IntroductionTop

Soybean [Glycine max (L.) Merr.], is one of the most cultivated legume worldwide due to its high protein content and important industrial by-products (Stacey et al., 2004; Masuda & Goldsmith, 2009). Nitrogen is one of the macro-nutrients required for growth and development of plants, so this nutrient must be supplied via chemical fertilization and/or biological N fixation (BNF) (de Carvalho et al., 2013; Gai et al., 2017). For soybean, the BNF is carried out in root nodules by symbiotic bacteria known as rhizobia. Rhizobia obtain carbon sources from plants and in return, the bacte­rial nitrogenase enzyme activity provides ammonia to plants (Santi et al., 2013). The symbiosis between soybean and rhizobia is a complex process involving activities of several genes (de Carvalho et al., 2013; Lira et al., 2015). Many legumes form root nodules that may stimulate plant growth when nodules are formed by compatible and functional rhizobial strains. In this regard, the bacterial establishment is highly specific, but strongly affected by genetic bacterial/plant interactions that determine an efficient symbio­tic relationship (Boonkerd & Singleton, 2002). This symbiosis is also largely dependent on environmental conditions (Valencia et al., 2010).

Most soybean varieties fail to nodulate efficiently in tropical soils even though plants are inoculated with competitive rhizobial strains (Kueneman et al., 1984). In this sense, native rhizobia most likely promote greater plant growth since they are more adapted to environmental conditions (Waluyo et al., 2005; Soe & Yamakawa, 2013). Thus, the isolation and re-introduction of highly competitive and effective native rhizobia are important for increasing soybean production.

Besides rhizobia, root nodules may host other endophytic non-symbiotic bacteria (Li et al., 2008; Saïdi et al., 2011; Aserse et al., 2013), and these bacteria are unable to form nodules nor perform N-fixation. However, these endophytic non-symbiotic bacteria may favor plant growth and nutrition, or assist the solubilization of insoluble forms of phosphates in the rhizosphere (Bai et al., 2002; Li et al., 2008; Liu et al., 2010; Stajkovic et al., 2011; Aserse et al., 2013). Co-inoculation of endophytic non-symbiotic bacteria along rhizobia has gained special interest as part of sustainable agriculture, since both bacteria may act synergistically for enhancing legume growth and performance, in comparison to the single inoculation of rhizobia (Bai et al., 2002). For instance, Rhizobium phaseoli co-inoculated with either Pseudomonas sp. or Bacillus sp. Bx, resulted in significant increase in the stem dry weight of beans (Stajkovic et al., 2011). Similarly, the co-inoculation of Mesorhizobium gobiense with Bacillus pumilus B402 resulted in increased number of nodules and growth of Sphaerophysa salsula (Pall.) DC. (Deng et al., 2011). Also, an increased root nodulation in Medicago sativa L. was reported due to the co-inoculation of Sinorhizobium meliloti and Agrobacterium tumefaciens (Wang et al., 2006). Similarly, an increase in nodulation of Wisteria sinensis (Sims) DC. was observed by combining S. meliloti and Agrobacterium sp. II CCBAU 21244 (Liu et al., 2010). In the case of soybeans, the co-inoculation of Bradyrhizobium japonicum with Bacillus subtilis or B. thuringiensis increased both plant weight and nodulation (Bai et al., 2002, 2003). In contrast, Camacho et al. (2001) reported decreases in root nodulation of soybean due to the co-inoculation of B. japonicum USDA110 and Bacillus sp. CECT450. The contrasting effects of bacterial co-inoculation on soybean indicate the necessity for identifying efficient combinations of rhizobia and other endophytic strains to promote plant growth and yield, thus, reducing the application of high doses of chemical fertilizers. So, it is important to explore endophytic bacterial strains cohabiting soybean nodules, as they may contribute to growth promotion in legumes. Consequently, the latter allows the reduction of environmental pollution and promotes sustainable agriculture of soybean in tropical regions.

