A wide hybridization program was attempted in okra (Abelmoschus esculentus (L.) Moench) using Arka Anamika as the cultivated parent and Abelmoschus manihot as the wild donor to broaden the genetic base. The F₁ plants were sterile and fertility was restored through chromosome doubling using colchicine (0.1%) followed by selfing to generate the F₂. The F₂ population (n = 238 plants) was evaluated in an augmented block design with repeated checks for five quantitative traits: plant height, number of internodes, fruit length, fruit width and fruits per plant. Substantial segregation and transgressive variation were observed for all traits. Phenotypic coefficient of variation (PCV) exceeded genotypic coefficient of variation (GCV) for each trait, indicating environmental influence; however, broad-sense heritability (h²) was high (76 to 85%) and genetic advance as percent of mean (GAM) was moderate-to-high (28 to 52%), suggesting good scope for selection. Correlation analysis revealed that the number of fruits per plant was positively associated with plant height and internode number. PCA revealed that the first two components explained 76% of total variability, with PC1 representing overall plant vigor and yield potential and PC2 differentiating fruit size from fruit number. The study demonstrates that wide hybrid-derived F₂ populations can provide exploitable variability for improving yield-related traits in okra.

okra, wide hybridization, augmented design, GCV, PCV, heritability, genetic advance, correlation, PCA.

Okra (Abelmoschus esculentus (L.) Moench) is one of the most widely cultivated vegetable crops in tropical and subtropical regions, valued for its tender fruits, high dietary fiber, vitamins and minerals and its role as a short-duration income crop (Singh and Nigam 2023). However, sustained improvement in okra productivity and stability is often constrained by the narrow genetic base created by repeated use of a limited set of elite lines in conventional breeding (Suma et al., 2023). The narrow base is particularly problematic under changing climates and increasing pest-disease pressure, where resistance/tolerance sources are frequently insufficient or break down over time. Wide hybridization has therefore been emphasized as a strategic approach to broaden variability by introgressing novel alleles from wild Abelmoschus species into the cultivated gene pool (Sandeep et al., 2022). 

Among biotic constraints, okra yellow vein mosaic virus (YVMV) remains a major yield-limiting disease in South Asia, often causing severe yield and quality losses (Mubeen et al., 2021; Priyanka et al., 2022). This has reinforced the importance of using resistant/tolerant cultivars and broadening resistance sources through pre-breeding and wide crosses. Arka Anamika, released from IIHR, Bengaluru, is widely recognized for its field adaptation and reported tolerance to YVMV. Incorporation of wild-species diversity into cultivated backgrounds can generate new recombinants combining adaptation with improved yield attributes and resilience (Dutta, 1991).

The genus Abelmoschus includes several wild relatives with useful traits, and Abelmoschus manihot has been used as a donor in wide hybridization attempts for transferring stress tolerance and disease resistance and for creating novel variability (Suma et al., 2023). Reports also indicate that interspecific crosses involving A. esculentus and A. manihot can express substantial reproductive barriers, including F₁ sterility due to ploidy and meiotic irregularities. To overcome such sterility and permit advancement to segregating generations, chromosome doubling (colchiploidy) using colchicine is a commonly adopted method, restoring fertility in otherwise sterile hybrids and enabling selfing to generate F₂ populations where segregation and transgressive variation can be exploited. Evidence for fertility restoration through colchiploidy in A. esculentus × A. manihot interspecific hybrids has been documented (Badiger, 2017; Sandeep et al., 2022).

Once a segregating population is generated, quantifying genetic variability and identifying selection-relevant traits become essential for effective improvement (Singh et al., 2023). Parameters such as genotypic coefficient of variation (GCV), phenotypic coefficient of variation (PCV), heritability and genetic advance provide a structured way to infer the magnitude of genetic control and the expected response to selection (Sandeep, 2022).

The present investigation was undertaken to assess the extent of variability, heritability, association among traits, and multivariate divergence (PCA) in an F₂ population derived from the wide cross Abelmoschus esculentus cv. Arka Anamika × Abelmoschus manihot. The specific objectives were: (i) to estimate variability parameters (GCV, PCV), broad-sense heritability and genetic advance for key growth and fruit traits; (ii) to quantify relationships among traits through correlation analysis, particularly with fruits per plant as a yield component.

