Sampling

Samples of rhizospheric soil from rice crops were taken in ten plots in the following areas: El 52 (area 1), Matecaña (area 2, 3, and 10), Limoncito (area 4), San Roque (area 5 and 6), Gorgona (areas 7 and 8) and Buena Esperanza (area 9); these areas are located in the Zulia River irrigation district (Norte de Santander). This district is located between 68 and 120 m altitude. It has a warm humid climate with temperatures between 27 and 34 °C, and a relative humidity of 70-90 %. According to Holdrige life zones, this region is classified as a tropical humid forest (bh-T) (^{IGAC, 2006}). The sampling areas were crossed in zigzag and 10 1-kg subsamples were taken from each plot (20-cm deep areas). This sampling was carried out 50-80 d after sowing. The samples were processed in the Applied Biology Research Laboratory of the University Francisco de Paula Santander.

Isolation of Azotobacter genus bacteria

The isolation was carried out spreading 10 to 30 rhizospheric granular soil on Ashby agar (^{Aquilanti et al., 2004}). Cultures were incubated at 32 °C until an abundant growth of viscous colonies with yellowish, green or brown pigment appeared around the soil grains. These colonies were cultured on new Ashby agar until pure cultures were obtained. Gram staining was performed in the isolations, to confirm the development of Gram-negative cells and cysts characteristic of this genus.

Isolation of Pseudomonas genus bacteria

The isolation was developed using the serial dilution methodology of rhizospheric soil in a saline solution (0.85 % NaCl). King B agar plates ^{(King et al., 1954}) were inoculated with 100 µl of 10^-2 to 10^-7 dilutions, and incubated at 32 °C until colonies surrounded with a soluble pigment formation appeared. The colonies were spread in a new agar King B until a pure culture was obtained. Later, Gram staining was performed to confirm the development of Gram-negative bacilli. The colonies on King B agar were exposed to ultraviolet light to observe the fluorescent pigment presence, a distinctive trait of the species of this genus.

The isolations were preserved using the method of vials under refrigeration with a sterile saline solution (0.85 % NaCl) (^{Sarmiento et al., 2013}).

Activation of microorganisms

Isolates preserved in a sterile saline solution were activated in Ashby and King B agar, for Azotobacter and Pseudomonas genera, respectively; they were incubated at 32 °C until the colonies grew. Plant growth promoting activities were carried out using a quantitative evaluation of the viable cultures obtained.

Preparation of bacterial inocula

From one strain of each isolation, an inoculum was prepared in 50 mL of DYGS culture broth (^{Radwan et al., 2005}) and incubated 24 h at 32 °C. A 10 mL suspension of this culture was transferred to a tube and centrifuged for 10 min at 4124 g. The supernatant was discarded, and 10 mL of a sterile phosphate buffer solution (0.06 M and pH 7.0) was added, in order to resuspend the cells and adjust the cell concentration to O.D._605nm=0.5 in a Spectroquant NOVA60 spectrophotometer. The cell suspensions obtained from each isolate were used to determine indoleacetic acid (AIA) production, solubilization of phosphates, biological nitrogen fixation, and siderophore synthesis.

Determination of indoleacetic acid (AIA) production

One-hundred µL were taken from the adjusted cell suspension, and they were inoculated into 25 mL of DYGS culture broth, plus a supplement of 0.15 g L^-1 tryptophan and 0.2 g L^-1 NH₄Cl. Cultures were incubated in the dark at 120 rpm and 32 °C. Each isolate was inoculated in three replicates and an uninoculated control was set up under the same conditions. After 72 h of incubation, 10 mL of the microbial broth were centrifuged for 10 min at 4124 g. Indolic compounds were determined by adding 1 mL of the supernatant and 4 mL of the Salkowsky reagent (5 mL of FeCL₃·6H₂O [0.5M] in 250 mL of HCLO₄ [35 %] for a 500 mL volume), until a pink coloration was obtained. The samples’ absorbance was recorded at 525 nm in the Spectroquant NOVA60 spectrophotometer. The concentration of indole compounds was calculated using the linear regression equation of the calibration curve constructed with seven concentrations of indoleacetic acid, AIA (25, 50, 100, 150, 200, 250 and 300 mM) (^{Cárdenas et al., 2010}).

