Introduction

Sugarcane (Saccharum officinarum L.) is a tropical plant belonging to the grass family, Gramineae, as well as maize, rice, sorghum, and wheat. Sugarcane is the world’s largest crop and is economically important in Thailand, which ranks fourth among sugarcane producers on the world market, after Brazil, India, and China (Food and Agriculture Organization of the United Nations 2015). Although sugar is the primary product from sugarcane, ethanol has become a new high-value product because of the worldwide interest in replacing gasoline with biofuel.

For several decades, boosting crop yields by chemical fertilizers and various nutrients and pesticides has dominated agronomy (Tilman 1998). Now, the use of environmentally friendly strategies for reducing the use of chemicals to enhance growth of sugarcane plants is an ultimate goal of sustainable agriculture. One of the alternative ways to promote plant growth and protect plants from diseases is the application of bacterial endophytes. These are free-living bacteria that colonize the inside of the plants and can affect plant growth by direct and indirect mechanisms (Gupta and others 2000; Glick 2012). They are termed plant growth-promoting endophytes (PGPE) (Taulé and others 2011) with various PGP traits including solubilization of rock phosphates, production of siderophores to scavenge iron, production of phytohormones, fixation of nitrogen, secretion of 1-aminocyclopropane-1-carboxylate (ACC) deaminase to reduce ethylene levels in plants, and production of specialized metabolites to protect plants from phytopathogens (Gupta and others 2000; Mendes and others 2007; Glick 2012). A range of PGPE including endophytic actinomycetes significantly increase vegetative growth and grain yield of several crops, including jatropha, maize, rice, sorghum, sugarcane, and wheat (Olivares and others 1997; Yanni and others 2001; Govindarajan and others 2007; Luna and others 2010; Rungin and others 2012; de Jesus Santos and others 2014; Alves and others 2015; Qin and others 2015) by increasing shoot and root lengths as well as biomass through the secretion of plant growth-promoting substances, increasing the availability of nutrients, or promoting growth by combined modes of action (Vessey 2003). Nevertheless, the application of single bacterial strains resulted in inconsistencies in the field (Bashan and Holguin 1997). Therefore, co-inoculation of compatible bacteria has been found to more significantly enhance the growth of plants than individual inoculation. For example, co-inoculation of mixtures of diazotrophic bacteria gave better growth of rice and sugarcane (Govindarajan and others 2007); and co-culture of diazotrophs and fungal endophytes revealed better growth and yield of chickpea (Verma and others 2014).

This study aimed to select powerful endophytic PGP diazotrophs and actinomycetes isolated from sugarcane plants in Thailand. The diversity of culturable root-associated endophytic bacteria and actinomycetes was systematically analyzed and their PGP traits and plant growth-promoting effects on sugarcane plants were thoroughly evaluated after individual and co-culture inoculation.

Materials and Methods

Isolation of Culturable Endophytic Bacteria and Actinomycetes from Sugarcane

Five-month-old healthy sugarcane plant cultivars LK92-11 and KK3, commercial canes mainly planted in Thailand for sugar industry, were obtained from Kud Chock and Nongkonthai sugarcane plantations, Phu Khiao, Chaiyaphum, Thailand through Mitr Phol Sugarcane Research Center. The plant samples were vigorously shaken to loosen attached soil from the roots and air-dried at room temperature (RT) overnight, then kept at 4 °C till use.

To isolate root-associated endophytic bacteria, sugarcane roots were surface sterilized using the method of Rachniyom and others (2015). The roots were first rinsed under running tap water for 2–3 min to remove soil particles, and then immersed with shaking in sterilized 0.1 % (v/v) Tween 20 for 5 min, followed by 70 % ethanol for 5 min, 1 % sodium hypochlorite solution for 10 min, 10 % (w/v) sodium hydrogen carbonate solution for 10 min, and finally washed three times with sterile distilled water.

Endophytic bacteria were isolated according to the method for isolation of nitrogen-fixing bacteria described by Cavalcante and Dobereiner (1988). The roots were crushed in 5 % sucrose solution and ten-fold serial dilutions were made. A 100 µl sample of each dilution was dropped into semisolid LGI medium (Baldani and others 2014) and incubated at 28 °C for 7–10 days. A veil-like pellicle near the surface of the medium appeared and was streaked onto LGI agar and nutrient agar (NA) supplemented with 50 µg ml−1 nystatin and 50 µg ml−1 cycloheximide and incubated at 28 °C for 3–7 days. The final rinsing water was also spread on both media to confirm surface sterilization. Colonies of bacteria were picked and purified on LGI agar and NA. The pure cultures were classified by Gram staining and morphological observation under the light microscope. Bacteria were stored in 40 % glycerol solution at −80 °C.

