Resistance of tomato genotypes to Ralstonia solanacearum in Ghana

1 Introduction

Tomato (Solanum lycopersicum L.) is one of the most widely cultivated vegetable crops globally and a key source of income and nutrition for smallholder farmers, particularly in sub-Saharan Africa. In Ghana, tomato production plays a critical role in food security and rural livelihoods; however, productivity remains relatively low. National average yields are about 7.85 t ha⁻¹ (Ghana Export-Import Bank, 2022; FAOSTAT, 2024) far below the global average of 37 t ha⁻¹ (FAOSTAT, 2024). Although yields vary across production systems (6.9–16.2 t ha⁻¹) (Melomey et al., 2022) this yield gap reflects multiple production constraints, including erratic rainfall, high temperatures, declining soil fertility, and the prevalence of economically important diseases (Melomey et al., 2019; Ayenan et al., 2021; Melomey et al., 2022).

Among these constraints, bacterial wilt caused by the Ralstonia solanacearum species complex (RSSC) is one of the most destructive diseases affecting tomato production. The pathogen is soil-borne and highly aggressive, entering plants through root wounds, colonizing the vascular system, and ultimately causing wilting and plant death. Under favorable environmental conditions, yield losses can reach up to 100% (Mansfield et al., 2012; Lowe-Power et al., 2018; Phiri et al., 2024). Since its first report in Ghana in 2012 (Subedi et al., 2014), the disease has become widespread across major tomato-growing regions. It is particularly severe in northern Ghana, including the Upper East Region, where high disease incidence and substantial yield losses are frequently observed.

Effective management of bacterial wilt is complicated by the extensive genetic and phenotypic diversity of RSSC. Traditionally classified into races and biovars based on host range and biochemical traits, the pathogen is now grouped into phylotypes linked to geographical origin (Fegan and Prior, 2005; Safni et al., 2014; Prior et al., 2016). In West Africa, including Ghana, RSSC populations are dominated by phylotypes I, II, and III, and biovars II and III (Adebayo and Ekpo, 2005; Mahbou Somo Toukam et al., 2009; Popoola et al., 2015; Sikirou et al., 2017; Shutt et al., 2018). This diversity has important implications for disease management, as strains vary in virulence, host range, and environmental adaptation, often resulting in inconsistent performance of resistant varieties (Genin and Denny, 2012; Lowe-Power et al., 2018; Hamilton et al., 2023).

Management of bacterial wilt using cultural practices and chemical control methods remain largely ineffective due to the pathogens’ persistence in soil, water, plant residues, and alternative hosts, as well as its broad host range (Genin and Denny, 2012; Lowe-Power et al., 2018; N’Guessan et al., 2012; Planas-Marquès et al., 2020). Consequently, the deployment of resistant cultivars is widely regarded as the most practical, cost-effective, and environmentally sustainable control strategy. Under Ghanaian conditions, most locally cultivated tomato genotypes are highly susceptible, and only a few have shown stable resistance under field and experimental conditions (B-Mensah et al., 2021; Adomako et al., 2024).

Although several studies have conducted resistance screening in controlled conditions, the expression of resistance is influenced by both pathogen diversity and environmental conditions. Greenhouse evaluations are valuable for preliminary screening of large numbers of genotypes under controlled conditions, but they may not fully capture the complexity of field conditions where multiple biotic and abiotic factors interact. As such, resistance identified under greenhouse conditions should be considered indicative and requires validation under field conditions.

Despite the increasing importance of bacterial wilt in Ghana, there is limited information on the diversity and distribution of RSSC populations in key production zones such as the Upper East Region, as well as the response of locally adapted tomato genotypes to these populations (B-Mensah et al., 2021; Sarfo et al., 2025). Addressing this gap is essential for the identification of durable sources of resistance and the improvement of breeding and disease management strategies under local conditions. Therefore, this study aimed to (i) isolate and characterize Ralstonia solanacearum strains associated with bacterial wilt of tomato in the Upper East Region of Ghana, and (ii) evaluate the response of selected tomato genotypes to locally prevalent strains under greenhouse conditions, with the goal of screening for promising sources of resistance for breeding and disease management.

2 Materials and methods

2.1 Study area

The study was carried out at the Biotechnology Centre and the greenhouse facility of the West Africa Centre for Crop Improvement (WACCI), University of Ghana, Legon–Accra. The site lies at 5°39′ N latitude and 0°10′ W longitude, at an altitude of approximately 97 m above sea level, within the coastal savannah agro-ecological zone of Ghana. The greenhouse provided controlled environmental conditions suitable for evaluating tomato responses to Ralstonia solanacearum. During the experimental period, temperatures in the greenhouse ranged from 30–35 °C (maximum) to 15–20 °C (minimum) with relative humidity maintained at 70–85%. The climatic conditions in the greenhouse approximated typical tropical environments favorable for bacterial wilt development.

