Influence of purple non-sulfur bacterial augmentation on soil nutrient dynamics and rice (Oryza sativa) growth in acidic saline-stressed environments

Experimental design

The two-factor experiment followed a completely randomized block design with 20 treatments and four replicates, each of which was a pot of soil. In detail, factor (A) was the salinity of the water (0‰, 2‰, 3‰, and 4‰) and factor (B) was the application of the EPS-producing PNSB (negative control; supplied with the single strain of EPS18, EPS37, or EPS54; and supplied with the mixture of the three strains EPS18, EPS37, and EPS54) (Fig. S2). The experiment was conducted at the greenhouse (10.029783 N, 105.767414 E) in the Research and Experiment Agriculture Station, College of Agriculture, Can Tho University from June 2022 to November 2022, which was the fall-winter rice cropping season, one of the two major rice cropping seasons in Vietnam. In the greenhouse, the light and dark hours per day were 11 and 13, respectively; the temperature was 36 °C and humidity was 60%.

Soil preparation

The soil from the cultivation system of integrated rice and shrimp was collected, dried naturally, smashed, removed from plant residues. Eight kg of dry soil was put into a pot whose size at large bottom diameter × small bottom diameter × height was 23 × 17 × 18 cm (80 pots in total), respectively, and added with 2 L of water into each pot to make the soil saturated in waterlogged rice cultivation. The properties of experimental soil was characterized in Table 1.

Seed treatment and sowing

The rice seeds were sterilized with ethanol and sodium hypochlorite (1%) and then rinsed with deionized water to ensure the seeds were fully sterilized. Next, the seeds were incubated to germinate for 24 h in the dark. Then, 500 rice seeds were divided into five equal portions into five beakers (100 seeds in each beaker). Four beakers contained 10 mL of bacterial suspension (1 × 10⁹ CFU mL⁻¹) of EPS18, EPS37, and EPS54 and their mixture for 1 h before sowing, and 1 beaker contained distilled water as the control treatment. Each pot was sowed with six seeds per pot, possessing a density of 1.2 × 10⁴ CFU g⁻¹ dry soil weight (DSW) to suit the amount of 8 kg of soil for soil not to be the restricitng parameter. Each treatment was performed with four replicates. Each treatment contained 24 seeds to meet the requirement for pot experiment as the guidance of the International Rice Research Institute (International Rice Research Institute, 2014).

Fertilizers

The fertilization following the recommendation included 100 N–60 P₂O₅–30 K₂O (kg ha⁻¹) (Tan & Hach, 2013). The P fertilizer was applied 100% three days before sowing, and the N fertilizer was utilized at the rates of 30%, 30%, and 40% at 10, 20, and 45 days after sowing (DAS). The K fertilizer was applied in halves at 10 and 45 DAS.

Watering

Tap water was kept from 3 to 5 cm height from the soil during the experiment, because the paddy rice cultivars require a submerging condition. Moreover, in the saline water irrigated treatments, 5 mL of saline water was poured into each pot at 20, 27, 34, 41, 48, and 55 DAS to maintain a sustainable saline stress on rice and bacteria there, while control was the tap water.

Bacterial inoculation

A total of 3 mL of 1 × 10⁹ CFU mL⁻¹ bacteria (for the mixed PNSB, each strain accounted for 1 mL) was applied three times at 13, 18, and 30 DAS, to maintain an acceptable density of the bacteria before applying saline water. However, further number of PNSB applications was not suitable for farmers to conduct. Thus, the supplied amount of bacteria was 1.8 × 10⁶ CFU g⁻¹ DSW, and the total amount from both seeds and liquid biofertilizers was 1.812 × 10⁵ CFU g⁻¹ dry soil. The functions of PNSB were to provide EPS (1.82–1.97 and 1.14–1.59 mg L⁻¹), to fix N (4.16–4.69 and 7.36–9.65 mg L⁻¹), and to solubilize P from Ca-P (11.8–13.7 and 15.6–18.9 mg L⁻¹), to dissolve K (1.56–2.16 and 1.44–1.73 mg L⁻¹), and to produce ALA (2.14–2.56 and 2.96–3.16 mg L⁻¹) under microaerobic light and aerobic dark conditions, respectively.