For isolation of endophytic actinomycetes, the surface-sterilized roots were macerated in ¼ Ringer’s solution (0.9 % NaCl, 0.042 % KCl, 0.048 % CaCl2, and 0.02 % NaHCO3). The root materials and the liquid solution were, respectively, placed and spread onto starch casein agar (SCA) (Küster and Williams 1964) and water agar (WA) supplemented with 50 µg ml−1 nalidixic acid, 100 µg ml−1 ampicillin, 50 µg ml−1 nystatin, and 50 µg ml−1 cycloheximide and incubated at 28 °C for 2–4 weeks. The final rinsing water was also spread on both media to confirm surface sterilization. Colonies of endophytic actinomycetes were picked and purified on mannitol soya (MS) agar (Hobbs and others 1989). The pure cultures were stored in 20 % glycerol solution at −80 °C.

Identification of Endophytic Bacteria

All the bacterial endophytes were grouped based on their colony morphology on agar plates. Endophytic bacteria were grown on NA and Tryptic Soy Agar (TSA) media at 28 °C for 2 days. Endophytic actinomycetes were grown on International Streptomyces Project (Shirling and Gottlieb 1966) Media No. 2, 3, and 4 at 28 °C for 14 days. The representatives of endophytic bacteria and actinomycetes were selected based on morphological characteristics for 16S rRNA gene sequencing and analysis. Genomic DNA of bacteria was prepared by a standard protocol (Green and Sambrook 2012). The 16S rRNA gene was amplified using universal primers and the PCR protocol of Lane (1991). For actinomycetes, genomic DNA was prepared according to Kieser and others (2000). 16S rRNA genes were amplified using primers and the PCR protocol of Rachniyom and others (2015). PCR products were purified using Gel/PCR DNA fragment extraction kit (Geneaid, Taiwan) according to the manufacturer’s protocol and subjected to DNA sequencing at Macrogen Inc. (Korea). All sequences obtained from the 16S rRNA genes were edited manually using BioEdit (Hall 1999), then compared with type strains using the EzTaxon–e server (Kim and others 2012a). The sequences of the isolates and the most similar species were aligned using CLUSTAL X program version 2.0 (Larkin and others 2007). Phylogenetic tree was constructed using neighbor-joining algorithm (Saitou and Nei 1987) in MEGA software version 6.0 (Tamura and others 2013). The topology of tree was evaluated with bootstrap analysis based on 1000 resamplings (Felsenstein 1985). A distance matrix was generated using Kimura’s two-parameter model (Kimura 1980).

Characterization of Plant Growth-Promoting Traits

Indole-3-Acetic Acid (IAA) Production

IAA production was determined by a colorimetric method (Pilet and Chollet 1970). Endophytes were inoculated into glucose-beef extract broth supplemented with 10 mM l-tryptophan and incubated at 28 °C for 7 days in the dark. The culture was then centrifuged and 1 ml of the supernatant was transferred into a vial containing 2 ml of Salkowski’s reagent (Pilet and Chollet 1970). The mixture was left at RT for 30 min in the dark. Development of a pink-red color indicated IAA production.

Siderophore Production

An overnight-NB culture of bacteria (107 CFU ml−1) or an agar plug of the 14-day growth of actinomycetes on MS agar was, respectively, dropped or placed on chrome azurol S medium (Schwyn and Neilands 1987) and incubated at 28 °C for 7 days for non-actinomycete bacteria and 14 days for actinomycetes. An orange halo indicated siderophore production.

Phosphate Solubilization

Phosphate solubilization was detected by the modified method of Rodríguez and Fraga (1999). An overnight-NB culture of bacteria (107 CFU ml−1) or an agar plug of 14-day growth of actinomycetes on MS agar was, respectively, dropped or placed on Pikovskaya’s agar (Pikovskaya 1948) containing 0.5 % (w/v) Ca3(PO4)2 and incubated at 28 °C for 7 days for non-actinomycete bacteria and 14 days for actinomycetes. Formation of a clear zone indicated solubilizing ability.

