Effect of Bacillus and Trichoderma species in the management of the bacterial wilt of tomato (Lycopersicum esculentum) in the field

Bacterial wilt caused by Ralstonia solanacearum is one of the most devastating diseases in tomato cultivation. This study aimed to evaluate the effect of Bacillus and Trichoderma isolates to manage the bacterial wilt disease. Field experiments were conducted in a randomized complete block design at Mwea and Kabete sites in Kenya. The treatments included 3 Trichoderma; 2 Bacillus isolates; a mixture of T1, T2, and T4; chemical standard; and distilled water as control. Trichoderma and Bacillus isolates were grown on sterilized sorghum grain and cow manure carriers respectively. Antagonist’s inoculation was carried out by dipping tomato plants for 30 min in each treatment suspension. Each treatment was then applied at a rate of 150 ml/plant hole and this was repeated after 35 days. All the treatments significantly reduced bacterial wilt incidence and severity at P ≤ 0.05 than the control at both sites. Trichoderma isolate T1, followed by Bacillus isolate CB64, was the best in reducing the disease incidence by more than 61.66 and 53%, respectively at both sites. Treatment CB64 and T1 had the highest reduction of R. solanacearum population in the soil by 93.17 and 92.07%, respectively. However, control had a pathogen increase of 20.40%. CB64 and T1 performed significantly better compared to the standard, while the mixture of isolates T1, T2, and T4 performed the poorest in all parameters. The treatments also increased the yield of tomato. Results from this study showed that Trichoderma and Bacillus isolates are effective biological control agents for use in management of bacterial wilt.


Background
Tomato (Lycopersicum esculentum) is one of the most consumed vegetables in Kenya (Smart Farm, 2016). Cultivation of tomato crop suffers high losses due to several viral, fungal, and bacterial diseases that affect the crop (Yuging, 2018). Among the diseases, bacterial wilt caused by Ralstonia solanacearum has been reported to be the most rampant disease in tomato production (Kago et al., 2019). In Kenya, bacterial wilt causes 64% losses on crops grown under open field conditions and up to 100% loss on the crops in the greenhouse (Mbaka et al., 2013).
Management of bacterial wilt is difficult since R. solanacearum pathogen has a wide host range and damages over 200 plant species in 50 different families that include tomato, potato, eggplant, pepper, and tobacco (Meng, 2013). The pathogen has a high destructive nature, ability to persist even in abandoned lands, and a wide geographical distribution (tropics, sub-tropics, warm temperate regions) (Mihovilovich et al., 2017). There are hardly known chemicals to manage bacterial diseases, except antibiotics that are used for animal and human diseases (Yendyo et al., 2017). These antibiotics are highly regulated to avoid development of resistance if carelessly used.
The effectiveness of available conventional management strategies for bacterial wilt is very limited (Aguk et al., 2018). Hence, biological control measures that use antagonistic fungal and bacterial agents are an attractive option (Mandal et al., 2017). Biological control of bacterial wilt disease of solanaceous crops caused by R. solanacearum, using antagonistic agents, has been reported earlier (Singh, 2013 andKumur, 2017). Plant growth promoting bacteria and fungi-like Bacillus and Trichoderma species have been reported to be promising biocontrol agents for management of R. solanacearum. It has been found that species of Bacillus and Trichoderma are able to reduce bacterial wilt incidence on tomato plants and increase the yields (Thongwai &Kunopakarn, 2007 andNarasimhamurthy et al., 2018).
Hence, the aim of this present study was to evaluate the efficacy of Bacillus and Trichoderma strains in managing the bacterial wilt of tomato under field conditions.

Experimental site
The experiments were conducted during the period March-August 2019. Tomato fields heavily infected with bacterial wilt disease were identified at Mwea and Kabete sites. Kabete field station is located in Kenya at an agro-ecological zone (AEZ) III, at 01°15′ S; 036°44′ E and an altitude of 1820 m above sea level. It has a bimodal rainfall of 1059 mm annually, and temperature ranges between 12.3 and 22.5°C. The soils are deep, dark brown to brown humic nitisols with kaolinite clay minerals and a good drainage ideal for tomato production. Mwea site, Kirinyaga County, is located in agroecological zones II at 0.5420°S, 37.2735°E and at an altitude of 1570 m above sea level. The region experiences a bimodal rainfall of 1470 mm annually, and temperature ranges between 15.6 and 28.6°C. The soils are deep humic nitisols that are moderately fertile with a pH of about 5 (Waiganjo et al., 2006).