2.2 Collection of disease plant samples

Twenty (20) infected tomato plants showing typical symptoms of bacterial wilt were collected from Soogo and Teshie communities in the Upper East Region of Ghana. Tomato fields were prospected in the target regions and samples were collected from fields with symptomatic plants. Samples were collected in December 2020 from five fields per location. In each field, two symptomatic plants were selected based on visible wilting, vascular discoloration, and bacterial oozing. Stem sections were cut and placed in clear water to confirm bacterial streaming as a preliminary diagnostic test. Samples were placed in sterile bags, properly labeled, and transported to the laboratory for further isolation and characterization.

2.3 Culture media composition and preparation

Ralstonia solanacearum was cultured on triphenyltetrazolium chloride (TTC) medium following the method described by Kelman (1954). The medium was prepared to a final volume of 1000 mL using a 1000 mL flat-bottom flask. It consisted of 18 g agar, 10 g peptone, 5 g glucose, and 1 g casamino acids (casein hydrolysate) dissolved in distilled water.

The pH of the medium was adjusted to approximately 7.0 before sterilization. The prepared medium was covered with cotton wool and aluminum foil and then autoclaved at 121 °C for 15 minutes. After autoclaving, the medium was allowed to cool to about 50–60 °C. A 1% TTC solution was prepared separately using distilled water and sterilized by passing it through a 0.2 µm membrane filter. Five milliliters (5 mL) of the sterile TTC solution were then aseptically added to the cooled medium and mixed gently to ensure even distribution. The medium was subsequently poured into sterile Petri dishes under aseptic conditions and allowed to solidify at room temperature for use. Prepared plates were stored at 4 °C and used within 3–5 days for bacterial isolation.

2.4 Isolation of Ralstonia solanacearum

Stem sections (approximately 10 cm long) were taken from the collar region of wilted tomato plants. The stem pieces were briefly surface sterilized using 1% Clorox solution prepared with sterile distilled water and then rinsed thoroughly to remove any residual disinfectant. The sterilized stem sections were cut and placed in 2 mL Eppendorf tubes containing sterile distilled water to observe bacterial streaming for about 10 minutes. Turbid suspensions indicating the presence of bacteria were collected. A loopful of the turbid suspension was streaked onto triphenyltetrazolium chloride (TTC) medium and incubated at 30 °C for 48 hours. Typical virulent colonies of Ralstonia solanacearum, characterized by fluidal, irregular, creamy white colonies with pink centers, were observed on the TTC plates. Single colonies were picked and subcultured onto fresh TTC medium to obtain pure cultures. This process was repeated twice to ensure uniformity in colony morphology and purity of the isolates. TTC medium was used for both isolation and maintenance of the bacteria. Pure isolates were preserved by mixing equal volumes of bacterial suspension and 50% (v/v) sterile glycerol solution, resulting in a final glycerol concentration of 25% (v/v), and stored at −20 °C for subsequent use.

2.5 Biovar differentiation

Biovar differentiation of Ralstonia solanacearum isolates was carried out based on their ability to oxidize selected disaccharides and sugar alcohols following the classification scheme described by Hayward (1964, 1991). The substrates tested included sucrose, lactose, maltose, mannitol, sorbitol, cellobiose and dulcitol. A basal mineral medium was prepared containing 1.0 g NH₄H₂PO₄, 0.2 g KCl, 0.2 g MgSO₄·7H₂O, 1.0 g Difco bacto peptone, 3.0 g agar, and 80 mg L⁻¹ bromothymol blue as a pH indicator. The pH of the medium was adjusted to 7.2 prior to sterilization. After sterilization and cooling, 10% (w/v) solutions of each carbohydrate substrate were prepared separately and incorporated into the medium. Approximately 200 µL of the medium supplemented with each substrate was dispensed into wells of sterile microtitre plates. Bacterial suspensions were prepared from pure cultures and adjusted to approximately 1 × 10⁸ CFU mL⁻¹. Each well was inoculated with 20 µL of the bacterial suspension. The inoculated plates were incubated at 29 °C under static conditions and observed at 3, 7, 14, and 28 days for changes in pH, indicated by a color shift of the medium from olive-green to yellow. A color change from olive-green to yellow was recorded as a positive reaction, indicating oxidation of the substrate, whereas no color change was considered a negative reaction (Schaad, 1988). Biovar classification was subsequently determined based on patterns of carbohydrate utilization (Table 1). Positive reactions were indicated by a color change, while negative reactions showed no observable change in the medium.

Table 1. Biovar differentiation of Ralstonia solanacearum isolates based on carbohydrate oxidation patterns.

+, Positive reaction (color change from olive-green to yellow); −, negative reaction (no color change). Adapted from Hayward (1964, 1991).

2.6 Extraction of DNA from Ralstonia solanacearum

Genomic DNA was extracted using a simple boiling lysis method for rapid preparation of bacterial genomic DNA as described by Chen and Kuo (1993). A single colony of R. solanacearum from each isolate was inoculated into casein hydrolysate, peptone, and glucose (CPG) broth and incubated overnight at 28 °C with constant shaking at 100 rpm. One milliliter (1 mL) of each bacterial culture was transferred into a 2 mL microcentrifuge tube and centrifuged at 12,000 × g for 3 minutes at 4 °C. The supernatant was discarded, and the pellet was washed twice with 500 µL of 1× TE buffer (Tris–EDTA, pH 8.0), followed by centrifugation at 12,000 × g for 3 minutes at 4 °C after each wash. The washed pellet was then resuspended in 200 µL of sterile distilled water and boiled at 100 °C for 8 minutes to lyse the cells. The lysate was immediately cooled on ice and centrifuged at 12,000 × g for 5 minutes to remove cell debris. The supernatant containing genomic DNA was carefully collected and stored at −20 °C until further use.