1-Aminocyclopropane-1-Carboxylate (ACC) Deaminase Activity

ACC deaminase activity was determined by the method of Glick and others (1995) using N-free medium (Nfb) (Dobereiner and Day 1976) for bacteria and minimal medium (MM) (Kieser and others 2000) for actinomycetes containing 0.3 m mol l−1 ACC (Sigma, USA) as a sole nitrogen source. MM with 0.1 % (w/v) NH4(SO4)2 was used as a positive control and cultivation without ACC was used as a negative control. After incubation at 28 °C for 7 days for non-actinomycete bacteria and 14 days for actinomycetes, colony growth on Nfb or MM with addition of ACC indicated ACC deaminase activity.

Nitrogen Fixation and nifH Gene Amplification

Nitrogenase activity was determined by the acetylene reduction assay (ARA) according to Hardy and others (1973). An overnight-NB culture of bacteria (107 CFU ml−1) was washed three times with distilled water to remove NB before transferring to 5 ml of semisolid Nfb medium and incubated at 28 °C for 2 days. Next, 10 % (v/v) of gas in the headspace of the test tube was replaced by acetylene and left at RT for 1 h. Gas was collected from the headspace and analyzed for ethylene by gas chromatography (Hewlett Packard HP Series II, Japan). The nitrogen-fixing Azospirillum brasilense TS13 (Meunchang and others 2004) was used as a positive control and an un-inoculated tube served as a negative control. All of the bacterial isolates and Azospirillum brasilense TS13 as positive control were subjected to nifH gene amplification using primers nifH-F (5′-AAAGGNGGNATCGGNAANTCCACCAC-3′) and nifH-R (5′-TTGTTNGCNGCNTACATNGCCATCAT-3′) and the PCR protocol described by Rösch and others (2002).

Evaluation for Antimicrobial Activity

To evaluate antimicrobial activity, seven test microbes, Bacillus cereus ATCC 11778, Escherichia coli ATCC 8739, Pseudomonas aeruginosa ATCC 15442, Staphylococcus aureus ATCC 259233, Aspergillus niger ATCC 6275, Colletotrichum falcatum DOAC 1655, and Fusarium moniliforme DOAC 1224, were used. Antibacterial activity was evaluated by an overlay method (Anand and others 2006). An overnight-NB culture of bacteria (107 CFU ml−1) or a 14-day-old MS agar plug of actinomycetes was, respectively, dropped or placed on NA plates and incubated for 24 h. The assay plates were overlaid with test bacteria (107 CFU ml−1). Appearance of an inhibition zone indicated positive activity.

Antifungal activity was evaluated by dual-culture assay (Fokkema 1976). An overnight-NB culture of bacteria (107 CFU ml−1) was spotted on potato dextrose agar and incubated for 24 h. Actinomycetes were streaked on one-third of the area of PDA and allowed to grow for 7 days. An agar plug of the test fungus was prepared using a sterile cock borer (diameter 8 mm), placed opposite the bacterial colony and incubated at RT for 7 days. Inhibition of fungal growth was compared to the growth of the fungal control.

Evaluation of Plant Growth Promotion by Individual and Co-Inoculation

Endophytic bacteria and actinomycetes were selected based upon their PGP traits to evaluate sugarcane growth enhancement by individual and co-culture inoculation. To abnegate the effect of antagonistic activity among them, the selected isolates were tested for antibacterial activity against each other, and only the strains showing no activity were used to inoculate sugarcane. Tissue cultured sugarcane plants of cultivar LK92-11 were obtained from the Plant Tissue Culture Unit at Central Laboratory and Greenhouse Complex, Kasetsart University (Kamphaengsaen Campus). Sugarcane plantlets were rinsed with sterile distilled water to remove remaining agar from the roots. The height of the plants was normalized to 6.5 cm by truncating shoots and roots. Plants were then inoculated with bacteria either individual or co-culture (107 CFU ml−1 of each strain) by the root dip method (Musson and others 1995) in a tissue culture bottle containing Murashige and Skoog (MS0) medium (Murashige and Skoog 1962) and immersed at RT for 24 h. Inoculated sugarcane plants were transferred to cleaned mini-pot (6 × 5.5 cm) containing sterilized hydroponic materials (perlite:vermiculite, 3:1). The experiment was set up in completely randomized design (CRD) using five replicate mini-pots per treatment and each treatment was replicated twice. Plants were maintained in the greenhouse and supplied with sterilized sucrose-free MS0 broth twice a day for 60 days. The plants were then harvested and examined for root and shoot lengths, and root and shoot fresh/dry weights. Plants were surface sterilized and the inoculated endophytes were re-isolated according to the procedure described above. Plant growth parameters were statistically analyzed by one-way ANOVA and Tukey’s multiple range tests (TMRT) using SPSS (version 16.0) at p = 0.05 to determine the efficacy of un-inoculated, individual, and co-culture inoculation to promote growth of sugarcane plants.