The objectives of this work were to: 1) isolate and characterize symbiotic and endophytic non-symbiotic bacteria from root nodules of Glycine max collected from two tropical cropping systems at Campeche, Mexico, and 2) evaluate the effects of the most prominent endophytic bacterial strains when co-inoculated with referential or native rhizobial strains on the growth of soybean plants. The results are expected to contribute on selecting an efficient combination of rhizobia and endophytic bacteria for being used as biofertilizers for soybean cultivation in tropical conditions.

Material and methodsTop

Sites, cultivars and nodule sampling

Soybean roots with nodules were collected during August 2015 from two soybean fields at Campeche state, Mexico, at the locations of San Antonio Cayal (19°39' N, 19°40' W) (municipality of Campeche), and Nuevo Progreso (19°40' N, 89°43' W) (municipality of Hopelchén). At San Antonio Cayal, three varieties of soybean already registered by the Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias (INIFAP) were collected: ‘Huasteca 100’ (SOY-014-251104), ‘Huasteca 200’ (SOY-015-251104), and ‘Huasteca 400’ (SOY-022-291105). At Nuevo Pro­greso, the ‘Huasteca 200’ variety, and the transgenic soybean resistant to herbicide-glyphosate (also known as Roundup Ready®, GR, or RR) were collected. In each location, the root system of five randomly selected healthy plants of each variety was harvested after 50 days after of sowing. Roots were transported in plastic sterile bags to the laboratory.

Bacteria isolation from root nodules

Nodules were dissected for each root system, especially those in which the presence of pink coloration in the cortex was observed, indicating their viability and potential N-fixation. Root nodules were kept at 4 °C and surface disinfected with 70% ethanol for 10 s, and with 0.4% sodium hypochlorite for 1 to 3 min, and then rinsed 5 to 6 times with sterile distilled water. Nodules were immediately placed in sterile test tubes containing 1 mL of sterile distilled water, and crushed for obtaining a bacterial suspension. An aliquot (100 µL) of the bacterial suspension of each soybean variety was spread on the surface of Petri dishes containing yeast extract-mannitol agar (YMA) medium (Vincent, 1970) modified to contain 5 g/L of mannitol and 0.00125% Congo red (w/v); then, incubated at 28°C for five days. The purification of bacteria was performed as indicated by Vincent (1970), and single bacterial colonies were selected by color and shape, and re-streaked for assuring the purification.

Morphological and biochemical characterization of bacterial isolates

Bacterial isolates were morphologically characterized by distinguishing form, color, margins, surface, and size of colonies. In addition, bacterial cells were stained with the Gram technique and microscopically examined (Leica CME) with a 100X objective lens.

All bacterial isolates were grown on glucose-peptone agar with bromocresol purple (GPA-BP) (Somasegaran & Hoben, 2012), and in litmus milk liquid medium (Litmus Milk®) (Ferrera-Cerrato et al., 1993). The isolates that did not show growth in these two culture media were considered as potential rhizobia. The isolates were also grown on YMA medium containing bromothymol blue (YMA-BTB) as indicator for determining their ability to produce alkaline (blue color) or acidic (yellow color) reactions (Ferrera-Cerrato et al., 1993), and tested for identify­ing the ability to solubilize Ca3(PO4)2 by streaking on Pikovskaya agar medium. The presence of a clear zone around the bacterial colony indicated a potential release of organic acids for inducing the phosphate solubilization (Sundara & Sinha, 1963).

Molecular identification of the isolates

The isolates were identified by partial sequencing of their 16S rRNA gene. The isolates were grown in YMA medium for 24 to 72 h, depending on the isolate. Total DNA was extracted using cetyl­trimethylammonium bromide 2% (CTAB) (Doyle & Doyle, 1990). Partial sequence of 16S rRNA gene was amplified by PCR using universal primers 8F (5'-AGAGTTTGATCCTGGCTCAG-3') and 1492R (5'-G­T­TACCTTGTTACGACTT-3') (Eden et al., 1991), u­sing the following conditions: one cycle of 95°C for 2 min, followed by 35 cycles of 95°C for 2 min, 59°C for 1 min, 72°C for 1.5 min, and finally at 72°C for 5 min.