Soil analysis

At the beginning and at harvest, the soil was drilled, collected, dried up, then removed from plant residues, smashed, and filtered via a sieve of 0.5 and 2.0 mm. All the soil analytic techniques were composed by Sparks et al. (1996) and briefly summarized as follows: pH_H2O or pH_KCl was extracted with a ratio of soil:water (1:2.5) or soil:1.0 M KCl (1:2.5) and measured by a pH meter. The solution derived from pH_H2O measurement was utilized for determining the EC value by an EC meter. The total N was digested by a mixture of saturated H₂SO₄-CuSO₄-Se, with a ratio of 100-10-1, determined by the Kjeldahl method, and titrated by 0.01 N H₂SO₄. The available N was measured by the blue phenol method at the 640 nm wavelength. The total P was converted into inorganic forms by a mixture of saturated H₂SO₄-HClO₄ and revealed in color by ascorbic acid at the 880 nm wavelength. The available P was determined by soil extraction with 0.1 N HCl+0.03 N NH₄F, where the soil : water ratio was 1:7. Then, ascorbic acid was utilized for exposing colors at 880 nm wavelength. The insoluble forms of P comprising of ferrous phosphate (Fe-P), aluminum phosphate (Al-P) and calcium phosphate (Ca-P), were extracted by corresponding solutions, including 0.10 M NaOH, 0.50 M NH₄F, and 0.25 M H₂SO₄, determined by ascorbic acid. Their colors were compared using a spectrophotometer at the 880 nm wavelength. The cation exchange capacity (CEC) was measured by extracting soil with 0.1 M BaCl₂ and titrated by 0.01 M EDTA. The concentrations of K⁺, Na⁺, Ca²⁺, and Mg²⁺ from the CEC extracted solution were measured using a spectrophotometer at wavelengths of 766.5, 589, 422.7, and 285.2 nm, respectively.

Bacterial densities

The parameter in the soil was determined according to the Most Probable Number (MPN) method of Naoki, Nishiyama & Matsumoto (2001). In brief, soil after diluted at 10⁻⁶ was spread on a dish, which was incubated in an anaerobic jar with gas-pak in 7 days. After incubation, the colonies appearing on the dish were counted.

Biochemical properties of the rice

The leaves and stem (straw) of rice were collected at 45 DAS, which is the stage that rice plants fully develop their leaves, in order to analyze proline by the ninhydrin method of Bates, Waldren & Teare (1973), which is briefly summarized as follows: the humidity of the samples was measured by drying up the samples of straw of rice seedling at 70 °C for 72 h. Then, 0.5 g of the rice samples was transferred to a tube whose size was 13 × 100 mm. A total of 10 mL of sulfosalicylic acid 3% was added to completely mill the samples, which were then shaken for 30 min and centrifuged at 3,000 rpm for 15 min. A clear solution was finally collected, and the debris was removed. The colorimetric analysis was conducted by reacting with 2.0 mL of the sample solution with 2.0 mL of ninhydrin and 2 mL of glacial acetic acid. The complex was well mixed and capped. The incubation was applied for 1 h at 100 °C. After that, the ice was used to cool it down, and 4 mL of toluene was added to the extracted solution, which was then shaken for 15–20 s and measured using a spectrophotometer at the 520 nm wavelength.

The concentrations of N, P, K, and Na in plants

The analysis of the concentrations of N, P, K, and Na in rice at harvesting (90 DAS) was conducted according to the method of Walinga et al. (1989). Rice straw and seeds of a pot at harvest were weighed, dried at 70 °C for 72 h, and measured for their dry weight. Then, they were then carefully milled and utilized for analyzing the concentrations of N, P, K, and Na in straw, and seeds. Samples were digested by a mixture of saturated H₂SO₄ and salicylic acid. The concentration of N in the sample was measured by the Kjeldahl method (Velp UDK129). The P was determined at 880 nm wavelength by a spectrophotometer (UV-VIS 1800; Shimadzu, Kyoto, Japan). Finally, K and Na were measured using an atomic spectrophotometer ICP–OES (Icap 6300 Duo; Thermo Fisher Scientific, Waltham, MA, USA).

