Estimation of genetic parameters in hybrid and F₂ generations of aromatic fine rice for breeding improvement

Abstract

Background

Genetic improvement in aromatic rice is crucial for enhancing its yield, quality, and resilience to environmental stressors. The present study was designed to analyze genetic parameters, heterosis, and inbreeding depression in F₁ and F₂ generations obtained from a crossing between fine and aromatic rice genotypes – Kataribhog and BRRI dhan50.

Results

A significant amount of variation was found from the analysis of variance among the genotypes of F₁, F₂, and their parents. Grain yield plant⁻¹ showed a significant positive correlation with the number of tillers hill⁻¹, number of effective tillers hill⁻¹, flagleaf length, panicle length, grains panicle⁻¹, filled grains panicle⁻¹, and grain yield panicle⁻¹. For all the traits of F₁ and F₂, the phenotypic coefficient of variation (PCV) was greater than the corresponding genotypic coefficient of variation (GCV), suggesting an influence of environment on the expression of these traits. Furthermore, high heritability along with high genetic advance as percentage of the mean (GAM) was observed for all the traits studied except days to first flowering and plant height in F₁ and for grains panicle⁻¹, filled grains panicle⁻¹, grain yield panicle⁻¹, and grain yield plant⁻¹ in F₂ generations which is an indication of additive gene control and selection for improvement could be effective. Both the cross and reciprocal cross had significant positive heterosis with subsequent inbreeding depression predominantly in the number of tillers hill⁻¹, grains panicle⁻¹, filled grains panicle⁻¹, grain yield panicle⁻¹, and grain yield plant⁻¹, excluding days to first flowering suggested the scope for exploitation of heterosis. A higher estimate of transgressive segregation for grain yield plant⁻¹ encouraged further breeding efforts.

Conclusions

This study highlights the potential for developing high-yielding aromatic rice through hybridization and selection. The observed genetic variation, high heritability, and heterosis confirm opportunities for yield improvement. By advancing segregating generations and integrating modern breeding tools, these findings pave the way for developing superior aromatic rice varieties with enhanced productivity and grain quality.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-025-07155-9.

Background

Rice (Oryza sativa L.), the most important dietary carbohydrate in the world, is a monocotyledonous angiosperm that belongs to the family Gramineae [1]. Asian countries comprise almost 90% of both the world’s rice producers and consumers [2]. In Bangladesh, food security is synonymous with rice security. Bangladesh is the third-highest rice producer worldwide, with a production of 38.4 million tons in 2022 [3]. As rice is a field-to-plate crop, customer demand and preference for rice are influenced by some grain quality attributes [4]. Among them, the aroma in rice is unique and a superior grain quality trait, based on which rice is categorized into two subgroups, i.e., aromatic and non-aromatic rice. Aromatic rice is a significant member of a tiny subgroup of rice [5].

Consumer demand for fine rice varieties is higher due to their good nutrition, palatability, taste, cooking quality, and fragrance [6]. Most consumers prefer fine rice varieties with good cooking quality and aroma. Due to its unique flavor and taste, aromatic rice is highly favored. Fine and aromatic rice has recently gained popularity in Bangladesh because of its massive demand for internal consumption and export [7]. Despite the mostly favorable agro-climatic conditions, aromatic rice is cultivated in less than 2% of the countrywide rice acreage of Bangladesh. The majority of aromatic rice cultivars are low-yielding, but their higher prices and low cost of cultivation generate higher profit margins compared to other varieties [8].

Rice yield is influenced by both environmental factors and genetic traits [9]. Studies have shown that traits like the number of tillers hill⁻¹, grains panicle⁻¹, and grain weight are closely related to yield. For instance, Zhang et al. [10] linked higher grain number to higher grain yield in hybrid rice, while Huang et al. [11] highlighted the significance of panicle number to higher yield. Additional agronomic factors such as plant height, panicle length, flagleaf length, and days to flowering are also important [12].

Developing varieties with higher production and desirable agronomic traits is the ultimate aim of crop breeding. Conventional hybridization is a commonly used method to increase the yield potential of rice [13]. Hybridization is either a natural or artificial process that results in the production of a hybrid. It is conducted to produce artificial variation in the population for selection and desired combinations of traits, to integrate the desired traits into a single individual, and to exploit and use the hybrid varieties properly. The amount of genetic variation available for usage and the heritability of the desired traits are the determinants of the success of a breeding program. Hence, plant breeders might be able to produce high-yielding and well-adapted rice varieties by making use of the good adaptation and stability of yield and yield-contributing traits in rice genotypes [14].