Results

Distribution of Sugarcane Root-Associated Bacterial Endophytes

One hundred and thirty-five culturable root-associated bacterial isolates were obtained from sugarcane. They comprised 52 isolates of endophytic bacteria and 83 isolates of actinomycetes. According to the preliminary dereplication by morphological criteria and cultural characteristics, 19 and 11 isolates of endophytic bacteria and actinomycetes were assigned to the genus level using 16S rRNA sequence analysis, respectively. Sequence similarity of 16S rRNA genes of these endophytes (ranging from 1282 to 1513 bp) to the closest type strains from the EzTaxon–e server was 98.4–100 % (Table 1). All of the representative strains displayed considerable diversity and were distributed under 11 families of Bacillaceae, Enterobacteriaceae, Micrococcaceae, Micromonosporaceae, Moraxellaceae, Paenibacillaceae, Pseudomonadaceae, Staphylococcaceae, Streptomycetaceae, Streptosporangiaceae, and Thermomonosporaceae comprised 10 genera of bacteria and 4 genera of actinomycetes (Fig. 1). Streptomyces was the most frequently isolated genus (n = 47, 34.81 %) followed by Microbispora (n = 29, 21.48 %), Enterobacter (n = 23, 17.04 %), Bacillus (n = 14, 10.37 %), Actinomadura (n = 5, 3.70 %), Pantoea (n = 4, 2.96 %), Acinetobacter (n = 2, 1·48 %), Kosakonia (n = 2, 1.48 %), Micromonospora (n = 2, 1.48 %), Paenibacillus (n = 2, 1.48 %), Staphylococcus (n = 2, 1.48 %), Lysinibacillus (n = 1, 0.74 %), Micrococcus (n = 1, 0.74 %), and Pseudomonas (n = 1, 0.74 %). 16S rRNA gene sequence similarity values of four strains, Actinomadura sp. GKU 822, Microbispora sp. GKU 823, GKU 898, and Pantoea sp. EN-29 with closest type strains were lower than 99 % (Table 1) and therefore probably represent new species.

Table 1 Colony morphology and 16S rRNA sequence similarity of plant growth-promoting bacterial endophytes from sugarcane to the closest type strains of validly described species
Full size table
Fig. 1
figure 1

Neighbor-joining tree based on 16S rRNA gene sequences showing relationships between the representative endophytic bacteria and actinomycetes isolated from sugarcane roots and the nearest type strains. Number of branch points indicates bootstrap value based on 1000 replications; only values above 50 % are shown. Bar, 0.01 substitutions per nucleotide position

Full size image

Plant Growth-Promoting Traits of Sugarcane Endophytes

The plant growth-promoting (PGP) traits, including antagonistic activity, were examined from 135 strains of sugarcane endophytes (Table 2). Among them, three strains of actinomycetes, Streptomyces sp. GKU 833, GKU 879, and GKU 895, and two strains of bacteria, Bacillus sp. EN-27 and Enterobacter sp. EN-21, showed the highest activity of PGP traits. The highest frequency of PGP trait possessed by both endophytic bacteria and actinomycetes in this study was siderophore production, followed by IAA production and phosphate solubilization (Table 2). Most of the siderophore-producing species belong to the genera Streptomyces, Microbispora, and Bacillus; while IAA producing dominant species were Streptomyces, Microbispora, and Enterobacter. Members of the genera Streptomyces and Microbispora were also the majority species capable of phosphate solubilization. However, the highest activity of phosphate solubilization was found in Pantoea sp. EN-29 and EN-39 (data not shown). Besides, ACC deaminase-production was mainly detected in species belonging to Streptomyces and Bacillus (Table 2).

