Detection and characterization of broad-spectrum antipathogen activity of novel rhizobacterial isolates and suppression of Fusarium crown and root rot disease of tomato

Journal of Applied MicrobiologyVolume 118, Issue 3 p. 685-703 Original ArticleFree Access Detection and characterization of broad-spectrum antipathogen activity of novel rhizobacterial isolates and suppression of Fusarium crown and root rot disease of tomato L. Zhang, L. Zhang South China Agricultural University, Guangzhou, China South Crop Protection and Food Research Center, Agriculture and Agri-Food Canada, London, ON, CanadaSearch for more papers by this authorS.E. Khabbaz, S.E. Khabbaz South Crop Protection and Food Research Center, Agriculture and Agri-Food Canada, London, ON, Canada Department of Biology, Western University, London, ON, CanadaSearch for more papers by this authorA. Wang, A. Wang South Crop Protection and Food Research Center, Agriculture and Agri-Food Canada, London, ON, CanadaSearch for more papers by this authorH. Li, H. Li South China Agricultural University, Guangzhou, ChinaSearch for more papers by this authorP.A. Abbasi, Corresponding Author P.A. Abbasi Atlantic Food and Horticulture Research Centre, Agriculture and Agri-Food Canada, Kentville, NS, Canada Correspondence Pervaiz A. Abbasi, Atlantic Food and Horticulture Research Centre, Agriculture and Agri-Food Canada, 32 Main Street, Kentville, Nova Scotia B4N 1J5, Canada. E-mail: Pervaiz.Abbasi@agr.gc.caSearch for more papers by this author L. Zhang, L. Zhang South China Agricultural University, Guangzhou, China South Crop Protection and Food Research Center, Agriculture and Agri-Food Canada, London, ON, CanadaSearch for more papers by this authorS.E. Khabbaz, S.E. Khabbaz South Crop Protection and Food Research Center, Agriculture and Agri-Food Canada, London, ON, Canada Department of Biology, Western University, London, ON, CanadaSearch for more papers by this authorA. Wang, A. Wang South Crop Protection and Food Research Center, Agriculture and Agri-Food Canada, London, ON, CanadaSearch for more papers by this authorH. Li, H. Li South China Agricultural University, Guangzhou, ChinaSearch for more papers by this authorP.A. Abbasi, Corresponding Author P.A. Abbasi Atlantic Food and Horticulture Research Centre, Agriculture and Agri-Food Canada, Kentville, NS, Canada Correspondence Pervaiz A. Abbasi, Atlantic Food and Horticulture Research Centre, Agriculture and Agri-Food Canada, 32 Main Street, Kentville, Nova Scotia B4N 1J5, Canada. E-mail: Pervaiz.Abbasi@agr.gc.caSearch for more papers by this author First published: 16 December 2014 https://doi.org/10.1111/jam.12728Citations: 23AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract Aims To detect and characterize broad-spectrum antipathogen activity of indigenous bacterial isolates obtained from potato soil and soya bean leaves for their potential to be developed as biofungicides to control soilborne diseases such as Fusarium crown and root rot of tomato (FCRR) caused by Fusarium oxysporum f. sp. radicis-lycopersici (Forl). Methods and Results Thirteen bacterial isolates (Bacillus amyloliquefaciens (four isolates), Paenibacillus polymyxa (three isolates), Pseudomonas chlororaphis (two isolates), Pseudomonas fluorescens (two isolates), Bacillus subtilis (one isolate) and Pseudomonas sp. (one isolate)) or their volatiles showed antagonistic activity against most of the 10 plant pathogens in plate assays. Cell-free culture filtrates (CF) of five isolates or 1-butanol extracts of CFs also inhibited the growth of most pathogen mycelia in plate assays. PCR analysis confirmed the presence of most antibiotic biosynthetic genes such as phlD, phzFA, prnD and pltC in most Pseudomonas isolates and bmyB, bacA, ituD, srfAA and fenD in most Bacillus isolates. These bacterial isolates varied in the production of hydrogen cyanide (HCN), siderophores, β-1,3-glucanases, chitinases, proteases, indole-3-acetic acid, salicylic acid, and for nitrogen fixation and phosphate solubilization. Gas chromatography–mass spectrometry analysis identified 10 volatile compounds from 10 isolates and 18 compounds from 1-butanol extracts of CFs of five