Experimental design
The experimental design was randomized complete block design (RCBD), with 8 treatments and 3 replicates. Each plot was 3.6 × 2 m, separated from other plot by 1 m of weed-free bare ground. The blocks were separated by 2-m paths. Rio Grande tomato variety that is moderately resistant to bacterial wilt disease was used. There were 20 plants in each plot with a spacing of 60 × 60 cm. The 8 treatments tested were Trichoderma isolates T1, T2, and T4 isolated from Kabete field station, purified and each multiplied in a sorghum carrier. Bacillus isolates CB64 and CA7 were sourced from the department of Plant Science and Crop Protection at the University, and each was multiplied in cow manure carrier. A chemical standard used by farmers (di-bromo di nitro propane 1, 3-diol) was also used. A mixture treatment of T1, T2, and T4 in the ratio of 1:1:1 was included, and distilled water was used as the control.
Growth and survival of Bacillus isolates in cow manure carrier Two Bacillus isolates CB64 and CA7 that had the highest activity against R. solanacearum in vitro were multiplied in cow manure used as a carrier for field experiments. Cow manure, which had decomposed for a period of 3 months, was sun dried, crushed, and sieved through 2-mm sieve. The C:N ratio of the manure was determined before sterilization at 121°C at 1.5 bars for 15 min. This process of sterilizing manure was repeated once for 3 days. Two hundred grams of the sterile manure was poured in sterile sandwich boxes, used as incubating chambers. Isolates CB64 and CA7 previously grown on nutrient agar at 27°C for 48 h were harvested by flooding the plates of each isolate with 5 ml sterile distilled water. Bacterial colonies were then scraped off, using sterile glass slides, and each Bacillus isolate suspension was emptied in different sterile conical flasks. The concentration of the bacterial suspension was adjusted to 1 × 10 9 CFU/ml by the serial dilution technique. Using a sterile pipette, 0.5 ml of sterile distilled water was added to 1 g of the dry sterile cow manure to make it moist and to facilitate bacteria multiplication. Using a sterile syringe, the sterile manure was inoculated by injecting 0.3 ml of each Bacillus suspension into 1 g of manure, following a modified protocol of Macharia (2002). Bacillus isolates were allowed to multiply in the carrier at 28°C for 48 h. The concentration of Bacillus isolates in the manure was determined by serial dilution technique.

Growth and survival of Trichoderma isolates in sorghum carrier
Three Trichoderma isolates, T1, T2, and T4, that had the highest activity against R. solanacearum in vitro were maintained by regular sub-culturing at intervals of 10 days on PDA. The isolates were multiplied in white sorghum grains following a modified protocol of Kumar (2017). The C:N ratio of the sorghum grains was determined before sterilization. Two hundred and fifty grams of sorghum grain was placed in 1-L conical flask and supplemented with 5% anhydrous dextrose in 250 ml distilled water. The sorghum was then parboiled and autoclaved at 121°C at 1.5 bars for 1 h for 2 consecutive days. Five-millimeter discs of 7-day old culture of each Trichoderma isolates were placed in each of the flasks with sterilized sorghum grains. The flasks were corked with cotton wool and aluminum foil then incubated at 28 ± 1°C for 18 days, with regular shaking after every 3 days. Colonized sorghum was air dried and ground into powder, using a grinding machine. Spore concentration per gram of the colonized sorghum grain carrier was determined on the 7th, 11th, 14th, and 18th day after inoculation, using the serial dilution technique of Muriungi et al. (2013).

Preparation and application of the treatments in the field
One gram of well-colonized manure that contained 5.86 × 10 15 CFU/g for Bacillus isolate CB64 and 6.73 × 10 15 CFU/g for isolate CA7 was mixed each in 10 ml distilled water. For each Trichoderma isolate, 1 g of wellcolonized ground sorghum that contained an average of 2.6 × 10 9 spores/g was mixed in 40 ml distilled water. The mixture of T1, T2, and T4 was prepared in a ratio of 1:1:1, and the concentration applied was 2.6 × 10 9 spores/g of sorghum carrier. One gram of the standard chemical was mixed in 3 L of distilled water, following the manufacturer's recommended application rates. Four-week-old tomato plants were uprooted from the nursery. The roots of tomato plants were first dipped for 30 min in the treatments to ensure that the isolates were given enough time for interaction with the plants before being transplanted into the field. Immediately after transplanting, a soil drench of 150 ml of each treatment was applied around the plant roots, following a modified protocol of Rosyidah et al. (2013). The soil drench with freshly prepared treatments was applied again after 35 days of transplanting.