Evaluating the growth, yield components, and grain yield

All the growth and agronomic parameters were measured following the description of International Rice Research Institute (2014).

The growth was measured by plant height and panicle length. The plant height of rice and their panicle length were checked at 90 DAS (harvesting). The plant height was determined from the ground to the peak leaf or the peak panicle, and four plants of a pot were put into measurement. The panicle length was determined from the neck of a panicle to its end, and eight panicles of a pot were checked.

Statistical analysis

The data was processed and analyzed with two-way ANOVA by SPSS software, version 13.0. The Duncan post-hoc test was applied to compared the differences between means of treatments at 5% significance.

The influences of PNSB and the saline irrigation on soil fertility

The pH extracted by water values in the treatments irrigated with saline water at 2‰, 3‰, and 4‰ concentration differed at 5% significance (p = 0.019), in comparison with that in the treatment without saline irrigation, with 3.77 compared to 3.95. On the other hand, the pH_H2O in the treatments supplied with a single strain of EPS18 or EPS37 or the PNSB mixture was 3.92, 3.95, and 3.88, respectively, while that in the treatment without bacteria was only 3.75. In the meantime, the pH_KCl did not change significantly under the influence of the saline water and the bacteria addition (Table 2). However, there were significant interactions between the two factors influencing on both pH_KCl and pH_H2O (Table S1). In detail, the pH_H2O increased in the treatments supplied with the single EPS37 strain at 2–3‰ salinity and with the single EPS54 strain at 3‰ salinity, in comparison to other salinity levels. For the pH_KCl, an identical trend as the pH_H2O was observed and recorded in Table S2.

The treatments irrigated with saline water had EC values that were significantly different from that in the treatment without saline water irrigation at 5% significance (p = 0.0014); that is, the values of 0.75, 0.76, 0.82, and 0.93 mS cm⁻¹ correspond to the salinity of 0‰, 2‰, 3‰, and 4‰. In addition, in the treatments supplied with the bacteria, the EC changed significantly at 5% (p = 0.000064) from that in the treatment without bacteria. To be more specific, the treatments supplied with a single strain of either EPS18, EPS37, or EPS54 or their mixture reduced the EC value, in comparison with that in the treatment without bacteria supplied, with 0.68–0.82 mS cm⁻¹ compared to 0.96 mS cm2⁻¹. Among those treatments, the treatment supplied with the mixture of EPS18, EPS37, and EPS54 reduced the EC the most (Table 2).

The total N and P contents did not differ significantly among the levels of saline water and the supplementation of the PNSB. However, for the available N content, although the salinity of the water resulted in equivalent outcomes of the available N concentration (17.7 mg NH₄⁺ kg⁻¹), in the treatments supplied with the PNSB individually or in mixture, the available N content was roughly 17.9–18.6 mg NH₄⁺ kg⁻¹, which was greater than that in the treatment without bacteria (15.6 mg NH₄⁺ kg⁻¹) (Table 2). For the available P content, the treatments supplied with a single strain of EPS18, EPS37, and EPS54 or their mixture had the available P content of 39.8–51.1 mg P kg⁻¹, remarkably greater than the available P content in the treatment without bacteria (35.6 mg P kg⁻¹). The concentrations of Al-P, Fe-P, and Ca-P were not statistically different at saline levels of the water. However, in the treatment supplied with the PNSB mixture, the concentrations of Al-P, Fe-P, and Ca-P were 59.4, 177.4, and 48.9 mg P kg⁻¹, respectively, and were lower at than those in the treatment without bacteria (p = 0.0044, p = 0.0519, p = 0.00003). Furthermore, although the treatments supplied with an individual strain of EPS18 and EPS37 reduced the Al-P content, supplying a single strain of EPS18, EPS37, and EPS54 did not contribute to the reduction of Fe-P content, with 195.6–198.1 compared to 200.7 mg P kg⁻¹ in the treatment without bacteria. Meanwhile, applying an individual strain of them reduced the content of Ca-P, with 54.3–57.5 compared to 69.4 mg P kg⁻¹ (Table 2).