isolates. Application of irradiated peat formulation of six isolates to tomato roots prior to transplanting in a Forl-infested potting mix and field soil provided protection of tomato plants from FCRR disease and enhanced plant growth under greenhouse conditions. Conclusions Five of the 13 indigenous bacterial isolates were antagonistic to eight plant pathogens, both in vitro and in vivo. Antagonistic and plant-growth promotion activities of these isolates might be related to the production of several types of antibiotics, lytic enzymes, phytohormones, secondary metabolites, siderophores and volatile compounds; however, any specific role of each needs to be determined. Significance and Impact of the Study Indigenous antagonistic bacterial isolates have the potential to be developed as biofungicides for minimizing early crop losses due to soilborne diseases caused by Fusarium and other soilborne pathogens. Introduction Fusarium crown and root rot (FCRR) is an important disease affecting tomato (Solanum lycopericum L.) plants worldwide in both greenhouses and open fields (Jarvis 1988; Hartman and Fletcher 1991). In Canada, the disease was first reported from Ontario in 1975 (Jarvis et al. 1975). FCRR is caused by a soilborne fungal pathogen, Fusarium oxysporum Schlechtend.:Fr. f. sp. radicis-lycopersici W. R. Jarvis & R. A. Shoemaker (Forl). The pathogen can attack tomato plants at any time from seedling to adult stage causing vascular discolouration of crown tissues at the base, stem and root rot, and wilting of the affected plants before fruit-ripening (Jarvis 1988). Infection usually starts through root tips or other underground wounded tissues after germination of spores in soil (Boland and Kuykendall 1998). Aerial dissemination of conidia can also play a role in the spread of FCRR (Rowe et al. 1977; Rekah et al. 1999). The pathogen may also be carried in plug trays through contaminated soilless growth media as a source of FCRR spread in the field (Brammall and McKeown 1989). Forl is also pathogenic to a wide range of other plant species from several different families (Rowe 1980; Menzies et al. 1990). The management of FCRR can be a real challenge as the pathogen has a wide host range and it can spread through multiple modes (Rekah et al. 2001). During the past many decades, chemical fungicides have been the main strategy to manage soilborne tomato diseases (Ajilogba and Babalola 2013). However, the extensive use of chemical fungicides can lead to nontarget and negative effects on humans and environment and can also trigger resistance development in pathogens. Biofungicides, which offer alternative strategies to chemical fungicides for managing soilborne diseases, generated enormous interest from researchers across the globe in the past few decades (Gerhardson 2002; Weller et al. 2002; Fravel 2005; Haas and Défago 2005; Lugtenberg and Kamilova 2009; Olubukola 2010). Many different strains of Bacillus and Pseudomonas bacteria have been reported to effectively reduce plant diseases caused by many soilborne pathogens in a number of different hosts (Weller 1988; Ramamoorthy et al. 2002; Ownley et al. 2003; De Vleesschauwer et al. 2008; Rashmi et al. 2010; Ardebili et al. 2011; Karimi and Amini 2012; Khabbaz and Abbasi 2014). These rhizobacteria are common soil inhabitants and may already be playing a role in natural biological control of plant diseases. The plant-growth promotion and disease suppression effectiveness of these rhizobacteria may be due to their ability to compete in the rhizosphere, fix nitrogen, solubilize phosphate, and produce antibiotics, siderophores and phytohormones (Neilands 1981; Nautiyal et al. 2002; Santoyo et al. 2012). Antagonistic bacteria are known to produce volatile compounds such as hydrogen cyanide (HCN), cyclohexanol, benzothiazole and n-decanal (Knowles 1976); antibiotics such as phenazine, 2,4-diacetylphloroglucinol (DAPG), pyrrolnitrin, pyoluteorin, bacillomycin, bacilysin, iturin A, surfactin and fengycin (Athukorala et al. 2009; Santoyo et al. 2012); and hydrolytic enzymes such as chitinases, β-1,3-glucanases and proteases (Kim et al. 2008), which all may be playing an important role in pathogen