Dissecting genetic variability and trait interactions underlying yield improvement in tomato (Solanum lycopersicum L.)

Introduction

Tomato (Solanum lycopersicum L.) is a globally cultivated vegetable crop of high economic and nutritional importance, serving as a rich source of vitamins, antioxidants, and dietary minerals (Rasheed et al. 2023) [5] . Despite its wide cultivation, improvement of fruit yield and associated quality traits remains a major breeding objective due to the quantitative nature of yield, which is controlled by multiple interacting traits influenced by both genetic and environmental factors (Zannat et al. 2023) [6] . A rigorous understanding of genetic variability and trait interrelationships is therefore essential to enhance selection efficiency in tomato breeding programmes. Genetic variability is the cornerstone of crop improvement, enabling breeders to exploit differences among genotypes for desirable traits. Estimation of genotypic and phenotypic coefficients of variation (GCV and PCV), heritability, and genetic advance (GA) provides insight into the degree of variability and the expected response to selection (Rasheed et al. 2023) [5] . High heritability combined with high genetic advance suggests the predominance of additive gene effects and greater probability of effective selection based on phenotypic performance (Eshwar et al. 2023) [3] . Conversely, traits with moderate heritability and low genetic advance may be influenced significantly by environmental factors or non-additive gene effects, limiting direct selection response, as reported earlier in tomato and other solanaceous crops (Podlesakova et al. 2019) [10] .

Interrelationships among yield components and quality traits offer additional guidance for indirect selection. Correlation analysis quantifies the magnitude and direction of associations between traits, which aids breeders in identifying key attributes that contribute positively to yield improvement (Eshwar et al. 2023) [3] . However, correlation coefficients alone do not distinguish between direct and indirect influences of component traits on yield. To further dissect the nature of these interrelationships, path coefficient analysis partitions correlation coefficients into direct and indirect effects, elucidating causal relationships among traits (Dewey & Lu 1959) [2] . This analytical approach has been widely employed in tomato genetic studies to prioritize traits that exert strong direct effects on yield, even when their phenotypic correlations may be weak (Kadyan et al. 2024; Zannat et al. 2023) [4,6] . Understanding these causal pathways is particularly important for breeding strategies aimed at simultaneous improvement of yield and quality in complex trait systems. Recent studies in tomato have reported significant genetic variability and strong inter-trait associations among yield and quality attributes, underscoring the potential for selection based on both primary and secondary traits (Kadyan et al. 2024; Eshwar et al. 2023) [3,4] . For instance, traits such as fruit number, fruit weight, and biochemical quality parameters have shown significant direct and indirect effects on yield performance, providing practical selection targets (Kadyan et al. 2024; Zannat et al. 2023) [4,6] . Nonetheless, trait contributions can be population-specific owing to genetic background and environmental interactions, necessitating population-specific studies for efficient breeding. In the present investigation, a diverse set of tomato genotypes was evaluated to estimate the magnitude of genetic variability, heritability and genetic advance, and genetic and phenotypic correlations among key yield and quality traits. Additionally, genotypic and phenotypic path coefficient analyses were conducted to elucidate direct and indirect effects of yieldcontributing traits. The findings provide a comprehensive genetic architecture of tomato yield components and inform evidence-based selection strategies for enhanced yield and quality improvement.

Materials and methods

Plant material and experimental layout

The experimental material comprised seventy-five tomato (Solanum lycopersicum L.) genotypes, representing a broad spectrum of genetic diversity. The field experiment was conducted during the Kharif-2023. The trial was laid out in a randomized complete block design (RCBD) with three replications to minimize environmental heterogeneity.

Trait phenotyping and data collection

Phenotypic evaluation was performed for four key yield and fruit quality traits selected for their direct relevance to productivity and market acceptability in tomato. All observations were recorded on fully developed, marketable fruits following standardized protocols to ensure accuracy and reproducibility.

