Two plant growth-promoting microorganisms, S2_8_1 and Shan2, were used in this study. Both strains were isolated from the experimental site of Henan University of Science and Technology (Luoyang, China) using a selective medium for ammonia-oxidizing bacteria. S2_8_1 was selected based on its strong ammonia-oxidizing capacity in preliminary screening. Shan2, a Streptomyces strain with cytokinin-producing potential, was subsequently selected from the isolate collection based on its hormone-related activity. These two strains represent distinct functional traits associated with nitrogen transformation and plant growth regulation. Molecular identification revealed that S2_8_1 belongs to the genus Rhizobium (GenBank accession: ON667919.1; CCTCC NO: M2021374), and Shan2 to Streptomyces (GenBank accession: PQ605713; CCTCC NO: M20242569). Both are preserved at the China Center for Type Culture Collection (Wuhan University, China). Strains were cultured in enrichment medium containing 3.79 mM (NH₄)₂SO₄, 5.51 mM KH₂PO₄, 1.81 mM NaH₂PO₄, 0.059 mM MnSO₄ 4H₂O, 0.12 mM MgSO₄ 7H₂O, and 50.0 mM CaCO₃ (pH 7.2). LB medium (5 g yeast extract, 10 g NaCl, and 10 g tryptone per liter, pH 7.0) was used in control treatments. Before application, bacterial suspensions were centrifuged and resuspended to standardize inoculum concentration. The final bacterial suspension contained approximately 2.16 × 10⁶ CFU mL⁻¹.

The experiment took place at the university experimental farm in Luoyang, which has a temperate continental monsoon climate, with average annual rainfall of 601 mm and a mean temperature of 14.2°C. The test plant was annual multiflora ryegrass (Lolium multiflorum) cultivar ‘Tetragold’, selected for its rapid growth, strong regenerative capacity, and high fresh biomass yield, traits ideal for forage production and root-soil interaction studies.

In September 2024, plastic pots (20 cm diameter × 25 cm height; 13.5 L) were filled with 5.80 kg of brown soil. Soil properties were: 23.8 g kg⁻¹ organic carbon, 2.1 g kg⁻¹ total nitrogen, 10.3 mg kg⁻¹ phosphorus, 120.5 mg kg⁻¹ potassium, 269.7 mg kg⁻¹ magnesium, and pH 7.7. Pots were kept in a naturally lit, open-sided greenhouse (25–30°C), sheltered from rain, and watered every 2–3 days to ensure sufficient soil moisture. Ryegrass seeds were sown and grown for three weeks. Uniform, healthy seedlings were then transplanted into 100 pots. Two parallel experiments were conducted: Experiment 1 (Exp-1) used soil-based cultivation with PGPM and nitrate treatments (54 pots), while Experiment 2 (Exp-2) used a hydroponic system with S2_8_1 and nitrate applied directly to the roots (36 pots). To reduce variability, only uniform seedlings were selected. All plants received adequate water and light throughout the study.

In Exp-1, 54 pots were randomly assigned to six treatment groups (n = 9 per group), each subdivided into three replicates (three pots per subgroup). Treatments included: (1) no addition (GW), (2) blank enriched medium (BT), (3) S2_8_1 bacterial suspension (JI), (4) LB medium (LB), (5) Shan2 bacterial suspension (JS), and (6) NO₃⁻–N combined with Shan2 (NS). The primary objective of this study was to evaluate plant physiological responses to microbial inoculation. Based on previous studies conducted by our group [29] indicating that plant responses are mainly associated with microbial activity, the present experiment focused on the integrated effects of functional strains. Therefore, a treatment group with only nitrate addition was not set up in the experiment.

On day 0 of regrowth, 200 mL of the designated solution was applied to each pot: blank medium (BT), S2_8_1 suspension (JI), LB medium (LB), Shan2 suspension (JS), or 35 mmol L-1 NO₃⁻–N (NS). The GW group received no treatment. The NO₃⁻–N concentration, previously shown to be optimal [29], was maintained throughout via replenishment. All pots received equal liquid volumes (200 mL), adjusted with deionized water. Soil moisture was maintained at 65–70% field capacity. After treatment, subgroup 1 was sampled for physiological and biochemical analyses. Subgroups 2 and 3 were clipped at a height of 5 cm, and biomass was dried for dry matter determination. On day 7, subgroup 2 was sampled; subgroup 3 was clipped again. Final measurements were taken from subgroup 3 on day 14.

