Auxin production of the selected strains was determined following Sarwar method (Sarwar et al., 1992). Freshly prepared bacterial cultures were inoculated in 15 ml of YEM medium amended with 50 μg mL^-1 of L-tryptophane as a precursor for auxin production. Different concentrations of PEG₆₀₀₀ (5%, 10%, 20%, 30%, 35%) were added to induce osmotic stress. Bacterial cultures were then incubated at 28 ± 2 °C for three days at a shaking speed of 120 rpm. Two milliliters of bacterial cultures were collected every 24 hours to measure bacterial growth and auxin production. The amount of IAA production in each isolate was measured using the Salkowski reagent (Ehmann, 1977). After incubation, the cultures were centrifuged at 8000 rpm for 15 min at 4 °C, and the supernatant was mixed with Salkowski reagent then incubated in the dark for 30 min for color development. Absorbance was measured at 530 nm and the IAA concentration was estimated by using a standard IAA curve with pure IAA concentrations ranging from 0 to 100 µg mL^-1 (Gordon and Weber, 1951).

The strains’ capacity to produce siderophores was tested using the CAS-shuttle assay method developed by Schwyn and Neilands (1987). Bacterial cultures were grown in iron-deficient minimal medium (El Attar et al., 2019), supplemented with the same PEG₆₀₀ concentrations as mentioned above. After seven days of incubation, the bacterial cultures were centrifuged to remove the pellet. An equal volume of supernatant and CAS reagent was then mixed, and the optical absorption at 630 nm was measured. Siderophore production was estimated using the formula of Payne (1993).

The inorganic phosphate (P) solubilization capacity was determined by using Pikovskaya’s medium (PVK) (Pikovskaya, 1948) supplemented with 0.05 g of Moroccan rock phosphate as a sole source of P. PEG₆₀₀₀ was added to the medium using the same concentrations mentioned above. Bacterial cultures were incubated at 28 °C for 3 days with shaking. After incubation, the cultures were centrifuged at 12 000 rpm for 15 min to collect the supernatant. Soluble P content in the supernatant was estimated following the Vanadate-molybdate method (Tandon et al., 1968). Absorbance was measured at 405 nm and soluble phosphorus in the medium was calculated from the standard curve of P concentration using KH₂PO₄.

To determine the host range of R. laguerreae strains, four grain legume hosts were used: Lentil (Lens culinars), Pea (Pisum sativum), faba bean (Vicia faba), and common bean (Phaseolus vulgaris). The seeds were kindly provided by INRA (Institut National de Recherche pour l’Agriculture) Institute of Rabat, Morocco.

Seeds of P. sativum, V. faba, and Phaseolus vulgaris were surface sterilized by immersion in 70% ethanol for 2 minutes, followed by thorough rinsing with sterile distilled water to remove any ethanol residues. The seeds were then immersed in a 20% sodium hypochlorite solution for 20 minutes and rinsed again with sterile distilled water to eliminate any remaining traces of salt. Lentil seed sterilization was done following the same protocol described by Taha et al. (2018). After surface sterilization, all seeds were germinated in 0.7% agar plates at 28 °C in the dark for 2 days.

Five germinated seeds were transferred to plastic pots containing a mixture of an equal quantities of sterilized sand and surface-applied beads (Haweison and dilworth, 2016), then thinned to 3 after successful seedling establishment.

R. laguerreae strains were grown in liquid YEM medium at 28°C for 24h and bacterial density was standardized to 10⁸ CFU mL^-1. Seedlings in each pot were inoculated with 1 mL of rhizobia culture.

Two sets of control plants were settled: non-inoculated non-fertilized seedlings (Control N0) and non-inoculated fertilized seedling supplied with mineral nitrogen (0.5 g L^–1 KNO₃) (Control N+). The experiment was carried out with three replicates using a Randomized Complete Block (RCB) design and all pots were watered twice a week with an N-free nutrient solution.

To ensure compatibility between rhizobial and PGPR strains intended for mixed inoculation, potential antagonistic interactions were assessed in-vitro following the procedure outlined by Mikiciński et al. (2016). The strains were co-cultured on nutrient agar plates, and zones of inhibition were monitored to detect antagonistic effects. Experiments were performed in triplicate to ensure reproducibility.

