Morphotype-based characterization of tomato (Solanum lycopersicum L.) accessions cultivated in Togo for conservation and genetic improvement

Introduction

Tomato (Solanum lycopersicum L.) is one of the most widely cultivated vegetable crops worldwide, valued for its nutritional composition, particularly its high content of vitamin C, lycopene, and other bioactive compounds that contribute to human health and food security (Zannou et al., 2024). Beyond its dietary significance, tomato plays a crucial role in agricultural economies, serving as both a subsistence and commercial crop (Grozeva et al., 2021). In sub-Saharan Africa, and particularly in Togo, tomato production is a major source of income for smallholder farmers and an important component of local diets (Kanda et al., 2014; Akintayo et al., 2023).

Globally, tomato production reached approximately 186 million tons on 4.9 million hectares in 2024, with an average yield of 38 t ha^-1. In contrast, Togo cultivates about 20,000–25,000 hectares, producing 150,000–200,000 tons annually, with yields of only 7–10 t ha^-1, far below the global average (FAOSTAT, 2024). This pronounced productivity gap is largely driven by biotic stresses, notably Tomato Yellow Leaf Curl Virus (TYLCV) and bacterial wilt (Bamazi, 2021). Bacterial wilt, caused by Ralstonia solanacearum, is among the most destructive diseases in tropical and subtropical regions and is widespread across West Africa (Sikirou et al., 2022; Paudel et al., 2025). The pathogen is a highly aggressive soil-borne bacterium capable of infecting a wide range of host plants and persisting in soil for extended periods, making its management particularly challenging (Genin and Denny, 2012; Lowe-Power et al., 2018). In Togo, field surveys indicate that the disease is prevalent in most production areas, with incidence exceeding 40% at the plant level and, under severe conditions, leading to total crop failure (Bamazi et al., 2022). Globally, bacterial wilt can cause yield losses ranging from 35% to 90% in tomato, depending on environmental conditions and pathogen pressure (Paudel et al., 2025). Its persistence in soil, broad host range, and high genetic variability further complicate effective control strategies (Genin and Denny, 2012; Lowe-Power et al., 2018). Similarly, TYLCV, a whitefly-transmitted begomovirus, is highly prevalent in West Africa, including Togo, with reports of up to 100% field prevalence and substantial yield losses (Atassé et al., 2018). These combined disease pressures underscore the urgent need to identify and characterize tomato germplasm with enhanced resistance to both bacterial wilt and TYLCV to support the development of resilient, locally adapted varieties.

In addition to these biotic constraints, tomato production in Togo is further limited by restricted access to improved varieties, suboptimal crop management practices, and inadequate knowledge of the genetic diversity of locally cultivated germplasm (Akintayo et al., 2023). Collectively, these factors reduce productivity, weaken system resilience, and pose significant challenges to the sustainability of tomato cultivation in the country (Bamazi et al., 2022).

Genetic variability is the cornerstone of crop improvement, providing the raw material for breeding programs aimed at enhancing yield, resilience, and adaptability (Rahman et al., 2018; Kumar and Lata, 2025). A comprehensive understanding of morphological diversity within cultivated germplasm is therefore essential for identifying promising genotypes and guiding the development of improved varieties (Badu-Apraku et al., 2015; Glato et al., 2017). Morphological descriptors, when systematically applied, offer a practical and cost-effective means of assessing genetic variation, particularly in resource-constrained settings where molecular tools may not be readily accessible (Rabbi et al., 2014; Kouassi et al., 2023). Such characterization not only facilitates varietal identification (Sattar et al., 2024) but also supports the conservation and sustainable utilization of genetic resources.