In Exp-2, 36 pots with ryegrass seedlings were transferred to a hydroponic system maintained at 25°C under greenhouse conditions. Pots were randomly assigned to four treatment groups (n = 9 each), subdivided into three replicates (three pots per subgroup, n = 3 independent pots per treatment at each sampling time point). Treatments included: (1) distilled water (GW), (2) 100 mL of 35 mmol L-1 NO₃⁻–N solution (NT), (3) 100 mL of blank enriched medium (BT), and (4) 100 mL of S2_8_1 bacterial suspension (JI).

Due to the slower regrowth rate of hydroponically grown ryegrass, a 10-day regrowth period was used. On day 0, subgroup 1 was sampled for physiological and biochemical measurements, while subgroups 2 and 3 were clipped at a height of 5 cm. Clipped biomass was weighed and dried to determine dry matter content. A 5-day regrowth cycle was adopted to detect early growth differences. On day 5, subgroup 2 was sampled; subgroup 3 was clipped again under the same conditions. Final sampling of subgroup 3 was conducted on day 10.

Roots were gently washed to remove adhering soil, and surface moisture was blotted before recording fresh weight. Roots, stems, and leaves were then oven-dried at 65°C for 72 h to determine dry biomass. Regrowth biomass was defined as the dry weight of newly grown leaves after clipping, and total biomass was calculated as the sum of root, stem, and leaf dry weights.

Transpiration rate (Tr) was measured between 11:00 and 12:00 using a LI-6400 portable photosynthesis system (LI-COR Inc., Lincoln, NE, USA) under controlled conditions: 1000 μmol m⁻² s⁻¹ light intensity, 28°C temperature, and 400 ppm CO₂. Due to the narrow leaf blades of ryegrass, two to three leaves were overlapped to fully cover the chamber area and ensure accurate measurements.

Ryegrass plants were gently removed from pots, and rhizosphere soil was collected by brushing soil particles from roots 5–10 times using a clean brush. Fine soil was retained, labeled, and stored in separate bags to avoid cross-contamination. Soil NO₃⁻–N concentration and net nitrification rate were determined using rhizosphere soil samples. Nitrate content was measured by the phenol disulfonic acid colorimetric method [30], using phenol disulfonic acid reagent, calcium sulfate, distilled water, a spectrophotometer, a constant-temperature water bath, and porcelain evaporating dishes. To assess net nitrification, soil samples were incubated in the dark at 25°C for 4 days. NO₃⁻–N was measured before and after incubation. The daily nitrification rate was calculated as: (1) N i t r i f i c a t i o n   r a t e = N O 3 − -N p o s t − N O 3 − -N p r e D a y (2) N O 3 − -N = C × V × V 2 V 1 × m where C is the NO₃⁻–N content in the test solution (mg mL⁻¹), determined using a standard calibration curve; V is the total volume of the sample solution (mL); V₁ is the volume of the filtrate used for color development (mL); V₂ is the volume of the color developed solution (mL); and m is the mass of the air-dried soil sample (g).

Zeatin riboside (ZR), a stable derivative of zeatin, serves as a key indicator of cytokinin signaling. For hormone extraction, 0.50 g of fresh leaf tissue was cut into small pieces and ground with quartz sand in a mortar precooled to −40°C using liquid nitrogen. During grinding, 5 mL of 80% methanol (with 1 mmol L⁻¹ butylated hydroxytoluene, BHT) was added three times. The homogenate was transferred to 10 mL centrifuge tubes and extracted at 4°C for 12 h. After centrifugation (10 min), the supernatant was filtered through a C18 column. A secondary extraction was performed with 1 mL of the same solvent, incubated for 1 h, centrifuged, and filtered again. Combined filtrates were vacuum-dried and stored at −80°C.

Xylem sap was collected by wrapping 0.200 g of absorbent cotton around the clipped stem, secured with a zip-lock bag and rubber band to prevent evaporation. After 12 h, sap volume was estimated by weight gain. Cotton was extracted twice with 1 mL of 80% methanol (1 mmol L⁻¹ BHT), first for 12 h at 4°C and again for 1 h. Extracts were combined, dried under vacuum, and stored at −80°C.