Table 2 Number of positive isolates of sugarcane endophytic bacteria and actinomycetes with plant growth-promoting traits
Full size table

To evaluate nitrogenase activity, endophytic bacteria were screened using the ARA method. Only six out of the 52 strains belonging to Bacillus, Enterobacter, Lysinibacillus, and Pantoea could reduce acetylene gas, ranging between 14.59 and 25.39 nmol C2H4 h−1 (107 cells)−1. The highest level of nitrogenase activity was found in Enterobacter sp. EN-21 (25.39) and Bacillus sp. EN-24 (20.11). To evaluate nitrogen fixation ability, nifH gene amplification was performed. Ten nifH-positive strains (19.23 % of the total) were obtained. There were five strains, namely Bacillus sp. EN-27, Enterobacter sp. EN-21, Enterobacter sp. EN-30, Lysinibacillus sp. EN-9, and Pantoea sp. EN-39 that exhibited both nitrogenase activity and a nifH gene (Fig. S1). However, the nifH gene was absent in an ARA-positive endophyte, Bacillus sp. EN-24.

Antagonistic potential tests conducted against seven test strains of bacteria and fungi revealed that the most susceptible fungi that interacted with the endophytes were Fusarium moniliforme, followed by Aspergillus niger and Colletotrichum falcatum, while the most susceptible bacterium was Bacillus cereus (Table 2). Surprisingly, none of the endophytes inhibited Pseudomonas aeruginosa. Six strains of endophytes, Bacillus sp. EN-8, Bacillus sp. EN-15, Bacillus sp. EN-24, Actinomadura sp. GKU 870, Streptomyces sp. GKU 833, and Streptomyces sp. GKU 878, effectively inhibited growth of A. niger and the pathogenic fungi, F. moniliforme and C. falcatum, causing red rot disease in sugarcane (Alexander and Viswanathan 2002). B. cereus and Escherichia coli were the test bacteria most frequently inhibited by the endophytic bacteria as well as actinomycetes (Table 2).

Sugarcane Plant Growth Enhancement

Based on PGP traits, two strains of endophytic diazotrophs, Bacillus sp. EN-24 and Enterobacter sp. EN-21, and two strains of actinomycetes, Microbispora sp. GKU 823 and Streptomyces sp. GKU 895, were selected to evaluate growth enhancement of sugarcane in pot experiments (Table 3). Antagonistic test between each strain was performed and reviewed no effect to each other; therefore, they were suitable for co-culture inoculation. Nine treatments included (T1) Enterobacter sp. EN-21, (T2) Bacillus sp. EN-24, (T3) Microbispora sp. GKU 823, (T4) Streptomyces sp. GKU 895, (T5) EN-21 + EN-24, (T6) GKU 823 + GKU 895, (T7) GKU 823 + EN-21 + EN-24, (T8) GKU 895 + EN-21 + EN-24, (T9) GKU 823 + GKU 895 + EN-21 + EN-24 and un-inoculated control were conducted to investigate individual and co-culture inoculation. Two-month-old sugarcane plants treated by either individual or co-inoculation gave statistically significantly (p ≤ 0.05) greater root and shoot lengths, and root and shoot biomass than un-inoculated plants (Table 4; Fig. 2). Either individual endophytic diazotroph or actinomycete inoculated plants (T1, T2, T3, and T4) gave measurable increments in root length (51–89 %), shoot fresh weight (67–106 %), and shoot dry weight (76–146 %) but remarkably increased in root fresh weight (147–278 %; Table 4). The increase of root fresh weight was due to increased lateral root production (data not shown). These parameters were not significantly different when two strains of diazotrophs (T5) and actinomycetes (T6) were inoculated except shoot fresh (126 %) and dry weight (180 %) that T6 gave notably increased. The percentage increase in shoot length of individual and co-inoculated sugarcane were higher than that of the un-inoculated controls, especially T4, T5, T7, and T9, which were significantly (44–59 %) greater; while significant increase in root length was noticed in T7, T8, and T9 (125–128 %; Table 4). The most significant enhancement of growth appeared in T7 and T8 in which each endophytic actinomycete was co-inoculated with both Enterobacter sp. EN-21 and Bacillus sp. EN-24. T7 and T8 caused drastic increases of 128–129 % in root length, 316–352 % in root fresh weight, 177–208 % in root dry weight, 37–58 % in shoot length, 116–126 % in shoot fresh weight, and 146–160 % in shoot dry weight over those of un-inoculated treatments. The results suggested that the enhancement of growth in sugarcane was due to multiple growth-promoting traits from the mixture of the bacterial endophytes (Table 3). Our results indicated that co-inoculation of either endophytic Microbispora or Streptomyces combined with the mixture of Enterobacter and Bacillus gave higher growth of sugarcane than individual inoculation.

