

Values are given as F of Fisher. *, and * indicate significant at p< 0.001, p< 0.01 and p< 0.05, respectively. ns, not significant; df, Degrees of Freedom; WQ, water quality; WR, water regime; MC, microbial consortium; A, net photosynthetic rate; E, transpiration rate; gs, stomatal conductance.

Net photosynthetic rate (A; Figure 5A) ranged between 32.4 and 16.0 µmol m^-2 s^-1. All treatments maintained similarly high assimilation rates, except for TWW irrigated plants at I100 and I70, which exhibited an approximately twofold reduction compared with the highest values. Regarding transpiration rate (E; Figure 5B), values ranged from 3.5 to 1.5 mmol m^-2 s^-1. The highest value was recorded in the FW × I100, whereas the lowest occurred in FW × I30. Intermediate values were observed in FW × I50 and FW × I70. In contrast, TWW-treated plants showed consistently lower and relatively stable transpiration rates across irrigation levels, with an average value of approximately 1.7 mmol m^-2 s^-1. Finally, stomatal conductance (g_s; Figure 5C) followed a pattern similar to transpiration, with the highest mean value recorded in FW × I100 (0.62 mol m^-2 s^-1) and the lowest in FW × I30 (0.16 mol m^-2 s^-1). In FW treatments, g_s decreased progressively with increasing water stress. In contrast, TWW-treated plants showed a different trend, with I100 and I70 exhibiting similar low values (~0.36 mol m^-2 s^-1), comparable to those observed under more limiting conditions.

3.3 Yield data

The analysis of variance (ANOVA) (Table 7) revealed highly significant effects (p< 0.01) of the interaction among WQ × WR × MC on yield parameters, including total aboveground biomass and grain yield. For yield-related variables, the I30 treatments were excluded from the analysis due to complete plant mortality before harvest.

Table 7. Analysis of variance (ANOVA) of main effects and interactions on yield parameters.

Values are given as F of Fisher. *, and * indicate significant at p< 0.001, p< 0.01 and p< 0.05, respectively. ns, not significant; df, Degrees of Freedom; WQ, water quality; WR, water regime; MC, microbial consortium.

The highest yield levels were recorded under full irrigation (I100), regardless of water quality and microbial consortium application, with biomass ranging from 66.4 to 77.9 q ha^-1 and grain yield from 17.9 to 24.3 q ha^-1 (Figure 6).

An exception was observed in the FW × I100 × M+ treatment, which showed lower values compared with the mean of all other I100 treatments, with reductions of approximately 35% and 50% in biomass and grain yield, respectively. This reduction was not associated with physiological stress symptoms, which remained within the expected range. Instead, it was attributed to an accidental reduction in plant density (fewer surviving/harvested plants per pot) uniquely observed in this specific treatment during sampling, whereas all other experimental pots across all treatments maintained a uniform density of 13 plants per pot until harvest. A progressive decline in productivity was observed with decreasing irrigation levels. The lowest yields were recorded under severe water deficit, particularly in FW × I50 × M+, with biomass and grain yields of 21.93 and 3.00 q ha^-1, respectively.

Overall, microbial consortium application did not result in yield enhancement; in several cases, M− treatments showed equal or higher productivity, particularly under FW irrigation. Reducing irrigation from 100% to 70% of full irrigation resulted in yield reductions exceeding 40% for biomass and 55% for grain yield, while further reduction to 50% led to losses of up to 60% in biomass and more than 70% in grain yield.

4 Discussions

The fate of PhACs and their transformation products in the soil–plant system after TWW irrigation is controlled by interconnected transport, retention, and degradation processes, which are influenced by soil properties, contaminant characteristics, crop species, and environmental conditions (5, 8, 11, 18, 34).

Against this background, the combined assessment of irrigation water quality, irrigation regime, and microbial inoculation allowed us to investigate the fate of PhACs along the water–soil–plant–grain continuum, highlighting the role of these factors in shaping contaminant distribution, persistence, and potential transfer in durum wheat, while simultaneously evaluating crop performance.