ZR content in leaves and xylem sap extracts were quantified using ELISA according to the manufacturer protocol (Shanghai Enzyme Linked Biotechnology Co., Ltd., Shanghai, China). Absorbance was measured at 490 nm ZR levels in leaves were expressed per gram of fresh weight, while xylem ZR (C_ZR) was expressed per milliliter of sap. Cytokinin transport rates were estimated under dark (B_ZR) and light (L_ZR) conditions using: (3) B Z R = C Z R × X r (4) L Z R = C Z R × T r × M × S L A where C_ZR is the ZR concentration in xylem sap (ng mL⁻¹), X_r denotes the rate of xylem sap transport in darkness (mL h⁻¹), T_r represents the transpiration rate (mL m⁻² s⁻¹), M is the biomass of regrown leaves (g), and specific leaf area (SLA, cm² g⁻¹) is the ratio of leaf surface area to leaf dry weight, which is determined beforehand.

All data are presented as means ± standard deviation (SD). One-way ANOVA was performed using SPSS (version 16.0; SPSS Inc., Chicago, IL, USA), followed by Duncan’s multiple range test for post hoc comparisons. Differences were considered significant at p < 0.05. To evaluate the combined effects of soil nitrification and leaf ZR content on biomass, multiple linear regression models were constructed using all-subsets regression. Both additive (nitrification and ZR as independent predictors) and interaction models (including a nitrification × ZR term) were tested. Model selection was based on the Akaike Information Criterion (AIC), with lower AIC indicating a better fit [31,32]. Model performance was assessed using the coefficient of determination (R²). Regression and visualization were conducted in R [33] (version 4.4.2), using the ‘lavaan’ package for regression, and ‘ggplot2’, ‘dplyr’, ‘tidyr’, and ‘plotly’ for data processing and plotting.

As shown in Fig. 1, the GW, BT, and LB treatments showed no significant differences in clipping or cumulative biomass during regrowth. In contrast, both JI and NS significantly enhanced ryegrass biomass compared to GW. Notably, the biomass increases in the JI treatment were comparable to those in the NS treatment, with JI achieving a 3.6-fold increase in clipping biomass and a 52% improvement in total biomass by day 14, indicating a strong growth-promoting effect. NS outperformed JS, suggesting that Shan2 impact is amplified when combined with nitrate. JS still led to moderate biomass gains, highlighting the role of cytokinin production in regrowth.

S2_8_1 also significantly increased soil NO₃⁻–N content and nitrification rate after 7 days (Fig. 2; p < 0.05), with values similar to those in NS by day 14. This indicates that S2_8_1 treatment was associated with higher net nitrification under the present experimental conditions. The JS treatment showed a smaller but significant increase in NO₃⁻–N (p < 0.05), potentially through indirect microbial interactions with nitrogen cycling.

As shown in Table 1, no treatment differences in leaf ZR, B_ZR, or L_ZR were detected at day 0. However, during regrowth, ZR transport from roots to shoots increased significantly in JI, JS, and NS treatments under both light and dark conditions (p < 0.05). By days 7 and 14, NS showed the highest ZR content, followed by JI. These results indicate that S2_8_1 promotes both nitrification and cytokinin translocation, supporting its dual functionality and potential as a microbial strategy for enhancing plant regrowth.

At day 0 of regrowth, no significant differences in total biomass were observed among treatments. However, by days 5 and 10, ryegrass treated with S2_8_1 (JI) showed significantly higher biomass than both the GW and NT controls (Fig. 3a), confirming the growth-promoting effect of S2_8_1 under hydroponic conditions. To explore the underlying mechanism, ZR content in leaves was assessed during regrowth. As shown in Fig. 3b, ZR levels did not differ among treatments at day 0 but increased significantly in the JI group by day 5, with further enhancement by day 10. By the end of the experiment, leaf ZR content in JI exceeded that of GW by 47%, indicating enhanced ZR accumulation and transport in treated plants. These results demonstrate that S2_8_1 remains metabolically active in a soilless environment, enhancing both cytokinin accumulation and its movement from roots to shoots. This dual function supports robust regrowth of hydroponically cultivated ryegrass and highlights the strain’s functional versatility beyond soil-based systems. The hydroponic system, by minimizing soil-mediated nitrification, provides a simplified framework to assess plant–microbe interactions independent of nitrogen transformation processes.