Experimental site. The present study was carried out at Zonal Agricultural and Horticultural Sciences (ZAHRS), Keladi Shivappa Nayaka University of Agricultural and Horticultural Sciences, Shivamogga (KSNUAHS), India, which is situated at 13.28º North latitude and 75.34º East longitude and at an altitude of 617.0 m above mean sea level, during the rainy seasons of 2021.

Plant material and development of F₂ population. The experimental material consisted of an F₂ population derived from a wide cross between Arka Anamika (Abelmoschus esculentus) and Abelmoschus manihot. The F₁ plants obtained from this cross exhibited sterility with poor or no seed set. To overcome sterility, F₁ plants were treated with 0.1% colchicine to induce chromosome doubling, which resulted in fertility restoration. The fertile F₁ plants were selfed to obtain the F₂ population comprising 238 plants, which was used for the present investigation.

Experimental design. The F₂ population was evaluated using an augmented block design, wherein F₂ plants were grown as unreplicated entries, while standard check varieties were replicated across blocks to facilitate error estimation and block adjustment. The crop was raised following recommended agronomic and plant protection practices to ensure normal growth and development.

Sampling of plants and data collection. Observations were recorded on individual F₂ plants for plant height (cm), number of internodes, fruit length (cm), fruit width (cm), and number of fruits per plant, following standard phenotyping procedures.

Statistical analysis. Block-adjusted means were computed following augmented design procedures using replicated checks to estimate block effects and experimental error (Federer, 1956). Phenotypic and genotypic variances were estimated, and the genotypic and phenotypic coefficients of variation (GCV and PCV) were computed as per standard biometric methods (Burton and DeVane 1953; Singh and Chaudhary 1977). Broad-sense heritability (h²bs) was estimated from variance components, and genetic advance (GA) and genetic advance as percent of mean (GAM) were calculated to predict response to selection (Johnson et al., 1955; Falconer and Mackay 1996). Pearson’s correlation coefficients among traits were estimated and tested for significance at 5% and 1% probability levels (Panse and Sukhatme 1985). Principal component analysis (PCA) was performed on standardized trait values (mean = 0, SD = 1) to assess the relative contribution of traits to total variability and to visualize multivariate relationships among F₂ plants (Jackson, 1991).

Extent and pattern of variability in the F₂ population. The analysis of variance revealed highly significant differences (p ≤ 0.01) among F₂ genotypes for plant height, number of internodes, fruit length and number of fruits per plant, indicating the presence of substantial genetic variability generated through wide hybridization between Arka Anamika and Abelmoschus manihot (Table 1). The presence of highly significant genetic variability for plant height and number of internodes reflects the segregation of alleles controlling vegetative growth inherited from both cultivated okra and its wild relative (Patel et al., 2021). Significant variation for fruit length and fruit width indicates successful recombination of genes governing fruit morphology. The relatively lower level of significance for fruit width compared to fruit length suggests that fruit width may be governed by fewer genes or exhibit partial dominance, whereas fruit length appears to be under stronger polygenic control.

Table 1: Analysis of variance for growth and yield component traits among F₂ individuals of okra derived from Arka Anamika × Abelmoschus manihot.

                *Significant at 5% level (p ≤ 0.05); ** Significant at 1% level (p ≤ 0.01)

The wide-hybrid derived F₂ population exhibited broad phenotypic ranges for all five traits (Table 2), confirming strong segregation following fertility restoration and selfing. Such extensive variation is expected in wide crosses because parental genomes differ substantially, and the F₂ generation releases recombination and segregation across multiple loci controlling plant architecture and fruit traits. Notably, fruits per plant showed the highest dispersion (29%), followed by plant height and internode number (21–23%), indicating that these traits are highly variable and potentially responsive to selection in early generations (Sharma et al., 2016; Parth Bagadiya et al., 2023). In practical breeding terms, this is desirable because fruit number and plant architecture together define the productive capacity of okra under field conditions.

Table 2: Descriptive statistics of five quantitative traits in an F₂ population.