Determination of bacterial wilt incidence and severity
Bacterial wilt incidence was assessed every week after treatment application up to the end of the experiment by counting the number of wilted plants in each plot. Disease incidence was assessed as percentage of wilted plants within each treatment, where I is the wilt incidence, NPSWS is the number of plant showing wilt symptoms, and NPPT is the number of plants per treatment (Ayana et al., 2011). Assessment of the disease severity at the end of the experiment, 126 days after transplanting, was based on stem browning and bacterial ooze (Elphinstone et al., 1998). Any browning observed on the split stem was recorded, using a score of 0-3 where 0-no browning, 1light brown color restricted to 2 cm from base of stem, 2-light brown color spread more than 2 cm from the base of the stem, and 3-dark brown color on the vascular tissue. Bacterial ooze was scored at 0-3 where 0-no ooze, 1-thin strands of bacteria ooze that stops in 3 min, 2-continuous thin flow that is unrestricted, and 3-heavy ooze turning the water turbid in 2 min. Percentage severity index (PSI) was calculated, using the method described by Cooke (2006): PSI = ∑ (sores × 100)/(number of plants rated × maximum scale of the scores).

Effect of Bacillus and Trichoderma isolates on Ralstonia solanacearum population in the soil
The pathogen's population was determined 3 times during the experiment. The first sampling was carried out before tomato plants were transplanted to determine the population of R. solanacearum in the experimental area. The second and third samplings were undertaken at 60 and 112 days after treatment application. Six plants were randomly sampled from each treatment to determine the population of R. solanacearum in the soil rhizosphere around the plant root zone. Scoops of about 50 g were randomly picked from a depth of 10 cm and were composited in polythene bags for laboratory analysis. Ten grams of the sampled soil was placed into 250-ml flasks containing 90 ml sterile water. After vigorous shaking for 30 min in a rotary shaker at 200 rpm, the suspension was serially diluted up to 7-fold as described by Xu et al. (2012). One millimeter from each of the 6th and 7th dilutions was plated on Kelman's TZC media. After 48 h of incubation at 28°C, colonies typical to R. solanacearum were counted. Colony-forming units were calculated per gram of soil.

Effect of Bacillus and Trichoderma isolates on tomato yields and fruit size
Tomato yields were assessed weekly from the first harvest of mature fruits at 85 days after transplanting by measuring the weight of the total fruit from each plot, using a weighing balance. The yields per plot were converted to tonnes per hectare (Diogo & Wydra, 2007). The quality of tomato fruits was determined by measuring fruit diameter per plot, using Vernier caliper.

Data analysis
The experiments in this study were laid out in a randomized complete block design, and each experiment was repeated thrice. Data collected on bacterial wilt incidence, severity, and population of R. solanacearum in the soil was subjected to analysis of variance (ANOVA), using the Genstat statistical software 15th edition. ANOVA test was conducted in a general treatment structure in randomized block design to evaluate the significant differences due to treatments (biological control agents), between the test location (Kabete and Mwea site) and among the days after treatment application. Means were separated using Fisher's protected LSD at 5% significance level.