Determination of phosphate solubilization

The determination of available P was carried out according to the Bray II method, adapted for bacterial cultures (^{Cárdenas et al., 2013}). 100 µL were taken from the cell suspension -adjusted to a O.D._605nm=0.5 from each isolation, as described above- and inoculated into 25 mL SRS broth (^{Sundara-Rao and Sinha, 1963}), with tricalcium phosphate as an insoluble P source. Cultures were incubated for 48 h at 30 °C at 120 rpm. Each isolation was inoculated in three replicates and an uninoculated control was set up under the same conditions. After incubation, 10 mL of microbial culture were centrifuged at 4124 g for 10 min. The available P was extracted from 10 mL of supernatant, plus 10 mL of the extraction solution (1.11 g of NH₄F and 25 mL of HCl [4N] in a volume of 1 L), and stirred 1 min. Then 1 mL of this solution was taken and 9 mL of a coloring solution were added (5 mL of solution A and 2 mL of solution B in 200 mL of distilled water) (Solution A: 30 g (NH₄)₆MO₇O₂₄·4H₂O in 100 mL of distilled water and 0.728 g of K(SbO)C₄H₄O₆·½H₂O in 160 mL of distilled water. Solution B: 3.05 g of ascorbic acid in 25 mL of distilled water.) The absorbance (605 nm) was recorded in a spectrophotometer 15 min after the reaction. The available P concentration (µg of P₂O₅ mL^-1) -which indicates the phosphate solubilization capacity- was calculated using the linear regression equation of the calibration curve constructed from concentrations of KH₂PO₄: 5, 10, 20, 30, 40 and 50 µg mL^-1.

Evaluation of biological nitrogen fixation

Nitrogen (N) fixation by rhizobacteria isolated strains was determined according to a modified version of the micro-Kjeldahl method (^{Kuss et al., 2007}). From the adjusted cell suspension to a O.D._605nm=0.5 (from each isolation, as described above), 1000 µL were inoculated into vials with 10 mL of a semi-solid Ashby medium for Azotobacter sp. and semi-solid JMV medium for Pseudomonas sp. The JMV medium had the following composition (g L^-1): Mannitol 5.0; KH₂PO₄ 0.6; MgSO₄·7H₂O 0.2; NaCl 0.1; CaCl₂·2H₂O 0.2; micronutrient solution 2 mL (g L^-1: CuSO₄·5H₂O 0.04; ZnSO₄·7H₂O 1.20; H₃BO₃ 1.40; Na₂MoO₄·H₂O 1.00; MnSO₄·H₂O 1.18); 2 mL of Bromothymol blue (0.5 % solution in 0.2N KOH); 4 mL of FeEDTA (1.64 % solution); 1 mL of vitamin solution (10 mg of biotin and 20 mg of pyridoxol-HCl in 100 mL); pH 4.0-4.5, agar 2.5. This procedure was performed in triplicate. After bacterial growth, cell lysis was carried out placing the vials in an autoclave for 1 min at 15 pounds of pressure. From the resulting suspension, 5 mL were added to a tube where the whole sample was digested, in order to transform all the organic N into a mineral form (NH⁺₃). The digestion was induced by adding 4 mL of concentrated sulfuric acid and 10 mL of 30 % hydrogen peroxide (V/V) to the tube containing the sample. The tubes were placed in the digestion equipment at an initial temperature of 180 °C for 120 min, and then, for 60 min at 375 °C. Subsequently, the tubes were removed from digester, and the content was gauged with distilled water into 50 mL. Samples were then recorded on the HACH DR/2000 spectrophotometer in order to quantify the total N (Nt) transformed into NH⁺₃. The quantizing of N_total was completed in: 1 mL of the digestion product, plus 2 mL of the Nessler reagent, 1 drop of catalyst, and 1 drop of polyvinyl alcohol. Yellow colorations were obtained, which were compared with the standard calibration curve of the equipment. The calculation of fixed Nt was showed in µg mL^-1.