Statistical Analysis Analysis of Variance

The recorded data were subjected to analysis of variance (ANOVA) following the linear model appropriate for RCBD:

Y ij is the observation of the I th genotype in the j th replication, μ is the overall mean, g i is the effect of the I th genotype, r j is the effect of the j th replication and e ij is the experimental error. Significance of genotypic effects was tested against the error mean square at 5% and 1% probability levels.

Estimation of Genetic Variability Parameters

Genotypic variance (σ²g), environmental variance (σ²e), and phenotypic variance (σ²p) were estimated from ANOVA mean squares. Genotypic coefficient of variation (GCV) and phenotypic coefficient of variation (PCV) were calculated to assess the magnitude of variability following Burton and DeVane (1953). Broad-sense heritability (h²) was estimated as the ratio of genotypic variance to phenotypic variance. Genetic advance (GA) and genetic advance as percentage of mean (GAM) were calculated following Johnson et al. (1955) [7] , assuming a selection intensity of 5%. Classes of heritability and genetic advance were categorized according to the criteria suggested by Singh and Narayan et al.,1993 [8] .

Correlation analysis

Genotypic and phenotypic correlation coefficients among traits were estimated to quantify the degree of association between yield and its component traits. Genotypic correlations were computed using variance and covariance components derived from ANOVA, whereas phenotypic correlations were calculated from observed trait values. The statistical significance of correlation coefficients was tested using a two-tailed t-test.

Covariance analysis

Analysis of covariance was performed to partition genotypic, phenotypic, and environmental covariances among trait combinations. This analysis enabled identification of shared genetic control and environmental influence underlying trait associations, thereby strengthening the interpretation of correlation results.

Path coefficient analysis

To elucidate the causal relationships among yield-contributing traits, genotypic and phenotypic path coefficient analyses were conducted following Wright (1921) [9] as outlined by Dewey and Lu (1959) [2] . Average fruit yield was considered the dependent variable, while pericarp thickness, total soluble solids, and per plant yield were treated as independent variables. Correlation coefficients were partitioned into direct and indirect effects, and residual effects were computed to estimate the contribution of unmeasured factors influencing yield.

Statistical software

All statistical analyses, including estimation of genetic parameters, correlation coefficients, covariance and path coefficients were performed using R statistical software with appropriate packages for quantitative genetic analysis.

Results

Detection of Significant Genetic Variability Among Tomato Genotypes

The analysis of variance revealed highly significant (p < 0.01) differences among genotypes for all the evaluated traits, including pericarp thickness (PT), total soluble solids (TSS), per plant yield (PPY) and average fruit yield (AFY), indicating the presence of substantial genetic variability in the experimental population (Table 1). The mean squares due to replications were non-significant for all traits, suggesting uniform experimental conditions and minimal environmental variation across replications. The significance of genotype effects confirms the adequacy of the experimental material for assessing genetic variability and selection potential.

Magnitude and Partitioning of Phenotypic and Genotypic Variability

Estimates of phenotypic and genotypic coefficients of variation revealed that phenotypic variability exceeded genotypic variability for all traits indicating environmental influence on trait expression (Table 2). Among the studied traits, per plant yield exhibited the highest magnitude of variability, with genotypic and phenotypic coefficients of variation of 56.71% and 62.87%, respectively. Moderate variability exhibited by average fruit yield (GCV: 34.56%; PCV: 37.41%). Low levels of variability were observed for pericarp thickness and total soluble solids. The relatively narrow differences between GCV and PCV for PPY and AFY indicate limited environmental influence and a greater contribution of genetic factors, suggesting favourable prospects for effective selection.

Genetic and phenotypic interrelationships among yield and quality components

Genotypic correlation coefficients were generally higher in magnitude than their phenotypic counterparts, reflecting stronger inherent genetic associations among traits (Table 3). At the genotypic level, pericarp thickness exhibited a significant and positive association with average fruit yield (r = 0.6877), indicating that increased pericarp thickness contributes favourably to yield improvement. Total soluble solids showed a strong positive genotypic correlation with per plant yield (r = 0.7618), whereas its association with average fruit yield was non-significant. At the phenotypic level, average fruit yield displayed a significant positive correlation with pericarp thickness but showed significant negative correlations with total soluble solids and per plant yield. The observed discrepancies between genotypic and phenotypic correlations suggest that environmental factors modify the expression of genetic associations among yield-related traits.