For S2_8_1, models that included both soil nitrification rate and leaf ZR content explained ryegrass biomass more effectively than single-factor models (Table 2). The interaction model yielded the best fit, with R² values of 0.91 for total biomass and 0.92 for aboveground biomass. Among individual predictors, nitrification rate showed a lower AIC than leaf ZR, indicating a stronger association with biomass variation. Importantly, the improved performance of the interaction model suggests a joint association and a positive statistical interaction between nitrification rate and cytokinin status in relation to biomass accumulation, rather than independent additive effects. In contrast, for Shan2, leaf ZR content alone best predicted total biomass (AIC = 34.26). For aboveground biomass, all models except the ZR-only model showed similar explanatory power (R² = 0.87). As illustrated in Fig. 4, total biomass under S2_8_1 treatment was positively correlated with both enhanced nitrification and elevated cytokinin levels (Fig. 4a). By comparison, in the Shan2 treatment, biomass was more strongly associated with ZR content than with nitrification (Fig. 4b). Regression analyses indicated that nitrification rate and leaf ZR jointly explained biomass variation under S2_8_1 treatment. This statistical interaction is consistent with, but does not by itself prove, a physiological synergy between nitrogen transformation and cytokinin-related responses.

This study shows that strain S2_8_1 promotes ryegrass regrowth in association with coordinated changes in rhizosphere nitrification and cytokinin-related plant responses: enhanced nitrogen cycling and cytokinin signaling. Compared with most PGPMs that influence either nutrient availability or hormone production, S2_8_1 boosts both soil nitrification and zeatin riboside accumulation and translocation. This dual effect highlights its potential as a promising candidate for next-generation PGPMs aimed at improving forage regrowth.

The concurrent increases in soil NO₃⁻ availability and cytokinin activity were consistently associated with improvements across multiple growth traits. Multiple linear regression revealed that the interaction between nitrification rate and leaf ZR content explained over 90% of biomass variation (Table 2; R² = 0.91–0.92). This result indicates a positive statistical interaction and joint association between the two variables in relation to plant growth, rather than independent additive effects. While previous studies have reported PGPMs enhancing either nitrogen cycling [19] or cytokinin production [24], S2_8_1 uniquely combines both. This trait is particularly valuable in intensively managed forage systems, where frequent cutting demands rapid nutrient uptake and shoot recovery [34,35]. By enhancing both rhizosphere nitrogen availability and cytokinin transport, S2_8_1 promotes faster regrowth and reduces reliance on synthetic nitrogen, aligning with sustainable agriculture goals [8,36].

Mechanistically, this synergy may involve co-expression of genes related to ammonia oxidation (amoA) and cytokinin biosynthesis (IPT like sequences) [37,38]. S2_8_1 may also activate root signaling in response to localized nitrate, triggering feedback that enhances both nutrient uptake and shoot development [29]. However, these interpretations remain hypothetical, and further genomic and transcriptomic analyses will be required to establish causal relationships and clarify the underlying regulatory mechanisms.

Although both S2_8_1 and Shan2 enhanced ryegrass regrowth, their mechanisms and overall efficacy differed markedly. Shan2, a Streptomyces strain, primarily promoted regrowth through cytokinin production, as evidenced by elevated ZR levels in treated plants. This aligns with previous reports highlighting the strong cytokinin-producing capacity of Streptomyces spp. [12,39]. However, Shan2 had smaller influence on soil nitrification and contributed less to biomass accumulation compared to S2_8_1. In contrast, S2_8_1 displayed broader functionality, enhancing both nitrogen cycling and cytokinin signaling. By day 7, S2_8_1 inoculated plants exhibited a 63.2% increase in soil NO₃⁻–N, a 19% increase in leaf ZR, and an 86% rise in L_ZR relative to controls (Table 1; Fig. 2). These results align with findings that nitrate availability can stimulate cytokinin biosynthesis and translocation [40,41], reinforcing the dual role consistent with physiological synergy of S2_8_1.