2.7 Molecular confirmation of Ralstonia solanacearum

Molecular confirmation of Ralstonia solanacearum isolates was carried out using polymerase chain reaction (PCR) with species-specific primers 759F (5′-GTC GCC GTC AAC TCA CTT TCC-3′) and 760R (5′-GTC GCC GTC AGC AAT GCG GAA TCG-3′), which amplify a 280 bp fragment, as described by Opina et al. (1997) and further applied by Mihovilovich et al. (2017) (Table 2). PCR reactions were performed in a total volume of 25 µL containing 4 µL of template DNA, 1× Taq Master Mix (containing PCR buffer, 1.5 mM MgCl₂, 200 µM of each dNTP, 50 mM KCl, 10 mM Tris-HCl, and 1.25 U Taq DNA polymerase), and 1 µL each of forward and reverse primers (20 pmol). The reaction volume was adjusted to 25 µL with sterile nuclease-free water. Amplification was carried out in a thermal cycler using optimized PCR cycling conditions for successful amplification of the expected 280 bp fragment as follows: initial denaturation at 96 °C for 2 minutes, followed by 35 cycles of denaturation at 94 °C for 20 seconds, annealing at 58 °C for 20 seconds, and extension at 72 °C for 20 seconds, with a final extension at 72 °C for 10 minutes. PCR products were separated on a 1.5% agarose gel prepared in 1× TAE buffer, stained with ethidium bromide, and electrophoresed at 100 V for approximately 60 minutes. A 100 bp DNA ladder was used as a molecular size marker. Bands were visualized under UV transillumination.

Table 2. PCR primers used for phylotype differentiation of Ralstonia solanacearum.

Based on primers described by Fegan and Prior (2005) and reported in Mihovilovich et al. (2017).

2.8 Phylotype differentiation using PCR

Phylotype identification of Ralstonia solanacearum isolates was carried out using PCR phylotype-specific primers as described by Fegan and Prior (2005). Separate PCR reactions were performed for each primer set targeting individual phylotypes. The primers used for differentiation of phylotypes I–IV, along with their expected band sizes, are presented in Table 2. PCR reactions were prepared as described in Section 2.7. Amplification was carried out in a thermal cycler under the following conditions: initial denaturation at 94 °C for 3 minutes, followed by 30 cycles of denaturation at 94 °C for 30 seconds, annealing at 58 °C for 30 seconds, and extension at 72 °C for 30 seconds, with a final extension at 72 °C for 10 minutes. PCR products were separated on a 1.5% agarose gel prepared in 1× TAE buffer, stained with ethidium bromide, and electrophoresed at 100 V for approximately 60 minutes. A 100 bp DNA ladder was used as a molecular size marker. Bands were visualized under UV transillumination. Phylotype identification was based on the presence of specific band sizes corresponding to each primer set (Table 2).

2.9 Test for pathogenicity

Isolates obtained from the two communities were cultured on TTC medium and incubated at 30 °C for 24 hours. Bacterial colonies were harvested into sterile distilled water to prepare suspension cultures. The bacterial suspensions were adjusted to an optical density (OD₆₀₀) of 0.3 using a spectrophotometer, corresponding to approximately 1 × 10⁸ colony-forming units (CFU mL⁻¹) (Lin et al., 2014). Thirty-day-old seedlings of the bacterial wilt–susceptible tomato variety ‘Petomech’, raised in sterile growth media under screen house conditions, were used for inoculation. The roots of the seedlings were gently wounded using a sterile razor blade and submerged in the bacterial suspension for 20 minutes prior to transplanting into pots in the greenhouse, following the method described by Kim et al. (2016) with slight modifications. Control plants were treated similarly but submerged in sterile distilled water. Each isolate was inoculated onto 20 seedlings of ‘Petomech’, a known susceptible genotype, replicated 3 times in a greenhouse, and the plants were monitored daily for symptom development.

2.10 Greenhouse experiment

Greenhouse experiments were conducted to evaluate the response of tomato genotypes to two Ralstonia solanacarum strains identified from Teshie and Soogo in Ghana. The experiment was carried out under controlled greenhouse conditions, with temperatures ranging from approximately 30–35 °C during the day and 15–20 °C at night. A 2 (bacterial strains) × 15 (tomato genotypes) (Table 3) factorial treatment design was arranged in a randomized complete block design (RCBD) with three replications, where each replication (block) represented a spatial unit within the greenhouse to account for environmental variability. Each experimental unit consisted of 10 plants. Seeds were sown in 72-cell styrofoam trays filled with cocopeat and maintained for four weeks before transplanting. Seedlings were transplanted into plastic buckets measuring approximately 30 cm in diameter and 40 cm in height, filled with a sterilized potting medium composed of topsoil, rice husk, and poultry droppings in a ratio of 3:1:1.