Results and discussion
Growth and survival of Trichoderma isolates in sorghum carrier The study showed that Trichoderma isolates were able to grow, survive, and multiply in sorghum grain carrier ( Table 1). Sporulation of each Trichoderma isolate in sorghum carrier started between 5 to 21 days after incubation. Colonization of the parboiled sorghum was characterized by the sorghum grains assuming whitish, yellow, and greenish color of the Trichoderma spp. Strains of Trichoderma are known to colonize substrates containing lignin and cellulose such as sorghum, rice, maize grains, and farmyard manure among others. They produce enzymes that degrade polysaccharides in sorghum and other substrates (Kumar, 2017).
Trichoderma isolate T4 gave the best growth as demonstrated by sporulation, followed by isolate T1 and then T2. Sporulation of all the Trichoderma isolates increased from 10 6 to 10 9 conidia per gram of sorghum carrier between 7 and 14 days of incubation. These findings are in agreement with those reported by Singh et al. (2014) that at 7 days after inoculation, sorghum grain substrate had the highest population of Trichoderma harzianum of 2.25 × 10 8 CFU/g. The results herein are also comparable to those reported by Rajput et al. (2014) that the highest population of 10.03 × 10 9 CFU/g of T. harzianum was recorded from sorghum grain carrier compared to other substrates.
The C:N ratio of the white sorghum was 13.4. This ratio showed that there was a high nitrogen content in the sorghum carrier. The availability of N in sorghum carrier increased sporulation and hypha growth essential for substrate colonization, and this was in agreement with reports by Okoth et al. (2009). Similar findings by Srivastava et al. (2010) also reported that out of 6 substrates used for mass multiplication of T. harzianum, sorghum grains that had high N content was the best carrier and gave significantly higher numbers of Trichoderma conidia than other substrates. Obtained results showed that sorghum grain was a good substrate for multiplication and use as a carrier for Trichoderma isolates for large application of the isolates in the field. These results herein confirm those reported by Kumar (2017). However, from 14 to 18 days of incubation, the concentration of Trichoderma isolates T4 and T1 remained constant at 10 9 conidia/g of sorghum, while for the isolate T2, the conidia concentration reduced slightly to 10 8 . These results are comparable to the findings reported by Muriungi et al. (2013) who found that there was an increase in Trichoderma isolates conidia per gram of sorghum carrier but the increase was not indefinite. These results indicated that 2 weeks was enough to give maximum spore concentration of the Trichoderma isolates in sorghum carrier. This is because with time, when the available nutrients in sorghum substrate are exhausted, the population of Trichoderma remains constant or falls rapidly (Singh et al. 2014).

Determination of bacterial wilt incidence and severity in Kabete and Mwea fields
All treatments showed significantly lower disease incidences and severity than the control at P ≤ 0.05 (Table 2). A minimum disease incidence (26.67%) was observed after 126 days of incubation in plots treated with Trichoderma isolate T1 at both Kabete and Mwea sites. This performance was significantly better than control plots that gave 88.33 and 93% disease incidence at Kabete and Mwea site, respectively. These results are in agreement with Kumar (2017) who reported tomato bacterial wilt disease incidence range of 30-40%, when plots were treated with T. harzianum. The best treatment, Trichoderma isolate T1, which had more than 61.66% reduction of bacterial wilt disease incidence also gave the lowest disease severity of 63.33% than the control (98%) severity index at both sites. The percentage severity observed in T1 was lower than control plots by more than 47%. These findings are in line