The sample collection and field trials were done at Agro Narayani Farm, Sukranagar, Chitwan District, Nepal (Latitude: 27.582016 and Longitude: 84.272259) where the bacterial wilt infection was recorded in the previous harvest (done 3 months before the test). Also, Ralstonia spp. were observed in the Casamino acid-Peptone-Glucose (CPG) Agar plates from the soil samples. The tests in the field were conducted from 11 March 2017 to 8 July 2017 for 120 days in new transplants. At the time of transplant, compost fertilizer at the rate 2.94 kg/m ², urea at the rate of 23.59 gm/m ², potash at the rate of 29.49 gm/m ², DAP 19.66 gm/m ², borax at the rate 1.97 gm/m ² and zinc at the rate 1.97 gm/m ² were applied. At 45 days and 90 days of transplant NPK 20:20:20 was applied at the rate 9.83/m ². The spacing of the plant was 50 cm in double row system of 50 × 50 cm. The plot size was 50 m ² and total of 8 plots were used having 100 plants per plot. Weather data were also collected from online resources as a reference.

Physical symptoms, such as wilting of young leaves, discolored tissue at the dissected part of the stem base, and white, slimy ooze when the dissected part of the plant was kept in the glass of water, were used for the identification of infected plants ²⁴ .

Bacterial wilt infection in tomato plants ( Solanum lycopersicum L. var. Manisha) grown in Chitwan, Nepal, was positively identified by S. Yendyo of Kishan Call Center (KCC). Six whole plants were brought to the Quality Control laboratory of Agricare Nepal Pvt. Ltd. in a sterile bag for the isolation of the pathogen.

Ralstonia spp. was isolated from dissected sections of the infected tomato plants on Casamino acid-Peptone-Glucose (CPG) Agar (casein hydrolysate 1 g/l, peptone 10 g/l, glucose 5g/l, agar 15g/l) which is preferred media for isolation of Ralstonia spp. The stems of the infected plant were washed three times with autoclaved distilled water and then blot dried. After drying the stem were washed with 80% ethanol solution, then 1% sodium hypochlorite (NaOCl) solution was applied for 2 minutes. Final washing was done with autoclaved distilled water three times. The xylem of 2–3 cm from the stem was dissected and the sap was rubbed in CPG medium and inoculated for 48h at 28°C. Identification was done on the basis of the morphology of the colony on CPG medium, Gram staining and microscopic examination ^{25, 26} .

A total of 13 strains were used in this study (see Table 1). Six strains of Trichoderma spp., two strains of Pseudomonas fluorescens and one strain of Bacillus subtilis (GenBank Accession No. MG952584) were isolated from soils and root soils collected from different sites in Nepal. ATCC 13525 strain of P. fluorescens from Microbiologics, St. Cloud, MN 56303, France was used as the reference for isolated Pseudomonas species. Trichoderma harzianum, Trichoderma virens and Trichoderma asperellum strains were provided by Tamil Nadu Agricultural University (TNAU), Coimbatore, Andhra Pradesh, India, and were used as reference species for isolated Trichoderma spp for two purposes. First, the reference species were used to identify the isolated species through morphological and microscopic analysis and second, the reference species were used to compare the antagonistic effects exhibited by the native isolates. All the species were collected from the tropical region except Bacillus subtilis which was collected from mid hill regions of Nepal. The reason behind this was that the site from where the samples were collected was unaffected by the wilt disease compared to nearby sites where the bacterial wilt was severely present. The testing of the soil samples from the unaffected site revealed the substantial presence of Bacillus subtilis which may have induced systemic resistance in the plants.

Trichoderma spp. were isolated by serial dilution method using TSM agar plate (K ₂HPO ₄: 0.9 g/l, MgSO ₄: 0.2 g/l, KCl: 0.15 g/l, NH ₄Cl: 1.05 g/l, Glucose 3 g/l, Rose Bengal: 0.15 g/l, agar 20 g/l, streptomycin: 100 mg/l, tetracycline: 50 mg/l) ²⁷ . 1 gm of each soil samples were suspended in 9 ml of sterile distilled water and vortexed (Accumax India, New Delhi-110058, India) for 5 min. The soil suspension was then serially diluted to 10 ^-3 and 10 ^-4. Pour plate technique was used by mixing 1 ml of the diluted soil suspension in 3 TSM agar plates for each sample and incubated at 25°C for 5 days. The strains were purified on TSM agar plates using sub-culture technique.

Pseudomonas fluorescens were isolated by serial dilution method using King’s B agar (Peptone: 20g/l, K ₂HPO ₄: 1.5 g/l, MgSO ₄: 1.5 g/l, glycerol: 10 ml/l, agar: 20g/l) ²⁸ . 1 gm of each soil samples were suspended in 9 ml of sterile distilled water and vortexed (Accumax India, New Delhi-110058, India) for 5 min. The soil suspension was then serially diluted to 10 ^-6. Pour plate technique was used by mixing 1 ml of the diluted soil suspension in 3 King’s B agar plates for each sample and incubated at 27°C for 48 h and the fluorescence was observed in UV light. The fluorescent strains were purified on King’s B agar plates using sub-culture technique.