Table 3. Tomato genotypes evaluated for bacterial wilt resistance, with Petomech as the susceptible check.

At transplanting, the two bacterial strains were used for inoculation. The strains were selected based on their frequency in the source populations and their virulence as determined by the pathogenicity test. Four-week-old seedlings were inoculated using an adapted root-wounding and root-immersion method with slight modifications from Kim et al. (2016). Roots were gently wounded and immersed in 20 mL of bacterial suspension adjusted to approximately 1 × 10⁸ CFU mL⁻¹ for 20 minutes, after which the seedlings were transplanted into the prepared containers. To support plant growth, NPK (15:15:15) fertilizer was applied at a rate of approximately 3 g per plant one week after transplanting and subsequently at weekly intervals throughout the experimental period. Whiteflies were monitored and managed using standard greenhouse practices. Yellow sticky traps were installed within the greenhouse and inspected weekly, and sanitation measures, including weed removal, were maintained to minimize pest pressure.

2.11 Assessment of plants for disease development

Plants were monitored regularly for disease development. Survival assessment was based on visual evaluation of plant condition, where plants that remained turgid and free from permanent wilting symptoms were considered alive, while completely wilted plants were considered dead. Plant survival was assessed as the percentage of surviving plants per experimental unit. Observations were first recorded at 3 days after inoculation (DAI) and subsequently at 7-day intervals until complete and permanent wilting of susceptible plants. Final percentage plant survival was calculated for each treatment.

At each observation time point, the number of surviving plants was recorded and expressed as a percentage of the total number of plants per experimental unit using the formula:

Plant survival (%)=Number of surviving plantsTotal number of plantsX 100

The final percentage plant survival per genotype was used for statistical analysis, consistent with standard screening approaches for evaluating resistance to bacterial wilt in tomato (Winstead and Kelman, 1952; Chellemi et al., 1994).

2.12 Disease assessment and progress quantification

Disease development was assessed by recording plant survival at each observation time and expressed as a percentage of the initial plant population. Disease incidence was derived as the inverse of survival using the relationship:

Disease Incidence (%)=100−Survival(%)

This approach is based on the definition of disease incidence as the proportion of plant units affected within a population, expressed as a percentage (Bock et al., 2021; Chiang and Bock, 2021). Similar percentage-based disease assessment approaches using wilted plant incidence have previously been reported in bacterial wilt studies (Hong Hai et al., 2008). Disease progression over time was quantified using the area under the disease progress curve (AUDPC), calculated using the trapezoidal method as described by Shaner and Finney (1977) and Campbell and Madden (1990), which was used in bacterial wilt studies by Hong Hai et al. (2008):

AUDPC=∑(yi+yi+1)2(ti+1−ti)

where y_i is the disease incidence at time t_i, and t_i+1 − t_i represents the time interval between successive observations.

2.13 Statistical analysis

Data on percentage plant survival at 42 days after inoculation were subjected to analysis of variance (ANOVA) using a 2 × 15 factorial experimental design to evaluate the effects of bacterial strain, tomato genotype, and their interaction. Prior to analysis, percentage data were arcsine square-root transformed to stabilize variance. Mean separation was performed using Tukey’s honestly significant difference (HSD) test at P ≤ 0.05 where significant differences were observed, and treatment effects were interpreted based on the final percentage plant survival of each genotype. The statistical model was fitted to the data following procedures described by Gomez and Gomez (1984):

Yijk= μ+Si+Ej+(S×E)ij+Rk+ϵijk

where Y_ijk is the observed response variable, including final percentage plant survival measured at 42 days after inoculation (DAI) or AUDPC calculated from disease assessments recorded between 3 and 42 DAI, μ is the overall mean, S_i is the effect of strain, E_j is the effect of genotype, (S × E)_ij is the interaction effect, R_k is the replication effect, and ϵ_ijk is the residual error. The reaction of tomato genotypes to Ralstonia solanacearum infection was classified according to the scale described by Sharma et al. (2006) as follows: susceptible (0–29% plant survival), moderately susceptible (30–59%), moderately resistant (60–79%), and resistant (80–100%). All statistical analyses were conducted using R statistical software (R version 4.6.0; R Core Team, 2026), and significance was determined at P ≤ 0.05.

3 Results

3.1 Biovar characterization

All putative bacterial wilt isolates were obtained from symptomatic tomato plants and initially confirmed using ooze diagnostic tests. The isolates produced characteristic colonies of Ralstonia solanacearum on TTC medium after 48 hours of incubation, appearing large, irregular, fluidal, and mucoid with white to pale red centers, typical of virulent strains (Supplementary Figure 1). Biovar differentiation based on the oxidation of disaccharides and sugar alcohols showed that all twenty isolates from Teshie and Soogo were able to oxidize maltose, lactose, cellobiose, mannitol, sorbitol, and dulcitol. Positive reactions were indicated by a color change from olive-green to yellow in the medium, as shown in Supplementary Figure 2. This pattern of carbohydrate utilization corresponds to biovar III according to the classification scheme presented in Table 1. Furthermore, all isolates were pathogenic on the susceptible tomato variety ‘Petomech’ and produced typical bacterial wilt symptoms within three days of inoculation. By 42 days after inoculation, 90–100% wilting was observed in the inoculated plants, while no symptoms were recorded in the control. Based on the combined characterization of carbohydrate utilization, pathogenicity identification, all isolates were confirmed as R. solanacearum belonging to biovar III and were all virulent (Table 4).