Studies across diverse agro-ecological contexts have consistently demonstrated that morphological characterization is a robust and informative approach for elucidating diversity patterns, discriminating morphotypes, and uncovering trait associations aligned with breeding objectives, particularly in settings where access to molecular tools remains constrained (Sinha et al., 2022; Alatawi et al., 2024). In Ghana, Osei et al. (2014) documented substantial phenotypic variability among cultivated tomato accessions, identifying plant vigor, leaf architecture, and fruit attributes as robust proxies for underlying genetic diversity. Consistent with this, Caramante and Rouphael (2024) demonstrated the discriminative power of morphological traits in resolving morphotypes and informing varietal improvement, thereby reinforcing their utility as cost-effective, field-applicable markers for enhancing yield potential, stress resilience, and environmental adaptability.

By contrast, in Togo, systematic germplasm characterization remains limited. Existing studies have largely prioritized agronomic performance and disease pressures; for instance, Akintayo et al. (2023) assessed a small set of tomato hybrids without a comprehensive evaluation of morphological diversity. This knowledge gap constrains the effective exploitation of local genetic resources for breeding. A rigorous documentation and analysis of morphological variation is therefore imperative to underpin tomato improvement programs and to support food security and rural livelihoods in the country.

Accordingly, the present study characterizes a panel of forty-five cultivated tomato accessions collected from diverse agro-ecological zones of Togo using standardized morphological descriptors. Variation in stem architecture, foliar traits, and disease responses was quantified, and multivariate approaches were employed to resolve diversity patterns and population structure. By generating an integrated phenotypic framework, this work provides actionable insights into the available genetic resources, facilitates the identification of superior and resilient genotypes, and informs evidence-based breeding strategies aimed at the sustainable intensification of tomato production in Togo.

Materials and methods

Plant material

A total of forty-five (45) cultivated tomato accessions, collected across the four agro-ecological zones of Togo by the Plant Genetic Resources Unit of the Togolese Institute of Agricultural Research (ITRA), were used as plant material. Detailed information on the accessions and their respective collection sites is presented in Table 1.

Table 1. Tomato accessions cultivated in Togo and their respective collection sites.

Experimental site

The experiment was conducted at the Agronomic Research Station of the Togolese Institute of Agricultural Research (ITRA) in Davié, located along National Road No. 1, approximately 30 km north of Lomé (6°23′5″ N, 1°12′18″ E; altitude: 76 m). Figure 1 presents maps depicting the location of Togo in West Africa, the country’s principal tomato production zones, and the experimental site at the Coastal Agricultural Research Centre (CRAL/Davié).

The site experiences a subequatorial guinean-type climate characterized by two rainy and two dry seasons. The major rainy season extends from mid-March to late July, while the minor rainy season occurs from September to mid-November (Detchinli et al., 2017). Annual rainfall ranges from 1,000 to 1,400 mm, with temperatures varying between 24 °C and 34 °C and relative humidity ranging from 70% to 90%.

The soil at the experimental site is predominantly sandy loam, with coarse sand accounting for more than half of the particle size distribution (Table 2) (Ezui et al., 2016). It is characterized by low organic matter content (1.05%), limited nitrogen (0.41%), and low available phosphorus (10 ppm), indicating poor fertility. The cation exchange capacity is moderate (2.89 cmol·kg^-1), with a base saturation of 66%, reflecting a relatively balanced exchange complex dominated by calcium and magnesium. The soil is moderately acidic, with pH values of 5.5 (H₂O) and 4.4 (KCl), which may restrict nutrient availability. Overall, the soil is deep and well-drained but nutrient-deficient, requiring fertilizer inputs to support optimal tomato growth (Ezui et al., 2016).

Table 2. Physicochemical properties of experimental site soil.

Experimental design

A randomized complete block design (RCBD) with three replications was employed. Each block comprised 45 plots. Each plot measured 2.5 m × 1.25 m and contained two rows of five plants. Plots were separated by 0.5 m, while adjacent replications were spaced 2 m apart. Each replication covered an area of 25.75 m × 8.5 m (218.9 m^²), resulting in a total experimental area of 759.6 m^². The experimental layout and plot arrangement are illustrated in Figure 2.

