Plasma-activated water reprograms the root–rhizosphere continuum to enhance nitrogen use efficiency and sustain maize productivity under reduced nitrogen input

Abstract

Improving nitrogen use efficiency (NUE) while sustaining crop productivity is a major challenge for sustainable maize production. Plasma-activated water (PAW), enriched with reactive oxygen and nitrogen species, has emerged as a promising biostimulant for enhancing nutrient acquisition and plant performance. This study evaluated the effects of PAW on root architecture, rhizosphere metabolism, nitrogen acquisition, agronomic nitrogen use efficiency (ANUE), and grain yield (GY) under four nitrogen application rates (0, 50, 75, and 100% of the recommended dose) in maize grown under controlled polyhouse conditions. PAW significantly increased root surface area (RSA), root hair density (RHD), root hair length (RHL), total organic carbon (TOC), organic acid secretion (OAS), free amino acid (FAA) accumulation, nitrogen uptake, GY, harvest index (HI), and ANUE compared with normal water irrigation (P ≤ 0.001). Field Emission Scanning Electron Microscopy (FESEM) revealed pronounced root hair proliferation and elongation under PAW, particularly at 75% nitrogen, while High-Performance Liquid Chromatography (HPLC) demonstrated greater accumulation of citric, malic, succinic, and oxalic acids, indicating enhanced rhizosphere carbon metabolism and nutrient mobilization. Multivariate analyses identified coordinated regulation of root and rhizosphere traits as the principal driver of productivity. Principal component analysis explained 95.2% of the total variation and clearly separated PAW + 75% N and PAW + 100% N from the remaining treatments. Pearson correlation and hierarchical clustering analyses demonstrated strong positive associations among RSA, rhizosphere metabolites, nitrogen uptake, GY, and HI, confirming an integrated root–rhizosphere functional network. Piecewise latent path analysis revealed that improved root architecture enhanced rhizosphere activity (β = 0.631), leading to improved nitrogen acquisition (β = 0.697), and ultimately higher grain productivity (β = 1.043). Random Forest analysis identified nitrogen uptake (27.99%), HI (22.95%), FAA (19.27%), TOC, and RSA as the strongest predictors of GY. PAW + 75% N achieved the highest Treatment Performance Index, outperforming PAW + 100% N and sustaining superior productivity with 25% lower fertilizer nitrogen application. These findings demonstrate that PAW reprograms the root–rhizosphere–nitrogen acquisition continuum to maximize NUE, providing a mechanistic framework for improving nitrogen acquisition under controlled polyhouse conditions; field validation is required before broad agronomic application.

1 Introduction

Improving nitrogen (N) use efficiency (NUE) while sustaining crop productivity is a major challenge for modern agriculture. Maize (Zea mays L.), one of the world’s most important cereal crops, requires substantial N inputs to support photosynthesis, protein synthesis, and grain development. However, crop recovery of applied fertilizer N rarely exceeds 50%, with considerable losses through leaching, volatilization, and denitrification. These losses reduce fertilizer efficiency, increase production costs, and contribute to environmental pollution and greenhouse gas emissions (Fageria and Baligar, 2005; Rafique, 2020). Therefore, improving the efficiency of N acquisition rather than increasing fertilizer inputs has become a key objective of sustainable crop production.

Nitrogen acquisition is regulated by the coordinated functioning of the root–rhizosphere continuum, where root architecture and rhizosphere metabolism collectively determine nutrient availability and plant uptake. Root system architecture (RSA), particularly root surface area and root hair development, enhances soil exploration and the absorptive interface for nutrient acquisition (Liu, 2021). Simultaneously, roots release organic carbon compounds, organic acids, and amino acids that stimulate microbial activity, enhance nutrient solubilization, and regulate nitrogen cycling in the rhizosphere (Ahmad et al., 2022). Together, these complementary processes improve nutrient acquisition and crop productivity (Škarpa et al., 2020). However, despite their close physiological association, root architectural traits and rhizosphere biochemical processes have rarely been investigated as an integrated mechanism regulating NUE.

Plasma-activated water is an emerging sustainable technology that offers new opportunities to regulate the root–rhizosphere system. Produced by non-thermal atmospheric plasma, PAW contains long-lived reactive oxygen species (ROS) and reactive nitrogen species (RNS), including hydrogen peroxide (H₂O₂), nitrite (NO₂^-), and nitrate (NO₃^-), which modify water chemistry and act as biologically active signaling molecules (Thirumdas et al., 2018; Miranda et al., 2023). These reactive species regulate cell division, root elongation, developmental plasticity, and nitrogen assimilation while maintaining cellular redox homeostasis (Lu et al., 2016; Mittler, 2017). ROS/RNS signaling also activates mitogen-activated protein kinase (MAPK) pathways that integrate growth, carbon metabolism, and stress adaptation (Jagodzik et al., 2018; Muhammad et al., 2021; Bao et al., 2025; Sozoniuk et al., 2026). Furthermore, plasma–water interactions generate plant-available nitrate and nitrite through atmospheric nitrogen fixation, establishing a direct connection between PAW chemistry and nitrogen nutrition (Volkov et al., 2021).

The beneficial effects of PAW on crop performance have been widely documented. Plasma treatment improves seed germination, seedling vigor, photosynthetic efficiency, nutrient uptake, and abiotic stress tolerance across diverse crops (Thirumdas et al., 2018; Sivachandiran and Khacef, 2017). In maize, PAW enhances root elongation, vascular differentiation, chlorophyll accumulation, enzyme activities, and plant growth (Mogo et al., 2022; Holubová et al., 2021; Lukacova et al., 2021). Dahal et al. (2024) further demonstrated improved nutrient acquisition through optimized PAW, while recent studies recognize PAW as a next-generation biostimulant that enhances plant growth, defense responses, and resource-use efficiency (Sozoniuk et al., 2026). However, previous studies have focused primarily on seed germination, vegetative growth, or aboveground physiological responses, with limited attention to the coordinated regulation of root architecture and rhizosphere biochemistry under different nitrogen regimes and their combined contribution to nitrogen uptake, agronomic nitrogen use efficiency (ANUE), and grain productivity (Ahmad et al., 2022; Bao et al., 2025).

We hypothesized that PAW enhances NUE by reprogramming the root–rhizosphere–nitrogen acquisition continuum, leading to coordinated improvements in root development, rhizosphere biochemical activity, and nutrient acquisition under reduced nitrogen supply. Accordingly, the objectives of this study were to (i) evaluate the effects of PAW on root architectural traits and rhizosphere biochemical characteristics under contrasting nitrogen application rates, (ii) determine their relationships with nitrogen uptake, ANUE, and grain yield, and (iii) identify the key physiological and biochemical drivers governing maize productivity. This study provides a mechanistic framework linking plasma-derived signaling, root function, rhizosphere metabolism, and nitrogen acquisition under controlled polyhouse conditions, providing a foundation for future field validation of PAW for improving nitrogen use efficiency in maize.