The PCR products of approximately 1500 bp were purified with EXO-SAP (Affymetrix, USA) following the manufacturer instructions. The fragments were sequenced in a Genetic Analyzer® Model 3130 (Ap­plied Biosystems, USA) but the consensus sequence was generated from forward and reverse sequence data using BioEdit v7.2.5 (Hall, 1999). It is important to note that all of the consensus sequences were analyzed with the Blastn algorithm from the BLAST/NCBI software (Altschul et al., 1997) and the Ribosomal Database Project release 11 (https://rdp.cme.msu.edu/). Sequences obtained in this study were compiled in a FASTA format along with sequences belonging to type strain (http://www.bacterio.net/). Sequences were aligned using the muscle option included in Mega X software (Kumar et al., 2018). In addition, they were trimmed at the ends for analyzing fragments with the same length. The phylogenetic reconstruction of all sequences was performed with Bayesian inference (BI) in MrBayes v3.2.6 (Huelsenbeck & Ronquist, 2001; Ronquist & Huelsenbeck, 2003), mega file was converted to nexus file for BI and the INVGAMMA substitution model was used with 1,000,000 generations and sampled every 1000 generations. The first 25% of generated trees were discarded as the burn-in phase option of each analysis and posterior probabilities were determined for the remaining trees. The construction of the phylogenetic tree considered the strain NR114653 of Thermococcus marinus as outgroup.

Effects of the endophytic isolates on soybean growth and nodule development

The bacterial isolates were evaluated for inducing nodule formation on roots of soybean ‘Huasteca 200’ plants, following a bioassay procedure (Ferrera-Cerrato et al., 1993). Seed surface was sterilized twice by using 0.2% sodium hypochlorite for 1 min, followed by 70% ethanol for 1 min (2 times), and 5 rinses with sterile distilled water. Seeds were placed on sterile filter paper in Petri dishes. After germination, seedlings were transplanted to 500 mL-pots with autoclaved perlite (121°C for 2 h).

Each isolate was grown in YMA for five days and then, adjusted to a concentration of 109 CFU/mL. One milliliter was used for inoculating each plant in accordance to the corresponding bacterial treatment, and ten individual plants were used as replicates per treatment. In total, 25 treatments were evaluated (corresponding to the 23 bacterial isolates) including a negative control (uninoculated plants), and a positive control (inoculated with the reference strain B. japonicum USDA110) (Abou-Shanab et al., 2017).

Plants were kept in growth chamber under light intensity of 136 μmol m-2 s-1 with a 12 h photoperiod, temperature of 27 ± 2°C, and 45% of relative humidity, and irrigated with sterile distilled water as needed and with 50 mL of the N-free Jensen’s nutrient solution every week (Vincent, 1970). Plants were harvested after 31 days of inoculation, and the plant height (PH), root dry weight (RDW), shoot dry weight (SDW), number of nodules (NN), and nodule dry weight (NDW) were evaluated. The RDW, SDW and NDW were determined after drying plant tissues at 70 ± 2°C for 48 h.

Effects of the bacterial co-inoculation on soybean growth and nodule development

In a second trial, three prominent bacteria, having been selected from the previous experiment, were used for the co-inoculation study. A total of 16 different inoculation treatments were applied to soybean plants var. ‘Huasteca 200’. The inocula were prepared with isolates CPO4.24C (Bradyrhizobium sp.), CPO4.13C (Rhizobium sp.) and CPO4.15C (Agrobacterium tumefaciens CPO4.15C), individually or in binary combinations with two referential strains as follows: Bradyrhizobium japonicum USDA110, and the plant growth promoting rhizobacteria Pseudomonas tolaasii P61 (Angulo-Castro et al., 2018). The referential strains were also applied individually and in combination. An uninoculated control group was also in place. The binary inoculum (1 mL) was prepared using a ratio 1:1 (v/v) of the constituent strains.