Trial establishment and management protocol

The experiment was conducted during the rainy season, a period characterized by high rainfall, elevated humidity, and favorable conditions for the proliferation of pests and diseases, thereby providing a relevant context for evaluating tomato accessions under natural disease pressure (Sinha et al., 2022). The subsequent cultural practices included substrate preparation, nursery establishment and maintenance, field preparation, seedling transplanting, and crop management (Alatawi et al., 2024).

Substrate preparation

This involved the collection, sieving, and sterilization of the substrate. The substrate was sterilized by heat treatment (Figure 3) for one hour (Osei et al., 2014) in iron trays, one week prior to nursery establishment.

Nursery establishment and maintenance

The nursery was established in trays filled with sterilized substrate and placed inside a greenhouse to protect the seedlings from pests, heavy rainfall, and strong winds. The trays were thoroughly watered one day prior to sowing. Seeds were treated with Montaz 45 WS, a seed treatment formulation containing imidacloprid (250 g/kg) and thiram (200 g/kg), to provide early protection against insect vectors and seed-borne fungal pathogens. Sowing was done in rows, with inter-row spacing used to separate the accessions. Seedlings were irrigated once daily and remained in the nursery for twenty one (21) days. Figure 4 illustrates the sequential steps involved in nursery establishment.

Experimental management

The experimental field was ploughed and harrowed to obtain a fine tilth. The field was then divided into three blocks, within which beds were raised manually using hoes. Raised beds were prepared according to the plot dimensions specified in the experimental design. Organic compost was applied at a rate of 20 tons per hectare during beds preparation. Compost was preferred over chemical fertilizer (NPK 15-15-15) to minimize the risk of plant burn under limited rainfall conditions.

Three weeks old healthy seedlings were transplanted in the plots at a plant spacing of 80 cm in between rows and 50 cm within the rows.

Crop maintenance primarily consisted of irrigation, weeding, fertilization, and phytosanitary treatments. Watering was carried out manually using watering cans to maintain adequate soil moisture, twice daily after transplanting, and once daily from the fifth day after transplanting (DAT), due to the low water-retention capacity of the soil (ferralitic soil). Basal fertilization was applied at 15 DAT using NPK (15-15-15), followed by top-dressing with urea at flowering and fruit set stages. Fertilizer rates were applied according to local agronomic recommendations. Weeds were controlled manually by hoeing at regular intervals to reduce competition and maintain field hygiene. The first weeding was performed at 15 DAT and subsequent ones as required. Plants were treated with BANKO PLUS^® at a rate of 3 liters par hectare. This broad-spectrum fungicidal formulation, containing carbendazim (12% w/w) and mancozeb (63% w/w), was applied to protect against seed-borne and soil-borne fungal pathogens during early crop establishment and at the onset of flowering. One month after planting, the plants were staked with wooden sticks to support upright growth and minimize lodging.

Phenotypic data collection

Table 3 summarizes the traits recorded during field observations, including growth habit, plant height, main stem pubescence density, plant vigor, foliage density, leaf attitude, and leaf type, as well as severity scores and percentage disease incidence for Tomato Yellow Leaf Curl Virus (TYLCV) and bacterial wilt. All traits were assessed for each plant within each plot at the vegetative stage (Osei et al., 2014; Dhyani et al., 2019).

Table 3. Plant trait assessment: parts, codes, methods, and rating scales.

TYLCV severity score was assessed at 30 days after transplanting under natural field conditions using a five-class visual scale (Segbefia et al., 2018; Zindović et al., 2025): 1 = no visible symptoms, 2 = mild infection, 3 = moderate infection, 4 = severe infection, and 5 = plant death. No artificial inoculation was performed; thus, the evaluation reflects field tolerance or symptom escape rather than confirmed genetic resistance. Bacterial wilt severity was assessed using the same scale (Kumar et al., 2025).