Bacillus subtilis was isolated by serial dilution method using Nutrient Agar (Peptic digest of animal tissue: 5 g/l, NaCl: 5 g/l, beef extract: 1.5 g/l, yeast extract: 1.5 g/l, agar: 20g/l) ²⁹ . 1 gm of each soil samples were suspended in 9 ml of sterile distilled water and vortexed (Accumax India, New Delhi-110058, India) for 5 min. The soil suspension was then serially diluted to 10 ^-6. Pour plate technique was used by mixing 1 ml of the diluted soil suspension in 3 NA plates for each sample and incubated at 27°C for 48 h. The strains were purified on NA agar plates by using sub-culture technique.

All thirteen isolates were screened against Ralstonia spp . by agar well diffusion technique ³⁰ . Ralstonia spp . on CPG agar plates were transferred to the nutrient broth and shaken in a rotary shaker (Talboys, Henry Troemner, LLC, USA) at 100 rpm at 27°C for 24 h. Similarly, the TSM, King’ B and NB were prepared for all Trichoderma spp., P. fluorescens and B. subtilis, respectively, and incubated for 7 days, 48 h and 48h, respectively. After incubation of the antagonists, 5 ml of broth suspension were centrifuged at 5000 rpm for 5 min and the supernatant was stored at 4°C for further procedure. Then, Ralstonia spp . suspension of 10 ⁸ cells/ml was prepared as per McFarland 0.5 turbidity method ³¹ and was swabbed on NA plates. Holes of 5 mm were punched into the agar plate and 40 µl of the supernatant prepared were added separately and the plates were incubated at 27°C for 48 h. Inhibition of Ralstonia spp . growth was assessed by measuring the radius in mm of the zone of inhibition (ZOI) after incubation.

For microscopic visualization of the inhibition, CPG agar plates were prepared to provide the most favorable growth to Ralstonia spp. , and the respective 5 mm mycelial discs of Trichoderma species were added in the center of the plate after cotton swabbing from the CPG broth of Ralstonia spp. For B. subtilis and P. fluorescens, the line was streaked parallel to the streak of Ralstonia spp. in two different CPG agar plates using dual culture technique. After 72 hr of incubation, live microscopic examination on the culture plate was done using a digital microscope (Olympus CX-43, Tokyo, Japan). Images were captured to visualize the interaction of the individual strains of biocontrol agents with Ralstonia spp .

There is common practice in Nepal of keeping bio-based products in 5–10% (w/v) sucrose solution for 2–4 h before application to the plant. Hence, to evaluate the effect of 5% (w/v) sucrose solution on the cell number (growth) of the biocontrol agents, 1 ml of concentrate containing 1 × 10 ^-9cells ml ^-1 was kept in 5% sucrose solution, made using autoclaved distilled water. Cell count was taken using a Hemocytometer (Reichert, Buffallo, NY, USA) with trypan blue at 1 h, 2 h, 3 h and 24 h to observe the effect on the microbial population.

For the field study, the concentrates containing 10 ⁹ cells/ml of the respective biocontrol agent were used. The densities of the cells were determined using Hemocytometer (Reichert, Buffallo, NY, USA) ³² . The TSM broth of all six isolated native Trichoderma species viz. AA2, AG3, AKD, A5, A9 and A10 were mixed in equal proportion to prepare 1 liter of concentrate containing 10 ⁹cells/ml. Similarly, two native P. fluorescens species viz. PFB and PFS were mixed in an equal proportion to prepare 1 liter of concentrate containing 10 ⁹ cells/ml. Concentrate of native B. subtilis viz. BS containing 10 ⁹ cells/ml was used to analyze the effect of B. subtilis as a possible biocontrol agent.

Before the application in the tomato plots at Agro Narayani Farm, the prepared concentrates of biocontrol agents were taken to the field and were further diluted at the rate of 2ml/l of tap water containing 5% (w/v) sucrose. After 2 h of incubation in 5% sucrose water, the diluted solutions were applied in the root of tomato plants at the rate of 100 ml per plant. The processes of applications were repeated every 7 days for 8 weeks (total of 8 applications) by preparing fresh dilutions in 5% sucrose solutions 2 h prior to applications. Effects after the 8 weeks of continuous application were measured in the field by identifying the number of plants that underwent recovery after treatment. 6 plots were treated with the biocontrol agent and 2 plots were used as controls. Chemical treatment was done in one plot (positive control plot) using the combination of Agricin (9% Streptomycin Sulphate and 1% Tetracycline Hydrochloride) at the rate of 100 ml of 0.1% (w/v) solution per plant from Agricare Nepal Pvt. Ltd., Nepal and Bavistin (50% carbendazim) at the rate of 0.2% (w/v) solution per plant from Crystal Crop Protection Pvt. Ltd., India. For negative control, no treatment methods were selected in one plot.

The treatment plots were designed such that the effects of the individual biocontrol agent and effects of combination treatment can be studied ( Table 2).