CEC and the concentration of Mg²⁺ were not statistically different among the levels of saline water and the supplementation of the PNSB. Increasing the salinity of the water resulted in a greater concentration of Na⁺ accumulated in soil, with salt concentration of 0‰, 2‰, 3‰, and 4‰ corresponding to 0.835 ~ 0.811 > 0.968 ~ 0.956 meq Na⁺ 100 g⁻¹. However, the treatments supplied with a single strain of EPS37 and EPS54 reduced the content of Na⁺, compared with the control treatment. Noticeably, the treatment supplied with the mixture of EPS18, EPS37, and EPS54 had the lowest amount of Na⁺ (0.796 meq Na⁺ 100 g⁻¹). Moreover, in this parameter, the two factors significantly interacted. To be more specific, the highest Na⁺ contents were in the treatments without bacteria at 2–4‰ salinity (1.150–1.250 meq Na⁺ 100 g⁻¹), while the lowest one was found in the treatment with only the bacterial mixture (0.643 meq Na⁺ 100 g⁻¹). Moreover, at the same salinity, even at 0‰ salinity, treatments with bacteria all resulted in lower concentration of Na⁺ in soil, except for the treatment with the single EPS18 strain at 0‰ salinity. For the content of K⁺, the difference was insignificant (p = 0.169) among saline levels of the water, approximately 0.143–0.160 meq K⁺ 100 g⁻¹. However, adding a single strain of EPS18 and EPS54 or the PNSB mixture led to a rise in the K⁺ content by 0.024–0.026 meq K⁺ 100 g⁻¹ compared with that in the treatment without bacteria. For the amount of Ca²⁺, in the treatments applied with 3–4‰ saline water, the concentration of Ca²⁺ was greater than that in the treatment with no saline water. However, the Ca²⁺ content did not change significantly, and was 0.256 meq Ca²⁺ 100 g⁻¹ on average (Table 2).

The increasing levels of salinity of the water resulted in a reduction in the density of PNSB, with 5.19 > 5.02 > 4.91 > 4.73 MPN g⁻¹ DSW at salinity of 0‰, 2‰, 3‰, and 4‰, respectively. Nevertheless, the supplementation of the bacteria increased the bacterial density. In detail, the bacterial density peaked in the treatment supplied with the mixture of the three PNSB strains (6.13 MPN g⁻¹ DSW), and the treatment without bacteria had the lowest bacteria density (2.11 MPN g⁻¹ DSW) (Table 2).

The influences of the PNSB and the saline irrigation on the proline production in rice

The concentration of proline increased along with the salinity of the water from 0‰, 2‰, 3‰ to 4‰, with a value of 3.64 < 3.96 < 4.87 < 5.28 µmol g⁻¹ DW. Furthermore, in the treatments supplied with PNSB, the proline content was lower than that in the treatment without bacteria. To be more specific, the accumulation of proline in the treatments supplied with the PNSB individually or in mixture was 4.02–4.32 µmol g⁻¹ DW, while that in the treatment without bacteria was 5.60 µmol ⁻¹ DW (Table 3).

The influences of the PNSB and the saline irrigation on the biomass, the N, P, K, and Na concentration, and uptake of rice

The influence of the PNSB and the saline irrigation on the rice growth and yield

Rice grain yield

The grain yield in the treatments irrigated with saline water was different at 5% significance (p = 0.00001) from that in the treatment without applied saline water. To be more specific, grain yield in the treatments applied with 0‰ and 2‰ saline water was 22.5 and 21.3 g pot⁻¹, respectively, and greater than those in the treatments applied with 3‰ and 4‰ saline water, with 19.6 and 17.5 g pot⁻¹, respectively. Meanwhile, the treatments applied with bacteria had significantly different grain yields from those without bacteria. In overall, the treatments supplied with the PNSB mixture had a greater grain yield (22.6 g pot⁻¹) than the treatments supplied with an individual strain of EPS18 and EPS37 (20.0 and 19.7 g pot⁻¹). The lowest grain yield was found in the treatments without bacteria (17.7 g pot⁻¹) (Table 5). In the case of no saline water irrigated, the application of a single strain of EPS37 and EPS54 or the PNSB mixture resulted in a greater grain yield than in the control treatment. Moreover, at 2‰ salinity, supplying the PNSB individually or in mixture provided a greater grain yield than the treatment without bacteria. However, at 3‰ salinity, only the treatment applied with the PNSB mixture had a greater grain yield (23.1 g pot⁻¹) than the control treatment (17.3 g pot⁻¹). Furthermore, when the salinity reached 4‰, treatments with or without the PNSB had statistically the same grain yield values (Fig. 1).