Disease incidence for both TYLCV and bacterial wilt was determined by recording the number of symptomatic plants per plot throughout the cropping period, based on characteristic symptoms. Percentage disease incidence (PDI) was calculated as (number of infected plants/total number of plants assessed) × 100, with final values recorded at peak disease expression and used for statistical analysis (Pandey et al., 2003).

Statistical analysis

All analyses were conducted in XLSTAT 2024. Raw data were screened for inconsistencies, outliers and missing values prior to analysis. Qualitative traits were encoded as categorical variables following the defined morphological descriptors.

The morphological diversity of the germplasm was assessed through two complementary approaches. First, descriptive analysis was performed to provide an overview description of the germplasm variation (Caramante and Rouphael, 2024). Frequencies were calculated for categorical variables. Disease severity scores and incidence were analyzed descriptively to characterize accession responses under natural infection conditions.

Subsequently, Multiple Correspondence Analysis (MCA) was applied to identify the most relevant traits explaining variability within the germplasm (Yadav et al., 2025). MCA was chosen because it is the most appropriate multivariate technique for analyzing categorical morphological descriptors, allowing us to capture patterns of association among traits more effectively than alternative methods such as Principal Component Analysis (PCA), which is better suited for continuous variables (Husson et al., 2008). Dimensions were retained based on eigenvalues and the proportion of explained inertia (Greenacre, 2007). Interpretation of axes relied on variable contributions, squared cosine values (cos^²), and v-test statistics. Categories withv-test| > 1.96 were considered significantly associated with a given dimension (Lebart et al., 1995). Variables contributing above the average expected contribution and exhibiting high cos^² values were retained for downstream analyses to define morphotype (Alatawi et al., 2024). Hierarchical clustering was performed on MCA-derived accession coordinates using Ward’s minimum variance method with Euclidean distance (Ward, 1963; Dhyani et al., 2019).

The optimal number of clusters (reflecting the intrinsic quality of the partition) was determined based on compactness (accessions within the same group being as similar as possible) and separability (accessions in different groups being as distinct as possible), using the Calinski-Harabasz index (Karna, 2021), which represents the ratio of between-group to within-group inertia. The appropriate number of clusters was selected by maximizing the Calinski-Harabasz criterion (Jain and Kumar, 2023). Finally, the morphological groups obtained were characterized by cross-referencing the typology with the variables included in the MCA factorial plan.

Results

Description of the germplasm based on morphological traits

The modalities of each trait observed are presented in Figure 5. All accessions (100%) exhibited a determinate growth type. With respect to plant height, nearly half of the accessions (47%) were of intermediate stature, while only 13% were tall. Plant vigor showed greater variability, with 29% of accessions classified as vigorous and 18% as weak (Figure 5).

Stem pubescence analysis revealed a balanced distribution, as 20% of accessions lacked pubescence, 60% displayed intermediate pubescence, and 20% exhibited dense pubescence. This variation may serve as a useful morphological marker for accession differentiation. Leaf attitude was predominantly semi-erect (49%), followed by horizontal (42%) and drooping (9%), reflecting a tendency toward upright foliage architecture. Foliage density assessment indicated that most accessions (64%) had intermediate foliage, while fewer exhibited sparse (24%) or dense (11%) foliage, suggesting moderate canopy coverage as the prevailing trait. Regarding leaf type, the majority (76%) exhibited the standard type, whereas 20% displayed the potato-leaf type and 4% the Peruvianum type, underscoring the predominance of conventional leaf morphology (Figure 5).

Disease severity scoring demonstrated strong field tolerance within the germplasm. A substantial proportion of accessions (76%) were tolerant to TYLCV, while only 9% were highly susceptible. Similarly, 91% of accessions were tolerant to bacterial wilt, with 9% classified as susceptible (Figure 5).