2 Materials and methods

2.1 Experimental site and growth conditions

The experiment was conducted during the Rabi 2025 season under controlled polyhouse conditions at the School of Agriculture, SR University, Warangal, Telangana, India (17.9689° N, 79.5941° E; 302 m above mean sea level). The polyhouse was equipped with automated ventilation and insect-proof screens to maintain a uniform growing environment throughout the experimental period. Mean day and night temperatures were maintained at 28 ± 3 °C and 20 ± 2 °C, respectively, with relative humidity ranging from 60 to 75%. Photosynthetically active radiation (PAR) during peak daylight hours ranged from 800 to 1200 μmol m^-² s^-¹, providing favorable conditions for maize growth. The controlled polyhouse grow-bag system was used to minimize environmental variability and facilitate mechanistic evaluation of PAW-induced changes in root architecture, rhizosphere processes, and nitrogen acquisition. However, because root growth was confined to a fixed soil volume and plant-to-plant interactions were absent, the findings should be regarded as a mechanistic proof-of-concept and require validation under diverse field conditions before broader agronomic application.

2.2 Experimental design and crop management

A widely cultivated single-cross maize hybrid, DHM 206 (Telangana Makka-3), was used as the experimental material collected from the Agricultural Research Station, Karimnagar, Professor Jayashankar Telangana Agricultural University (PJTAU), Telangana, India. Seeds were surface sterilized with 1% sodium hypochlorite for 5 min, thoroughly rinsed with distilled water, and subsequently subjected to their respective priming treatments before sowing. Plants were grown individually in UV-stabilized polyethylene grow bags (45 × 50 cm; approximately 60 L capacity) containing 20 kg of air-dried sandy-loam soil sieved through a 2-mm mesh. The initial soil properties were pH 7.2, electrical conductivity 0.32 dS m^-¹, organic carbon 0.54%, available nitrogen 218 mg kg^-¹, available phosphorus 24 mg kg^-¹, and available potassium 286 mg kg^-¹.

The experiment was arranged in a factorial randomized complete block design (RCBD) with two irrigation-water sources and four nitrogen application levels; each replicated five times. The irrigation treatments consisted of normal water (NW) and plasma-activated water (PAW), whereas nitrogen was supplied at 0, 50, 75, and 100% of the recommended dose, corresponding to 0, 100, 150, and 200 kg N ha^-¹, respectively. The resulting treatment combinations were: NW + 0% N, NW + 50% N, NW + 75% N, NW + 100% N, PAW + 0% N, PAW + 50% N, PAW + 75% N, and PAW + 100% N. Three seeds were sown in each grow bag and thinned to one healthy and uniform plant ten days after emergence. Phosphorus and potassium were uniformly applied as basal fertilizers at 80 kg P₂O₅ ha^-¹ and 60 kg K₂O ha^-¹ using single superphosphate and muriate of potash, respectively. Nitrogen was supplied as urea in three equal splits at sowing, the V6 (six-leaf) stage, and the VT (tasseling) stage to ensure adequate nutrient availability throughout vegetative and reproductive growth. Representative maize plant phenotypes under different irrigation-water and nitrogen treatments at the V6, VT, and R6 growth stages are presented (Figure 1).

Figure 1

2.3 Preparation of plasma-activated water and irrigation management

Plasma-activated water was generated using an atmospheric-pressure dielectric barrier discharge (DBD) cold plasma reactor operated under ambient air conditions following the procedure described by Thirumdas et al. (2018) and Miranda et al. (2023). One liter of distilled water was exposed to plasma discharge at 18–20 kV, 20 kHz, for 20 min, producing a solution enriched with reactive oxygen and reactive nitrogen species (RONS). Fresh PAW was prepared immediately before each irrigation event to preserve its physicochemical properties and biological activity. PAW was applied immediately after preparation without pH adjustment, additional cooling, or ozone purging. Under the operating conditions employed, no noticeable increase in water temperature was observed, and residual ozone dissipated naturally during the brief interval between PAW preparation and irrigation. Immediately after plasma activation, PAW was characterized for pH, electrical conductivity (EC), oxidation–reduction potential (ORP), H₂O₂, NO₃^-, and NO₂^- using standard analytical procedures (Table 1).

For seed priming, seeds assigned to PAW treatments were immersed in freshly prepared PAW for 6 h, whereas control seeds were soaked in distilled water for the same duration. Following seedling establishment, PAW served as the exclusive irrigation source for the designated treatments, while control plants received normal water throughout the experiment. Soil moisture was maintained at 75–80% field capacity using gravimetric monitoring, and equal irrigation volumes were applied to all treatments to ensure that observed responses resulted from water quality rather than differences in water quantity. Based on the measured concentrations of NO₃^- (81.5 mg L^-¹) and NO₂^- (7.6 mg L^-¹), PAW contained approximately 20.7 mg inorganic N L^-¹. Assuming a cumulative irrigation volume of approximately 10 L plant^-¹ during the experiment, PAW supplied approximately 0.207 g inorganic N plant^-¹, representing less than 1% of the fertilizer nitrogen applied in the 100% N treatment. Therefore, this contribution was considered negligible and was not included in the fertilizer nitrogen budget or agronomic nitrogen use efficiency (ANUE) calculations.

2.4 Root morphological and rhizosphere biochemical analysis

Root morphological and rhizosphere biochemical traits were evaluated at the V6 (six-leaf) and VT (tasseling) growth stages to characterize the effects of plasma-activated water and nitrogen supply on belowground development and nutrient acquisition.

2.4.1 Root morphological analysis

Prior to exudate collection, root systems were gently washed with distilled water to remove adhering soil particles while minimizing mechanical disturbance. Only visually intact root systems without visible tissue injury or root breakage were selected for exudate collection to minimize cytoplasmic leakage and ensure that the collected solution primarily represented root exudates. Root surface area (RSA; cm² plant^-¹) was quantified using an Epson Expression 12000XL root scanner coupled with RhizoVision Explorer/WinRHIZO image analysis software. Entire root systems were scanned at 600 dpi, and RSA was automatically calculated from root length and diameter distributions.

Root hair characteristics were determined using field-emission scanning electron microscopy (FESEM). Approximately 1cm root segments were collected from actively growing lateral roots located 3–5 cm behind the root apex and fixed in 2.5% glutaraldehyde prepared in 0.1 M phosphate buffer (pH 7.2) for 24 h at 4 °C. Samples were subsequently post-fixed in 1% osmium tetroxide, dehydrated through a graded ethanol series, critical-point dried, sputter-coated with gold–palladium, and examined using FESEM operated at 3–5 kV. Root hair density (RHD; hairs mm^-²) was determined by counting the number of root hairs within a defined root surface area from FESEM micrographs. Root hair length (RHL; µm) was measured using ImageJ software, and at least 20 root hairs were analyzed per biological replicate. Mean values were used for subsequent statistical analyses.

2.4.2 Root exudate collection and biochemical analysis

Root exudates were collected immediately after root sampling. Intact root systems were transferred to sterile glass containers containing 100 mL of 0.5 mM CaCl₂ solution and incubated in darkness for 4 h at 25 ± 2 °C. The collected exudates were filtered through 0.22µm membrane filters and stored at −80 °C until analysis. The solution-based exudate collection method was used to estimate the potential exudation capacity of maize roots under standardized conditions. It does not directly quantify metabolite concentrations in the soil rhizosphere, where adsorption, microbial turnover, and diffusion influence exudate dynamics.