Table 4. Classification of R. solanacearum according to pathogenicity, phylotype, and biovar.

Positive reaction (+); Disease symptoms were realized 3 days after inoculation on 30 days old Petomech tomato seedlings.

3.2 Molecular characterization of Ralstonia solanacearum

All twenty isolates produced the expected amplification product of approximately 280 bp when screened using the Ralstonia solanacearum-specific universal primers 759/760, confirming their identity as R. solanacearum (Supplementary Figure 4). Phylotype differentiation using polymerase chain reaction (PCR) revealed variation among the isolates. None of the isolates produced the expected amplicon sizes of 372 bp and 213 bp corresponding to phylotypes II and IV, respectively. However, thirteen (65%) of the isolates generated an approximately 144 bp fragment and were identified as phylotype I (Supplementary Figure 5), while the remaining seven isolates (35%) produced a 91 bp fragment and were classified as phylotype III (Supplementary Figure 6). The distribution of phylotypes across the isolates from Teshie and Soogo is summarized in Table 4, where both phylotype I and phylotype III were detected, with phylotype I being predominant.

3.3 Assessment of tomato genotypes for final plant survival

Analysis of variance revealed a highly significant effect of genotype on disease response (P < 0.001), indicating substantial variation among tomato genotypes (Table 5A). In contrast, neither strain (P = 0.404) nor the strain × genotype interaction (P = 0.275) had a significant effect, suggesting that genotype performance was consistent across both Ralstonia solanacearum strains. To further examine variation among tomato genotypes, mean plant survival values were grouped according to resistance classes under each strain and when averaged across strains (Table 6). Under P1B3, susceptible genotypes (Petomech and Lebombo) recorded very low survival, ranging from 0.00–6.67%. Moderately resistant genotypes showed a wide range of survival values between 56.67% and 73.33%, including CLN4018C, CLN4018B, CLN4018D, Manica, CRA66 Prior, F7-80-465-10-pink, CLN4018A, Kwando, and Hawaii 7996. Resistant genotypes, represented by TmL114-42-N-H.T.P early and CRA66 WorldVeg, recorded higher survival values, although these were not statistically distinct from other moderately resistant genotypes (Table 6). Mean separation using Tukey’s HSD test under P1B3 indicated partial differentiation among genotypes, with Petomech clearly separated as highly susceptible, while most of the remaining genotypes showed overlapping groupings and were not significantly different from one another (Table 6). Under P3B3, susceptible genotypes again recorded low survival values ranging from 0.00–3.33%, confirming consistent susceptibility. Moderately resistant genotypes showed a wide range of survival values between 56.67% and 73.33%; however, most of these genotypes were not significantly different from one another based on Tukey’s HSD test. Resistant responses were limited, with only CRA66 WorldVeg maintaining survival at 80.00%, while TmL114-42-N-H.T.P early (73.33%) fell slightly below the resistance threshold under this strain (Table 6).

Table 5. Analysis of variance for plant survival at 42 days after inoculation (42 DAI) (A) and area under the disease progress curve (AUDPC) (B) of tomato genotypes evaluated against Ralstonia solanacearum strains.

P-values are based on analysis of variance (ANOVA); values< 0.05 indicate statistical significance. Plant survival data (42 DAI) were arcsine square-root transformed prior to analysis.

Table 6. Mean plant survival (%) at 42 days after inoculation (42 DAI) for tomato genotypes evaluated under two Ralstonia solanacearum strains (P1B3 and P3B3), including combined means across strains.

Values are presented as mean percentage plant survival, with arcsine square-root transformed values shown in parentheses. Means within a column followed by the same letter are not significantly different according to Tukey’s honestly significant difference (HSD) test at p ≤ 0.05. Reaction classes were assigned based on mean survival as follows: Resistant (≥ 80%), Moderately Resistant (60–79%), Moderately Susceptible (30–59%), and Susceptible (< 30%).

When averaged across both strains, susceptible genotypes maintained survival between 0.00% and 5.00%, while moderately resistant genotypes ranged from 60.00% to 76.67%, indicating relatively stable performance across strains. Resistant genotypes showed survival values between 80.00% and 83.33% but were not statistically distinct from moderately resistant genotypes (Table 6). Mean separation based on combined data showed that Petomech and Lebombo were significantly different from the remaining 13 genotypes, which were grouped within the same statistical category, indicating that differences among the moderately resistant and resistant genotypes were not statistically significant. Overall, the classification ranges indicate that most genotypes clustered within the moderately resistant category across both strains, with relatively narrow variation at the susceptible and resistant extremes. The consistency of survival ranges and the absence of a significant strain × genotype interaction further indicate that genotype effects dominated over strain effects in determining host response.