with Narasimhamurthy et al.'s (2018) reports that T. asperellum-treated plots showed a reduction of bacterial wilt disease by 51.06%. Rosyidah et al. (2013), Tinatin and Saykal (2016), and Yendyo et al. (2017) confirmed the same results. The reduction of bacterial wilt incidence and severity may be due to Trichoderma strains producing antagonistic compounds against R. solanacearum. These findings are in line with Tapwal et al. (2011) who reported that different strains of Trichoderma produce various metabolites and secondary compounds that have antagonistic activity against R. solanacearum. Bacillus isolate CB64 was the second best treatment that showed bacterial wilt incidence of 30% at Kabete and 40% at Mwea after 126 days of transplanting. These findings are comparable to those reported by Singh et al. (2012) that a minimum disease incidence of 12.67% at 50 days after incubation was observed in Bacillus BS-5 treated soil. Similarly, in the present study, a disease incidence of 6.7% was recorded 60 days after plots were treated with Bacillus isolate CB64. Treatment CB64 had an average of 50% reduction in bacterial wilt incidence than the control. This was similar to earlier reports by Lemessa and Zeller (2007) that Bacillus subtilis strain B 2 G reduced bacterial wilt incidence by 60%. Akintokun et al. (2019) also reported similar findings. Plots treated with CB64 had the lowest disease severity at an average of 59.46% than the control. Comparable findings were reported by Wei et al. (2011) and Huang et al. (2014) of the ability of Bacillus strains to produce various metabolites and compounds that are able to reduce bacterial wilt incidence and severity in tomato plants.
Plots treated with Trichoderma isolate T1 and Bacillus isolate CB64 performed better than plots treated with the standard chemical in reducing bacterial wilt disease incidence and severity (by more than 20%). These findings were similar to those reported by Yendyo et al. (2017) that the best treatment outperformed the chemical control by more than 3% in reducing the disease. Plots treated with Trichoderma isolates T1, T2, and T4 had significantly lower (by more than 32%) disease incidence and severity compared to plots treated with the mixture of the same treatments at both sites. These results agree to those reported by Akrami et al. (2011) that the mixture of T. asperellum, virens, and T. harzianum performed significantly lower in reducing Fusarium oxysporum disease severity compared to the same isolates applied individually. According to Xu et al. (2010), application of Trichoderma isolates as a mixture gave poor results in reducing disease incidence and severity of Botrytis cinerea compared to the same isolates applied individually. This is because each Trichoderma isolate produces antimicrobial compounds that might inhibit any BCA that they are applied in combination with, hence, may cause interference among the BCAs leading to the observed results. These findings support the results observed in the present study. However, further studies are needed to be carried out to test possible synergistic and antagonistic effects of the mixture of Trichoderma isolates.
Bacterial wilt disease incidence and severity progressed significantly faster at Mwea than Kabete site during the experiments. The mean difference between the 2 sites showed that Mwea site had higher disease incidence and severity by more than (7.26 and 8.30%), respectively than at Kabete site. This is because Mwea site located in AEZ II had higher temperatures than at Kabete site located at AEZ III. Higher maximum temperatures that ranged from 26 to 31.3°C were recorded at Mwea site and those of Kabete site ranged from 20.41 to 26°C. High temperatures and moisture favor the growth of R. solanacearum. This explains the significant difference in the results observed in the two sites. These findings are similar to those reported by Yendyo et al. (2017).