Total organic carbon (TOC; mg C L^-¹) was determined using a catalytic combustion total organic carbon analyzer equipped with a non-dispersive infrared detector following acidification to remove inorganic carbon. Carbon concentration was quantified using potassium hydrogen phthalate standards following the methods of Jones and Willett (2006).

Total organic acids (TOA; µg mL^-¹) were quantified by high-performance liquid chromatography (HPLC) using a reverse-phase C18 column with 25 mM KH₂PO₄ (pH 2.5) as the mobile phase at a flow rate of 1.0 mL min^-¹. Organic acids were detected at 210 nm, identified using authentic standards, and expressed as the combined concentration of oxalic, malic, citric, and succinic acids (Cawthray, 2003). The HPLC analytical characteristics, including the retention times of the identified organic acids, are provided in Supplementary Table 1.

Free amino acids (FAA; µg mL^-¹) were determined using the ninhydrin colorimetric method. Filtered exudates were reacted with freshly prepared ninhydrin reagent, incubated in a boiling water bath for 20 min, and absorbance was recorded at 570 nm. Amino acid concentration was calculated from an L-leucine standard curve following Yemm et al. (1955).

2.5 Yield, nitrogen uptake and agronomic nitrogen use efficiency

Plants were harvested at physiological maturity (R6 stage) and separated into grain, shoot, and root fractions. Plant materials were oven-dried at 70 °C until constant weight, and grain yield (GY; g plant^-¹) was recorded after shelling and standardized to 14% moisture content.

Harvest Index (HI; g g^-¹) was calculated following the method of Donald and Hamblin (1976):

Total nitrogen concentration in grain, shoot, and root tissues was determined using the Kjeldahl digestion method (Bremner, 1965). Nitrogen Uptake (N Uptake; g plant^-¹) was calculated as the product of tissue dry weight and nitrogen concentration and subsequently summed across all plant components:

Agronomic Nitrogen Use Efficiency (ANUE; g grain g^-¹ N applied) was estimated according to Fageria and Baligar (2005):

Where, is the grain yield under nitrogen fertilization, is the grain yield of the corresponding zero-nitrogen treatment, and is the quantity of nitrogen fertilizer applied.

2.6 Statistical analysis

All statistical analyses were performed in R (version 4.5.0; R Core Team, 2025) unless otherwise specified. Root morphological and rhizosphere biochemical traits measured at the V6, and VT growth stages were analyzed using Type III three-way ANOVA. Growth stage, irrigation-water source [normal water (NW) and plasma-activated water (PAW)], Nitrogen level (0, 50, 75, and 100% of the recommended N dose), and their interactions were considered fixed effects (Piepho et al., 2003). Yield and nitrogen-use traits measured at physiological maturity (R6) were analyzed separately using Type III two-way ANOVA, considering irrigation-water source, nitrogen level, and their interaction as fixed factors. Treatment means were compared using Tukey’s honestly significant difference (HSD) test at P ≤ 0.05. Treatment effect sizes were quantified using partial eta-squared (η²p) according to:

where SS_effect and SS_error represent the treatment and residual sums of squares, respectively. Effect sizes were interpreted as small (η²p ≥ 0.01), medium (η²p ≥ 0.06), and large (η²p ≥ 0.14) according to Cohen’s guidelines.

Multivariate treatment responses were first characterized by principal component analysis (PCA) using centered and standardized variables, and the first two principal components explaining the largest proportion of total variance were retained for interpretation (Lê et al., 2008; Jolliffe and Cadima, 2016). Relationships among root architectural, rhizosphere biochemical, yield, and nitrogen-use traits were subsequently examined using Pearson correlation analysis. To identify treatment similarity and coordinated trait responses, hierarchical clustering analysis (HCA) was performed on Z-score standardized treatment means using Euclidean distance and Ward’s minimum variance (Ward.D2) method (Legendre and Legendre, 2012; Murtagh and Legendre, 2014). Finally, piecewise latent path analysis was conducted using standardized variables and latent composite scores derived from the first principal component (PC1) of each biologically related trait group to elucidate the mechanistic relationships among root architecture, rhizosphere biochemical activity, nitrogen acquisition, and grain productivity following the piecewise structural equation modeling framework of Lefcheck (2016).

Random Forest (RF) regression was employed to identify the principal root morphological, rhizosphere biochemical, and nitrogen-use traits governing maize grain yield (Breiman, 2001). Grain yield was used as the response variable, and all remaining measured traits were included as predictors. Model performance was evaluated using repeated 10-fold cross-validation, and predictive accuracy was assessed using the coefficient of determination (R²), root mean square error (RMSE) and mean absolute error (MAE). Variable importance was estimated using permutation importance, and partial dependence plots were generated for the most influential predictors.

Overall treatment performance was evaluated using a Treatment Performance Index (TPI), calculated as the average of the Z-score standardized values of root architectural, rhizosphere biochemical, yield, and nitrogen-use:

where Z_{RSA to} Z_ANUE represent the standardized values of the corresponding traits. Higher TPI values indicate superior overall treatment performance.

Graphical visualization, PCA, Pearson correlation, HCA, and TPI calculations were performed in R (version 4.5.0) using the tidyverse, ggplot2, emmeans, FactoMineR, factoextra, ComplexHeatmap, pheatmap, and circlize packages, whereas piecewise latent path analysis and Random Forest regression were implemented in Python using the pandas, scikit-learn, and statsmodels libraries.

3 Results

3.1 Physico-chemical characteristics of PAW

Plasma activation markedly modified the physicochemical properties of distilled water, generating a reactive and nitrogen-enriched solution (Table 1). Compared with distilled water (pH 6.82), PAW exhibited a lower pH (3.18), together with substantially higher electrical conductivity (472 µS cm^-¹) and oxidation–reduction potential (546 mV). In addition, PAW contained appreciable concentrations of hydrogen peroxide (21.7 mg L^-¹), nitrate (81.5 mg L^-¹), and nitrite (7.6 mg L^-¹), whereas these species were absent or detected only at trace levels in distilled water. These results confirm the successful generation of a chemically stable and highly oxidative irrigation medium for evaluating its effects on maize root development, rhizosphere activity, and nitrogen acquisition.

3.2 Analysis of variance

Type III three-way ANOVA showed that growth stage (V6 and VT), irrigation-water source [normal water and plasma-activated water], and nitrogen level (0, 50, 75, and 100% of the recommended N dose) significantly affected root morphological and rhizosphere biochemical traits (Table 2). Growth stage and nitrogen level had highly significant effects on abbreviated words RSA, RHD, RHL, TOC, TOA and FAA should be there (P ≤ 0.001). Irrigation-water source also significantly influenced all traits. Significant growth stage × nitrogen level (G × N) interactions across all variables indicate that nitrogen responses varied with plant development. The water source × nitrogen level (W × N) interaction significantly affected RHL, TOC, TOA, and FAA, whereas the growth stage × water source (G × W) and G × W × N interactions significantly influenced RSA, TOC, and FAA, with TOA additionally affected by the three-way interaction. These findings demonstrate that root growth and rhizosphere metabolism were jointly regulated by growth stage, irrigation-water source, and nitrogen availability. Low coefficients of variation (4.03–7.07%) confirmed high experimental precision.