Determination of the siderophore synthesis

From the cell suspension adjusted to an O.D._605nm=0.5 (from each isolation, as described above), 2.5 mL were taken and inoculated into 50 mL of Simon and Tesmann Broth (g L^-1 broth: malic acid 10; NaCl 5.8; KCl 3.7; CaCl₂·2H₂O 0.15; Tris 12.1; magnesium chloride 0.1; ammonium chloride 1.1; Na₂SO₄ 0.142; K₂HPO₄ 0.272; pH 6.8). The bacterial cultures were incubated at 120 rpm at 32 °C during 5 d. Each isolate was inoculated in three replicates and an uninoculated control was set up under the same conditions. From the cultures, 10 mL of broth were taken and carried to centrifuge tubes in order to determine the siderophores produced. Centrifugation was performed at 4124 g for 10 min. Siderophores quantizing was completed in 1 mL of the supernatant, plus 1 mL of a ferric perchlorate and perchloric acid solution (5 mM and 0.14 M, respectively). After the reaction was obtained, a 525-nm absorbance was recorded in the spectrophotometer. The siderophore concentration was calculated using the linear regression equation of the calibration curve constructed with five concentrations (5, 20, 50, 80, 100 µg mL L^-1) of a hydroxamate siderophore (deferoxamine mesylate, SIGMA) with the ferric perchlorate solution, and recorded with the same spectrophotometer.

Production of indoleacetic acid (AIA)

Synthesis of AIA in Azotobacter isolates ranged from 6.34 to 158.27 µg mL^-1. The values had great variability, but they showed this bacteria genus’ capacity to produce indolic compounds. The maximum value was obtained by the Azotobacter RzA044 isolate (158.27 µg mL^-1) -which only showed similarity to the Azotobacter strain RzA042 (145.30 µg mL^-1) -and was close to the maximum value recorded by ^{Sahoo et al. (2013)}, who report a production close to 180 µg mL^-1 -from A. vinelandii strain isolated from Indian rice soil. The other isolates were similar to each other, forming groups where the Azotobacter RzA040 and Azotobacter RzA041 isolates stood out. ^{Hussein and Joo (2015)} also reported -in Azotobacter strains isolated from Panax ginseng- a 32.29 and 46.68 µg mL^-1 production, which are close to those obtained by Azotobacter RzA038 and RzA045 isolates. ^{Barua et al. (2012)} report a 23.16 µg mL^-1 of indole compound production from an Azotobacter vinelandii strain (isolated from forest soils in India), which is similar to the values obtained by Azotobacter RzA043 and RzA046 strains.

The AIA production variability among same genus bacteria is also reported in other researches. ^{Escobar et al. (2011)} reported an 11.99 to 57.99 µg mL^-1 range for indoleacetic acid production in Azotobacter strains isolated from vegetables roots and rhizospheric soil.

The AIA synthesis of Pseudomonas isolates ranged from 7.62 to 25.6 µg mL^-1; RzA014, RzA027, and RzA035 strains stood out, but several isolates were not different (p>0.05). These values are close to those reported by ^{Malik and Sindhu (2011)} with a indoleacetic acid production range of 10.2 and 31.2 µg mL^-1 in fluorescent Pseudomonas species isolated from chickpea crops. ^{Lavakush et al. (2014)} also report 10 to 18 µg mL^-1 of AIA in Pseudomonas putida and Pseudomonas fluorescens strains isolated from rice soils. In other researches, the AIA production is higher than the one obtained in our study: ^{Habibi et al. (2014)} reported 37 µg AIA mL^-1 in a P. mandelii isolate obtained from soil cultivated with rice, while ^{Hussein and Joo, (2015)} reported 35.9 to 47.54 µg AIA mL^-1 of fluorescent Pseudomonas isolated from Panax ginseng.

The wide ranges of indole compounds production reported show the importance of quantifying this PGPR activity in the isolates obtained because their synthesis capacity may vary due to their genetic diversity and their environmental conditions (^{Karagöz et al., 2012}).

Solubilization of phosphates

The isolates with the highest available P values in the SRS culture broth were those of the Pseudomonas genus, with 5.97 to 46.43 µg mL^-1 values of available P; the RzA028 (33.13 µg mL^-1), RzA012 (33.46 µg mL^-1), and RzA035 (46.43 µg mL^-1) isolates -which were statistically similar (Table 1)- stood out. The results recorded for all Pseudomonas isolates were similar to those reported for: P. putida (40 µg mL^-1 available P) and P. pseudoalcaligens (45 µg mL^-1 available P) isolated from marginal soils in India with scarce vegetation (^{Patel et al., 2012}); Pseudomonas sp. strains isolated from cotton crops with less than 20 µg mL^-1 of available P; P. otitidis isolated from olive (6.55 µg mL^-1) and sesame plants (7.63 µg mL^-1) (^{Saber et al., 2015}). However, these values are lower than other rhizospheric Pseudomonas species’ such as: P. chlororaphis with 154.1 µg mL^-1 (^{Yu et al., 2012}) and P. putida with 140 and 150 µg mL^-1 (^{Lavakush et al., 2014}).