The soil used in this experiment was non-sterilized agricultural soil collected from a local site. The soil’s pH was 7.64 ± 0.05, the cation exchange capacity (CEC) was found to be 0.21 ± 0.02 mS/cm, the potassium content was 120.02 ± 12.50 ppm, while phosphorus was present at 58.45 ± 0.71 ppm. Nitrogen was measured at 0.038 ± 0.004%, and the organic matter content was 2.88 ± 0.91%. The dry matter content was 91.13 ± 0.26%, and magnesium was present at 7.25 ± 0.04%. The carbon content in the soil was measured at 1.81 ± 0.005%, with sodium present at 1.29 ± 0.06 m eq 100 g^-1.

A preliminary irrigation test was conducted to determine the soil’s maximum water retention capacity, from which 30% and 70% moisture levels were selected to represent drought stress and well-watered conditions, respectively. The experimental setup included a total of 72 pots, randomly distributed across three blocks, with six replicates per treatment. The six treatments comprised: inoculated plants under drought stress, inoculated plants with normal irrigation, non-inoculated plants under drought stress, non-inoculated plants with normal irrigation, and two non-inoculated controls, negative (N0) and positive (N+), that was supplemented with N₂ fertilizer (0.5 g L^-1 KNO_3-). Both were subjected to the same irrigation regime. Non-inoculated controls received no microbial treatment, while inoculated plants were treated with the bacterial mixed inoculum.

The plants were initially watered to meet their growth requirements, and drought stress was applied starting in the third week of growth, continuing until the flowering stage. After 8 weeks of growth, all plants were harvested for further analysis of growth parameters, including root and shoot biomass, chlorophyll content, and drought tolerance indicators.

The multi-strain mixed inoculum was prepared by combining equal proportions of fresh cultures of each strain, which were separately grown in liquid YEM medium. The mixed inoculum suspension was then adjusted to a final concentration of 10⁸ cfu mL^–1 and applied immediately after transplantation, with a follow-up application after 15 weeks. The experiment was conducted following the RCB design, with six replicates per treatment.

Six weeks into the experiment, the plants were harvested, and the measurements of nodule number, shoot, and root system lengths were measured. Plants shoot and root parts were oven-dried at 70°C for 48 hours and dry weights were recorded. For chlorophyll (a), (b), and carotenoid measurements, leaf segments of each treatment were immersed in 4 ml of 80% v/v acetone solution and incubated for 48 hours at 4°C (Amine-Khodja et al., 2022). Photosynthetic pigments were quantified according to Lichtenthaler and Wellburn’s (1983) equation.

The proline content of all tested plants was extracted by using a cold extraction procedure by mixing plant samples with ethanol and left to incubate overnight at 4°C, then they were centrifuged for 5 min at 8000 rpm to retrieve the pellet and optical density of the supernatant was measured at 520 nm (Gibon et al., 2000). The detection of proline was related to a standard curve prepared as described by Carillo et al. (2011).

Anthocyanin biosynthesis was measured following the protocol of Mustilli et al., 1999. Fresh shoot biomass was incubated in a 1% HCl methanol solution in the dark for two nights at +4°C. After incubation, chloroform and distilled water were added to the supernatant retrieved by centrifugation. The absorbance of the supernatant was measured at 535 and 650 nm, using the extraction solution as the blank. Anthocyanin content was calculated as the difference in absorbance (A535 - A650) per mg of fresh material.

Total soluble sugar content was estimated using the method described by DuBois et al. (1951). Plant material was homogenized with 80% ethanol and incubated for two nights. After centrifugation, the collected supernatant was added to a reaction mixture containing 5% phenol and 96% sulfuric acid. The mixture was then incubated at 30°C for 20 minutes. The optical density of the resulting solution was measured at 490 nm, and the total soluble sugar content was estimated per gram of fresh weight, referring to a standard curve prepared with varying concentrations of glucose.

Protein production was calculated by grinding fresh leaves with a cooled phosphate buffer (pH=7.8). The supernatant obtained after centrifugation and pellet removal was incubated with Bradford reactant at 37°C for 5 min. The mixture absorbance was measured at 595 nm after incubation, and the concentration of soluble protein was assessed by using the Bovine serum albumin standard curve following the method described by Bradford (1976).

Bacterial growth was assessed simultaneously with the measurement of auxin production. The results presented in Figure 1 revealed the negative effect of water deficiency on bacterial growth, which showed a graduate decrease as PEG₆₀₀₀ concentrations increased. R. laguerreae LMR575, LMR597, and Bacillus sp. LMR 698, strains exhibited moderate growth compared the control, at a concentration of 30% maintaining an optical density above 0.5 at 620 nm. In contrast, R. laguerreae LMR 655 was able to tolerate up at 20% PEG₆₀₀₀. Meanwhile, E. aerogenes LMR 696 was highly sensitive to osmotic stress. Notably, all strains were found to be vulnerable to the high osmotic pressure applied by PEG₆₀₀₀ (35%).