Effect of Bacillus and Trichoderma isolates on Ralstonia solanacearum population in the soil
At the onset of the experiments, the amount of R. solanacearum inoculum in the soil was an average of 9.08 × 10 6 CFU/ml at Kabete and 1.05 × 10 8 CFU/ml at Mwea sites. This level of inoculum was able to cause 23.30 and 46.67% disease incidence on control plots at Kabete and Mwea sites, respectively 60 days after transplanting. These findings confirm those reported by Pradhanang  (2003) who found that R. solanacearum initial inoculum containing 4.5 × 10 6 CFU/ml was able to cause 100% bacterial wilt incidence in control plots at the end of the experiment. However, at 60 days after treatment application, all plots treated with Bacillus, Trichoderma, and standard chemical showed a significant reduction in R. solanacearum population in the soil at P ≤ 0.05 than in the control at both Kabete and Mwea sites (Table 3). Just like in bacterial wilt incidence and severity reduction, plots treated with Bacillus isolate CB64 and Trichoderma isolate T1 gave the highest percentage reduction of the pathogen population in the soil of 93.17 and 92%, respectively at Kabete site. At Mwea site, the reduction was 92% by Trichoderma T1 and 88.78% by CB64. Towards the end of the experiment, after 112 days of treatment application, all the treated plots gave more than 50% reduction of R. solanacearum population in the soil. Control plots gave 20% increase of the pathogen population at both sites. The results herein are similar to those reported by Sharma and Kumar (2009) that plots treated with Trichoderma viride reduced R. solanacearum population in the soil by 29% at 90 days of treatment application. Similar studies of Rosyidah et al. (2013) have reported that soil application of Trichoderma strains significantly reduced the population of R. solanacearum in the soil. The pathogen reduction in the soil may be attributed to Trichoderma strains that have the ability to adapt in extreme soil conditions. This adaptation allows Trichoderma spp. to colonize the soils, outcompete, and suppress the pathogen. According to studies by Sharma and Kumar (2009) strains of Trichoderma produces antibiotics like dermadine and gliotoxins that are able to suppress the development of bacterial pathogens such as R. solanacearum.
The antagonism between bacterial inoculum of Bacillus strains and R. solanacearum that occurs in the rhizosphere of tomato plant explains why there was reduction of the pathogen population in plots treated with Bacillus isolates. Comparable findings have reported more than 64% reduction of the pathogen population, when the soils were treated by Bacillus fortified organic fertilizer (Wei et al., 2011). Also, studies of Huang et al. (2014) reported that application of Bacillus amyloliquefaciens strongly reduced R. solanacearum population in the soil rhizosphere.
Plots treated with Trichoderma isolate T1 and Bacillus CB64 recorded more than 25% reduction of R. solanacearum population in the soil than the chemical standard at both sites. These results are in line with Sharma and Kumar (2009) findings that soil drench with chemical treatment reduced bacteria pathogen population in the soil by only 36.2% at 90 days. Plots treated with the mixture of T1, T2, and T4 gave an average increase of R. solanacearum population of 2% at both Kabete and Mwea sites. This mixture of Trichoderma isolates performed poorly than the same isolates applied as individual treatments. These results are comparable to Guo et al. (2004) findings who reported that mixing strains of Trichoderma provided significantly lower suppression of bacterial wilt disease in the soil than the same isolates applied individually.
The trend of the results observed in the performance of all the treatments in reducing the pathogen population in the soil was in harmony with the findings recorded for the disease incidence and severity reduction. This was especially in plots treated with Trichoderma isolate T1 and Bacillus isolate CB64 that performed best in all the parameters. Plots that had a high percentage reduction of R. solanacearum population in the soil also had low bacterial wilt incidence and severity. This is because there is a positive and significant correlation between the disease incidence and the pathogen population in the soil. These findings are in agreement to those reported earlier (Wei et al., 2011) that plots treated with Bacillus-fortified organic fertilizer had significantly reduced R. solanacearum populations in the soil and hence the plots had decreased disease incidence compared to the control. Cow manure and sorghum grains were used as carriers for Bacillus and Trichoderma isolates, respectively. These carriers are known to provide beneficial strains in the soil with abundant nutrients that aid in their survival, production of antibiotics, and in suppressing soil pathogens such as R. solanacearum. This could explain why isolates of Bacillus and Trichoderma significantly reduced the pathogen population in the soil compared to control. These findings are comparable to those reported by Tan et al. (2013) that use of organic substrates like manure, rice husks, and sorghum or rice grains as carriers for BCAs can provide nutrients for the desired antagonists. This, therefore, increases their opportunity for establishment in the soil, which then increases their effectiveness.
Effect of Bacillus and Trichoderma isolates on tomato yields and fruit size All treatments had significantly higher yields and fruit sizes than the control at P ≤ 0.05. The average yields observed in plots treated with Bacillus isolate CB64 were 231.76 and 266.67% more than the control at Kabete and Mwea site, respectively. Seleim et al. (2011) reported similar findings that Bacillus subtilis and Pseudomonas fluorescence achieved 91 and 348% increase of tomato yields per plant, respectively than the control. These yield results showed that the treatment with Bacillus isolate CB64 had the highest yield than the control. Comparable findings of Bacillus strains ability to promote growth and yields of tomato plants have been reported (Lemessa &Zeller, 2007 andYendyo et al., 2017).
In Trichoderma T1-treated plots, the average yields observed were 345.45 and 153.53% more than the control at Kabete and Mwea site, respectively. These results agree with that of Sharma and Kumar (2009) who reported a 142.1% increase of tomato yield than the control after application of Trichoderma viride in the soil. Kumar (2017) and Narasimhamurthy et al. (2018) also confirmed the results given herein. Not only were the fruit yields significantly higher, the fruit sizes in plots treated with CB64 and T1 were significantly higher by more than 62.60% than the control (27.46 mm) at both sites. Strains of Bacillus and Trichoderma are known to produce plant growth hormones that are responsible for increased yields and fruit size (Seleim et al. 2011 andSingh et al., 2012). Isolate CB64 and T1 also outperformed the standard chemical in yields and fruit sizes. The mixture of Trichoderma isolates was among the treatments that recorded the least yields and fruit size than the same isolates applied individually.

Conclusion
Bacillus isolates CB64 and Trichoderma isolate T1 had the highest percentages of reduction in bacterial wilt incidence and severity in the field. The isolates were also able to reduce the population of R. solanacearum in the soil by more than 90% than the control. Therefore, they are recommended for use as biological control agents in management of bacterial wilt disease caused by R. solanacearum of tomato. As well, cow manure and sorghum grains can be used as carriers for Bacillus and Trichoderma isolates, respectively.