Morphological variability of accessions

Table 4 summarizes detailed variable contributions derived from the Multiple Correspondence Analysis (MCA). The analysis identified five principal axes that together explained 80.86% of the total variability. Seven descriptors plant height, vigor, stem pubescence, foliage density, leaf type, and tolerance/susceptibility to TYLCV and bacterial wilt emerged as the most distinctive traits differentiating the accessions (Tables 5, 6).

Table 4. Eigenvalues, explained variability, and contributions of variables to the first five multiple correspondence analysis (MCA) dimensions.

Traits that contributed most to the variation of a particular component are in bold.

Table 5. Squared cosine (cos^²) values of variable categories on the first two dimensions of the MCA.

Cos^² values indicate the quality of representation of each category on the axes, with higher values reflecting better representation.

Table 6. V-test statistics of variable categories across the first two dimensions of the Multiple Correspondence Analysis (MCA).

Values indicate the strength and direction of association between categories and axes; categories withv-test| > 1.96 are significantly associated (p < 0.05). Bold values indicate statistical significance at α = 0.05.

Axis F1, accounting for 38.57% of the variability, was defined by plant height, vigor, stem pubescence, and foliage density, underscoring the dominant role of vegetative traits. Axis F2 (18.62%) further captured variation in stem pubescence and vigor, with smaller plant size negatively correlated, reinforcing the importance of vegetative development. Axis F3 (9.87%) was associated with bacterial wilt tolerance, effectively distinguishing tolerant from susceptible accessions. Axis F4 (8.48%) separated leaf types, highlighting morphological diversity across standard, potato-leaf, and peruvianum forms. Axis F5 (5.30%) was linked to TYLCV susceptibility, indicating viral tolerance as a secondary but relevant source of variability.

Identification and characterization of morphotypes

Hierarchical clustering, evaluated using the Calinski-Harabasz index (Table 7) as a measure of partition quality, revealed five morphotypes (Figures 6, 7). Each morphotype represents a group of accessions sharing similar morphological profiles. Morphotype 2 was the largest cluster, comprising 18 accessions, while morphotype 3 contained only a single accession (TGTA006), underscoring its distinct morphological profile. Morphotypes 1, 4, and 5 represented intermediate clusters with 6, 13, and 7 accessions, respectively (Figure 6). The distribution of variance and centroid distances confirmed the robustness of the five-class partition, with morphotype 3 showing perfect homogeneity (variance = 0), highlighting its uniqueness within the germplasm (Table 8).

Table 7. Evolution of statistical indices assessing partition quality across cluster mumbers (K).

Calinski–Harabasz index increased steadily with the number of classes, reaching its maximum at five, thereby supporting the selection of five morphotypes as the optimal partition. Although Silhouette index values remained relatively low, they were consistent across K, while the Hartigan index exhibited fluctuations. Taken together, these indices confirm that the five-class solution provides the most robust and biologically meaningful grouping of accessions.

Table 8. Morphotypes composition derived from cluster analysis.

Accessions in bold represent the central objects of each morphological group.

Integration of clustering results with the MCA factorial plan (Figure 8) enabled detailed characterization of these morphotypes. Morphotype 1 comprised large, vigorous accessions highly susceptible to TYLCV but tolerant to bacterial wilt (Figures 6, 8).

Morphotype 2 comprised intermediate-sized, vigorous accessions with sparse pubescence, intermediate foliage density, and standard leaf type, tolerant to TYLCV but susceptible to bacterial wilt. Morphotype 3, highly tolerant to bacterial wilt, was represented by a single accession, reflecting its unique profile. Morphotype 4 included short accessions with moderate vigor, sparse pubescence, sparse foliage density, potato-leaf type, and tolerance to both TYLCV and bacterial wilt. Morphotype 5 consisted of short, weak accessions with sparse pubescence, sparse foliage density, drooping leaf attitude, and tolerance to both TYLCV and bacterial wilt.