The L. sphaeroides improved the properties of the acidic salt-affected soil

Acidity and salinity are considered major obstacles to rice cultivated on salt-affected acid-sulfate soil (Huang et al., 2017; Syed et al., 2021). Moreover, low pH reduces the availability of macronutrients but increases that of toxins, resulting in changes in plants’ absorption of nutrients (Tsai & Schmidt, 2021). Thus, acidity adversely affects the growth and the capability of accumulating nutrients of the plants (Huang et al., 2017; Tsai & Schmidt, 2021). Furthermore, high Na⁺ concentration also interferes with the process of taking nutrients and the growth of rice (Irakoze et al., 2021; Syed et al., 2021), while rice is one of the major crops in Vietnam, whose soil has been damaged by salinity and acidity. In the current study, the soil salinity was positively correlated with the saline levels in the water applied for rice cultivation, shown via the EC value and the amount of exchanging Na⁺ in soil. Therefore, an approach to ameliorate the pH, and to decease the EC and the Na⁺ concentration in paddy soil is required. The results in the current study indicated that supplying a single strain of EPS18 and EPS37 or the mixture of the three PNSB EPS18, EPS37, and EPS54 improved pH_H2O, and reduced the EC and the Na⁺ concentration, with 3.88–3.99 compared to 3.75, with 0.68–0.82 compared to 0.96 mS cm⁻¹, and 1.144% compared to 0.769% in the treatment without bacteria, respectively (Table 2). This is because these bacteria belonged to PNSB which has been proven to be able to improve soil functions by increasing carbon and nitrogen contents, soil aggregation and stability and soil water status in adverse environments (Mager & Thomas, 2011; Chamizo et al., 2012). In the meantime, this bacteria group can produce EPS, which is considered to be a compound that can cut off the salinity in soil via functional groups within (Nunkaew et al., 2015b; Panwichian et al., 2011). According to Panwichian et al. (2011), the PNSB are capable of reducing Na⁺ content by secreting EPS to immobilize Na⁺ or to accumulate Na⁺ in bacterial biomass. The correlation analysis of the current study indicated the interaction between the bacterial density in soil and the soil Na⁺ content (r = 0.7635) (Fig. 2A). This should show that the appearance of the PNSB changed the Na⁺ concentration in soil. Other bacteria, such as Alcaligenes sp., are found to be able to reduce saline stress on rice plants by producing EPS (Fatima et al., 2020). In the current study, the ability to produce EPS by the strains of EPS18, EPS37, and EPS54 was almost the same, with 1.86, 1.97, and 1.82 mg L⁻¹ in microaerobic light and 1.14, 1.59, and 1.24 mg L⁻¹ in aerobic dark conditions (our preliminary study, LNT Xuan, ND Trong, VM Thuan, TC Nhan, LT Quang, LTM Thu, NHM Anh, NQ Khuong, 2023, Unpublished data). However, the capability of reducing Na⁺ content of these bacterial strains was insignificantly different. For further evaluation, the more saline it is, the more EPS the PNSB produce (Zeng et al., 2016; Ji et al., 2020). The bacteria, Lactobacillus plantarum VAL6, have been found to produce exopolysaccharide 6.4 times greater in stress caused by low pH (pH 3.0). However, being stressed by low pH and high Na⁺, the L. plantarum VAL6 produces exopolysaccharides with charged components (Nguyen et al., 2021). Therefore, these strains should be further evaluated under different stresses for EPS production.