Soybean plants were kept in a growth chamber at light intensity of 138 μmol m-2 s-1, temperature: 26 ± 2°C, and 44% relative humidity, and irrigated as previously described. Plants were harvested after 32 days of inoculation for determining the NN, and their total dry biomass. The N fixation was measured using the acetylene reduction assay (ARA) as described by Ferrera-Cerrato et al. (1993). Roots were placed in 1000 mL hermetically sealed flasks, and 100 mL of air (10% volume of the flask) was pulled out with a syringe and replaced with 100 mL of acetylene. After 1 h of incubation, 5 mL of the gaseous mixture of each flask were recovered and placed in Vacutainer® tubes for further analysis in gas chromatograph (Chrompack, model 5890, series II, USA) using an Agilent J&W Capillary Poraplot Q column® (25 m/0.32 mm) (Agilent Technologies, Santa Clara, CA, USA). In addition, PH, NN, RDW, SDW and NDW were also recorded. The relative chlorophyll content was measured by taking SPAD readings (Minolta SPAD-502) from 15 plants randomly chosen. Readings were taken from full developed leaves (3rd and 4th position leaves) of each plant. For this determination an average value from 15 measurements per plant were taken.

Statistical analysis

A completely randomized design was applied in both experimental assays. Collected data were subjected to an analysis of variance (ANOVA), and to a Least Significant Difference test (LSD, α = 0.05) with the assistance of the Statistical Analysis System (SAS) v. 9.

ResultsTop

Morphological and biochemical characterization of endophytic bacterial colonies obtained from soybean nodules

A total of 23 culturable isolates were obtained from field grown soybean nodules collected from the test site at Nuevo Progreso, Hopelchén (7 from the ‘Huasteca 200’ variety, and 5 from the transgenic variety), and from San Antonio Cayal (6 from ‘Huasteca 100’, 4 from ‘Huasteca 400’, and 1 from ‘Huasteca 200’) (Table 1); all bacterial isolates represented bacilli based on microscopic examinations. The colonial characterization showed that 59% of them had circular shape, 33% were pointed, and only 8% had an amoeboid shape.

The isolates labelled as CPO4.24C, CPO4.13C, CPO4.17TA and CPO4.18T showed scarce growth on GPA-BP medium and no growth on litmus milk, with slight acidification on the YMA-BTB medium, indicating that these four strains may correspond to symbiotic bacteria. The isolates CPO4.1T and CPO4.19T produced alkaline reactions, and the remaining bacterial isolates showed strongly acidic reactions in culture medium. Furthermore, the CPO4.10C, CPO4.7CA, CPO4.8CA, CPO4.9CA, CPO4.11C, CPO4.12C, CPO4.14C, CPO­4.15C, CPO4.35C, CPO4.45, CPO4.22S and CPO­4.2TA were able to solubilize tricalcium phosphate (Table 1). Isolates were identified as gram-negative bacteria, and only the isolate CPO4.23C was identified as gram-positive.

Molecular identification of the isolates

All of the nucleotide sequences of the bacterial isolates from this study were deposited in the GeneBank, NCBI (USA) to obtain the corresponding accession numbers (Table 2). Bacterial isolates were identified by using 16S rRNA gene sequence analysis. Using the BLAST analysis, the bacterial sequences showed a maximum identity of 99-100% with genera like Agrobacterium (CPO4.15C), Enterobacter (CPO4.10C, CPO4.7CA, CPO­4.9CA, CPO4.8CA, CPO4.12C, CPO4.14C, CPO­4.35, CPO4.58, CPO4.45, CPO4.65, and CPO4.22S), Bradyrhizobium (CPO4.24C), Ensifer (CPO4.17TA and CPO4.18T), Chryseobacterium (CPO4.20S and CPO­­­4.21S), Massilia (CPO4.1T), Microbacterium (CPO­4.23C), Rhizobium (CPO4.13C), Serratia (CPO­4.11C, CPO4.2TA), and Xanthomonas (CPO4.19T) (Table 2).