Table 3 Identification and PGP traits of selected sugarcane diazotrophic bacteria and endophytic actinomycetes
Full size table
Table 4 Effect of individual and co-inoculation of endophytic diazotrophs and actinomycetes on growth of sugarcane plants at 60 days after inoculation
Full size table
Fig. 2
figure 2

Effects of individual and co-inoculation of endophytic bacteria and actinomycetes on growth promotion of sugarcane plants at 60 days after inoculation. T1, Enterobacter sp. EN-21; T2, Bacillus sp. EN-24; T3, Microbispora sp. GKU 823; T4, Streptomyces sp. GKU 895; T5, EN-21 + EN-24; T6, GKU 823 + GKU 895; T7, GKU 823 + EN-21 + EN-24; T8, GKU 895 + EN-21 + EN-24; T9, GKU 823 + GKU 895 + EN-21 + EN-24 and C, un-inoculated control

Full size image

Remarkably, when four strains of endophytes were co-inoculated onto sugarcane plants (T9), some growth parameters such as root and shoot fresh/dried weights were significantly lower than those of T7 and T8 and some (root fresh/dried weight) were even lower than single-strain inoculation (Table 4). Re-isolation experiments of the endophytes from two-month-old sugarcane plants were then performed and the identity of isolates was proved by examination of colony morphology and 16S rRNA sequencing. Inoculated endophytic diazotrophs and actinomycetes abundantly appeared in roots but less in stems in every treatment (data not shown). About 104 CFU g fresh weight−1 were found in sugarcane roots inoculated with individual (T1, T2, T3, T4) and combinations of two bacterial strains (T5 and T6; Table 4). However, the viable cell numbers of the endophytes decreased 10 times when more than two strains of bacteria were co-inoculated (T7, T8, and T9); particularly T9 gave the lowest. This might be explained by a reduction of plant biomass in T9 (Table 4) due to the low population of the endophytes. In addition, this re-isolation experiment confirmed that Enterobacter sp. EN-21, Bacillus sp. EN-24, Microbispora sp. GKU 823, and Streptomyces sp. GKU 895 are true endophytes of sugarcane.

Discussion

Plant growth-promoting endophytes (PGPE) are heterogeneous groups of bacteria that beneficial to host plants (Gaiero and others 2013). Several PGPE can encourage plant growth through the release of phytohormones, increment of nutrients, and protection from phytopathogens. In this work, culture-dependent approach was used to isolate culturable root-associated bacteria including actinomycetes from sugarcane cultivated in Thailand. The sugarcane roots used for isolation in this study were successfully surface disinfected in which no colonies were detected from the final rinses on any medium agar used; thus, the isolated bacteria were true endophytes. From the isolation procedures used in this work, 38.52 % of endophytic bacteria and 61.58 % of actinomycetes were obtained. Fourteen genera were systematically identified and the most dominant species were Bacillus, Enterobacter, Microbispora, and Streptomyces.

Diazotrophs appeared to be the most numerous bacteria so far isolated from sugarcane including Azospirillum spp., Burkholderia spp., Enterobacter spp., Gluconacetobacter diazotrophicus, Herbaspirillum spp., Klebsiella spp., Pantoea spp., and Pseudomonas spp. (Asis and others 2000; Suman and others 2001; Perin and others 2006; Govindarajan and others 2007; Mendes and others 2007; Castro-González and others 2011; Taulé and others 2011; Lin and others 2012). Surprisingly, the well-known sugarcane diazotrophic species, including Azospirillum spp., Burkholderia spp., G. diazotrophicus, and Herbaspirillum spp. (Govindarajan and others 2007; Taulé and others 2011), were not found in this work. A plausible explanation for their notable absence could be that the composition of the endophytic bacterial community depends to some degree on a specificity of particular bacterial genotypes for particular sugarcane cultivars, different geographic origins, and level of fertilizers used (Reis-Junior and others 2000; Taulé and others 2011). Moreover, the sugarcane root-associated nitrogen-fixing bacteria in this work were isolated without using the enrichment method of reseeding in the N-free medium (Castro-González and others 2011), the amount of diazotrophic bacteria could, therefore, be affected and resulted in less population. The majority of endophytic bacteria isolated in this study were mainly found as members of the genera Bacillus and Enterobacter. The results were in agreement with several reports of which Bacillus and Enterobacter were dominant sugarcane endophytes (Velazquez and others 2008; Magnani and others 2010; Taulé and others 2011).