Phenotypic coefficient of variation (PCV) and genotypic coefficient of variation (GCV) are used to measure variation in crop traits [15]. GCV indicates genetic variability, while PCV accounts for both genotypic and environmental variations. Traits with high GCV respond well to selection. A large difference between PCV and GCV shows strong environmental influence, suggesting a need to improve growing conditions [16].

Broad-sense heritability is the ratio of genetic variance to total phenotypic variance, reflecting the proportion of traits transmitted to offspring [17, 18]. Estimating heritability helps understand trait inheritance [19, 20], but combining it with genetic advance gives a better prediction of breeding progress [21]. For example, Abebe et al. [22] in rice showed that use of both measures identifies traits with high selection potential. Genetic advance estimates the expected improvement in traits like yield or disease resistance in subsequent generations [23], and is used to evaluate selection efficiency in crops such as wheat and rice [24, 25]. Accurate heritability estimation requires separating genetic effects from environmental influence [26], as emphasized by Adhikari et al. [27] in rice studies.

Heritability guides aromatic rice breeding by identifying traits like panicle length, aroma, and amylose content that respond well to direct selection [28]. Complex traits such as yield and stress tolerance require advanced methods like recurrent and genomic selection [29]. Additionally, heritability helps choose better parents for hybrids [30]. Heterosis, or hybrid vigor, occurs when hybrids outperform their parents in yield, biomass, and stress tolerance, boosting crop productivity. First described by Shull [31], it is widely used in breeding. In rice, F₁ hybrids show greater grain and culm yield than parents [32]. Heterosis can increase yields by 30 to 400% and enhance traits like disease resistance and drought tolerance [33, 34]. Both positive and negative heterosis play key roles in crop improvement –– positive heterosis boosts yield and agronomic traits, while negative heterosis benefits traits like early maturation. In rice, hybrid varieties have shown enhanced yield and stress tolerance across environments [35, 36]. For instance, Paril et al. [37] reported significant yield and stress resistance gains in hybrid rice, while Borah et al. [38] highlighted negative value of heterosis in promoting early maturity, useful for short growing seasons [36].

Heterosis in aromatic rice breeding boosts yield, stress tolerance, and grain quality while preserving aroma. Hybrids show improved biomass, tillering, and grain filling [30]. Crossing elite aromatic and high-yielding, stress-tolerant lines addresses yield gaps [39], with careful parent selection maintaining aroma [40]. This approach produces resilient, high-quality aromatic rice. In contrast, inbreeding depression refers to reduced vigor in inbred individuals due to increased homozygosity. Heterosis counteracts this by enhancing traits through crosses between inbred lines. Both phenomena result from non-additive gene action, as explained by the theories of quantitative genetics [41, 42]. In addition, the existence of transgressive segregation is common in rice when the segregating generations are produced by selfing hybrids [43].

Improving fine aromatic rice varieties is crucial for sustainable agriculture and food security. Aromatic rice varieties can be bred for better nutritional profiles, including higher levels of essential vitamins and minerals. This can help combat malnutrition in populations that rely heavily on rice as a staple food. Moreover, developing improved varieties can lead to higher yields, making it possible to produce more food on the same amount of land. This is especially important as the global population grows and arable land becomes more limited. Furthermore, improved rice varieties can be more resilient to adverse climate conditions, such as drought or flooding. Breeding for traits like drought tolerance can help farmers maintain productivity in the face of climate variability. Finally, by diversifying rice varieties and improving their quality and yield, countries can enhance their food security, reducing reliance on imports and increasing self-sufficiency.

The rice sector is indeed a cornerstone of the economy of Bangladesh, contributing 70% to agricultural GDP and one-sixth of national income, while providing 48% of rural employment [44]. Aromatic rice varieties, such as Kalijira, Kataribhog, Rasulbhog, Badshabhog, Chinigura, Basmati, Dulabhog, and Radhunipagol are of high demand but suffer from the lower yields (2–3 tons/ha) [45]. Developing high-yielding aromatic varieties could boost production, increase farmer incomes, and improve competitiveness in the global market.

The study aimed to investigate genetic variations in morphological traits among the F₁ and F₂ generations, which were obtained from a hybridization program between two fine and aromatic rice genotypes. Additionally, the study estimated heterosis, inbreeding depression and key genetic parameters, such as phenotypic and genotypic variances, phenotypic and genotypic coefficients of variation, heritability, genetic advance, and genetic advance as a percentage of the mean in both F₁ and F₂ generations. Furthermore, the F₂ population provides information on trait segregation and recombination. It is possible to identify superior and versatile genotypes by evaluating F₂ families throughout the Boro and Aman seasons. Correlation analysis supports indirect selection for yield-related traits. This study thus contributes not only to early-generation selection but also lays the groundwork for future molecular breeding efforts in rice.