The consensus sequences were pooled by phylo­genetic analysis to determine the identity at the species level. The clustering of the 16S rRNA sequences (Fig. 1) showed that Enterobacter was the most predominant clade in which the strains CPO4.7CA, CPO4.8CA, CPO4.12C, CPO4.14C, CPO4.35, and CPO4.22S were placed in the group of E. cloacae, whereas the strains CPO4.10C and CPO4.9CA were in the group of E. ludwigii, and the strain CPO4.45 to E. hormaechei. On the other hand, the strains CPO4.17TA and CPO4.18T corresponded to Ensifer adhaerens, whereas strains CPO4.11C and CPO4.2TA had a maximum identity with Serratia marcescens, and the strain CPO4.15C belonged to Agrobacterium tumefaciens. The remaining strains were identified only at genus level because they belonged to a cohort of undescribed species most likely to the genera Bradyrhizobium, Chryseobacterium, Enterobacter, Massilia, Microbacterium, Rhizobium, and Xanthomonas.

Figure 1. Phylogenetic tree constructed with Bayesian inference using 1,000,000 generations. The sequences correspond to the amplification of the 16S rRNA gene of endophytic and symbiotic bacteria (bold names) associated with root nodules of Glycine max (L.) Merr. The species Thermococcus marinus was used as outgroup. Superscript ‘T’ indicates type strain. The scale bar indicates the number of substitutions per site.

Effects of inoculation produced by the endophy­tic isolates on soybean growth and nodule de­velopment

Out of the 23 strains evaluated, only Bradyrhizobium sp. CPO4.24C produced 36 nodules/plant on average, which was significantly higher than those obtained by inoculating the referential strain B. japonicum USDA110 (21 nodules/plant) (Table 3). Plants inoculated with Rhizobium sp. CPO4.13C or A. tumefaciens CPO4.15C resulted in significantly increased PH (51.2 and 49.5 cm, respectively) and SDW (0.47 and 0.41 g/plant, respectively) when compared to the negative control (uninoculated plants) (35.1 cm and 0.26 g/plant, respectively) (Table 3). In addition, the inoculation of Rhizobium sp. CPO4.13C resulted in significantly higher RDW (0.17 g/plant) when compared to plants of the negative control (0.10 g/plant).

Based on these results, the strains Bradyrhizobium sp. CPO4.24C, Rhizobium sp. CPO4.13C, and A. tumefaciens CPO4.15C were selected for their appli­cation as co-inoculants with the referential strain B. japonicum USDA110 or with the plant growth promoting bacterium P. tolaasii P61.

Effects of bacterial co-inoculation on soybean growth and nodule development

The co-inoculation of B. japonicum USDA110 and the strain Bradyrhizobium sp. CPO4.24C (T2) resulted in significantly greater plant height, NN and NDW than the negative control (T16) (Table 4). Moreover, no significant differences (PH, NN and NDW) were observed between plants individually inoculated with each Bradyrhizobium strain (T11 and T12). However, the single inoculation of Bradyrhizobium sp. CPO4.24C (T11) yielded greater SDW and RDW (0.69 and 0.27 g/plant, respectively) than the negative control (0.30 and 0.15 g/plant), but no significant differences were observed with the remaining treatments (Table 4).