as well as disease suppression. To develop the most effective biofungicides for managing specific plant diseases, novel isolates of beneficial bacteria with broad-spectrum antipathogen activities and multiple modes of action are constantly searched for. Recently, we identified some antagonistic bacterial isolates which showed broad-spectrum biocontrol effect in suppressing seedling diseases (Khabbaz and Abbasi 2014). There are also potential bacterial isolates that have antifungal activity against F. oxysporum f. sp. radicis-lycopersici and can suppress FCRR disease of tomato. The objectives of this study were to (i) identify antagonistic bacterial isolates with broad-spectrum antipathogen activity, (ii) study the effects of culture filtrates (CF), volatile substances, and secondary metabolite compounds of the selected bacterial isolates against important plant pathogens, (iii) identify different types of volatile compounds produced by the selected bacterial isolates, (iv) identify the most effective bacterial isolates by sequence analysis of 16S rRNA gene and (v) evaluate the effects of the selected bacterial isolates as bioformulations in irradiated peat (a sterile carrier material used for adsorbing the bacteria for developing bioformulations) on plant growth and suppression of FCRR disease of tomato under greenhouse conditions. Materials and methods Cultures of antagonistic bacterial isolates and plant pathogens Eight isolates (five characterized in this study and three from a previous study (Khabbaz and Abbasi 2014)) were obtained from field plots established in a commercial potato field located near Delhi, Ontario, Canada, to establish the disease suppressive conditions (Abbasi 2013). Two isolates were obtained from soya bean plants grown in field soil in a greenhouse. Three previously uncharacterized isolates from our culture collection were also included in this study (Table 1). The bacterial isolates were stored as glycerol stocks in a −80°C freezer. The cultures of eight fungal (Alternaria alternata, Botrytis cinerea, Colletotrichum coccodes, Cylindrocarpon destructans, F. oxysporum, Rhizoctonia solani, Thielaviopsis basicola and Verticillium dahliae) and two oomycete (Phytophthora capsici and Pythium ultimum) pathogens used in this study were also stored in a −80°C freezer. These cultures were maintained on agar plates throughout the course of this study. Table 1. Antagonistic bacterial isolates used in this studyaa The bacterial isolates Pf-5 and PA23 were used as reference strains in PCR assays for detection of genes for biosynthesis of antibiotics. Isolate number Bacteria References Origin 8B-1 Bacillus subtilis Khabbaz and Abbasi (2014) Potato soil 8D-45 Pseudomonas sp. Khabbaz and Abbasi (2014) Potato soil 9A-14 Pseudomonas fluorescens Khabbaz and Abbasi (2014) Potato soil 1A-48 Bacillus amyloliquefaciens This study Potato soil 9A-31 B. amyloliquefaciens This study Potato soil 1B-14 B. amyloliquefaciens This study Potato soil 1B-23 B. amyloliquefaciens This study Potato soil 1B-26 Pseudomonas chlororaphis This study Potato soil SL5 Ps. chlororaphis This study Soya bean leaf SL23 Paenibacillus polymyxa This study Soya bean leaf # 50 P. polymyxa This study Culture collection # 53 P. polymyxa This study Culture collection PEF-5 #18 Ps. fluorescens This study Culture collection Pf-5 Ps. fluorescens Howell and Stipanovic (1979) Cotton rhizosphere PA23 Ps. chlororaphis Savchuk and Fernando (2004) Soya bean root a The bacterial isolates Pf-5 and PA23 were used as reference strains in PCR assays for detection of genes for biosynthesis of antibiotics. Detection of broad-spectrum antagonistic activity of bacterial isolates The broad-spectrum antagonistic activity of bacterial isolates against 10 model pathogens was determined on agar plates in co- or dual-culture assays. Mycelial discs (8 mm diameter) from 7-day-old pathogen cultures were placed in the centre of Petri dishes containing potato dextrose agar (PDA), and 48-h-old cultures of the bacterial isolates were streaked along a straight line on one side of the Petri dish (1 cm from the edge). Three replicate plates per pathogen were used, and the plates were placed