Type III two-way ANOVA revealed that irrigation-water source and nitrogen level at R6 stage significantly influenced grain yield, harvest index, nitrogen uptake, and agronomic nitrogen use efficiency (Table 3). Nitrogen level was the major source of variation, while irrigation-water source also exerted highly significant effects on all harvest traits (P ≤ 0.001). The significant water source × nitrogen level (W × N) interaction indicates that crop productivity and nitrogen-use efficiency depended on the combined effects of irrigation strategy and nitrogen supply. The low coefficients of variation (0.97–2.48%) further demonstrate the robustness and reliability of the experimental data.

Partial eta-squared (η²p) analysis confirmed that growth stage (η²p = 0.63–0.96) and nitrogen level (η²p = 0.65–0.98) exerted large effects on root morphological and rhizosphere biochemical traits, with the G × N interaction also showing substantial effects (η²p = 0.14–0.76). For yield and nitrogen-use traits, nitrogen level was the dominant source of variation (η²p = 0.97–0.98), while the W × N interaction exhibited moderate-to-large effects (η²p = 0.24–0.84), highlighting the critical role of nitrogen management and its interaction with irrigation water source in determining crop performance.

The large partial eta-squared (η²p) values observed for the Water × Nitrogen (W × N) interaction (up to 0.84) indicate that the response of maize to PAW was strongly dependent on nitrogen availability. These pronounced interaction effects demonstrate that the magnitude of PAW-induced improvements in root development, nitrogen uptake, ANUE, and grain productivity varied across nitrogen application rates, with the greatest benefits consistently observed under the 75% N treatment. This finding highlights the synergistic interaction between PAW and nitrogen availability in improving crop performance.

3.3 Mean performance

Distinct treatment-dependent differences in root morphology, rhizosphere biochemistry, and harvest performance were observed under PAW and nitrogen application (Figures 2–4). Representative root system and FESEM images demonstrated enhanced root proliferation, greater root hair density, and longer root hairs under PAW irrigation than under NW, with the most pronounced responses observed under PAW + 75% N and PAW + 100% N at both the V6 and VT growth stages (Figure 2). These qualitative observations were consistent with the quantitative mean performance illustrated in the bar charts (Figure 3).

Figure 2

Figure 3

Figure 4

The bar charts further confirmed that PAW significantly promoted belowground development (Figure 3). Root surface area increased progressively with increasing nitrogen supply and attained its highest value under PAW + 100% N, whereas root hair density and root hair length were maximized under PAW + 75% N. Similarly, rhizosphere biochemical activity was markedly enhanced by PAW, with total organic carbon and total organic acids reaching their highest levels under PAW + 75% N, while free amino acid concentration was greatest under PAW + 100% N. In contrast, NW + 0% N consistently exhibited the lowest values for all root morphological and rhizosphere biochemical traits.

Representative HPLC chromatograms supported the quantitative results by showing higher peak intensities of the major organic acids under PAW + 75% N than under NW + 100% N (Figure 4). Citric acid was the predominant component, followed by malic, succinic, and oxalic acids, with all peaks being more pronounced at the VT stage than at the V6 stage, indicating increased rhizosphere carbon exudation during reproductive growth. In contrast, NW + 0% N consistently showed the lowest root morphological and rhizosphere biochemical responses, highlighting the beneficial effect of PAW on root function and rhizosphere metabolism.

Harvest traits followed a similar trend (Figure 3). Grain yield, harvest index, and nitrogen uptake increased with increasing nitrogen application and reached their highest values under PAW + 100% N, with PAW consistently outperforming the corresponding NW treatments. In contrast, agronomic nitrogen use efficiency was maximized under PAW + 75% N, exceeding both PAW + 100% N and all corresponding NW treatments. While NW + 0% N exhibited the lowest overall harvest performance.

3.4 Principal component analysis

The first five principal components explained 99.81% of the total variation in maize physiological, growth, yield, and nitrogen-use traits (Table 4). PC1 (eigenvalue = 8.273) accounted for 72.39% of the total variance, while PC2 (eigenvalue = 2.606) explained an additional 22.81%, together capturing 95.20% of the overall variability. The remaining components contributed only marginally (<4%), indicating that the first two principal components adequately represented the multivariate response of maize to plasma-activated water and nitrogen treatments.

The PCA biplot (Figure 5) clearly separated the treatments along the first two principal components, with PAW + 75% N and PAW + 100% N exhibiting a strong positive association with yield, nitrogen-use, carbon metabolism, and root system traits, while NW + 0% N, PAW + 0% N, and NW + 50% N clustered on the negative side of PC1. Trait contribution analysis revealed that RSA, TOC, GY, HI, TOA, NU, FAA, and ANUE were the primary contributors to PC1 (each ~ 10-13%), whereas RHL (39.1%) and RHD (35.3%) predominantly defined PC2 (Supplementary Figure 1). These findings indicate that treatment discrimination was primarily governed by coordinated improvements in productivity, nitrogen acquisition, and carbon-related traits, while root hair characteristics represented the major source of secondary variation.

Figure 5

3.5 Correlation, hierarchical clustering, and clustered heatmap analysis

Pearson correlation analysis showed predominantly strong and significant positive relationships among the evaluated traits (Figure 6). RSA was strongly associated with FAA, NU, GY, and HI (P < 0.001). Free amino acids also exhibited significant positive correlations with NU, GY, and HI (P < 0.001). Nitrogen uptake and grain yield displayed an almost perfect positive relationship, whereas grain yield and harvest index showed the strongest association (P < 0.001). TOC and TOA were positively correlated with RSA, GY, HI, and ANUE, indicating significant coordination among growth, productivity, and antioxidant traits. Ward hierarchical clustering separated RHD and RHL into one cluster, while RSA, FAA, NU, GY, HI, ANUE, TOC, and TOA formed a second cluster, reflecting similar response patterns across treatments.

Figure 6

Hierarchical clustering and clustered heatmap analysis consistently differentiated the irrigation-water source × nitrogen treatments based on their integrated root, rhizosphere, yield, and nitrogen-use efficiency responses (Figure 7). The dendrogram (Figure 7A) separated the treatments into three major clusters, indicating distinct multivariate response patterns. NW + 100% N and NW + 75% N exhibited the highest similarity and clustered together with PAW + 50% N, whereas PAW + 0% N, NW + 50% N, and PAW + 100% N formed a separate cluster. NW + 0% N grouped with PAW + 75% N, demonstrating that treatment similarity was determined by the combined expression of measured traits rather than irrigation-water source or nitrogen level alone.

Figure 7

The clustered heatmap (Figure 7B) further revealed clear treatment-specific trait patterns. PAW + 75% N showed consistently higher standardized values for all studied traits, indicating superior overall performance. PAW + 100% N also exhibited positive responses for most productivity- and nitrogen-related traits, whereas NW + 0% N displayed uniformly lower standardized values across the measured variables. Trait clustering identified a close association among RSA, FAA, NU, GY, and HI, while TOC and TOA formed a separate cluster, suggesting coordinated regulation of root development, nitrogen acquisition, and rhizosphere biochemical activity. Collectively, these multivariate analyses identified PAW + 75% N as the treatment with the most favorable integrated trait profile, supporting enhanced nitrogen-use efficiency and crop productivity under reduced nitrogen input.