In regards to the relative leaf chlorophyll content, the individual inoculation with the two Bradyrhizobium strains (CPO4.24C or USDA110), and the combination of B. japonicum USDA110+P. tolassi P61 (T1) resulted in the highest chlorophyll content (28.4 SPAD units) which was significantly higher than the negative control (T16) (13.28 SPAD units) (Table 4). In contrast, the co-inoculation of Bradyrhizobium sp. CPO4.24C with either B. japonicum USDA110 (T2) or P. tolassii P61 (T5) resulted in significantly higher activity of acetylene reduction (3.10 and 2.91 μmol C2H4 h-1 plant-1, respectively) than the remaining treatments. Co-inoculation of Bradyrhizobium sp. CPO4.24C with either A tumefaciens CPO4.24C or Rhizobium sp. CPO4.13c also induced significantly higher ARA values than those obtained with the single bacterial inoculation.

DiscussionTop

The legume-rhizobia symbiosis is an important biological process for plant productivity (Bai et al., 2003; Stajkovic et al., 2011). Many studies have shown that the co-inoculation of rhizobia and some endophytic bacteria may contribute on plant growth promotion and yield (Bai et al., 2003; Liu et al., 2010; Deng et al., 2011); thus, such co-inoculation may improve the effectiveness of the symbiotic relationship. In this context, the nodules are colonized by several non-rhizobial endophytes (Bai et al., 2002; Palaniappan et al., 2010; Saïdi et al., 2011; Li et al., 2012; Aserse et al., 2013) which influence the growth and yield of legumes by different mechanisms such as mineral solubilization, or enhanced root nodulation and N fixation activity (Bai et al., 2003; Palaniappan et al., 2010; Deng et al., 2011; Stajkovic et al., 2011; Li et al., 2012). In the present study, 23 strains (predominantly Gram-negative bacteria) were isolated. Of them, 19 were found as non-symbiotic endophytes, and 4 showed symbiotic features. In this respect, Li et al. (2008) isolated a high number of Gram-negative endophytes from soybean nodules. In contrast, other studies reported low number of endophytic bacteria with the predominance of Gram-positive bacteria (Bai et al., 2002; Hung & Annapurna, 2004; Aserse et al., 2013). Therefore, our study besides isolating three genera of potentially nodule-forming rhizobia, also reports the proliferation of 7 genera of non-symbiotic bacteria harbored in root nodules.

The analysis of the 16S rRNA gene sequence of the isolated strains (symbiotic and non-symbiotic) indicated that they belonged to 10 different genera (Table 2). Among the isolated bacteria, Agrobacterium tumefaciens is known as a soil-borne phytopathogen previously reported from soybean nodules (Li et al., 2008). Same authors also reported the genus Serratia in the soybean nodules, and this bacterium has been shown to stimulate the growth and development of soybean (Zhang et al., 1996); however, the two Serratia strains isolated in this work did not show significant effects on growth nor on soybean nodulation. Our report revealed that bacterial genera contrast with other findings in which other genera of endophytic bacteria, such as Acinetobacter, Bacillus, Burkholderia, Deinococcus, Rhodococcus, Pantoea, Staphylococcus and Tsukamurella, were found in soybean nodules (Bai et al., 2002, 2003; Hung & Annapurna, 2004; Li et al., 2008; Aserse et al., 2013).

Our study also noted the isolation of other endophytic bacteria not previously described as inhabitants of soybean nodules, such as Enterobacter, Chryseobacterium, Massilia, Microbacterium, and Xanthomonas. However, these bacteria were reported as nodule inhabitants of other legume species; for example, Enterobacter was identified from nodules of Abrus precatorius and Vigna unquiculata (Ghosh et al., 2015; Leite et al., 2016), Chryseobacterium from V. unquiculata (Leite et al., 2016), Massilia from nodules of Hedysarum flexuosum (Ezzakkioui et al., 2015), Microbacterium from Medicago sativa and Sphaerophysa salsula (Stajkovic et al., 2009; Deng et al., 2011), Serratia from nodules of Sphaerophysa salsula and Hedysarum flexuosum (Deng et al., 2011; Ezzakkioui et al., 2015), and Xanthomonas was iden­tified in nodules of Medicago hispida (Arone et al., 2014). More importantly, the influence of these bacteria on the symbiosis between rhizobia and soybean has been rarely described.