The maximum amount of auxin produced by bacteria in YEM medium supplemented with 50 μg mL^-1 of L-tryptophan, was reached after 72 hours of incubation ( Figure 2 ). The bacterial strains exhibited a greater capacity to produce auxin under normal conditions compared to stressful ones. For R.laguerreae LMR 597, the auxin production dropped from 24.22 µg/ml in normal conditions to 12.03 µg mL^-1 at PEG₆₀₀₀ (35%). In contrast, auxin production by Bacillus sp. LMR 698 and R. laguerreae LMR 655 was only observed under normal conditions, with concentrations of 25.24 and 12.87 µg mL^-1, respectively. Additionally, R. laguerreae LMR 575 exhibited the highest auxin synthesis after 72 hours of incubation, while E. aerogenes LMR 696 showed a gradual decrease in auxin synthesis as the PEG concentration increased. It is noteworthy, that except for R. laguerreae LMR 597, the activity of all strains, was significantly inhibited at PEG₆₀₀₀ (35%), with auxin synthesis falling below 5 µg mL^-1.

Under normal conditions, E.aerogenes LMR 696 exhibited the highest level of siderophore production of 82%, followed by R. laguerreae LMR 597 with 27%, Bacillus sp. LMR 698 with 18%, and finally R. laguerreae LMR 655 with 17%. However, these strains showed no siderophore activity under varying water potential levels. Conversely, R. laguerreae LMR 575 showed a high level of siderophores production up to 88% at 5% PEG₆₀₀₀, but this production declined significantly with the increase in PEG₆₀₀₀ concentrations, reaching 16.78% at 35% of PEG₆₀₀₀ ( Figure 3 ).

All the bacterial strains tested demonstrated the ability to solubilize inorganic phosphate ( Figure 4 ). The highest level of production was recorded by R. laguerreae LMR 655, which produced 113.85 µg mL^-1. In comparison, R. laguerreae LMR 575 and LMR 597 showed moderate production levels of 48.5 and 61.57 µg mL^-1, respectively. Conversely, E.aerogenes LMR 696 and Bacillus sp. LMR 698 exhibited lower solubilization rates, producing 22.5 and 20.5 µg mL^-1, respectively. The ability of the studied strains to solubilize phosphate was significantly diminished compared to the control, and this decrease was correlated with increasing PEG concentrations. While R. laguerreae LMR 597 and LMR 655 could produce PO₄²^- at a lower level of 30 µg mL^-1 at 20% and 30% PEG₆₀₀₀, respectively. In contrast, R. laguerreae LMR 575, E. aerogenes LMR 696, and Bacillus sp. LMR 698 showed no phosphate solubilization capacity starting at 10% PEG₆₀₀₀.

The in-vitro compatibility test revealed that there are no antagonistic interactions between the selected strains.

Under non-stress conditions with inoculation, P. sativum and V. faba demonstrated high nodulation, with P. sativum producing over 80 nodules per plant and V. faba yielding an average of 67 nodules with minimal variation (SD = 0.52), indicating consistent nodule formation. In contrast, both species showed no nodulation under non-inoculated or KNO₃-supplemented treatments, regardless of stress conditions. Under stress conditions, however, inoculated plants of P. sativum and V. faba still formed nodules, though in reduced numbers (22.67 and 41.33 nodules per plant, with SD values of 0.90 and 0.45 respectively) ( Table 1 ).

As shown in Figure 5A, B , V. faba and P. sativum exhibited distinct responses in root and shoot length under various treatments. In V. faba, the non-stressed, inoculated plants showed the highest shoot length (84.33 cm), which was significantly greater than the non-stressed plants supplemented with KNO_3- and the stressed, non-inoculated plants, both of which did not differ significantly from the other treatments. Root length in V. faba remained relatively consistent across all treatments. In P. sativum, root length was significantly higher in stressed, inoculated plants (28 cm) compared to the stressed and non-stressed plants treated with KNO_3- (19 cm and 22 cm, respectively). The shoot length of stressed, non-inoculated P. sativum plants was notably lower (72 cm), differing significantly from stressed, inoculated plants and non-stressed plants supplemented with KNO_3- or inoculated, which all showed higher shoot lengths averaging around 113 cm.