3.4 Area under the disease progress curve

Significant differences in AUDPC were observed among tomato genotypes for both Ralstonia solanacearum strains (Table 5B), confirming substantial variation in disease progression and resistance responses (Figures 1, 2, 3). However, the interaction between genotype and strain was not significant, indicating consistent genotype performance across strains. Under strain P1B3, ‘Petomech’ and ‘Lebombo’ recorded the highest AUDPC values and were grouped together within the same statistical category, indicating rapid disease progression and high susceptibility (Figure 1). In contrast, ‘CRA66 WorldVeg’ and ‘TmL114-42-N-H.T.P early’ recorded the lowest AUDPC values and were assigned to the lowest statistical group, reflecting strong resistance. ‘LE 415 Anagha’ also clustered within this low-AUDPC group and did not differ significantly from the most resistant genotypes. Intermediate AUDPC values were observed for ‘CLN4018A’, ‘Kwando’, and ‘Hawaii 7996’, which formed overlapping statistical groups between resistant and susceptible extremes. Other genotypes, including ‘CLN4018B’, ‘CLN4018C’, ‘CLN4018D’, ‘Manica’, ‘CRA66 Prior’, and ‘F7-80-465-10-pink’, occupied intermediate-to-high AUDPC groupings, indicating moderate resistance.

Under strain P3B3, a similar overall ranking was observed; however, notable shifts in statistical groupings occurred (Figure 2). ‘Petomech’ and ‘Lebombo’ again recorded the highest AUDPC values and remained in the most susceptible group. ‘CRA66 WorldVeg’ maintained the lowest AUDPC values and remained statistically distinct from susceptible genotypes. ‘TmL114-42-N-H.T.P early’ and ‘LE 415 Anagha’ also exhibited low AUDPC values but showed slight overlap with intermediate groups. Notably, ‘CLN4018D’ and ‘CRA66 Prior’ shifted to lower AUDPC groupings compared to P1B3, indicating improved performance under P3B3. In contrast, ‘BL989’ shifted toward higher AUDPC values and grouped with more susceptible genotypes, reflecting increased disease progression under this strain. Genotypes such as ‘CLN4018A’, ‘Kwando’, ‘Hawaii 7996’, ‘CLN4018B’, and ‘CLN4018C’ remained within intermediate statistical groupings, showing moderate and relatively stable responses across strains.

When averaged across both strains, clear and consistent separation among genotypes was maintained (Figure 3). ‘CRA66 WorldVeg’ and ‘TmL114-42-N-H.T.P early’ consistently recorded the lowest AUDPC values and were grouped within the most resistant statistical category. ‘Petomech’ and ‘Lebombo’ exhibited the highest AUDPC values and remained statistically distinct as the most susceptible genotypes. The remaining genotypes were distributed across intermediate statistical groups, with overlapping Tukey groupings indicating gradual transitions in resistance levels rather than sharp boundaries. Genotypes such as ‘Hawaii 7996’, ‘Kwando’, and ‘CLN4018A’ consistently occupied intermediate-low AUDPC groupings, while ‘CLN4018B’, ‘CLN4018C’, and ‘BL989’ showed relatively higher AUDPC values within the intermediate range.

Disease progression following inoculation with Ralstonia solanacearum strains P1B3 and P3B3 resulted in a consistent decline in plant survival across all tomato genotypes, with clear differences in the timing, rate, and extent of decline (Figures 4, 5), while Figure 6 depicts the survival dynamics of the genotypes across both strains. Wilting symptoms were first observed at 3 days after inoculation (DAI) on the susceptible control, after which survival trajectories diverged distinctly among genotypes. Supplementary Figure 3 shows representative susceptible and resistant genotype responses at 28 DAI.

Under strain P1B3, ‘Petomech’ and ‘Lebombo’ exhibited the most rapid decline in survival, dropping sharply shortly after symptom onset and reaching complete wilting earlier than all other genotypes (Figure 4). In contrast, ‘CRA66 WorldVeg’ and ‘TmL114-42-N-H.T.P early’ maintained consistently high survival throughout the assessment period, with only a slight and gradual reduction. ‘LE 415 Anagha’ and ‘BL989’ also sustained relatively high survival over time, showing slower disease progression compared to most genotypes. Intermediate responses were observed for ‘Hawaii 7996’, ‘Kwando’, and ‘CLN4018A’, which displayed gradual but continuous declines. In contrast, ‘CLN4018B’, ‘CLN4018C’, ‘CLN4018D’, ‘Manica’, and ‘CRA66 Prior’ showed steeper declines than the intermediate group, indicating greater susceptibility under P1B3. Differences among genotypes were already evident early in the disease progression period, with susceptible genotypes separating rapidly from the more tolerant group.