Of the 11 compounds detected in TWW, including both parent compounds and transformation products, a marked reduction in the number of detectable contaminants was observed across the analyzed matrices at the end of the growing season. In contrast, the residual concentrations exhibited greater variability and were influenced by the different treatments. Among carbamazepine (CBZ), clarithromycin (CLR), climbazole (CLZ), diclofenac (DCF), and metoprolol (MTP), only CBZ, CLR, and their respective transformation products were detected in the soil. In plant biomass, only CBZ and its transformation products were detected, whereas the concentrations of all investigated compounds were below their respective LOQs in the grain. The absence of DCF, CLZ, and MTP in the sampled matrices suggests that their physicochemical properties, together with the characteristics of the experimental soil, limited their persistence within the system. The sandy loam soil used in this study, characterized by a low organic matter content (1.2%), has a relatively limited sorption capacity for many organic contaminants, thereby promoting their dissipation through degradation, transport, and leaching processes (18, 35).

Despite being present in TWW at relatively high concentrations (0.41 μg L^-1), DCF was not detected in either soil or plant tissues at the end of the experiment. This finding is consistent with the predominantly anionic nature of the molecule under the pH conditions typically found in agricultural soils, which reduces its interaction with soil particles, particularly in soils with low clay and organic matter contents (18, 36). In addition, DCF is characterized by relatively rapid degradation in soil, with reported half-lives (DT₅₀) ranging from a few days to several weeks depending on environmental conditions (5, 37, 38).

Similarly, CLZ, which was detected in TWW at particularly low concentrations (0.03 μg L^-1), was likely affected by the combined effect of a limited initial input and relatively rapid degradation, with a reported half-life of approximately 5.3 days under aerobic conditions (39), resulting in residual concentrations below the analytical detection limit. Likewise, although MTP occurred in TWW at intermediate concentrations (0.23 μg L^-1), it was not detected in either soil or plant tissues. Its high biodegradability, together with its relatively high mobility in the soil matrix, may have promoted its transport within the soil solution while reducing its residence time in the root zone, ultimately contributing to its complete dissipation over the growing season (4, 19).

In contrast to these compounds, CBZ and CLR remained detectable in the soil until the end of the experiment. However, they exhibited distinct patterns of plant transfer. While both compounds persisted within the soil–plant system, only CBZ was detected in plant tissues. This difference is likely attributable to their contrasting physicochemical properties, which govern their mobility, bioavailability, and ability to cross root membranes.

The higher abundance of CBZ-related residues in the soil, considering both the parent compound and its transformation products, likely reflects both the greater initial input through TWW and the ability of CBZ and its transformation products to persist in the soil environment. However, the initial concentration alone does not appear sufficient to explain the contrasting behavior observed between CBZ and CLR. Carbamazepine is predominantly present in its neutral form under typical soil pH conditions and is characterized by a relatively low adsorption coefficient (Kd = 12–20 L kg^-1) and limited susceptibility to biodegradation. These properties favor its persistence in the soil solution and facilitate subsequent root uptake (5, 32). In contrast, although CLR persisted in the soil, it exhibited a substantially higher adsorption coefficient (Kd = 262–400 L kg^-1), lower mobility within the soil solution, and a larger molecular size, all of which are likely to reduce its bioavailability for plant uptake (5, 37, 40).

The absence of CLR in plant tissues is not necessarily inconsistent with previous studies reporting its uptake by different crop species. In most cases, however, these studies were conducted under controlled conditions or involved PhACs concentrations substantially higher than those typically detected in TWW. For example, Tian et al. (41) reported CLR accumulation in lettuce irrigated with solutions containing 10 μg L^-1 of the compound, a concentration several orders of magnitude higher than those commonly found in TWW. These findings suggest that, in addition to the intrinsic physicochemical properties of the molecule, the concentration effectively available in the soil is also a key determinant of plant uptake. Supporting this interpretation, Gatta et al. (37) recently reported that residual concentrations of contaminants of emerging concern remaining in the soil after TWW irrigation during a previous cropping cycle, although still detectable (0.25 ng g^-1 for CBZ and 5.54 ng g^-1 for CLZ), were insufficient to promote root uptake and subsequent translocation into wheat tissues. In the present study, the combination of low initial CLR concentrations and its limited bioavailability likely reduced plant transfer to levels below the analytical detection limit.

The combined effects of irrigation regime and microbial inoculation were associated with an overall reduction in the residual concentrations of CLR and its transformation product, N-desmethylclarithromycin (NCLR), in the soil at the end of the growing season. In particular, compared with the highest irrigation treatment, residual concentrations of both compounds decreased by approximately 50% and 30% under the I50 and I70 treatments, respectively, whereas no quantifiable residues were detected under the I30 regime. A similar trend was also observed for CBZ and its transformation products.