in an incubator at 24 ± 2°C for 3–14 days (P. capsici and P. ultimum for 3 days; B. cinerea, C. coccodes and R. solani for 5 days; A. alternata, C. destructans, F. oxysporum and V. dahliae for 7 days; and T. basicola for 14 days). The inhibition zones were recorded by measuring the radial growths on control plates (no bacteria) and on treatment plates (with bacteria). The per cent inhibition was calculated as: (growth radius of control − growth radius of treatment) × 100/growth radius of control. The effect of volatile compounds produced by antagonistic bacterial isolates was also evaluated against these 10 model pathogens in double-plate assays. Mycelial discs (8 mm diameter) from 7-day-old pathogen cultures were placed in the centre of PDA plates, and the bacterial isolates were streaked on LB agar plates (Fernando et al. 2005). The PDA plates were inverted on the LB agar plates facing each other but not touching agar surfaces, and the two plates were sealed together with parafilm and incubated at 24 ± 2°C for 3–14 days (P. capsici and P. ultimum for 3 days; B. cinerea, C. coccodes and R. solani for 5 days; A. alternata, C. destructans, F. oxysporum and V. dahlia for 7 days; and T. basicola for 14 days). The PDA plates with the pathogen inverted on LB agar plates with no bacteria served as the control. Three replicate plates per pathogen were used. The measurement of radial growth and per cent inhibition was calculated as described above. The cell-free CF of five bacterial isolates were selected based on growth inhibition in a pilot study and tested against 10 plant pathogens on PDA plates. The bacterial isolates were cultured in 250-ml flasks containing 150 ml NBY broth (8 g nutrient broth, 2 g yeast extract, 2 g K2HPO4, 0·5 g KH2PO4, 2·5 g glucose, 1 ml 1·0 mol l−1 MgSO4·7H2O and 1000 ml distilled water (DW)), grown for 48 h on a rotary shaker at room temperature (24 ± 2°C). Cells of bacteria were removed by centrifuging the broth for 15 min at 12 100 g, and the CFs were collected and passed through 0·22-μm sterile filters (Corning, NY, USA). CF saturated filter discs (0·5 cm; Whatman, Baie d'Urfe, Quebec, Canada) were prepared by saturating with 100 μl of CF of each bacterium. CF was applied five times in aliquots of 20 μl each time between air-drying the discs in a laminar flow hood. Each saturated disc was placed on PDA plate containing 8-mm mycelium disc from 7-day-old pathogen cultures in the centre of the Petri dish. There were three replicate plates for each test pathogen. Sterile filter disc saturated with 100 μl NBY broth served as the control. Plates were incubated at room temperature (24 ± 2°C). When the mycelia growth in the control plates reached the filter discs, the radial growth measurements were taken and the inhibition zones were calculated, as described above. The CF of antagonistic bacteria was also extracted with petroleum ether, cyclohexane, 1-butanol, ethyl acetate and methanol. These organic solvent extracts were then assayed against the selected eight pathogens on agar plates, as described above. In brief, 100 ml CF of each bacterial isolate was extracted with equal volume of each solvent in a 500-ml separating funnel for 10 min by mixing and was kept overnight at room temperature (24 ± 2°C). The top phase of the solvent extract was assayed (100 μl) on agar plates, as described above. Identification of antagonistic bacteria Three of the 13 bacterial isolates showing broad-spectrum antipathogenic activities against most pathogens in plate assays have been identified previously (Khabbaz and Abbasi 2014), and the remaining 10 isolates were identified in this study based on sequence analysis of 16S rRNA gene as described previously (Khabbaz and Abbasi 2014). For DNA extraction, purified colonies of the bacterial isolates were grown overnight in NBY broth at room temperature (24 ± 2°C) on a rotary shaker. DNA was extracted using the Gen Elute Bacterial Genomic DNA Kit (Sigma-Aldrich St. Louis, MO, USA), dissolved in 50 μl of TE buffer, and used as a template for PCR reactions. The amplification of 16S rRNA gene fragment was performed with universal primers 8F (5′-AGAGTTTGATCCTGGCTCAG) and 1492R (5′-GGTTACCTTGTTACGACTT) (Galkiewicz and Kellogg 2008) in a 50 μl final volume containing 5 μl DNA, 3 μl 10 μmol l−1 forward/reverse primers, 4 μl 10 mmol l−1 dNTPs, 2·5 μl 50 mmol l−1 MgCl2, 5 μl 10× PCR buffer and 0·5 μl 5 U μl−1 Taq DNA polymerase (Invitrogen, Burlington, ON, Canada). PCR mixture without DNA template was used as a negative control. The PCR conditions included the following: 94°C for 5 min followed by 35 cycles at 94°C for 1 min, 55°C for 45 s and 72°C for 2 min, followed by a final extension at 72°C for 5 min. The PCR products were separated in 1·0% (w/v) agarose gel containing Red Safe™ (nucleic acid staining solution, 20 000×; Intron Biotechnology, Toronto, ON, Canada). The PCR products were purified using QIA quick PCR purification kit (Qiagen, Gaithersburg, MD, USA) and sequenced at Eurofins MWG Operon, Huntsville AL, USA. Analysis of sequences was carried out with basic local alignment search tool (blast) against the database from the National Center for Biotechnology Information (NCBI) (www.ncbi.nlm.nih.gov/BLAST). Sequencing results were submitted to GenBank to get the accession numbers, and a phylogenetic tree was constructed using mega 6.0. Greenhouse FCRR bioassay Plant material, potting mixtures and soil Tomato 'Bonny Best' seeds were purchased from OSC Seeds (Waterloo, ON, Canada). A commercial peat-based potting mix, Pro-Mix BX®, was used in all greenhouse pot assays. This potting mix consists of sphagnum peat moss 75–85% by volume, horticultural grade perlite and vermiculite, macro and micro nutrients, dolomitic and calcitic limestone, and a wetting agent (Premier Horticulture Inc., Rivière-du- Loup, Quebec, Canada). The moisture content of potting mix was kept around 50%, and the pH was 7·4. Potting mix for the first FCRR bioassay was sterilized in an oven at 85°C for 5 h. Furthermore, an alkaline loam soil (pH 8·0) with a moderate level of organic C was collected from the Agriculture and Agri-Food Canada Research Farm (London, ON) and used in greenhouse FCRR bioassays. Controlled slow-release fertilizer Osmocote 14-14-14 (N-P-K) (Scotts-Sierra Horticultural Products Co., Marysville, OH) was added to the potting mix or soil at 4 g kg−1 mix or soil. Preparation of pathogen inoculum and bioformulations A culture of a Forl isolate, pathogenic to tomato plants, was maintained on PDA plates. For inoculum production, the pathogen was grown on autoclaved (twice on two consecutive days) moist pearl millet seed (100 g seed and 100 ml DW) in 500-ml flask, as described previously (Borrego-Benjumea et al. 2014). The flask was inoculated with 10–15 pathogen mycelial discs (8 mm diameter) obtained from the edge of fresh culture of Forl and kept at 24 ± 2°C for 2 weeks. The concentration of spores was determined by dilution plating on PDA agar plates, and a final concentration of 106 colony-forming units (CFU) g−1 mix was used in pot experiments. The formulations of six bacterial isolates showing broad-spectrum antipathogenic activities against most pathogens in plate assays as well as disease suppression in a previous study (Khabbaz and Abbasi 2014) were prepared with sterile irradiated peat (Becker Underwood Company, Saskatchewan Canada) as described previously (Khabbaz and Abbasi 2014). The bacterial isolates were grown in flasks containing 150 ml NBY broth and incubated on a rotary shaker (150 rev min−1) for 48 h at room temperature (24 ± 2°C). The bacterial cells were centrifuged at 12 100 g, resuspended in sterile DW, and the concentration was determined by a spectrophotometer and also by dilution plating on LB agar plates. The concentration 8 × 109 CFU ml−1 was used for the preparation of irradiated peat formulation. To the 70 ml LB broth, 120 g of irradiated peat and 5 ml of respective bacterial culture were added and mixed under sterile conditions. The formulated products were air-dried in a laminar flow hood to a workable (15–20%) moisture level and kept in polyethylene bags and used immediately or within a week of preparation. Bioformulation application, planting and disease evaluation Irradiated peat formulations of the six bacterial isolates were investigated as root treatments for suppressing FCRR