Regression analysis further highlighted this distinction (Table 2; Fig. 4). In S2_8_1-treated plants, biomass positively correlated with both nitrification rate and ZR content, supporting a synergistic effect. In contrast, under Shan2 treatment, soil nitrification explained only 79% of biomass variation, lower than the 91% in S2_8_1, and ZR content alone was a weak predictor. This likely reflects the limited role of cytokinins in root development, which depends more on nutrient availability and auxin signaling [42,43,44].

These contrasting mechanisms underscore the importance of selecting PGPM based on functional complementarity. While cytokinin-producing strains like Shan2 can support shoot regrowth, their limited contribution to nutrient cycling may reduce their effectiveness in nutrient-deficient systems. In contrast, multifunctional strains like S2_8_1 offer coordinated promotion of both shoot and root growth, outperforming single-function PGPMs [45,46]. Similar synergy has been observed in Pseudomonas fluorescens, where quorum sensing regulates both nitrogen and hormone pathways [5], suggesting a comparable regulatory mechanism may operate in S2_8_1.

Many soil derived microbes fail to function effectively in hydroponic systems due to the absence of soil structure and microbial networks [47,48]. In contrast, S2_8_1 retained its growth-promoting activity under hydroponic conditions, significantly increasing both total biomass and leaf ZR content. By day 10, ZR levels in S2_8_1 treated plants were 47% higher than in water controls, alongside substantial biomass gains (Fig. 3), demonstrating that the strain remains metabolically active in soilless environments. However, because nitrification-related variables were not quantified in the hydroponic system, these results do not directly support a nitrification-mediated mechanism under hydroponics. Instead, they demonstrate that the growth-promoting effect of S2_8_1 extends beyond soil-grown systems. S2_8_1 consistent performance across soil and hydroponic systems suggests adaptive traits such as efficient root colonization, secretion of bioactive compounds, and osmotic stress tolerance [16]. However, these interpretations remain speculative and require further investigation to elucidate the physiological and molecular basis of its broad-spectrum functionality.

This functional robustness is particularly relevant to emerging soilless agriculture systems, including vertical farming and greenhouse-based forage production, where biologically compatible inputs are increasingly sought [49,50]. While hydroponics offers advantages like high nutrient availability, reduced water use, and faster growth [27,28], the success of PGPMs in these systems depends on their ability to function independently of soil-based interactions. The ability of S2_8_1 to do so highlights its potential as a reliable bioinoculant for controlled-environment agriculture.

The dual functionality of S2_8_1 positions it as a promising biofertilizer that supports both productivity and ecological sustainability. By enhancing ryegrass regrowth in both soil and hydroponic systems, S2_8_1 improves nitrate availability, reduces reliance on synthetic nitrogen, and accelerates biomass recovery, benefits especially relevant to intensively managed grasslands where nutrient depletion and yield instability are persistent challenges [51,52].

Despite these promising outcomes, several limitations warrant consideration. This study was conducted under controlled conditions, which, while valuable for mechanistic insights, may not fully represent field complexity. Field-scale trials are essential to evaluate the consistency and robustness of S2_8_1 under variable environmental and edaphic conditions. Additionally, the genetic and molecular basis of its dual function remains to be elucidated. Genomic studies could reveal key genes involved in nitrification enhancement and cytokinin biosynthesis. The proposed dual-function mechanism was inferred from physiological measurements and statistical associations rather than direct molecular evidence. We did not quantify amoA, IPT-like genes, nitrifier abundance, gene expression, or enzymatic activity. Therefore, our results should be interpreted as evidence that S2_8_1 treatment was associated with increased net nitrification and elevated ZR-related responses, rather than direct proof of the underlying genetic mechanism. Future work should combine genome annotation, targeted qPCR, transcriptomics, enzyme assays, and direct metabolite profiling to verify the molecular basis of these effects. As interest in microbial consortia grows, S2_8_1 may serve as a foundational strain in engineered inoculants that combine nutrient cycling and hormone-mediated growth regulation [46,53]. Together, these findings contribute to the broader understanding of multifunctional PGPMs and highlight the potential of S2_8_1 for future exploration as a bioformulation component targeting improved crop performance and sustainability.