3.6 Piecewise latent path analysis

Piecewise latent path analysis identified significant positive relationships among the four latent constructs (Table 5). Root architecture positively influenced rhizosphere activity (β = 0.631, P < 0.001), explaining 27.6% of its variation. Rhizosphere activity subsequently enhanced nitrogen acquisition (β = 0.697, P < 0.001) with an R² of 0.781. Nitrogen acquisition exhibited the strongest positive effect on productivity (β = 0.943, P < 0.001), accounting for 93.2% of the observed variation. The results demonstrate a hierarchical pathway linking improved root morphological traits with enhanced rhizosphere functioning, greater nitrogen acquisition, and ultimately higher productivity.

3.7 Random forest analysis

The Random Forest model showed excellent predictive performance for maize grain yield, achieving a training R² of 0.9989 with an RMSE of 1.18 and MAE of 0.77, while repeated 10-fold cross-validation yielded an R² of 0.9911, RMSE of 3.31, and MAE of 2.23, indicating high predictive accuracy and minimal overfitting (Figure 8). Permutation importance analysis identified nitrogen uptake (NU) as the most influential predictor, contributing 27.99% of the total model importance, followed by harvest index (HI; 22.95%), free amino acids (FAA; 19.27%), total organic carbon (TOC; 12.79%), and root surface area (RSA; 9.18%) (Figure 8A). In contrast, total organic acids (TOA; 5.10%), agronomic nitrogen-use efficiency (ANUE; 2.26%), root hair length (RHL; 0.26%), and root hair density (RHD; 0.21%) contributed minimally to grain yield prediction. The close agreement between observed and predicted grain yield confirmed the strong predictive capability of the RF model across treatments (Figure 8B). Partial dependence analysis further revealed positive relationships of grain yield with NU, HI, FAA, and RSA, indicating that increases in these traits consistently enhanced predicted grain yield under plasma-activated water and nitrogen treatments (Figure 8C).

Figure 8

3.8 Treatment performance index

The TPI effectively integrated multiple response variables into a single composite score and clearly differentiated the irrigation–nitrogen treatments (Figure 9). PAW + 75% N achieved the highest TPI ranking (1.38), followed by PAW + 100% N (0.88), indicating superior overall performance. In contrast, NW + 0% N recorded the lowest TPI (−1.31), while the remaining treatments exhibited intermediate performance. The corresponding standardized heatmap illustrating treatment-wise trait responses has been provided as Supplementary Figure 2. This map further confirmed same trend, with PAW treatments displaying predominantly positive standardized responses and NW treatments showing comparatively lower responses across the evaluated variables. Notably, the higher TPI of PAW + 75% N relative to PAW + 100% N demonstrates that plasma-activated water enhanced overall plant performance while sustaining superior productivity with 25% lower fertilizer nitrogen application, identifying PAW + 75% N as the most efficient treatment under the present experimental conditions.

Figure 9

4 Discussion

4.1 Root architecture and rhizosphere responses to PAW

Plasma-activated water markedly improved maize performance by coordinately enhancing root development, rhizosphere metabolism, and nitrogen acquisition across contrasting nitrogen application rates. Compared with NW, PAW consistently increased RSA, RHD, RHL, TOC, TOA, FAA, NU, ANUE, GY, and HI. Among all treatments, PAW combined with 75% of the recommended fertilizer nitrogen produced the most balanced physiological response, demonstrating that enhanced root function and nutrient acquisition can partially compensate for reduced fertilizer application. These findings indicate that PAW strengthens the functional integration of root architecture and rhizosphere metabolism, enabling maize to maintain productivity with 25% lower fertilizer nitrogen application under controlled polyhouse conditions (Fageria and Baligar, 2005; Rafique, 2020; Sadras and Lawson, 2013).

The improved root development observed under PAW is closely associated with plasma-generated reactive oxygen and reactive nitrogen species (ROS/RNS). Plasma activation produces hydrogen peroxide, nitrate, nitrite, and other long-lived reactive molecules that function as signaling compounds regulating root growth, hormonal interactions, and nitrogen assimilation rather than simply inducing oxidative stress (Lu et al., 2016; Thirumdas et al., 2018; Miranda et al., 2023). Although PAW contained plasma-derived nitrate and nitrite, the calculated inorganic nitrogen input represented less than 1% of the fertilizer nitrogen applied, confirming that its primary role was biostimulatory rather than a direct nitrogen source. These reactive molecules also activate MAPK-dependent signaling pathways involved in growth regulation, carbon metabolism, developmental plasticity, and stress adaptation (Mittler, 2017; Jagodzik et al., 2018; Muhammad et al., 2021; Bao et al., 2025; Sozoniuk et al., 2026). Consequently, PAW significantly increased RSA, RHD, and RHL across all nitrogen levels, resulting in a larger and more efficient absorptive system. Earlier studies mainly demonstrated improved seed germination and early seedling growth following plasma treatment (Sivachandiran and Khacef, 2017; Holubová et al., 2021; Lukacova et al., 2021; Mogo et al., 2022). In contrast, the present study demonstrates that PAW-induced root remodeling persists throughout crop development and directly contributes to improved nitrogen acquisition under reduced fertilizer nitrogen application.

Root architecture plays a central role in nitrogen acquisition because it determines soil exploration and the effective absorptive interface between roots and soil. Greater RSA increased the soil volume available for nutrient interception, whereas increased RHD and RHL expanded the contact area for nutrient absorption. Root hairs are particularly important under moderate nitrogen availability because they reduce diffusion limitations and enhance nutrient foraging efficiency. The superior root hair development observed under PAW + 75% N therefore reflects enhanced adaptive plasticity, enabling maize to compensate for reduced fertilizer nitrogen through more efficient nutrient acquisition. Similar relationships between root architectural traits and nitrogen recovery have been reported in cereals grown under nitrogen-limited conditions (Škarpa et al., 2020; Ahmad et al., 2022; Liu, 2021). Collectively, these findings indicate that PAW regulates root morphology through ROS/RNS-mediated signaling, thereby increasing the functional capacity of the maize root system.

Enhanced root development was accompanied by pronounced changes in rhizosphere metabolism. PAW significantly increased TOC, TOA, and FAA at both the V6 and VT growth stages, while HPLC analysis revealed greater concentrations of citric, malic, succinic, and oxalic acids, particularly under the PAW + 75% N treatment. These responses indicate increased carbon allocation belowground and enhanced rhizosphere metabolic activity during periods of high nutrient demand. Root exudates provide essential carbon substrates for microorganisms involved in nutrient mineralization, nitrogen cycling, and plant growth promotion (Upadhyay et al., 2022; Wankhade et al., 2025). They also regulate microbiome recruitment and community assembly, thereby improving nutrient availability and plant health (Yang et al., 2025; Zhalnina et al., 2018). Therefore, the simultaneous increases in TOC, TOA, and FAA suggest that PAW promotes coordinated biochemical processes that complement improved root development and facilitate nitrogen acquisition.