Although endophytic actinomycetes have been isolated from several different kinds of plants including a variety of medicinal and crop plants (Coombs and Franco 2003; Tian and others 2007; Zhao and others 2011), this is the first report to describe the diversity of endophytic actinomycetes from sugarcane. The dominance of Streptomyces spp. in the culturable diversity of root-associated endophytes in this report has been noticed in several previous studies from various plant species, that is, wheat, rice, and herbaceous and medicinal plants (Coombs and Franco 2003; Tian and others 2007; Zhao and others 2011; Kim and others 2012; Li and others 2012). However, we found different genera of non-streptomycetes (Actinomadura, Microbispora, and Micromonospora) associated with roots of sugarcane (Fig. 1). The presence of Actinomadura, Microbispora, and Micromonospora was in agreement with previous reports that they are common genera found in root tissues of plants (Qin and others 2011; Mingma and others 2014). Although Streptomyces spp. was predominantly associated with sugarcane roots, we observed a large number of colonies of Microbispora spp. (21.48 %). The results suggested that Microbispora spp. also largely occupied in root tissues of sugarcanes. This work indicated that sugarcanes are rich bio-resources for diversity of endophytic bacteria including diazotrophs and actinomycetes.

PGP traits were examined in sugarcane endophytes and revealed that strains belong to genera Streptomyces, Bacillus, and Enterobacter showed the highest activity of PGP traits. Bacillus and Enterobacter have been previously reported to promote the growth of maize, sugarcane, and tobacco (Kloepper and others 2004; Lin and others 2012; Naveed and others 2014). The highest frequency of PGP trait possessed by sugarcane endophytes in this study was siderophore production, followed by IAA production and phosphate solubilization. It has been suggested that siderophores are involved in both plant growth and health protection by chelating iron (Jaber and others 2002; Rungin and others 2012; Radzki and others 2013). Recent work of Rungin and others (2012) also confirmed that a siderophore-producing endophytic Streptomyces increased root/shoot lengths and biomass of rice and mung bean compared to the deficient mutant. It was demonstrated that bacterial endophytes carrying IAA trait could increase root elongation and could affect the development of lateral roots, which improves the plant’s nutrient uptake from the rhizosphere (Idris and others 2007; Goudjal and others 2013). A recent report showed that Microbispora spp. and Streptomyces spp. exhibited a high variability in IAA production and resulted in significant promotion of shoot length of mandarin seedlings (Shutsrirung and others 2013). Members of PGPE in the genera Bacillus, Enterobacter, Pseudomonas, Micromonospora, and Streptomyces are known as phosphate solubilizers (Hamdali and others 2008; Bashan and others 2013) with the ability to convert insoluble compounds of phosphorus into available phosphates that enhance nutrient availability to plants (Son and others 2006). However, the highest activity of phosphate solubilzation in this work was found in Pantoea spp. Recently, phosphate-solubilizing Pantoea has been reported as a plant growth promoter which increased height and dry weight of Lotus tenuis cv. Pampa INTA (Castagno and others 2011). PGPE harboring ACC deaminase in sugarcane could facilitate plant growth by conversion of ACC to ammonia and α-ketobutyrate, which bacteria can consume and consequently lower the ethylene level in plants (Glick 2005).

Most of the endophytic bacteria isolated from sugarcane have been reported as diazotrophs, with effects on plant growth promotion (Govindarajan and others 2007; Lin and others 2012). In this work, only 11.53 % out of sugarcane endophytic bacteria showed nitrogenase activity belonging to Bacillus, Enterobacter, Lysinibacillus, and Pantoea (Table 2). Although the standard protocol for isolation of diazotrophic bacteria was employed (Cavalcante and Dobereiner 1988), fewer ARA positives were often obtained from plants such as rye grass (Habibi and others 2014) and rice (Rangjaroen and others 2015). Our ranges of ARA activity were in accordance with previous reports of Enterobacter spp. and Bacillus spp. (Habibi and others 2014). When nifH gene amplification was investigated to evaluate nitrogen fixation ability of ARA-positive endophytes, 19.23 % of the total was detected (Table 2). There were only five strains which exhibited both nitrogenase activity and a nifH gene (Fig. S1). Our results were in agreement with Yim and others (2009) in which the nifH gene could not be amplified from all ARA positives. Although, ARA is a common method to detect nitrogenase activity of microbial cultures and has been widely used to consequently identify nitrogen-fixing bacteria (Hardy and others 1973), it was suggested that such activity varies with growth stage, culture condition, and media composition. Hence, amplification of the nifH gene, encoding dinitrogenase reductase, was additionally used to perform the possible nitrogen-fixing ability. Although the nifH gene is the most widely used marker gene for identification of nitrogen-fixing bacteria, nucleotide sequences of nifH genes were diverse in many microorganisms (Waugh and others 1995; Zehr and others 2003). In this experiment, degenerate primers were used for amplification of nifH gene (Rösch and others 2002), nevertheless, it was yet absent in an ARA-positive Bacillus sp. EN-24 (Fig. S1). The absence of the nifH gene could therefore be explained by the variability of this gene (Waugh and others 1995; Zehr and others 2003). It is suggested that the other nitrogenase genes, nifD and nifK, could be used to confirm the nitrogen-fixing ability since nifHDK genes are known to encode the components of the nitrogenase enzyme complex (Howard and Rees 1996).