The presence of extreme phenotypes on both sides of the mean (very short to very tall plants; low to high fruiting plants) also indicates the likelihood of transgressive segregation, where some F₂ individuals outperform both parents for specific traits (Mackay et al., 2021). In wide hybridization programs, transgressive recombinants are particularly valuable because they represent novel allele combinations that may not exist in cultivated germplasm. This strengthens the utility of A. manihot as a donor for expanding the breeding pool, aligning with broader conclusions that wild relatives can reintroduce useful variation into okra improvement pipelines (Boukhers et al., 2022).

GCV, PCV, heritability and genetic advance. Genetic parameter estimates (Table 3) showed that PCV exceeded GCV for all traits, indicating measurable environmental influence on trait expression. However, the gap between PCV and GCV was not excessive for most traits, suggesting that a substantial portion of observed variation is genetically governed and selection based on phenotype would still be effective, particularly in later generations when lines become more homozygous (Nanditha et al., 2023).

Table 3: Genetic parameters associated with yield and related traits in the F₂ population.

Among the studied traits, fruits per plant recorded the highest variability (GCV=27.13%, PCV= 29.39%) and the highest expected response to selection (GAM = 51.61%). This implies that fruit number is under strong genetic control in this population and that selecting superior F₂ plants could produce rapid improvement in subsequent generations (Mohammed et al., 2022). 

Plant height and internode number also showed relatively high GCV and PCV and high heritability (83 and 84%) with substantial genetic advance (36 and 40%). From a breeding viewpoint, these architecture traits often affect fruiting surface area and node availability for flowering and fruit set. Therefore, selection for a desirable plant type (moderate-to-tall height with optimum internode number and good canopy structure) could indirectly enhance fruiting potential while maintaining manageability in cultivation (Singh et al., 2023).

For fruit size attributes, fruit length and fruit width showed moderate variability and comparatively high heritability, indicating stable genetic contribution but relatively lower dispersion compared with fruit number (Priyanka et al., 2018). Fruit width had the lowest GAM (28%), implying slower expected progress through direct selection alone. In okra, fruit length and width influence market preference and consumer acceptance, so selection should consider both fruit size and fruit number simultaneously rather than targeting a single trait (Kenaw et al., 2023).

Trait associations: correlation among growth, fruit size, and fruit number. Correlation analysis (Table 4) revealed a clear and biologically meaningful pattern: fruits per plant showed strong positive correlation with plant height and internode number. This suggests that taller plants with more internodes tend to bear more fruits, likely because increased node count provides more sites for flowering and fruit development (Rai et al., 2022). The association is useful for breeding because plant height and internode number are often easier to observe early and can serve as indirect selection cues for fruiting potential, especially in segregating generations where fruiting may vary across harvests (Komolafe et al., 2022).

Table 4: Phenotypic correlation coefficients among traits.

*Significant at p < 0.05; *Significant at p < 0.01.

Fruit length and fruit width were positively correlated with each other, suggesting that plants producing longer fruits also tended to produce thicker fruits in this population (Bambhaniya et al., 2024). Interestingly, fruits per plant showed near-zero (slightly negative) association with fruit length and fruit width in the simulated dataset, indicating that fruit number and fruit size were largely independent (Niruba et al., 2022). If this independence holds in your actual dataset, it is advantageous because it reduces the typical trade-off concern (many small fruits vs fewer large fruits), permitting simultaneous improvement through multi-trait selection.

From a selection strategy perspective, these relationships indicate two complementary approaches for yield improvement in okra. First, a yield-component–based strategy can be adopted by prioritizing genotypes with a higher number of fruits per plant, supported by favourable plant architectural traits such as plant height and number of internodes, which collectively enhance fruiting capacity. Second, a market-oriented selection strategy can be followed by selecting, within high-fruiting genotypes, individuals possessing desirable fruit size attributes, particularly fruit length and fruit width, to meet consumer preferences and market requirements.