The predominance of citric and malic acids provides further insight into the mechanisms underlying enhanced nitrogen acquisition. Organic acids acidify the rhizosphere, chelate mineral nutrients, stimulate microbial activity, and increase nutrient solubility, thereby improving nutrient availability. Jiang et al. (2024) demonstrated that rhizosphere metabolites establish functional links between microbial communities and nutrient acquisition, whereas Jacoby et al. (2017) emphasized the indispensable role of soil microorganisms in plant mineral nutrition. Increased rhizosphere carbon allocation also promotes microbial activity and expands the biologically active zone surrounding the root system (Ahmad et al., 2022). Together, these structural and biochemical responses demonstrate that PAW enhances the functional efficiency of the root–rhizosphere continuum, thereby improving the capacity of maize to acquire nitrogen under reduced fertilizer input.

4.2 PAW improves nitrogen acquisition and productivity

The coordinated improvements in root architecture and rhizosphere metabolism were directly reflected in enhanced nitrogen acquisition and crop productivity. PAW consistently increased nitrogen uptake across all fertilizer levels, while the PAW + 75% N treatment achieved the highest ANUE. Although plasma–water interactions generate nitrate and nitrite through atmospheric nitrogen fixation (Volkov et al., 2021), the cumulative inorganic nitrogen supplied through PAW irrigation was estimated to be approximately 0.207 g N plant^-¹, representing less than 1% of the fertilizer nitrogen applied. Therefore, the enhanced nitrogen uptake and ANUE primarily reflected improved root function and rhizosphere-mediated nutrient acquisition rather than direct nitrogen supplementation by PAW. These findings indicate that PAW increased the intrinsic capacity of maize to acquire and utilize soil nitrogen through coordinated physiological and biochemical responses. Similar improvements in nitrogen acquisition following PAW application have been reported in several crop species (Dahal et al., 2024; Andrade et al., 2025; Priatama et al., 2025), while a recent meta-analysis confirmed the consistent benefits of PAW for crop productivity and resource-use efficiency across diverse cropping systems (Ali and Dong, 2026). The present study extends these observations by demonstrating that sustained improvements in nitrogen acquisition are associated with coordinated changes in root morphology and rhizosphere metabolism under reduced fertilizer nitrogen application.

The improved nitrogen acquisition ultimately translated into greater grain yield and harvest index. Grain yield increased progressively with increasing nitrogen supply under both irrigation regimes; however, PAW consistently outperformed the corresponding NW treatments at every fertilizer level. The large Water × Nitrogen (W × N) interaction effect sizes (η²p up to 0.84) further demonstrate that the response to PAW was strongly dependent on nitrogen availability rather than being uniform across all nitrogen treatments. The greatest improvements were consistently observed under the 75% N treatment, indicating that PAW was most effective under moderate nitrogen supply, where enhanced nutrient acquisition compensated for reduced fertilizer application. Notably, PAW + 75% N maintained grain yield and harvest index comparable to PAW + 100% N while achieving the highest ANUE, demonstrating that improved nutrient acquisition rather than increased fertilizer input was the primary driver of productivity. Enhanced root development and rhizosphere activity likely sustained nitrogen availability throughout crop growth, supporting chlorophyll synthesis, protein accumulation, photosynthesis, and assimilate partitioning required for grain filling. Similar improvements in productivity following PAW application have been reported in tomato and other agricultural systems (Priatama et al., 2025; Andrade et al., 2025; Ali and Dong, 2026).

The superior performance of PAW + 75% N further highlights the importance of coordinated structural and biochemical responses under moderate nitrogen availability. Increased RSA, RHD, and RHL expanded the physical capacity for nutrient interception, whereas greater secretion of TOC, TOA, and FAA maintained an active rhizosphere capable of sustaining nutrient mobilization throughout vegetative and reproductive growth. Root-derived carbon compounds stimulate microbial mineralization and nutrient transformation (Upadhyay et al., 2022; Wankhade et al., 2025), while root exudates regulate microbiome recruitment and substrate utilization (Yang et al., 2025; Zhalnina et al., 2018). These complementary responses enabled PAW-treated plants to sustain productivity with 25% lower fertilizer nitrogen application, demonstrating that improving nutrient acquisition efficiency can be more effective than increasing fertilizer input alone.

4.3 Multivariate analysis identifies key productivity drivers

The multivariate analyses consistently demonstrated that the beneficial effects of PAW resulted from coordinated regulation of multiple physiological processes rather than changes in any single trait. PCA clearly separated PAW + 75% N and PAW + 100% N from the remaining treatments, with RSA, TOC, TOA, FAA, NU, ANUE, GY, and HI contributing most strongly to treatment discrimination. Hierarchical clustering further grouped PAW + 75% N with the highest standardized trait values, indicating that treatment performance depended on the integrated enhancement of root development, rhizosphere metabolism, and nitrogen utilization rather than fertilizer quantity alone. Pearson correlation analysis supported these findings by revealing strong positive associations among root architectural traits, rhizosphere metabolites, nitrogen uptake, grain yield, and harvest index. Collectively, these analyses demonstrate that PAW enhances crop performance through coordinated physiological regulation rather than isolated trait responses, consistent with previous reports highlighting the close association between root morphology, rhizosphere metabolism, and nutrient recovery in cereals (Škarpa et al., 2020; Ahmad et al., 2022; Jiang et al., 2024).

Latent path analysis provided further mechanistic insight into the relationships among these traits. Nitrogen acquisition explained more than 93% of the variation in grain yield, identifying nutrient uptake as the principal physiological link between belowground processes and productivity. The sequential relationships among root architecture, rhizosphere metabolism, nitrogen acquisition, and grain yield support the concept that PAW improves productivity indirectly by strengthening belowground processes that enhance nutrient acquisition and utilization. This systems-level interpretation is consistent with the integrated physiological responses observed across all measured traits.

Random Forest regression independently confirmed these relationships by identifying nitrogen uptake, harvest index, free amino acids, total organic carbon, and root surface area as the most influential predictors of grain yield. The strong agreement among PCA, hierarchical clustering, correlation analysis, latent path analysis, and Random Forest regression demonstrates the robustness of the proposed mechanistic framework and highlights nitrogen acquisition as the central integrative process linking PAW-induced root remodeling with improved productivity. The application of complementary multivariate approaches for identifying key productivity drivers is increasingly recognized for analysing complex biological systems (Breiman, 2001; Lê et al., 2008; Jolliffe and Cadima, 2016; Lefcheck, 2016). Nevertheless, although the Random Forest model showed high predictive accuracy, the relatively small experimental dataset (eight treatments × five biological replicates) may increase the risk of model overfitting despite repeated 10-fold cross-validation. Therefore, these findings should be considered exploratory and require validation using larger, independent datasets.

Taken together, the multivariate analyses provide convergent evidence that PAW functions as a systems-level biostimulant by integrating structural, biochemical, and nutritional processes to improve nitrogen use efficiency and sustain grain productivity under reduced fertilizer nitrogen application.