The lower residual concentrations observed under reduced irrigation volumes can be explained, at least in part, by the lower overall contaminant load introduced into the system. Indeed, assuming similar contaminant concentrations in the irrigation water and the same number of irrigation events, reducing the applied water volume proportionally decreases the total mass of contaminants delivered to the soil.

Further insight into the transformation processes affecting CLR during the experiment was provided by the CLR: NCLR ratio. In TWW, the CLR: NCLR ratio was approximately 5:1, indicating the predominance of the parent compound over its main transformation product. By the end of the growing season, this ratio had markedly decreased, approaching more balanced values and suggesting the progressive transformation of CLR into NCLR during its residence in the soil. This pattern was particularly evident in the non-inoculated soils, where the residual concentrations of NCLR closely matched those of the parent compound. These findings indicate that the formation of NCLR likely represented an important transformation pathway governing the environmental fate of CLR.

In soils inoculated with beneficial microorganisms, CLR concentrations were comparable to those in the non-inoculated treatments, whereas NCLR concentrations were significantly lower, suggesting that microbial inoculation may have influenced the fate of NCLR rather than that of the parent compound. Although information on the environmental fate of N-desmethylclarithromycin remains limited, previous studies have shown that CLR undergoes microbial transformation in both soils and aquatic environments, leading to the formation of metabolites that may persist in the environment for extended periods (15).

Unlike CLR, carbamazepine exhibited a more complex behavior, characterized by the occurrence of several transformation products in both soil and plant tissues. Comparison between the contaminant profile detected in TWW and that observed at the end of the growing season revealed that the composition of CBZ transformation products changed during their residence in the soil–plant system. In particular, in addition to the compounds already detected in TWW, acridine (ACD) was identified, indicating that both the parent compound and its transformation products continued to undergo transformation after TWW application.

The distribution of the sum of CBZ and its transformation products (ΣCBZ) across the different compartments of the system revealed a clear concentration gradient (roots > straw > soil), with concentrations ranging from 29.10 to 69.23 ng g^-1 in roots, 8.40 to 31.23 ng g^-1 in straw, and 5.87 to 22.57 ng g^-1 in soil. Overall, these results indicate that, in durum wheat, CBZ accumulation occurs predominantly in the roots, whereas its transfer to aboveground tissues is strongly restricted. This interpretation is supported by the translocation factor (TF), which ranged from 0.24 to 0.60, indicating limited mobility of ΣCBZ from the roots to the aboveground tissues regardless of the experimental treatments.

Although greater water inputs may theoretically enhance xylem transport by increasing the transpiration stream (18, 21), CBZ translocation appears to be primarily controlled by physiological barriers at the root level, which regulate its entry into the vascular system and restrict its movement to aboveground tissues (21).

In this context, the observation that quantifiable residues under the I30 treatment were detected exclusively in roots further supports the role of the root system as the primary sink for carbamazepine in durum wheat. It also suggests that translocation to the aboveground tissues is governed not only by the intrinsic physicochemical properties of the molecule but also by its availability within the soil–root compartment, becoming appreciable only once a minimum concentration threshold is exceeded (5).

More generally, this behavior was reflected in the bioconcentration factor (BCF), which ranged from 4.16 to 7.46 and was generally higher in the inoculated treatments. The increase in BCF indicates a greater reduction in contaminant concentrations in the soil than in plant tissues, reflecting a higher relative accumulation efficiency in the roots.

Differences in microbial inoculation treatments were also observed for the translocation factor (TF) of ΣCBZ, with a reduction in TF of approximately 40% under the I50 and I70 irrigation regimes, consistent with greater root retention. Conversely, under the I100 treatment, microbial inoculation was associated with a slight increase in TF despite lower absolute ΣCBZ concentrations, suggesting that its effect may have been related to the interaction among irrigation regime, contaminant bioavailability, and microbial activity.

Although microbial inoculation was associated with lower overall concentration of ΣCBZ across all irrigation regimes, the difference was more pronounced for the transformation products than for the parent compound. Indeed, whereas CBZ concentrations varied only slightly among treatments, its transformation products showed a marked decline. This pattern may indicate a greater sensitivity of the transformation products to processes occurring in the soil–plant system under microbial inoculation. However, because degradation and microbial activity were not directly measured, alternative processes, such as changes in sorption, contaminant availability, plant uptake, or water-mediated transport, cannot be excluded. In addition, dilution associated with greater plant growth may also have contributed to the lower residual concentrations observed, rather than microbial attenuation alone.