Under strain P3B3, a similar overall pattern was observed, but with notable genotype-specific differences in survival trajectories (Figure 5). ‘Petomech’ and ‘Lebombo’ again showed the fastest and most pronounced decline, confirming their high susceptibility. ‘CRA66 WorldVeg’ and ‘TmL114-42-N-H.T.P early’ maintained the highest survival throughout the observation period, with minimal reduction over time. ‘CLN4018D’ and ‘CRA66 Prior’ showed improved survival under P3B3 compared to P1B3, maintaining higher survival for longer durations. In contrast, ‘BL989’ and ‘CLN4018B’ exhibited a steeper decline under P3B3 than under P1B3, indicating greater sensitivity to this strain. ‘Hawaii 7996’, ‘Kwando’, ‘CLN4018A’, ‘CLN4018C’, and ‘Manica’ maintained intermediate responses, with steady but moderate declines in survival over time.

When averaged across both strains, clear and consistent separation among genotypes was maintained throughout the disease progression period (Figure 6). ‘Petomech’ and ‘Lebombo’ showed rapid and continuous decline, reaching near-zero survival earlier than all other genotypes. In contrast, ‘CRA66 WorldVeg’ and ‘TmL114-42-N-H.T.P early’ maintained high survival across all observation points, showing the slowest disease progression. ‘LE 415 Anagha’, ‘Hawaii 7996’, ‘Kwando’, and ‘CLN4018A’ exhibited relatively stable survival with gradual declines, while ‘BL989’, ‘CLN4018B’, ‘CLN4018C’, ‘CLN4018D’, ‘Manica’, and ‘CRA66 Prior’ showed moderate rates of decline, falling between the resistant and susceptible extremes. Differences among genotypes were reflected not only in final survival but also in the rate of decline, with distinct curve patterns observed across genotypes.

4 Discussion

4.1 Pathogen diversity and characterization

The present study confirmed that bacterial wilt in the Upper East Region of Ghana is caused by genetically diverse populations of the Ralstonia solanacearum species complex, with all isolates identified as biovar III belonging to phylotypes I and III. The pathogenicity of all isolates on the susceptible cultivar ‘Petomech’ further confirms their virulence and is consistent with earlier reports from Ghana and West Africa, which indicate that RSSC populations are highly aggressive and capable of causing severe disease outbreaks (Subedi et al., 2014; Sarfo, 2018).

The predominance of phylotype I observed in this study is consistent with previous findings in Ghana and across West Africa. Subedi et al. (2014) reported that the majority of isolates collected from different regions in Ghana belonged to phylotype I, with only a small proportion classified as phylotype III. Similar trends have been reported in neighboring countries, where phylotype I strain dominates due to their wide host range and high genetic diversity (Mahbou Somo Toukam et al., 2009; Popoola et al., 2015; Sikirou et al., 2017; Shutt et al., 2018).

The detection of both phylotypes I and III within the same production region highlights the complexity of pathogen population structure and its implications for resistance breeding and disease management. The coexistence of multiple phylotypes within tomato production systems in Ghana has been reported previously, with disease incidence ranging from 5.9% to 18.5% under greenhouse conditions (Sarfo et al., 2025). The higher prevalence of phylotype I in the present study may be associated with its greater adaptability and genetic diversity (Chesneau et al., 2018). These findings underscore the importance of evaluating tomato genotypes against locally prevalent pathogen populations to ensure durable, stable resistance.

4.2 Resistance response and genotype performance

The significant effect of genotype observed in this study indicates that host genetic factors play a major role in determining resistance to bacterial wilt. Although neither strain nor the strain × genotype interaction was statistically significant, slight variations in genotype performance across P1B3 and P3B3 were observed. These differences, however, were not large enough to result in significant interaction effects, suggesting that most genotypes exhibited relatively stable responses across strains, with only minor strain-specific variation. The classification of tomato genotypes based on plant survival and AUDPC revealed clear differentiation among genotypes, with only a few demonstrating strong resistance across both strains, while most exhibited moderate resistance. Genotypes such as CRA66 WorldVeg and TmL114-42-N-H.T.P early consistently recorded low disease progression and high survival across both P1B3 and P3B3 strains, indicating stable resistance. In contrast, ‘Petomech’ and ‘Lebombo’ exhibited high AUDPC values and low survival, reflecting rapid disease development and high susceptibility.

This distribution of resistance levels is consistent with previous studies showing that high-level resistance to bacterial wilt in tomato is relatively rare. B-Mensah et al. (2021) reported that only a few genotypes exhibited stable resistance across environments, while most were moderately resistant. Similarly, Adomako et al. (2024) observed significant variation in disease incidence and severity among Solanum accessions, with most tomato genotypes showing susceptibility. Comparable trends have been reported in broader germplasm evaluations, where resistance is often quantitative and only a small proportion of accessions exhibit stable resistance (Kim et al., 2016; Carmeille et al., 2006; Dossoumou et al., 2021).