Discussion

Morphological diversity of the germplasm

Genetic variability constitutes the cornerstone of plant breeding, providing the essential raw material for selection, genetic improvement, and adaptation (Yadav et al., 2025). Demonstrating diversity in agro-morphological traits is therefore a critical prerequisite for genetic enhancement programs (Caramante and Rouphael, 2024). Comprehensive knowledge of accession characteristics not only facilitates germplasm management and conservation but also enhances its effective utilization in varietal development (Alatawi et al., 2024).

The present study revealed substantial variability among accessions in traits such as plant height, vigor, stem pubescence, foliage density, leaf orientation, and disease severity scores. These polymorphic attributes serve as reliable descriptors for varietal identification and registration, thereby strengthening the genetic basis for breeding initiatives. Comparable findings have been reported in tomato germplasm from Burkina Faso, where agromorphological diversity was similarly pronounced (Sondo, 2017). Studies in Ghana revealed substantial genetic variability in tomato germplasm using morphological traits (Osei et al., 2014) and SSR markers (Gongolee, 2016), highlighting the importance of varietal identification for breeding and conservation. In Nigeria, evaluations of growth and yield diversity among local tomato germplasm demonstrated clear differentiation in phenotypic traits (Adeyemo and Olukanmi, 2014), supporting the role of morphotype classification in guiding selection strategies. Research in Asia has shown both inter and intra genetic diversity among tomato genotypes, integrating morpho-molecular and biochemical analyses (Alatawi et al., 2024). These findings emphasize the versatility of tomato as a model species and confirm the global relevance of morphotype-based diversity studies.

Interestingly, all accessions evaluated in this study exhibited determinate growth, characterized by restricted vegetative development. This uniformity is particularly advantageous for agronomic evaluation across tomato production systems in Togo. In contrast, Sondo (2017) documented 69.7% indeterminate growth in Burkina Faso, underscoring differences in allelic richness and production practices between regions. Plant height variation was notable, with 47% of accessions classified as intermediate and 13% as tall. Such variation may be attributed to the genotype of accessions.

Variation in vigor was also evident, with 29% of accessions displaying high vigor and 18% exhibiting low vigor. Stem pubescence demonstrated polymorphism, with 20% of accessions lacking pubescence and another 20% exhibiting dense pubescence. These traits are of particular interest, as pubescence has been associated with pest resistance and microclimate regulation, thereby influencing adaptation to humidity and temperature fluctuations. This observation is consistent with previous reports linking pubescence to insect deterrence and enhanced resilience under variable climatic conditions (Dhyani et al., 2019).

Leaf orientation exhibited considerable diversity, with 49% semi-erect, 42% horizontal, and 9% drooping. Such variation may directly influence photosynthetic efficiency and water-use dynamics. Foliage density was predominantly intermediate (64%), a desirable trait that optimizes photosynthesis, fruit productivity, weed suppression, and reduced evapotranspiration (Dhyani et al., 2019). Collectively, these morphological attributes highlight the adaptive potential of the germplasm under diverse agro-ecological conditions.

Tolerance levels against major biotic stresses were particularly promising. These results emphasize the overall resilience of the germplasm to major biotic stresses, reinforcing its potential value for breeding programs targeting disease resistance. A majority of accessions (91%) demonstrated field tolerance to bacterial wilt, while 76% exhibited tolerance to Tomato Yellow Leaf Curl Virus (TYLCV), a disease transmitted by whiteflies. These findings align with recent studies reporting genetic tolerance to TYLCV in tomato hybrids and improved lines (Alatawi et al., 2024). The early manifestation of bacterial wilt symptoms corroborates observations by Bamazi et al. (2022). Furthermore, genome-wide association studies have identified SNP markers linked to bacterial wilt resistance, confirming its heritable basis (Bhunchoth et al., 2025).