Apart from that function, there are two types of EPS in the soil produced by bacteria. The first one is the loosely bound EPS, which is less condensed, and easily released into the environment; the other one is the tightly bound EPS, which is a more condensed, and firmly bonds to the cells and soil particles (Rossi & De Philippis, 2015; Rossi, Mugnai & De Philippis, 2018). The EPS matrix performs as a repository for nutrients and metabolites produced by microorganism (Swenson et al., 2018), and cations (De Philippis, Colica & Micheletti, 2011; Rossi & De Philippis, 2015), it can also hold water, and increase water uptake and retention in soils (Colica et al., 2014; Adessi et al., 2018). Thereby the matrix greatly enhances the nutrient contents in the soil. This contributed to explaining why nutrient contents in the soil in treatments with the bacteria were greater than those without bacteria. Consequently, these strains indirectly contributed to the improvement in the availability of macronutrients and limited the harm from soil toxins. This helped in interpreting why the amount of available N, P, and K increased in the treatments supplied with bacteria. Previous studies have also revealed that the bacteria of R. palustris possessing abilities to fix N and solubilize P and K were applied in order to enhance plant growth (Khuong et al., 2018, 2021). Similarly, the PNSB strains EPS18, EPS37, and EPS54 were found to be capable of fixing N and solubilizing P and K. As in the outcome, their application increased the concentrations of the NH₄⁺ from 2.1 to 3.0 mg NH₄⁺ kg⁻¹, soluble P from 4.2 to 15.5 mg P kg⁻¹, and exchangable K from 0.013 to 0.026 meq K⁺ 100 g⁻¹, in comparison with those in the treatments without bacteria (Table 2). This can be interpreted that the PNSB can fix N, due to the genes they have, including nifH, vnfG, and anfG (Sakpirom et al., 2017), and solubilize insoluble P forms in soil because of their ability to secrete acid phosphatase enzyme and siderophores (Khuong et al., 2018, 2020). Previous studies indicated that the PNSB belonging to Rhodopseudomonas and Luteovulum genera are capable of solubilizing P in saline and tolerating acidic conditions, and aimed to ameliorate the availability of P in soil (Khuong et al., 2018, 2021). This can be seen in the reduction in the amount of insoluble P compounds in soil in the treatments supplied with the mixture of EPS18, EPS37, and EPS54, whose concentrations of Al-P, Fe-P, and Ca-P valued at 59.4, 177.4, and 48.9 mg P kg⁻¹, respectively, and lower than those in the treatments without bacteria, with 67.5, 200.7, and 69.4 mg P kg⁻¹, respectively. However, the P solubilization was different from each strain. In particular, the strain of EPS18 and EPS37 had a lower concentration of Al-P than the control treatment, while applying the strain EPS54 had an equivalent amount of Al-P, compared to the control treatment. Nevertheless, supplying the PNSB strain EPS18, EPS37, or EPS54 did not reduce the Fe-P significantly, while the combination of them made a drop in the concentration of Fe-P in soil. Moreover, using the three strains EPS18, EPS37, and EPS54 separately or in mixture all had a Ca-P concentration of 48.9–57.5 mg P kg⁻¹, lower than those in the control treatment (Table 2). Thus, the improvements in the nutrient availability and limitation in Na element by the PNSB application, the performance of rice under saline condition Therefore, the N fixation and P solubilization of these strains should be further specifically investigated in the future studies.

The L. sphaeroides improved the tolerance of rice under acidic saline condition

Abdelaziz et al. (2018) assumed that proline is synthesized to maintain cellular osmosis and to protect the structure of enzymes, proteins, membrane, and the cell itself in dry condition; that is, proline helps plants to tolerate stress. Rice plants accumulate greater proline concentration in a saline stress condition than in a normal one to adjust the osmosis, enhance water absorption, restrict Na⁺ and Cl⁻ uptake, and transport from roots to stem (Wu et al., 2003; Bhusan et al., 2016). Consistently, the proline concentration in the treatments applied with 2–4‰ saline water was 3.96–5.28 µmol g⁻¹ DW, while in the treatment without saline water, the amount of proline was 3.64 µmol g⁻¹ DW. It can be observed that the increase in the salinity of the water resulted in the increase in the accumulation of proline in plants. Therefore, the amount of proline produced is an expression of stress in rice plants. Interestingly, in the treatments without bacteria, supplying an individual strain of EPS18, EPS37, and EPS54 or their mixture, the proline concentration was 5.60 > 4.32 ~ 4.18 ~ 4.05 ~ 4.02 µmol g⁻¹ DW (Table 3). This is in consistent with the experiment where the Na⁺ concentration in soil was also reduced by the application of the PNSB. In other words, it can be interpreted that the application of the PNSB immobilized the exchanging Na⁺ amount in soil, leading to rice that was less influenced by saline stress and produced less proline as well. The result also demonstrated the interaction between the amount of proline produced and the Na⁺ concentration in soil (r = 0.6220) and in plants (r = 0.7462) (Figs. 3A and 3B). That is, the greater the Na⁺ concentration appeared in soil, the greater Na content accumulation in plants, and the greater proline was produced. Figure 4A illustrated that the treatments supplied with bacteria had a low proline concentration, with a correlation coefficient of r = 0.6293, i.e., the bacteria reduced the saline stress in rice plants. Overwise speaking, the PNSB strains EPS18, EPS37, and EPS54 in the current study were the factor that helped rice plants to overcome saline stress. Other strains of PNSB have been proven for reducing saline stress on rice via proline content in plants (Khuong et al., 2023a).