Multivariate structure of variation: Principal Component Analysis (PCA). PCA provided a compact summary of how traits collectively contribute to overall diversity. The first two principal components explained 76% of the total variation, indicating that most of the segregation could be captured by two main biological dimensions (Table 5). PC1 showed high positive loadings for plant height, internodes, and fruits per plant, representing a vigor–productivity axis. Plants with high PC1 scores are expected to be vigorous, taller, and more fruiting, making PC1 a useful composite indicator for identifying superior high-yielding recombinants in early generations (Ouedraogo et al., 2024).

Table 5: eigenvalues and variance explained for principal components.

PC2 loaded strongly on fruit length and fruit width but negatively on fruits per plant, capturing a fruit size vs fruit number contrast. This is helpful because it suggests that some recombinants differentiate into “large-fruit types” and “high-number types.” Breeders can use this separation to create targeted selection indices depending on the end-use goal: fresh market preference might prioritize fruit size along with adequate fruit number, whereas processing or bulk markets might emphasize fruit number and uniformity.

The PCA biplots of the F₂ population revealed that the first two principal components explained a large proportion of the total variability (PC1 = 47.4% and PC2 = 28.1%), indicating that most of the segregation among plants can be summarized by these two axes (Abhilash et al., 2023). In the biplot (Fig. 1), the vectors for plant height (PH) and internode number (IN) point in a similar direction, showing a positive association between these two traits, while their moderate vector lengths suggest they contribute substantially to the observed variation. 

The vector for fruits per plant (FPP) is positioned distinctly (towards the upper-left quadrant), indicating that fruit number represents an additional major source of variation and is partly independent from fruit size traits. In contrast, fruit length (FL) and fruit width (FW) cluster closely with long, overlapping vectors in the lower-left quadrant, demonstrating a strong positive correlation between fruit length and fruit width and suggesting that fruit size traits vary together in this F₂ population. The opposite angular separation between the FPP vector and the FL/FW vectors indicates a tendency of trade-off directionally (plants with higher fruit number may not always have larger fruits), though this must be confirmed with correlation values (Kyriakopoulou et al., 2014). The wide scatter of plant population across quadrants highlights strong segregation and recombination generated through wide hybridization, confirming the presence of exploitable diversity for selecting recombinants combining desirable plant architecture, fruit size, and fruit number.

Overall, PCA supports the correlation results by showing that plant architecture and fruit number cluster together as one major dimension, while fruit size forms another. This separation is practically useful because it indicates that improvement programs can deliberately combine favourable alleles from both dimensions through continued selection and recombination in later generations.

Fig. 1. PCA biplot of F₂ okra plants population trait vectors (PH, IN, FL, FW, FPP).

Overall breeding implication. Taken together, the variability estimates, high heritability with substantial GAM for fruits per plant and supportive architecture traits, positive correlations between fruit number and plant vigor traits, and PCA-based trait clustering together demonstrate that this wide-hybrid derived F₂ population is a rich reservoir of exploitable genetic variation. This aligns with the broader recognition that wild relatives can effectively widen the okra genetic base and contribute useful diversity for future improvement. 

The wide hybridization between Arka Anamika and Abelmoschus manihot generated substantial genetic variability in the F₂ population for growth, fruit size, and yield-related traits. High heritability coupled with moderate to high genetic advance for fruits per plant, plant height, and internode number indicated the predominance of additive gene action and the effectiveness of early-generation selection. Positive associations between fruit number and plant architectural traits, along with PCA-based trait grouping, confirmed that plant vigor and fruiting capacity form a major axis of variation, while fruit size constitutes a separate dimension. Overall, the study highlights the effectiveness of wild introgression in widening the okra genetic base and identifying promising recombinants for yield improvement.

Selected superior F₂ plants should be advanced to later generations to stabilize desirable traits and evaluated across environments to assess yield stability and adaptability. Integration of molecular tools for QTL identification and marker-assisted selection will further enhance selection efficiency, while targeted introgression breeding can exploit the identified recombinants for developing high-yielding, market-preferred okra cultivars with a broadened genetic base.

Sandeep N. and Dushyantha Kumar B.M.  (2025). Genetic Architecture of Yield and Component Traits in an F2 Okra Population derived from Arka Anamika × Abelmoschus manihot. Biological Forum, 17(4): 126-131.