disease of tomato. Tomato seedlings were grown in vermiculite plug trays in a growth room (15 h of fluorescent light at 22°C and 9 h of darkness at 19°C) for 4 weeks. Millet seed inoculum of Forl (30 g kg−1 mix) was mixed with potting mix in plastic bags and incubated overnight at 24 ± 2°C in the dark prior to planting. The inoculated potting mix from each bag was thoroughly mixed and dispensed into 10-cm-diameter plastic pots (eight replicate pots per treatment). Each pot received one 4-week-old tomato seedling according to the following treatments: tomato seedling treated with irradiated peat formulation of six bacterial isolates (Pf 9A-14, Bs 8B-1, Psp. 8D-45, Pnp #53, Pf PEF-5 #18 and Pc SL5); tomato seedling treated with benomyl (0·6 g l−1) as a fungicide control; and tomato seedling treated with irradiated peat or sterile DW as positive control. Tomato seedlings were removed from the plug trays and their roots washed and coated with the control and inoculated irradiated peat. Treated seedlings were immediately planted in pots and placed in a greenhouse for 15 h of fluorescent light at 22°C and 9 h of darkness at 19°C. The pots were watered daily or as required. Tomato plants were rated 25 days after planting for FCRR severity using a 0–3 scale (Mihuta-Grimm et al. 1990): 0 = no visible symptoms; 1 = slight discoloration of taproot; 2 = moderate discoloration of vascular tissues not extending above soil line; and 3 = excessive discolouration of vascular tissues extending well above soil line, or plant dead. Disease severity was calculated as follows: (DS = ∑(Number of disease level × Numbers of disease plant/Numbers of plant in total). Growth parameters, such as plant height, root length, fresh and dry weights, were determined as well. The experiment was repeated twice. Another set of experiments was performed in a natural field soil artificially infested with millet seed inoculum of Forl as described above. Preparation of pathogen inoculum and bioformulations of bacterial isolates were as described above for potting mix experiments. For each treatment, 2·5 kg of soil (enough soil to fill five replicate 10-cm-diameter plastic pots per treatment) was infested with the millet seed inoculum of Forl (30 g kg−1 of soil at 6 × 106 CFU g−1) by mixing in plastic bags. The infested soil was incubated for 24 h at room temperature (24 ± 2°C) in the dark before dispensing into five replicate plastic pots and planting tomato seedlings treated with irradiated peat formulation of six bacterial isolates and benomyl as described above. The plants were kept in a greenhouse and rated for FCRR severity and growth parameters as described above. The experiment was repeated twice. Characterization of antagonistic bacterial isolates The following characterization of three bacterial isolates Pf 9A-14, Bs 8B-1 and Psp. 8D-45 was completed in a previous study (Khabbaz et al. 2014). The rest of the 10 bacterial isolates were characterized in this study based on following various tests. Production of β-1,3-glucanase, chitinase and protease The β-1,3-glucanase activity of antagonistic bacterial isolates was determined based on their ability to digest glucagon in plate assays (Bang et al. 1999). The bacterial isolates were grown on the azurine cross-linked (AZCL) agar medium (3·0 g tryptic soy broth, 1·0 g chromogenic AZCL substrate, 15·0 g agar, 1000 ml DW) and incubated at 24 ± 2°C for 3 days. A dark blue halo around the bacterial colonies indicates a positive result for the production of β-1,3-glucanase by the bacterial isolates. The production of β-1,3-glucanase was scored as none (–, no blue), weak (+, light blue), strong (++, blue) and very strong (+++, dark blue). The chitinolytic activity of antagonistic bacterial isolates was determined as described previously (Chernin et al. 1995). The cultures were spotted onto chitin agar medium (1·62 g nutrient broth, 0·5 g NaCl, 6·0 g M9 salts, 2·0 g colloidal chitin, 0·1 mmol l−1 CaCl2, 1·0 mmol l−1 MgSO4, 3·0 nmol l−1 thiamin-HCl, 15·0 g agar, 1000 ml DW) and incubated at 24 ± 2°C for 10 days. A clear zone around a spotted culture indicates chitinase production by the bacterial isolates. The production of