The chlorophyll and carotenoid content in V. faba and P. sativum demonstrated a clear response to inoculation and stress treatments. In both species, the non-stressed, inoculated plants exhibited the highest concentrations of chlorophyll (a), chlorophyll (b), total chlorophyll, and carotenoids, significantly surpassing non-inoculated and KNO_3-supplemented plants. Specifically, non-stressed inoculated plants had a chlorophyll (a) concentration of 144.39 mg g^-1, which was significantly higher than both non-inoculated and KNO_3-supplemented plants. Regarding chlorophyll (b), inoculated plants under non-stressed conditions showed a concentration of 35.12 mg g^-1, again surpassing non-inoculated and stressed non-inoculated plants. Under drought stress, inoculation mitigated the reduction in chlorophyll and carotenoid content more effectively than KNO_3- supplementation, with stressed inoculated plants in V. faba maintaining higher chlorophyll (a) and total chlorophyll levels ( Figure 6A ). In P. sativum, stressed inoculated plants similarly retained relatively higher chlorophyll (a) and carotenoid concentrations compared to KNO_3- treated plants. For total chlorophyll content, non-stressed inoculated plants displayed a significantly high concentration of 179.51 mg g^-1, while under stress, inoculation increased total chlorophyll concentration to 84.90 mg g^-1 compared to KNO_3-supplemented plants. Interestingly, stressed non-inoculated plants retained relatively high chlorophyll (a) concentrations (115.85 mg g^-1) compared to KNO_3-treated plants ( Figure 6B ).

Proline production in P. sativum and V. faba under drought stress is shown in Figure 7 . In both leguminous species, the level of proline produced under drought stress in plants treated with a mixed inoculum is higher with values of 48 and 38 mg g^-1 for P. sativum and V. faba respectively, differing significantly with the other treatment. In V. faba ( Figure 7A ), no proline production was detected in plants under other treatment; Meanwhile the production of proline in P. sativum ( Figure 7B ) was not only related to the stress response but also to inoculation and KNO₃ addition, where there was no significant difference in the amount of proline produced in stressed and non-stressed plant amended with KNO₃ and non-stressed plant inoculated with the mixed inoculum. Whilst the lowest proline production was observed in non-stressed non-inoculated plants (6.5 mg g^-1) and stressed non-inoculated plants (13.5 mg g^-1).

Anthocyanins, pigments produced in response to stress, were also measured in our study ( Figure 8 ). The results showed that concentrations did not exceed 1 mg g^-1. In V. faba, anthocyanin synthesis did not appear to be associated with either stress response or the effect of inoculation ( Figure 8A ). Conversely, in P. sativum, the highest anthocyanin concentration was observed in the non-inoculated, stressed plants, reaching 0.65 mg g^-1 which is significantly higher than in the inoculated plants and those supplemented with KNO₃, where concentrations decreased to 0.176 and 0.116 mg g^-1, respectively ( Figure 8B ).

The effect of inoculation on plant physiological responses varies between leguminous species. Examining the behavior of V. faba ( Figure 9A ), we observed a highly significant accumulation of soluble sugar in stressed inoculated plants, reaching 3.29 mg g^-1, compared to 1.57 mg g^-1 in stressed non-inoculated and 1.60 mg g^-1 in the non-inoculated non stressed plants. Broadly speaking, both KNO₃ treatment and inoculation helped maintain soluble sugar levels in stressed plants similarly to those under normal condition treated with a mixed inoculum and KNO₃. Regarding P. sativum ( Figure 9B ), the highest significant concentration of soluble sugar was observed in the stressed non-inoculated plants reaching 3.44 mg g^-1, compared to the other treatments. Simultaneously no significant difference was detected between the plants irrigated with KNO₃ and those inoculated under the same stress condition, where the sugar concentration decreased to a range of 2.14 and 1.98 mg g^-1 respectively.

The protein content in V. faba plants ( Figure 10A ) was impacted by the applied stress. The non-inoculated plants exhibited the lowest protein content of 25.74 mg g^-1, which differed significantly from the plants amended with KNO_3, reaching a concentration of 49.03 mg g^-1, whilst under normal conditions, no significant difference was observed among the applied treatments. Conversely, in P. sativum plants ( Figure 10B ), a significant effect of the inoculation was observed in the non-stressed plants achieving a concentration of 66.79 mg g^-1, differing significantly from the plants supplemented with KNO₃ (52.70 mg g^-1) and the control (34.71 mg g^-1). Meanwhile, the inoculated plants and those irrigated with KNO₃ under drought stress displayed protein concentrations of 40.52 and 43.23 mg g^-1 respectively, which were significantly higher than the stressed non-inoculated plants.