In the present study, the co-inoculation of non-symbiotic endophytic bacteria did not produce significant effects on the growth of soybean plants. In contrast, both dry weight and root nodulation were increased due to the co-inoculation of B. japonicum 532C with the endophytic Bacillus subtilis and B. thuringien­sis (Bai et al., 2002). Similarly, the coinoculation of S. meliloti with endophytic bacteria like A. tumefaciens (Wang et al., 2006) or Rhizobium sp. II CCBAU21244 (Liu et al., 2010) resulted in increased nodulation of Melilotus dentatus and Wisteria sinensis. Nevertheless, in our work, the co-inoculation of A. tumefaciens CPO4.15C did not affect root nodulation which is opposite to results obtained by Camacho et al. (2001). Overall, the endophytic bacteria isolated in the present study did not influence plant growth, however, they may be involved in creating an ecological micro-niche suitable for both survival and proliferation of symbiotic bacteria, as discussed by Deng et al. (2011).

On the other hand, the symbiotic strain Brady­rhizobium sp. CPO4.24C was able to form nodules in the soybean plants. On the contrary, the absence of nodules in the ‘Huasteca 200’ variety inoculated with Rhizobium sp. CPO4.13C and Ensifer adherensis CPO4.2TA or CPO4.18T can be explained due to the specificity between legumes and rhizobia; in this regard, these two bacterial genera are not soybean symbionts (Wu et al., 2011; Zhang et al., 2011; Yan et al., 2014). Similarly, some non-nodulating Rhizobium and Bradyrhizobium bacteria were reported in the rhizosphere of legumes (Segovia et al., 1991; Pongsilp et al., 2002; Aserse et al., 2013).

As mentioned, the inoculation of Bradyrhizobium sp. CPO4.24C resulted in greater nodulation in comparison to the inoculation of the reference strain B. japonicum USDA110 (Table 3). This reference strain has induced abundant nodulation at low temperatures ranging between 17 to 23ºC (Ando & Yokoyama, 1999; Suzuki et al., 2014). Nevertheless, the average temperature recorded in the present study was 27 ± 2ºC by which the growth and infectivity of the reference bacterial strain might have been affected. However, the co-inoculation of both strains resulted in high PH and nitrogenase activity (consequently greater N fixation can be expected) when compared to the single inoculation of each strain. This demonstrates a synergistic effect produced by both of the aforementioned bacteria. In this respect, Htwe & Yamakawa (2016) reported lower ARA (1.15 C2H4 h-1 plant-1) due to the co-inoculation of soybean plants with B. japonicum SAY3-7 and Streptomyces griseoflavus P4. In addition, these authors reported that SDW and RDW were significantly higher in co-inoculated plants (0.42 and 0.25 g/plant) than uninoculated controls (0.39 and 0.25 g/plant). Similarly, in this study the shoot and root biomass were enhanced with the co-inoculation of Bradyrhizobium sp. CPO4.24C and B. japonicum USDA110 in comparison to uninoculated control (Table 4).

Our results show that the native strain Brady­rhizobium sp. CPO4.24C has good potential for being introduced as biofertilizer for soybean cultivation in the tropical regions of Mexico. Nevertheless, further research should be conducted for evaluating the effects of this bacterium on the growth and yields of soybean plants under appropriate field conditions.

Overall, this study isolated twenty-three endophytic bacterial strains belonging to ten different genera from nodules of four varieties of Glycine max grown at field conditions. Furthermore, co-inoculation of the three prominent bacterial endophytes with either native or referential Bradyrhizobium strains did not enhance plant growth nor root nodulation. The native Bradyrhizobium sp. CPO4.24C showed high potential for being inoculated alone or combined with the referential strain B. japonicum USDA110, since PH, nodulation, dry weight, relative chlorophyll content, and nitrogenase activity were significantly enhanced.