Statistical analysis of the data was done using IBM SPSS Statistics (ver 23) and figures and data were made through Microsoft Excel 2007 and Microsoft Word 2007.

Weather data of Chitwan District for temperature, humidity and rainfall were collected from www.worldweatheronline.com from March 2017 to July 2017 (see Table 3).

The bacterial wilt outbreak was reported in the late May, 2017, when the temperature and humidity level increased. Higher temperature and moisture favors the growth of Ralstonia spp. ³³

From the field examination of the tomato plants, observation revealed that the leaves were flaccid, adventitious roots started to appear on the stem and ooze appeared after dipping the stem in water. Also, field experts from KCC confirmed the presence of bacterial wilt infection, due to their years of experience in plant disease diagnosis.

Infected plant saps from six xylems showed similar bacterial colonies on CPG medium. All the colonies were similar to avirulent type, as the appearance was white or cream-colored, irregularly-round, fluidal, and opaque on CPG medium ^{34, 35} . Gram staining and observation using a microscope showed that the bacteria were gram negative, rod-shaped and non-spore forming, which further confirmed that the bacteria was Ralstonia spp.

Strains tested showed antagonistic effect against Ralstonia spp., with inhibition zone radii ranging from 13 to 21.33 mm ( Table 4). P. fluorescens PFS isolated from Parthenium hysterophorus L. rhizoplane soil was most potent compared to other P. fluorescens strains. Trichoderma virens ATV and Trichoderma harzianum ATH provided by TNAU were the least and most potent species. Among six natively isolated Trichoderma sp., AA2 isolated from Parthenium hysterophorus L. rhizoplane soil was most potent and AKD and A9 were least potent. However, the activities of native Trichoderma spp. were satisfactory in the term of inhibition zone shown. Bacillus subtilis BS isolated from Bhaktapur top soil did not show satisfactory inhibition activity. The complete photographs showing zone of inhibitions can be retrieved from data availability section.

The effect of biocontrol agents against Ralstonia spp. was analyzed at a microscopic level in dark phase using image analyzer (Olympus CX-43, Tokyo, Japan). Figure 1– Figure 3 show the distinct inhibitory effect on the growth and survival of Ralstonia spp. caused by different biocontrol agents. From the figures, it can be seen that the most of the pathogenic cells ( Ralstonia spp.) were either killed or growth was retarded or limited in or towards the region of growth of antagonists as compared to the region away from the growth of antagonists.

The digital images from Figure 1 reveal that the population of Ralstonia spp. is significantly less and most of the cells are dead in the region of growth of samples treated with Trichoderma spp., compared to the region 4 cm far from the growth of Trichoderma spp as the Ralstonia is a rod-shaped bacteria but near the growth region of Trichoderma, most of the bacterial cells clearly seems to be round and distorted in shape with much lower population density than the region far away This indicates the fact that the Ralstonia cells that got inoculated in the plate during the cotton swab were unable to multiply and grow in the zone where Trichoderma was growing.

The images from Figure 2 reveal that PF tended to grow on the side of RS, whereas RS tended to restrict the growth towards the PF species. RS positive control (without PF streak nearby) tended to spread, which confirms the spreading pattern of RS. The microscopic study was to verify a fact that the incompatible species does not grow towards each other and the growth of dominant microbes always surpasses the growth of other recessive microbes. The same phenomenon was observed in the microscopic analysis as Ralstonia spp. formed a clear boundary or showed restricted spreading in the region of the streak as compared to the Positive control but the streak of Pseudomonas fluorescence was easily growing towards the Ralstonia spp.

The effect of sucrose (5% w/v) in the concentrate mixture of individual biocontrol agents was analyzed ( Figure 4), which showed that there was a profound increase in the number of cells of biocontrol agents after 2h of incubation compared with the initial population. The cell count between 2h and 3 h of incubation was not significant as compared with 2 h and 24 h of incubation. Thus, 2 h of incubation in sucrose solution can be considered as optimal time, as lengthier time can result in the growth of contaminant in the solution whilst using tap water in the field.

Before applications, 2ml of respective biocontrol agents of (10 ⁹ cells/ml) were incubated in 5% (w/v) of sucrose solution and incubated for 2 hr. The prepared dilution was thus applied at the rate of 100 ml per plant in the root region every week. 8 applications were done over 8 weeks. 8 plots (100 plants/plot) were selected, out of which one plot was used as positive control/chemical treatment plot (Agricin+Bavistin) and one plot as negative control/zero treatment plot (no treatment given).

The results are displayed in Table 5 and show that the application inhibited the bacterial wilt infection ( Ralstonia spp.) in tomato plants by and the highest rate of plant recovered was 97% from treatment using antagonists, which was comparable with the plant recovery of 94% using the chemical treatment (Agricin and Bavistin). Only 37% of plants were recovered in the plot where no treatment methods were applied. Field Images ( Figure 5) show a clear visualization of growth of plants and severity of infection in treated and untreated plot. The complete photographs of the field trial can be retrieved from the data availability section of this manuscript.