The observed variability underscores the immense potential of this germplasm for breeding programs aimed at enhancing agronomic performance and resilience to TYLCV and bacterial wilt. However, a limitation of this study is that agronomic data related to fruit yield and quality were not evaluated. As a result, while morphotype diversity and field tolerance to diseases were documented, the findings cannot be directly linked to productivity or market traits. To strengthen conservation and breeding programs, future studies should integrate comprehensive agronomic evaluations including yield, fruit quality, and post-harvest traits alongside morphotype characterization. This combined approach will provide a more holistic understanding of accession performance and ensure that promising morphotypes also meet productivity and market demands. Future research should also prioritize controlled screening under artificial infestation, coupled with molecular characterization through marker-assisted selection. Such integrative approaches will be essential to validate resistance traits and accelerate their deployment in tomato improvement programs.

Identification and characterization of morphotypes

The morphotype-based classification of tomato accessions highlights substantial phenotypic and functional diversity within the germplasm, with important implications for both conservation and genetic improvement in Togo.

Morphotype 1 made of six accessions is characterized by large, highly vigorous plants, indicative of strong vegetative growth and potential for high biomass and yield. However, these genotypes were susceptible to Tomato Yellow Leaf Curl Virus while exhibiting tolerance to Bacterial wilt. This contrasting response suggests a possible trade-off between growth-related traits and viral resistance, a phenomenon often reported in crop improvement studies where resource allocation to rapid growth may compromise defense mechanisms against certain pathogens (Foolad, 2007). The tolerance to bacterial wilt may be associated with root system architecture or physiological adaptations that limit pathogen proliferation in vascular tissues (Scott et al., 2005). Consequently, Morphotype 1 represents a valuable genetic pool for improving yield potential and bacterial wilt tolerance, although the incorporation of TYLCV resistance genes remains necessary.

Morphotype 2 composed of 18 accessions, the most represented group, comprises genotypes with intermediate plant size but sustained vigor. These accessions exhibited tolerance to TYLCV but were susceptible to bacterial wilt, indicating an inverse disease response compared to Morphotype 1. The prevalence of this morphotype may reflect farmer-driven selection for TYLCV tolerance, given the widespread and economically significant impact of this viral disease in tropical and subtropical regions (Hanssen et al., 2010). The observed TYLCV tolerance suggests the presence of resistance loci, such as Ty genes, which have been widely deployed in breeding programs (Prasanna et al., 2015). However, the susceptibility to bacterial wilt underscores the need to integrate resistance to soil-borne pathogens, which remain a major constraint to tomato production in humid environments.

Morphotype 3, represented by a single accession (TGTA006), exhibited a particularly high level of tolerance to bacterial wilt, distinguishing it from all other groups. This strong tolerance suggests the presence of robust and potentially unique genetic factors conferring effective defense against Ralstonia solanacearum. Given that bacterial wilt is one of the most destructive soil-borne diseases affecting tomato in tropical environments, the identification of such a highly resistant genotype is of considerable importance for breeding programs.

The singularity of morphotype 3 further indicates that this accession may harbor rare alleles or novel resistance mechanisms that are absent in the broader germplasm collection. Resistance to bacterial wilt is often quantitatively inherited and influenced by multiple genes, making the discovery of highly resistant sources particularly valuable for gene pyramiding strategies (Scott et al., 2005). Therefore, this morphotype should be prioritized for conservation and advanced characterization, including molecular analysis to identify resistance loci and validate its stability across environments. From a breeding perspective, morphotype 3 represents an elite donor parent for the introgression of bacterial wilt resistance into high-yielding but susceptible backgrounds. Its integration into crossing programs, particularly with morphotypes combining desirable agronomic traits or resistance to other diseases such as Tomato Yellow Leaf Curl Virus, could accelerate the development of improved tomato cultivars with durable, broad-spectrum resistance.