The moderate performance of ‘Hawaii 7996’, a widely recognized source of bacterial wilt resistance, further supports the quantitative nature of resistance. Lebeau et al. (2011) demonstrated that Hawaii 7996 possesses broad but incomplete resistance, while N’Guessan et al. (2012) reported that certain virulent strains can overcome its resistance. Wang et al. (2013) further showed that resistance in this genotype is controlled by quantitative trait loci (QTLs), including Bwr-6 and Bwr-12. The response observed in the present study suggests that resistance expression is influenced by both host genetics and pathogen diversity, consistent with the complex genetic control of bacterial wilt resistance reported in tomato (Acharya et al., 2018b).

4.3 Strain-specific interactions and resistance stability

Variation in genotype performance between P1B3 and P3B3 strains observed in this study indicates the influence of strain-specific interactions on resistance expression. While some genotypes maintained consistent performance across both strains, others exhibited differential responses, suggesting that resistance mechanisms may be partially strain dependent. This observation is supported by Planas-Marquès et al. (2020), who demonstrated that resistance to R. solanacearum involves multiple infection bottlenecks, including root colonization and vascular spread. Similarly, Chachar et al. (2025) highlighted the role of pathogen effectors in overcoming host defenses, which may contribute to variation in disease severity among genotypes. The variability observed among moderately resistant genotypes in this study is therefore likely associated with differences in pathogen virulence and host defense responses.

The susceptibility of ‘Lebombo’, despite being regarded as a resistant commercial variety, underscores the significance of local validation of resistance. Similar observations have been reported in Ghana, where commonly cultivated varieties such as ‘Petomech’ and ‘Roma VF’ have been found to be highly susceptible under local conditions (Subedi et al., 2014; Sarfo, 2018). This suggests that resistance identified in one region may not be effective against different pathogen populations, particularly in areas with diverse RSSC strains.

4.4 Limitations and implications for breeding

Although this study provides important insights into pathogen diversity and host resistance, it was conducted under controlled greenhouse conditions, which may not fully represent field environments where multiple biotic and abiotic factors interact. Previous studies have noted that resistance responses observed under controlled conditions may differ under field conditions due to environmental variability and pathogen dynamics (Lebeau et al., 2011; B-Mensah et al., 2021).

In addition, resistance was evaluated using two bacterial strains isolated from the major disease hotspots identified during the study period. These strains represent the most prevalent and virulent populations in the target production areas at the time of sampling, thereby ensuring that the resistance screening reflects realistic disease pressure under local conditions. However, they may not fully capture the broader diversity of Ralstonia solanacearum populations across regions or seasons, as RSSC populations are known to exhibit considerable genetic variability (Chesneau et al., 2018; Sossou et al., 2024).

Future research should incorporate multi-location field trials, and a broader range of pathogen isolates to further evaluate the stability of resistance observed in this study. Such approaches have been suggested as important for identifying durable resistance across diverse environments and pathogen populations (N’Guessan et al., 2012). Despite these limitations, the consistent performance of CRA66 WorldVeg and TmL114-42-N-H.T.P early across strains highlights their potential as valuable genetic resources for breeding programs aimed at improving bacterial wilt resistance in tomato under Ghanaian conditions. Notably, TmL114-42-N-H.T.P early carries both the Bwr-12 and Bwr-6 resistance QTLs (Kunwar et al., 2020), which likely underpin its stable performance across the phylotype I and III strains tested. The genetic basis of CRA66 resistance remains less defined and merits further investigation, including molecular characterization of potential novel resistance loci.

Integrating these donor lines into active breeding pipelines is critical for introgressing robust resistance alleles while maintaining the elite genetic background required for competitive yield and fruit quality (Acharya et al., 2018a). While variety recommendation will necessitate multi-environment trials across the disease’s hotspots to validate yield stability and the durability of resistance to RSSC, breeding efficiency can be optimized by strategically identifying representative Target Population of Environments (TPEs). By leveraging environmental characterization, including knowledge on RSSC strain diversity and growing conditions, to subset these representative sites, tomato breeding programs can significantly reduce resource allocation to phenotyping.

5 Conclusion

In this study, we confirmed the presence of genetically diverse Ralstonia solanacearum populations in the Upper East, a major tomato-producing region of Ghana, with all isolates classified as biovar 3 and belonging to phylotypes I and III. Significant variation in resistance among tomato genotypes was observed, while strain and strain × genotype effects were not significant, indicating stable genotype performance across both strains. Resistant genotypes such as CRA66 WorldVeg and TmL114-42-N-H.T.P early maintained high survival rates (≥80%) and low disease progression, whereas susceptible genotypes such as Petomech and Lebombo exhibited 0–5% survival and high AUDPC values, reflecting rapid disease development. The majority of genotypes were classified as moderately resistant, confirming that high-level resistance to bacterial wilt is uncommon in the evaluated germplasm. These findings highlight the importance of host resistance in managing bacterial wilt and identify promising genotypes for breeding and deployment in Ghana. Further multi-environment field validation is recommended to ensure durability of resistance under field conditions.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fhort.2026.1874910/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Muhammad Amjad Ali, University of Agriculture, Faisalabad, Pakistan

Reviewed by: Bhuvaneswari Shivashankar, Indian Institute of Vegetable Research (ICAR), India

Habibeh Hajian Maleki, Tarbiat Modares University, Iran