Partitioning of Genotypic and Environmental Covariances Among Traits

Covariance analysis revealed significant genotypic covariance between pericarp thickness and per plant yield, as well as between per plant yield and average fruit yield, indicating shared genetic control among these traits (Table 4). Positive genotypic covariance between pericarp thickness and average fruit yield further substantiates their favourable genetic association. Conversely, negative environmental covariance observed for certain trait combinations suggests antagonistic environmental effects on trait expression.

Direct and indirect genetic effects on fruit yield revealed by path analysis

Genotypic path coefficient analysis demonstrated that per plant yield exerted the highest positive direct effect on average fruit yield (2.4156), followed by average fruit yield itself (2.0462), total soluble solids (0.7173), and pericarp thickness (0.5792) (Table 5). Despite some traits exhibiting weak or negative correlations with average fruit yield, their substantial direct effects indicate that indirect effects mediated through other traits may obscure their true contribution to yield. The relatively high residual effect suggests that additional yield-determining traits not included in the present study may also influence fruit yield expression.

Phenotypic causal relationships governing fruit yield expression

Phenotypic path coefficient analysis revealed that average fruit yield had the highest positive direct effect (0.9041) on itself, followed by pericarp thickness (0.2260) (Table 6). In contrast, total soluble solids (-0.3808) and per plant yield (-0.2887) exhibited negative direct effects on average fruit yield at the phenotypic level. The low residual effect (0.0959) indicates that a substantial proportion of phenotypic variation in fruit yield was explained by the traits included in the analysis The integrated interpretation of variability parameters, heritability estimates, correlation coefficients, covariance structure, and path analysis clearly identifies per plant yield and pericarp thickness as priority traits for yield improvement in tomato. Their high heritability, substantial genetic advance, favourable genetic correlations, and strong direct effects on yield suggest that simultaneous selection for these traits would be highly effective in tomato breeding programmes.

Discussion

The significant genotypic variation observed for all studied traits confirms the presence of substantial genetic diversity among the evaluated tomato genotypes, providing a strong basis for selection. Although phenotypic coefficients of variation exceeded genotypic coefficients for all traits, the relatively small differences between PCV and GCV for per plant yield and average fruit yield indicate predominant genetic control of these yield attributes, a trend consistently reported in recent tomato variability studies (Rasheed et al. 2023; Eshwar et al. 2023) [3,5] . Similar observations have been documented across diverse genetic backgrounds, emphasizing the reliability of yield-related traits as effective targets for selection (Kadyan et al. 2024) [4] . High heritability coupled with high genetic advance for per plant yield and average fruit yield suggests the predominance of additive gene action, indicating that direct phenotypic selection for these traits would be highly effective. Comparable findings have been reported in recent global studies, where yield components showed rapid response to selection owing to their strong additive genetic control (Rasheed et al. 2023; Kadyan et al. 2024) [4,5] . In contrast, total soluble solids exhibited moderate heritability with relatively low genetic advance, implying the involvement of non-additive genetic effects and greater environmental influence, a pattern commonly observed for fruit quality traits in tomato (Eshwar et al. 2023) [3] . Genotypic correlations were generally stronger than phenotypic correlations, highlighting the masking effect of environmental variation on observable trait associations. The significant positive genotypic association between pericarp thickness and average fruit yield suggests that genetic factors enhancing pericarp thickness may simultaneously contribute to increased fruit weight and yield stability. Similar positive relationships between pericarp thickness, fruit firmness, and yield stability have been reported in recent tomato studies, underscoring the importance of this trait for both productivity and post-harvest quality (Kadyan et al. 2024; Rasheed et al. 2023) [5] . The strong positive genotypic correlation observed between total soluble solids and per plant yield is particularly noteworthy, as yield and quality traits are often reported to be negatively associated. This favourable association indicates the presence of genotypes combining high productivity with improved soluble solids content, a finding also reported in population-specific studies where yield and quality improvement were not mutually exclusive (Eshwar et [3, ] . Pericarp thickness and total soluble solids also contributed positively, either directly or indirectly, reinforcing their importance as secondary selection traits. In contrast, phenotypic path analysis showed negative direct effects of per plant yield and total soluble solids on average fruit yield, reflecting environmental influences and compensatory trade-offs such as fruit number versus fruit size under field conditions. Such discrepancies between genotypic and phenotypic path coefficients have been widely reported and highlight the importance of relying on genotypic parameters for selection decisions (Rasheed et al. 2023) [5] . The low phenotypic residual effect observed in the present study indicates that the included traits collectively explained most of the phenotypic variation in yield.