4.4 Agronomic implications and study limitations

From an agronomic perspective, the ability of PAW + 75% N to sustain grain productivity while reducing fertilizer nitrogen application by 25% has important implications for sustainable maize production. Conventional maize cultivation is frequently constrained by low nitrogen recovery and substantial fertilizer losses through leaching, volatilization, and denitrification, resulting in economic losses and environmental pollution (Fageria and Baligar, 2005; Rafique, 2020; Sadras and Lawson, 2013). The present findings demonstrate that improving the efficiency of nitrogen acquisition, rather than simply increasing fertilizer application, represents an effective strategy for enhancing crop productivity and resource-use efficiency. By simultaneously promoting root development, rhizosphere function, nitrogen uptake, and agronomic nitrogen use efficiency, PAW reduced dependence on fertilizer nitrogen while maintaining grain yield, supporting the principles of climate-smart agriculture, precision nutrient management, and sustainable intensification.

The controlled polyhouse grow-bag system used in this study enabled precise mechanistic evaluation of PAW-induced changes in root architecture, rhizosphere processes, and nitrogen acquisition by minimizing environmental variability. However, root growth was confined to a fixed soil volume and did not fully represent the complexity of field conditions, including unrestricted root exploration, heterogeneous soil environments, and plant-to-plant competition. Therefore, the present findings should be regarded as a mechanistic proof-of-concept under controlled conditions, and validation through multi-location field experiments across diverse soil types and environmental conditions is required before recommending PAW for commercial maize production.

Another limitation of the present study is that root exudates were expressed as solution concentrations rather than normalized to root dry weight and incubation time because root dry weight was not determined at the time of exudate collection. Furthermore, the solution-based exudate collection method estimates the potential exudation capacity of roots under standardized conditions rather than the in-situ rhizosphere metabolite pool, where adsorption, microbial turnover, and diffusion continuously influence metabolite dynamics. Therefore, the measured total organic carbon, organic acids, and free amino acids should be interpreted as indicators of root exudation potential rather than absolute rhizosphere concentrations. Future studies should quantify exudation rates on a root biomass basis and integrate rhizosphere metabolomics, microbiome analyses, and field validation to provide a more comprehensive understanding of PAW-mediated regulation of root–soil interactions under diverse production environments.

5 Conclusion

This study demonstrates that plasma-activated water enhances maize nitrogen use efficiency by coordinately improving root development, rhizosphere metabolism, and nutrient acquisition. Rather than acting as a direct nitrogen source, PAW primarily functions as a biostimulant that strengthens the plant’s intrinsic capacity for nitrogen acquisition and utilization under reduced fertilizer nitrogen application. The superior performance of the PAW + 75% N treatment confirms that the improved agronomic nitrogen use efficiency resulted from enhanced root and rhizosphere function rather than the negligible amount of inorganic nitrogen supplied through PAW. By integrating root phenotyping, rhizosphere metabolite profiling, and complementary multivariate analyses, this study provides a mechanistic framework linking plasma-derived signaling with improved nitrogen acquisition and crop productivity. Although these findings represent a mechanistic proof-of-concept under controlled polyhouse conditions, they provide a strong foundation for developing PAW-based precision nutrient management strategies and warrant validation through multi-location field experiments across diverse soil and environmental conditions before broad agronomic application.

References

• 1

  AhmadH. M.FiazS.HafeezS.ZahraS.ShahA. N.GulB.et al. (2022). Plant growth-promoting rhizobacteria eliminate the effect of drought stress in plants: A review. Front. Plant Sci.13, 875774. doi: 10.3389/fpls.2022.875774

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  AliN.DongY. (2026). Global meta-analysis of plasma-activated water for improving crop establishment, productivity, and health. Front. Plant Sci.17, 1754073. doi: 10.3389/fpls.2026.1754073

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  AndradeP. E.SaviP. J.AlmeidaF. S.CarciofiB. A.PaceA.ZouY.et al. (2025). Plasma-activated water as a sustainable nitrogen source: Supporting the United Nations Sustainable Development Goals in controlled environment agriculture. Crops.5, 35. doi: 10.3390/crops5030035

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  BaoL.WangW.LiM.LiuJ.LiuJ.AlifuG.et al. (2025). Reactive oxygen species–post translational modifications–central carbon metabolism regulatory loop: Coordination of redox homeostasis and carbon flux allocation in plants under abiotic stress. Front. Plant Sci.16, 1637328. doi: 10.3389/fpls.2025.1637328

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 5

  BreimanL. (2001). Random forests. Mach. Learn.45, 5–32. doi: 10.1023/A:1010933404324

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  BremnerJ. M. (1965). “ Total nitrogen,” in Methods of Soil Analysis: Part 2 Chemical and Microbiological Properties. Ed. NormanA. G. ( American Society of Agronomy, Madison), 1149–1178. doi: 10.2134/agronmonogr9.2.c32

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  CawthrayG. R. (2003). An improved reversed-phase liquid chromatographic method for the analysis of low-molecular mass organic acids in plant root exudates. J. Chromatogr. A1011, 233–240. doi: 10.1016/S0021-9673(03)01129-4

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  DahalR.DhakalO. B.AcharyaT. R.LamichhaneP.GautamS.ChaliseR.et al. (2024). Investigating plasma-activated water as a sustainable treatment for improving growth and nutrient uptake in maize and pea plants. Plant Physiol. Biochem.216, 109203. doi: 10.1016/j.plaphy.2024.109203

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 9

  DonaldC. M.HamblinJ. (1976). The biological yield and harvest index of cereals as agronomic and plant breeding criteria. Adv. Agron.28, 361–405. doi: 10.1016/S0065-2113(08)60559-3

  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  FageriaN. K.BaligarV. C. (2005). Enhancing nitrogen use efficiency in crop plants. Adv. Agron.88, 97–185. doi: 10.1016/S0065-2113(05)88004-6

  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  HolubováĽ.ŠvubováR.SlovákováĽ.BokorB.Chobotová KročkováV.RenčkoJ.et al. (2021). Cold atmospheric pressure plasma treatment of maize grains: Induction of growth, enzyme activities and heat shock proteins. Int. J. Mol. Sci.22, 8509. doi: 10.3390/ijms22168509

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 12

  JacobyR.PeukertM.SuccurroA.KoprivovaA.KoprivaS. (2017). The role of soil microorganisms in plant mineral nutrition—Current knowledge and future directions. Front. Plant Sci.8, 1617. doi: 10.3389/fpls.2017.01617

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  JagodzikP.Tajdel-ZielińskaM.CieślaA.MarczakM.LudwikówA. (2018). Mitogen-activated protein kinase cascades in plant hormone signaling. Front. Plant Sci.9, 1387. doi: 10.3389/fpls.2018.01387

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  JiangP.WangY.ZhangY.FeiJ.RongX.PengJ.et al. (2024). Intercropping enhances maize growth and nutrient uptake by driving the link between rhizosphere metabolites and microbiomes. New Phytol.243, 1506–1521. doi: 10.1111/nph.19906

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  JolliffeI. T.CadimaJ. (2016). Principal component analysis: A review and recent developments. Philos. Trans. R. Soc A.374, 20150202. doi: 10.1098/rsta.2015.0202

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  JonesD. L.WillettV. B. (2006). Experimental evaluation of methods to quantify dissolved organic nitrogen (DON) and dissolved organic carbon (DOC) in soil. Soil Biol. Biochem.38, 991–999. doi: 10.1016/j.soilbio.2005.08.012

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 17

  LêS.JosseJ.HussonF. (2008). FactoMineR: An R package for multivariate analysis. J. Stat. Software25, 1–18. doi: 10.18637/jss.v025.i01

  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  LefcheckJ. S. (2016). piecewiseSEM: Piecewise structural equation modelling in R for ecology, evolution, and systematics. Methods Ecol. Evol.7, 573–579. doi: 10.32614/cran.package.piecewisesem

  • CrossRef
  • Google Scholar
  • View reference in article

• 19

  LegendreP.LegendreL. (2012). Numerical Ecology. 3rd English ed (Amsterdam: Elsevier).