A similar pattern was observed for clarithromycin, where inoculation mainly affected the transformation product N-desmethylclarithromycin (NCLR) rather than the parent compound. This pattern suggests that transformation products may represent more readily degradable molecules than their parent compounds, likely because they have already undergone structural modifications during primary transformation processes.

Among all the transformation products analyzed, 3-hydroxycarbamazepine (3OH-CBZ) was the predominant compound in both soil and plant tissues, consistent with its high initial concentration in TWW. The highest concentrations were observed in the non-inoculated I70 and I100 treatments, whereas a marked reduction was detected in the corresponding inoculated treatments, indicating that 3OH-CBZ accounted for a substantial proportion of the reduction in ΣCBZ associated with microbial inoculation. Accordingly, considering ΣCBZ as a whole, residual concentrations were consistently lower in inoculated treatments than in their respective controls, with average reductions of approximately 35%, 26%, and 22% under the I100, I70, and I50 irrigation regimes, respectively. Taken together, these results suggest that microbial inoculation was associated with lower residual concentrations of ΣCBZ in the soil–plant system. The extensive root system of durum wheat and its well-documented ability to release root exudates may have favored the establishment and activity of the inoculated microbial consortium, including arbuscular mycorrhizal fungi (AMF), which may have influenced the fate of organic contaminants (18, 42–44). Furthermore, the more pronounced effect observed under the I100 treatment suggests that greater water availability, together with the higher contaminant input associated with TWW irrigation, may have affected contaminant bioavailability and microbial activity, potentially contributing to the observed differences in residual contaminant concentrations (15).

The distribution pattern observed in the present study, when compared with previous reports, highlights the strong influence of crop species on the uptake and partitioning of PhACs. Carter et al. (32) reported higher concentrations in leaves than in roots of radish and ryegrass, whereas Herklotz et al. (45) observed greater root accumulation in cabbage. Similarly, Gatta et al. (37) detected carbamazepine residues exclusively in the roots of durum wheat, with no evidence of accumulation in leaves or grain. In contrast, Tanoue et al. (46) documented efficient translocation of carbamazepine to cucumber shoots, while Dodgen et al. (47) found that other contaminants of emerging concern, such as diclofenac and naproxen, accumulated predominantly in the root system.

Finally, the concentrations of all analyzed PhACs in the grain were below their respective LOQs. This finding reinforces previous evidence obtained under environmentally realistic exposure conditions, indicating that, at the concentrations applied in this study, contaminant transfer to the grain is negligible (5, 37, 48). Nevertheless, the possibility of contaminant accumulation under higher exposure levels cannot be excluded, as recently reported by Denora et al. (31).

Regarding crop performance, the physiological response of durum wheat showed that gas exchange parameters, including net photosynthetic rate (A), transpiration (E), and stomatal conductance (gs), measured at the heading stage (BBCH 51), were influenced by the interaction between irrigation water quality and irrigation regime, whereas final grain yield was also affected by microbial inoculation. In particular, under the highest irrigation levels with TWW (I100 and I70), net photosynthetic rate was reduced by approximately 50% compared with the corresponding freshwater (FW) treatments (≈16 μmol m^-2 s^-1). This reduction was accompanied by lower stomatal conductance (gs< 0.30 mol m^-2 s^-1) and transpiration rates comparable to those observed under the most severe water-deficit conditions.

Previous studies have shown that exposure to PhACs may impair plant physiology by reducing root activity, chlorophyll content, and photosynthetic efficiency (13, 14, 49). These effects have been associated with alterations in hormonal homeostasis (50) and interference with electron transport at photosystem II (51, 52). However, such responses have generally been reported at exposure concentrations several orders of magnitude higher than those investigated in the present study.

Given the substantial difference between the exposure levels used in previous studies and those applied here, it is likely that the physiological responses observed were also influenced by other factors associated with the use of TWW. Therefore, although a transient contribution of PhACs exposure cannot be ruled out, the pronounced reduction in gas exchange observed under the I100 and I70 TWW treatments was more likely the result of the combined effects of multiple water quality-related factors. The repeated application of large volumes of TWW, characterized by a non-negligible organic load (COD and BOD) and high microbiological indicators (E. coli > 4,000 CFU 100 mL^-1), may have temporarily altered rhizosphere conditions, affecting root functioning and water transport during the heading stage (BBCH 51), when atmospheric evaporative demand is particularly high.