The Azotobacter genus isolates presented a 7.64 to 23.38 µg mL^-1 soluble P production in the liquid culture medium. The values obtained by the Azotobacter isolates were similar to those recorded in the cultured medium with 13 of the 32 Pseudomonas genus fluorescent rhizobacteria isolates. They also matched the P availability in the culture medium recorded by the phosphatesolvent activity of A. chrococcum strains (18.63 µg mL^-1 available P) (^{Abd El-Fattah et al., 2013}), and were higher than those reported by ^{Escobar et al. (2011)} in Azotobacter strains isolated from tomato (0.46-2.46 µg mL^-1 available P).

The Pseudomonas genus is set apart by its high phosphate-solubilizing capacity, although there are significant values in this biofertilization activity in Azotobacter (161.11 and 151.32 µg mL^-1, ^{Rojas-Tapias et al., 2012}), and A. chrococcum isolates (41.3 and 93.72 µg mL^-1, ^{López-Ortega et al., 2013}). Therefore, rhizobacterium are very important microorganisms for the plant, because they can directly increase P availability by solubilization, mineralization mechanisms, and, in alkaline soils, the reduction of pH, caused by the production of the following acids: lactic, acetic, oxalic, malonic, maleic, pyruvic, propionic, gluconic, citric, succinic, tartaric, fumaric, and trans-aconitic (^{Park et al., 2011}; ^{Marschner et al., 2011}).

Biological nitrogen fixation

The isolates evaluated presented a variation in the amount of N_total from 0.46 to 11.79 µg mL^-1 N-NH₃. The isolate with the highest biological nitrogen fixation value was Azotobacter RzA042. Higher values were recorded for other Azotobacter isolates (9.54.-11.79 µg mL^-1 N-NH₃) regarding Pseudomonas (0.46-9.28 µg mL^-1 N-NH₃). ^{Abdel-Aziez et al. (2014)} reported a biological nitrogen fixation of 11.11-200 µg mL^-1 N-NH₃ by Azotobacter genus bacterium. The Pseudomonas genus also fixes nitrogen, especially the species P. putida PS9 (^{Ahemad and Khan, 2012}), P. fluorescens and P. mendocina (^{Karagöz et al., 2012}).

Azotobacter is the main group of heterotrophic free-living nitrogen-fixing bacterium associated with plants. The species of this genus can fix, at least, 10 mg N g^-1 of carbohydrate, and N-fixing activity is higher in strains associated with rice than with other similar crops (^{Sahoo et al., 2013}). Therefore, we can infer that this genus has a great biofertilizer potential in rice crops. In our study, although N_total values varied among the isolate, all showed N-fixing activity, which is one of the main characteristics of plant growth promoting rhizobacteria.

Siderophore production

There were no significant differences in siderophore production. The isolates with higher numerical production were Pseudomonas sp. RzA034 and RzA035 (685.96 and 670.23 µg mL^-1 respectively), while the values of Azotobacter genus isolates were lower than 174.09 to 469.80 µg mL^-1. Pseudomonas genus species produce hydroxamate-type siderophores (ferribactin, pseudobactin) and catechol-type siderophores (pioverdine) that sequester Fe III from their environment, when its conditions are deficient; in this way, they limit the phytopathogens’ growth, acting as a biocontrol mechanism (^{Aguado-Santacruz et al., 2012}). Therefore, selecting isolates with this physiological feature is important in order to obtain an indirect benefit from plant growth promoting rhizobacteria. ^{Abdel-Aziez et al. (2014)} report a production of 11.3-59 µg mL^-1 of siderophores per Azotobacter strain. Even though quantitative evaluations of siderophores synthesis by this genus are not frequent, the production of several biscatecholato-type siderophores were reported: azotoquelin (with a high affinity for Mo, V, Cu, Zn) (Bellenger et al.,) and protoquelin triscatecholato (with affinity for Fe III, Mn II, and Cu II) (^{Harrington et al., 2012}). These authors indicate that the said siderophores are associated with the capture of Mo and V ions, which are required by the nitrogenase enzymatic complex in the atmospheric nitrogen fixation process.