Some recent reports have indicated that the root nodule-associated Micromonospora has the possible ability to fix atmospheric nitrogen (Trujillo and others 2010; Lorena and others 2012), but endophytic Micromonospora from medicago plants apparently had no functional nitrogenase and consequently no nitrogen fixation occurred (Martunez and others 2014). Because nitrogen fixation in endophytic actinomycetes has not yet been definitively explained and there is very little information about it, ARA assays were not performed on actinomycetes in this work.

Antagonistic tests revealed susceptible bacteria and fungi including F. moniliforme and C. falcatum causing red rot disease in sugarcane (Alexander and Viswanathan 2002) that interacted with the endophytes. B. cereus and E. coli were most frequently inhibited (Table 2), which was consistent with a previous report (Hassan and others 2010). The results of this study indicated that endophytic bacteria and actinomycetes from sugarcane carried multiple traits of PGP and are therefore considered to be safe agents to apply for plant growth enhancement and to control phytopathogens in sugarcane plantations.

Based on PGP traits, two endophytic diazotrophs, Bacillus sp. EN-24 and Enterobacter sp. EN-21, and two strains of actinomycetes, Microbispora sp. GKU 823 and Streptomyces sp. GKU 895, were selected to evaluate growth enhancement of sugarcane plants by individual and co-inoculation. Two-month-old sugarcane plants treated by either individual or co-inoculation gave significantly greater biomass and lengths of root and shoot than un-inoculated control (Table 4; Fig. 2). The increase of root fresh weight was due to increased lateral root production, which might be the effect of plant hormone production by the endophytes (Idris and others 2007; Goudjal and others 2013). The most significant enhancement of growth appeared in the treatment that each endophytic actinomycete was co-inoculated with both Enterobacter sp. EN-21 and Bacillus sp. EN-24. Because some Enterobacter species are known to be human pathogens, Enterobacter sp. EN-21 should be further investigated concerning human health if it is to be applied at the field level. It has been demonstrated that several diazotrophic bacteria, including Bacillus and Enterobacter, could promote growth of sugarcane (Lin and others 2012) as well as jatropha, maize, rice, sorghum, sugarcane, and wheat (Olivares and others 1997; Yanni and others 2001; Govindarajan and others 2007; Luna and others 2010; Rungin and others 2012; de Jesus Santos and others 2014; Alves and others 2015; Qin and others 2015). Furthermore, co-inoculation of five diazotrophs (Azospirillum, Burkholderia, G. diazotrophicus, Klebsiella, and Pseudomonas) was reported to give better growth of rice (Govindarajan and others 2007); while, co-culture of diazotrophs and a fungal endophyte, Trichoderma, revealed better growth and yield of chickpea (Verma and others 2014). It was suggested that the combination of phosphate-solubilizing, nitrogen-fixing, and phytohormone producing bacteria could provide a more balanced nutrition for plants and stimulate more growth of plants. For example, the combination of a phosphate-solubilizing B. megaterium, nitrogen-fixing B. subtilis, and Rhizobium leguminosarum bv. phaseoli caused a significant enhancement of seed yield and an uptake of macronutrients and micronutrients elements in common bean (Elkoca and others 2010). Mixed inoculation of Rhizobium, a phosphate-solubilizing B. megaterium subsp. phospaticum strain-PB, and a biocontrol fungus Trichoderma sp. was reported to increase seed germination, nutrient uptake, plant height, number of branches, nodulation, pea yield, and total biomass of chickpea compared to either individual inoculation or an un-inoculated control (Rudresh and others 2005). Therefore, the plant growth-promoting properties of endophytic bacteria and actinomycetes suggest that these bacteria merit further investigation for potentially safe and environmentally friendly biofertilizers which can help us limit the use of chemical fertilizers in agriculture.