The absence of yield reductions under the I100 and I70 treatments suggests that the observed physiological responses were largely transient. During this growth stage, increased translocation of contaminants to aboveground tissues may have resulted in temporary accumulation in the leaves, leading to short-term alterations in gas exchange. This effect may subsequently have been mitigated by the progressive transformation and dissipation of contaminants within the soil–plant system, without causing permanent biochemical damage to the photosynthetic apparatus. This interpretation is supported by the absence of negative effects on final grain yield, which remained comparable to that observed in the corresponding FW and inoculated treatments.

Microbial inoculation, however, exhibited a differential response depending on irrigation water quality, negatively affecting yield primarily under FW conditions across irrigation regimes, while showing no yield penalties under TWW application. Rather than implying that native–inoculated microbial interactions occur exclusively under FW, these findings suggest that the functional outcome of such interactions is strongly modulated by the chemical and biological matrix of the irrigation water. In FW systems, where nutrient inputs are lower, the establishment of the inoculated consortium, comprising AMF, Trichoderma, and rhizosphere bacteria, may have imposed a transient metabolic or carbon cost on the host plant, or faced functional antagonism with native soil microorganisms, ultimately leading to non-positive yield effects. Conversely, the continuous supply of organic carbon, nitrogen, and nutrients provided by TWW may have mitigated these competitive trade-offs by providing additional substrates, thereby supporting both native and introduced microbial populations without impairing plant biomass and grain production. Nevertheless, because these responses can be highly specific to the commercial formulation used, further research is needed to unravel the specific strain-level interactions driving these crop responses.

Although soil microbial dynamics, including interactions between the inoculated consortium and native microbial communities, the detailed behavior of contaminants of emerging concern within the rhizosphere, and leachate composition were not investigated, the overall experimental evidence provides a robust basis for interpreting contaminant fate within the soil–plant system and for assessing the role of durum wheat in contaminant accumulation and potential transfer along the food chain. The absence of these measurements represents a limitation of the present study; however, it does not compromise the overall interpretation of the results, which is consistent with the available literature.

5 Conclusions

Among the contaminants detected in TWW, only CBZ and CLR persisted in the soil at the end of the cropping cycle, whereas CBZ was the only compound detected in plant tissues. The contrasting behavior observed confirms that contaminant persistence and transfer are compound-specific and are modulated by the management conditions of the system.

Durum wheat showed a marked ability to retain carbamazepine and its transformation products in the root compartment, limiting their translocation to aboveground tissues, while their concentrations in grain remained below the respective LOQs at the exposure levels investigated.

This pattern was also observed under microbial inoculation, which was associated with greater root retention and lower residual concentrations of transformation products, potentially reflecting a complementary role of plants and microorganisms in influencing contaminant fate within the soil–plant system. The lower residual concentrations observed under microbial inoculation may also reflect dilution associated with improved plant growth or changes in contaminant sorption, rather than microbial attenuation alone.

Contaminant transfer to aboveground tissues was strongly species-dependent but was also influenced by the overall contaminant load within the system, indicating that irrigation management may represent a useful tool for modulating contaminant persistence and reducing their potential for bioaccumulation.

At the exposure levels investigated, the use of TWW did not result in significant negative effects on crop growth and productivity, despite the transient alterations in gas exchange observed under treatments receiving the highest irrigation volumes.

Further studies are needed to elucidate the mechanisms underlying the retention and transformation of persistent contaminants, particularly CBZ, and to evaluate the potential application of phytomanagement strategies based on the selection of genotypes with enhanced contaminant containment capacity.

Overall, the results obtained under the controlled conditions of this single-season pot experiment support the potential of TWW reuse, combined with appropriate irrigation management, microbial inoculation strategies, and the selection of crop species capable of limiting contaminant transfer to edible fractions, as a sustainable approach for the valorization of non-conventional water resources. However, the short-term and controlled nature of the experiment, including the use of a rainout shelter, limited pot volume, and absence of mineral fertilization, should be considered when extrapolating these findings to field conditions. Long-term monitoring and field-scale studies remain essential to confirm the environmental sustainability and food safety implications of TWW reuse under operational agricultural conditions.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fsoil.2026.1954651/full#supplementary-material.

Supplementary material Supplementary material at the source

Footnotes

Edited by: Jianxu Wang, Chinese Academy of Sciences (CAS), China

Reviewed by: Davide Staedler, Université de Lausanne, Switzerland

Elida Nora Ferri, University of Bologna, Italy