The L. sphaeroides improved N, P, K, and Na uptake

The N concentration in straw and the K concentration in seeds decreased when the salinity of the water increased to 0‰, 2‰, 3‰, and 4‰. Meanwhile, the N and P concentrations in seeds and the P and K concentrations in straw remained statistically unchanged among the saline water levels. Nevertheless, the Na concentration in straw and in seeds increased positively to the saline levels of the water. This showed the influence of the soil salinity on the nutrient contents within the plants. However, the supplementation of the mixture of the three PNSB boosted the amount of N, P, and K and reduced the Na concentration in straw and in seeds. This can be inferred that supplying the PNSB strains promoted soil pH, improving the availability of N, P, and K (Table 2). This played a role in increasing N, P, and K uptake and concentrations within the plant. However, the nutrient providing capacity among the strains were different. In particular, supplying the single strain of EPS54 shared the same trend as supplying the mixture of the three PNSB for the amount of N, P, K, and Na, except for N concentration in seeds. Supplying the single strain of EPS37 increased the N and P concentrations in seeds and the P and K concentrations in straw. Meanwhile, it reduced Na content in both stovers. Supplying the single strain of EPS18 increased P and K content and reduced Na content. However, it did not change the N content both in straw and in seeds (Table 3). The bacterial density in soil and the Na content in seeds were closely correlated (r = 0.7635) (Fig. 2B). This demonstrated that the treatments supplied with either a single strain of EPS18, EPS37, and EPS54 or their mixture were effective in decreasing the amount of Na in seeds. The different functions possessed by different strains in the same mixture were also observed in the study by Khuong et al. (2023a). However, the current strains should be further investigated for K solubilization due to the increases in K content and K uptake found in plants. This has been also found in the study by Khuong et al. (2023b).

When the salinity reached 4‰, the N, P, and K uptake dropped and the Na uptake rose both in straw and in seeds. Applying the saline water also lowers the nutrient uptake of plants (Fageria, Gheyi & Moreira, 2011). In addition, supplying a single strain of EPS18, EPS37, and EPS54 or their mixture increased the N, P, and K content in straw and in seeds, except for supplying the single strain of EPS18 for N uptake in straw. This can be explained that the bacteria produced EPS limiting the movement of Na⁺ ion (Nunkaew et al., 2015b; Panwichian et al., 2011), which inhibited the transportation of Na⁺ into rice. Bacterial strains of Bacillus aerius ATHM 35, Arthrobacter luteolus ATHM 36, Halomonas eurihalina ATHM 37, B. tequilensis UPMRB9, and B. aryabhattai UPMRE6 have been isolated from a hypersaline environment containing functional groups, including hydroxyl, alkyne, amides, and carboxyl, in order to bind and chelate Na⁺ in the soil (Isfahani et al., 2018; Shultana et al., 2020b). Thus, Na accumulated less within the rice (Table 4). This result was in accordance with the study by Khuong et al. (2021), where the supplementation of L. sphaeroides W03 and W11 contributed to reducing Na uptake in rice. Furthermore, L. sphaeroides secreted ALA, promoting rice tolerance to overcome saline stress and their nutrient uptake (Khuong et al., 2023a). However, using a single strain of EPS18, EPS37, and EPS54 or their mixture did not reduce the Na content in straw and did not significantly affect the Na content in seeds (Table 3). This could be interpreted that, before flowering, the saline accumulation occurs due to the control of roots and the Na distribution to plant parts, but, after flowering, it is caused by the transpiration in the panicles (Asch et al., 1999). Nevertheless, supplying a single strain of EPS18, EPS37, and EPS54 or their mixture followed the opposite trend. That is, the total N, P, and K uptake increased, and the total Na uptake decreased (Table 4).