Morphotypes 4 (13 accessions) and 5 (seven accessions) consist of short-statured genotypes that exhibited combined resistance to both TYLCV and bacterial wilt. These morphotypes are particularly valuable for the development of resilient varieties adapted to low-input systems. Reduced plant height may confer agronomic advantages, including improved lodging resistance, better resource-use efficiency, and suitability for high-density planting systems (Bai and Lindhout, 2007). The differentiation between morphotypes 4 and 5, based on plant vigor and leaf architecture, suggests additional layers of genetic variability within resistant groups. Variations in leaf morphology may influence canopy structure, light interception, and microclimate, thereby affecting disease dynamics and overall productivity.

Collectively, the observed structuring indicates a strong association between plant architecture, vigor, and disease response, potentially reflecting genetic linkage or pleiotropic effects. These findings emphasize the importance of conserving representatives from all morphotypes to maintain the full spectrum of genetic diversity. From a breeding perspective, strategic hybridization between complementary morphotypes such as crossing high-yielding but TYLCV-susceptible genotypes (Morphotype 1) with dual-resistant, short-statured genotypes (Morphotypes 4 and 5), or combining TYLCV-tolerant genotypes (Morphotype 2) with bacterial wilt-resistant lines could facilitate the development of improved varieties that integrate productivity with broad-spectrum disease resistance. Such approaches align with contemporary breeding strategies aimed at pyramiding resistance genes while maintaining desirable agronomic traits (Foolad, 2007; Prasanna et al., 2015).

Comparative studies across West Africa (Segbefia et al., 2018) and Asia (Yadav et al., 2025) confirm that tomato germplasm exhibits significant morphotype diversity, reinforcing the relevance of the five morphotypes identified in Togo as part of a broader global pattern of genetic variability. This strengthens the case for their conservation and targeted use in breeding programs.

Collectively, the five morphotypes provide a broad spectrum of opportunities for selection and improvement of tomato traits. They offer potential to address diverse breeding objectives, including enhanced resistance to pathogens, improved productivity, and adaptation to varying environmental conditions. Moreover, the classification of morphotypes contributes to a deeper understanding of the interactions between morphological traits, thereby informing both conservation strategies and applied breeding programs.

Future research should integrate molecular characterization, such as SNP-based diversity analysis and marker-assisted selection, to complement morphological clustering. This integrative approach will strengthen the predictive power of morphotype classification and accelerate the deployment of improved tomato varieties tailored to local and regional production systems.

Conclusion

This study provides a comprehensive characterization of the morphological diversity within local tomato accessions cultivated in Togo. The substantial variability observed across traits including plant size, vigor, stem pubescence, foliage density, and leaf architecture together with contrasting responses to Tomato Yellow Leaf Curl Virus and Bacterial wilt, underscores the agronomic relevance of this germplasm. These traits constitute informative descriptors for the preliminary characterization of tomato germplasm in Togo. Five distinct morphotypes were identified, spanning highly vigorous but disease-susceptible accessions to short-statured, disease-tolerant forms. Notably, morphotypes 4 and 5 (comprising 13 and 7 accessions, respectively) emerged as particularly promising groups, as they include accessions expressing dual field tolerance. These morphotypes therefore represent valuable reservoirs of resistance that could be strategically exploited as donor parents in breeding programs targeting durable resistance under tropical production conditions. However, it is important to note that the present study did not assess key agronomic and market-related traits such as yield, fruit size, firmness, taste, and shelf life. Consequently, the overall breeding value of these resistant morphotypes remains only partially validated. Future work should integrate morphotype-based classification with comprehensive evaluations of yield components, fruit physicochemical quality, and market-preferred traits across multiple environments. Coupling these phenotypic assessments with high-resolution molecular characterization will enable the dissection of the genetic architecture underlying both adaptive and productivity-related traits, facilitate marker-assisted selection, and support the development of high-performing, climate-resilient tomato varieties adapted to the agro-ecological conditions of Togo.

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.

Footnotes

Edited by: Sudip Kumar Dutta, ICAR Research Complex For NEH Region, India

Reviewed by: Wang Xiukang, Yan’an University, China

Ehtisham Hussain, University of Sargodha, Pakistan