chitinase was scored as none (–, no clear zone), weak (+, small clear zone), strong (++, medium clear zone) and very strong (+++, large clear zone). The protease activity of antagonistic bacterial isolates was determined as described previously with some modifications (Abo Aba et al. 2006). The cultures were spotted onto the centre of LB agar plates containing 3% skim milk and incubated at 24 ± 2°C for 3–5 days. A formation of transparent halo surrounding the spotted culture indicates protease production by the bacterial isolates. The production of protease was scored as none (–, no halo), weak (+, small halo), strong (++, medium halo) and very strong (+++, large halo). Detection of the phosphate-solubilizing and nitrogen-fixing activities The ability of antagonistic bacterial isolates to solubilize phosphate was determined in a plate assay, as described previously (Nautiyal 1999). Bacterial isolates were spot-inoculated on National Botanical Research Institute's phosphate (NBRIP) growth medium (10·0 g glucose, 5·0 g Ca3 (PO4)2, 5·0 g MgCl2·6H2O, 0·25 g MgSO4·7H2O, 0·2 g KCl, 0·1 g (NH4)2SO4, 15·0 g agar, 1000 ml DW, pH 7·0) and incubated at 24 ± 2°C for 2 weeks. The appearance of clearing zones surrounding the spotted cultures indicates phosphate solubilization by the bacterial isolates. The capacity of phosphate solubilization was scored as none (–, no clear zone), weak (+, small clear zone), strong (++, medium clear zone) and very strong (+++, large clear zone). The nitrogen-fixing activity of antagonistic bacterial isolates was determined by the ability of bacteria to grow on the nitrogen-free medium (NFM; containing 5·0 g malic acid, 0·5 g K2HPO4, 0·2 g MgSO4·7H2O, 0·1 g NaCl, 20·0 mg CaCl2·2H2O, 2·0 mg NaMoO4·2H2O, 5·0 ml 0·5% bromothymol blue, 4·5 g KOH, 10·0 mg biotin, 15·0 g agar, 1000 ml DW, pH 7·5). The cultures were streaked onto NFM plates, incubated at 24 ± 2°C for 10 days and observed for growth which is an indication of nitrogen-fixing activity by the bacterial isolates. The nitrogen-fixing capacity of bacterial isolates was scored as none (–, no growth), weak (+, poor growth), strong (++, moderate growth) and very strong (+++, vigorous growth). Production of HCN and siderophore The ability of antagonistic bacterial isolates to produce HCN was determined according to the method described previously (Millar and Higgins 1970). The cultures were streaked onto trypticase soy agar (15·0 g animal peptone, 5·0 g soya peptone, 5·0 g sodium chloride, 4·4 g glycine, 15·0 g agar, 1000 ml DW) plates. Sterile filter paper saturated with an alkaline picric solution (2·5 g l−1 picric acid, 12·5 g Na2CO3, pH 13·0) was placed in the upper lid inside of the Petri plate containing bacterial culture. The plates were sealed with parafilm and incubated at 24 ± 2°C for 4 days. A change of colour of the filter paper from yellow to light brown, brown or reddish brown was recorded as a weak (+), moderate (++) or strong (+++) reaction, respectively. The production of siderophores by antagonistic bacterial isolates was determined using a plate assay, as described previously (Schwyn and Neilands 1987). The tertiary complex chromeazurol S (CAS)/Fe3+/hexadecyl trimethyl ammonium bromide (HDTMA) served as an indicator dye. The 48-h cultures were streaked onto the succinate medium (4·0 g succinic acid, 3·0 g K2HPO4, 0·2 g (NH4)2SO4·7H2O, 15·0 g agar, 1000 ml DW, pH 7·0) amended with indicator dye. To prepare 1 l of the blue agar, 60·5 mg CAS was dissolved in 50 ml of DW and mixed with 10 ml Fe III solution (1·0 mmol l−1 FeCl3.6H2O in 10 mmol l−1 HCl). While constantly stirring, this solution was slowly added to 72·9 mg of HDTMA dissolved in 40 ml water. The plates were incubated at 24 ± 2°C for 5–7 days. Development of yellowish fluorescent to orange halo around the bacterial growth indicated siderophore production by the antagonistic bacterial isolates. The change in colour from blue to orange indicated as hydroxamate or to purple as catechol siderophore. The production of siderophore was scored as none (blue), little (light yellow), strong (yellow) and very strong (orange). The quantity of siderophores produced by bacterial isolates was estim