The EPS providing L. sphaeroides improved the growth, and the yield of rice

Irrigating saline water from 2‰ above, the panicle length, the number of panicles per pot, and the percentage of filled grains remarkably decreased. This resulted in a reduction in rice yield by 12.9–22.2% in the treatments irrigated with 3–4‰ saline water (Fig. 1). This was in accordance with the study by Irakoze et al. (2021), where rice yield decreased by 50% when being applied with 3‰ saline water. As demonstrated by Aguilar et al. (2017), applying saline water at the vegetative stage of rice lowered plant height, number of panicles, 1,000-seed weight, number of filled seeds, harvest index, and grain yield due to the osmosis and toxicity of Na⁺ ion (Mensah et al., 2006; Rengasamy & Olsson, 1993). However, the average 1,000-seed weight was insignificantly different under the influence of both saline water irrigation and the supplementation of the PNSB (Table 5). On the other hand, supplying a single strain of EPS18, EPS37, and EPS54 increased plant height, number of panicles per pot, and number of seeds per panicle. Meanwhile, supplying the mixture of EPS18, EPS37, and EPS54 enhanced agronomic parameters and yield components previously mentioned. In addition, supplying a single strain of EPS18, EPS37, and EPS54 or their mixture did not change the weight of 1,000 seeds (Table 5). Therefore, the application of a single strain of EPS18, EPS37, and EPS54 increased grain yield by 11.3–19.8%. Meanwhile, combining the three PNSB strains altogether increased the grain yield by 27.7% (Fig. 1) due to the positive correlation between the bacterial density and the rice grain yield (Fig. 4B). This was consistent with the study by Nunkaew et al. (2015a), where supplying the bacteria of R. palustris TN114 and PP803 stimulated rice growth in saline stress conditions. Moreover, in the current study, since the strains of EPS18, EPS37, and EPS54 excreted EPS, soil salinity was decrease (Table 2) and, thereby, grain yield was enhanced (Fig. 1). L. sphaeroides provides EPS, a key compound for metabolic processes, and helps bacteria to colonize plant roots and the rhizosphere, resulting in an improvement of the growth and yield of plants (Banerjee et al., 2019). Furthermore, when supplying with EPS providing saline-tolerating bacteria, including B. tequilensis UPMRB9 and B. aryabhattai UPMRE6, the photosynthesis, transpiration, and stomatal conductance are better, resulting in better plant yield (Shultana et al., 2020a). Ultimately, supplying the PNSB strains of EPS18, EPS37, and EPS54 individually or in mixture all improved rice grain yield in saline conditions. Nevertheless, although rice growth and yield benefited from the EPS18, EPS37, and EPS54 strains by their ability to reduce salinity, and to increase nutrient availability, other mechanisms may also involve, such as pyrroloquinoline quinone (Lo et al., 2023), IAA (Yusof & Hasmoni, 2023), siderophores (Nookongbut et al., 2018), and ALA (Khuong et al., 2023a). Therefore, the strains investigated in the current study should be latter on tested for producing the above plant growth promoting substances. Moreover, the PNSB are sensitive and behave differently to different concentrations of nutrients. For instance, in the study by Shaikh et al. (2023), the growth of PNSB is restricted during nutrient deficiency. Moreover, when N is deficient, biofilm formation is enhanced; when Mg is not applied, PNSB produce protein the most; and the least protein content is produced when Ca is deficient. Therefore, an appropriate amount of nutrient fertilizers should be further considered along with a suitable density of bacteria used according to their specific functions. Finally, a field test should be made, when the three strains have been investigated for their specific functions and their response to different nutrient rates, so as to clarify their true performance and their possibility to be commercialized and transferred to farmers.