  • Google Scholar
  • View reference in article

• 20

  LiuD. (2021). Root developmental responses to phosphorus nutrition and nutrient acquisition. J. Integr. Plant Biol.63, 1065–1090. doi: 10.1111/jipb.13090

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  LuX.NaidisG. V.LaroussiM.ReuterS.GravesD. B.OstrikovK. (2016). Reactive species in non-equilibrium atmospheric-pressure plasmas: Generation, transport, and biological effects. Phys. Rep.630, 1–84. doi: 10.1016/j.physrep.2016.03.003

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  LukacovaZ.SvubovaR.SelvekovaP.HenselK. (2021). The effect of plasma-activated water on maize (Zea mays L.) under arsenic stress. Plants.10, 1899. doi: 10.3390/plants10091899

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  MirandaF. S.TavaresV. K. F.GomesM. P.NetoN. F. A.ChiappimW.PetraconiG.et al. (2023). Physicochemical characteristics and antimicrobial efficacy of plasma-activated water produced by an air-operated coaxial dielectric barrier discharge plasma. Water15 (23), 4045. doi: 10.3390/w15234045

  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  MittlerR. (2017). ROS are good. Trends Plant Sci.22, 11–19. doi: 10.1016/j.tplants.2016.08.002

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  MogoJ. P. K.FovoJ. D.Sop-TamoB.MafouassonH. N. A.NgwemM. C. N.TebuM. J.et al. (2022). Effect of gliding arc plasma activated water (GAPAW) on maize (Zea mays L.) seed germination and growth. Seeds.1, 230–243. doi: 10.3390/seeds1040020

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  MuhammadI.ShalmaniA.AliM.YangQ. H.AhmadH.LiF. B. (2021). Mechanisms regulating the dynamics of photosynthesis under abiotic stresses. Front. Plant Sci.11, 615942. doi: 10.3389/fpls.2020.615942

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  MurtaghF.LegendreP. (2014). Ward's hierarchical agglomerative clustering method: Which algorithms implement Ward's criterion? J. Classif.31, 274–295. doi: 10.1007/s00357-014-9161-z

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 28

  PiephoH. P.BüchseA.EmrichK. (2003). A hitchhiker's guide to mixed models for randomized experiments. J. Agron. Crop Sci.189, 310–322. doi: 10.1046/j.1439-037X.2003.00049.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  PriatamaR. A.BeakH. K.ParkS.SongI.ParkS. J.KimS. B.et al. (2025). Tomato yield enhancement with plasma-activated water as an alternative nitrogen source. BMC Plant Biol.25, 668. doi: 10.1186/s12870-025-06701-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  RafiqueS. (2020). “ Drought responses of Zea mays in relation to nitrogen nutrition and source–sink relationships,” in Physiological Responses of Plants Under Abiotic Stress ( IntechOpen, London). doi: 10.5772/intechopen.91288

  • CrossRef
  • Google Scholar

• 31

  R Core Team (2025). R: A Language and Environment for Statistical Computing (Vienna: R Foundation for Statistical Computing). Available online at: https://www.R-project.org/ (Accessed May 28, 2026).

  • Google Scholar
  • View reference in article

• 32

  SadrasV. O.LawsonC. (2013). Nitrogen and water-use efficiency of Australian wheat varieties released between 1958 and 2007. Eur. J. Agron.46, 34–41. doi: 10.1016/j.eja.2012.11.008

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 33

  SivachandiranL.KhacefA. (2017). Enhanced seed germination and plant growth by atmospheric pressure cold air plasma: Combined effect of seed and water treatment. RSC. Adv.7, 1822–1832. doi: 10.1039/C6RA24788B

  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  ŠkarpaP.KlofáčD.KrčmaF.ŠimečkováJ.KozákováZ. (2020). Effect of plasma activated water foliar application on selected growth parameters of maize (Zea mays L.). Water.12, 3545. doi: 10.3390/w12123545

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  SozoniukM.ŚwiecaM.BohatáA.BartošP.BedrníčekJ.LorencF.et al. (2026). Plasma activation of plant-based biostimulants enhances growth and defense responses in soybean (Glycine max L.). BMC Plant Biol.26, 552. doi: 10.1186/s12870-026-08372-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  ThirumdasR.KothakotaA.AnnapureU.SiliveruK.BlundellR.GattR.et al. (2018). Plasma-activated water (PAW): Chemistry, physico-chemical properties, applications in food and agriculture. Trends Food Sci. Technol.77, 21–31. doi: 10.1016/j.tifs.2018.05.007

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  UpadhyayS. K.SrivastavaA. K.RajputV. D.ChauhanP. K.BhojiyaA. A.JainD.et al. (2022). Root exudates: Mechanistic insight of plant growth-promoting rhizobacteria for sustainable crop production. Front. Microbiol.13, 916488. doi: 10.3389/fmicb.2022.916488

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  VolkovA. G.BookalA.HairstonJ. S.RobertsJ.TaengwaG.PatelD. (2021). Mechanisms of multielectron reactions at the plasma/water interface: Interfacial catalysis, reactive oxygen and nitrogen species, nitrogen fixation, and plasma-activated water. Electrochim. Acta389, 138441. doi: 10.1016/j.electacta.2021.138441

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 39

  WankhadeA.WilkinsonE.BrittD. W.KaundalA. (2025). A review of plant–microbe interactions in the rhizosphere and the role of root exudates in microbiome engineering. Appl. Sci.15, 7127. doi: 10.3390/app15137127

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 40

  YangC. X.ChenS. J.HongX. Y.WangL. Z.WuH. M.TangY. Y.et al. (2025). Plant exudates-driven microbiome recruitment and assembly facilitates plant health management. FEMS Microbiol. Rev.49, fuaf008. doi: 10.1093/femsre/fuaf008

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 41

  YemmE. W.CockingE. C.RickettsR. E. (1955). The determination of amino acids with ninhydrin. Analyst.80, 209–214. doi: 10.1039/AN9558000209

  • CrossRef
  • Google Scholar
  • View reference in article

• 42

  ZhalninaK.LouieK. B.HaoZ.MansooriN.Da RochaU. N.ShiS.et al. (2018). Dynamic root exudate chemistry and microbial substrate preferences drive patterns in rhizosphere microbial community assembly. Nat. Microbiol.3, 470–480. doi: 10.1038/s41564-018-0129-3

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article