Overall, the integrated interpretation of variability parameters, heritability estimates, correlation coefficients, and path analysis identifies per plant yield as the most reliable primary selection criterion for yield improvement in tomato, with pericarp thickness serving as an important secondary trait contributing to yield stability and fruit quality. [4,5] .

Conclusion

The evaluated tomato germplasm exhibits substantial genetic variability for yield-related attributes, especially PPY and AFY, which display high heritability and genetic advance and act as major causal determinants of average fruit yield. Pericarp thickness and total soluble solids are promising secondary targets the former for yield stability and processing suitability, the latter requiring integrated Table 3. Genotypic (above diagonal) and phenotypic (below diagonal) correlations and molecular selection. These results provide a robust, population-specific basis for trait prioritization and breeding strategy design aimed at simultaneous yield and quality improvement.

References

1. Al-Araby AA. Genetic variability, heritability, genetic advance, and correlation analysis in F₂ segregating population of tomato (Solanum lycopersicum L.). Ann Agric Sci Moshtohor. 2021. doi:10.21608/assjm.2021.194267.
2. Dewey DR, Lu KH. A correlation and path-coefficient analysis of components of crested wheatgrass seed production. Agron J. 1959;51(9):515-518. doi:10.2134/agronj1959.00021962005100090013x.
3. Eshwar O, Sridevi O, Mahmadshafi AJ, Revanasiddayya C, Raghavendra VC, Prem Sagar SP, et al. Genetic variability, correlation and path coefficient analysis in multi-parent derived F₂ population of tomato (Solanum lycopersicum L.). Int J Environ Clim Change. 2023;13(11):2372-2379. doi:10.9734/ijecc/2023/v13i113402.
4. Kadyan D, Srivastav P, Kumar J, Singh H. Genetic variability, correlation and path coefficient analysis in tomato (Solanum lycopersicum L.). Pharma Innov J. 2024;12(8):2325-2329.
5. Rasheed A, Ilyas M, Khan TN, Mahmood A, Riaz U, Chattha MB, et al. Study of genetic variability, heritability, and genetic advance for yield-related traits in tomato (Solanum lycopersicum L.). Front Genet. 2023;13:1030309. doi:10.3389/fgene.2022.1030309.
6. Zannat A, et al. Exploring genotypic variability and interrelationships among growth, yield, and quality characteristics in tomato. Heliyon. 2023;9(11):e06166. doi:10.1016/j.heliyon.2023.e06166.
7. Johnson HW, Robinson HF, Comstock RE. Estimates of genetic and environmental variability in soybeans. Agron J. 1955;47(7):314-318. doi:10.2134/agronj1955.00021962004700070009x.
8. Singh P, Narayanan SS. Biometrical techniques in plant breeding. New Delhi: Kalyani Publishers; 1993.
9. Podlesakova K, Ugena L, Spichal L, Dolezal K, De Diego N. Phytohormonal regulation of γ-aminobutyric acid accumulation in tomato fruits. Plant Physiol Biochem. 2019;141:176-184. doi:10.1016/j.plaphy.2019.05.028.