Materials and methods

Time and experimental site

The current investigation was conducted in the Field Laboratory of the Department of Genetics and Plant Breeding, Bangladesh Agricultural University, Mymensingh-2202 from July 2020 to December 2021, which included three rice growing seasons i.e. Aman 2020, Boro 2021, and Aman 2021. The site is located in the Old Brahmaputra Floodplain (AEZ-9) and features non-calcareous dark grey floodplain soil [46]. The experimental field’s soil has an organic matter content of 1.29% and a pH of 6.8. The land is classified as medium-high land [47]. In Aman 2020, temperatures averaged 28.33 °C, humidity 83.9%, and rainfall 224.2 mm; Boro 2021 was slightly cooler (26.8 °C), less humid (78.28%), and drier (85 mm); while Aman 2021 was warmer (28.86 °C), similarly humid (84.15%), and had moderate rainfall (184 mm) [48].

Parent materials

Two aromatic rice varieties viz. Kataribhog and BRRI dhan50 were selected as genetic material and used as parents in hybridization. The agronomic characteristics of the parental materials are presented in Table S1. These two varieties were chosen for their complementary traits, with Kataribhog contributing exceptional aromatic grain quality and BRRI dhan50 providing high yield potential, early maturity, and improved stress tolerance [49]. Kataribhog, a traditional aromatic rice, is valued for its premium grain quality but suffers from low yield and poor stress resistance, while BRRI dhan50, a modern high-yielding variety, is highly adaptable to different agro-ecological zones of Bangladesh except for the salt-affected regions [50]. Both BRRI dhan50 and Kataribhog are photo-insensitive [51].

By crossing these two, the authors aim to combine the desirable grain quality and aroma of Kataribhog with the high yield and resilience of BRRI dhan50, ultimately developing a variety that meets both consumer preferences and farmer needs. Compatibility between the two parents was determined through genetic diversity analysis using phenotypic evaluation for complementary traits, and successful hybridization followed by the assessment of F₁ and F₂ generations for heterosis and segregation patterns.

Experimental design

In Aman 2020, the pre-germinated seeds of parental varieties were sown at different dates with 10-day intervals for synchronization of flowering. After 25 days from sowing, the seedlings of Kataribhog and after 15 days of sowing, the seedlings of BRRI dhan50 were transplanted to the main field, divided into 10 m² plots (4 m × 2.5 m), maintaining the spacing of 20 cm × 15 cm. Different transplanting times were used to synchronize flowering of Kataribhog and BRRI dhan50 for effective hybridization. Due to differences in growth duration, staggering transplanting ensured simultaneous anthesis, maximizing cross-pollination and genetic recombination for successful hybrid seed formation. Both cross (Kataribhog × BRRI dhan50) and reciprocal cross (BRRI dhan50 × Kataribhog) were conducted at the field when there were plenty of flowers of both parents. Hand emasculation and pollination methods were used, followed by proper bagging and tagging. Hybrid seeds were harvested at maturity and stored for the next season. A total of thirty plants were crossed, however, all crosses were not successful. Eventually, 36 F₁ seeds were harvested from two cross combinations. Each F₁ plant was treated as an independent genetic unit.

In Boro 2021, hybrid seeds were germinated in the laboratory, providing optimum temperature (30˚C) and moisture. After germination, they were sown in the field to develop seedlings. The F₁ seedlings were transplanted in two different lines and raised with normal agronomic practices to obtain F₂ seeds via the selfing of F₁s. A total of 11 F₁ plants were successfully raised — five from Kataribhog × BRRI dhan50 and six from BRRI dhan50 × Kataribhog. Each F₁ gave rise to an F₂ family after selfing, resulting in 11 F₂ families (designated as S1 to S11). The F₂ seeds obtained through selfing each F₁ were kept separate.

In the next Aman season 2021, the collected F₂ seeds were sown in the seedbed. In the F₂ generation, after germination, seedlings from 11 segregating progenies (11 F₂ families) were transplanted into 22 rows, with two randomized rows allocated per segregant family, following a completely randomized design (see Supplementary Appendix I). Each of the 22 rows had 100 plants obtained from random seeding that accounted for a total of 200 plants per family. Data were collected from 10 selected plants per F₂ family - five plants from each row for advancing the generation based on superiority in performance. Elements of the vegetation cycle of rice genotypes for each of the three seasons are presented in Table S2. A flowchart showing the experimental steps is presented as Fig. 1.

Fig. 1.

A flowchart showing the pipeline of experimentation and data collection

To minimize environmental variation across the seasons, all the experiments on three seasons were carried out on the same location. However, Aman season is called wet season due to the prevalence of frequent rain. But Boro season needs supplementary water for the growth of rice seedlings. So, irrigation was provided in the Boro field when necessary.

Differences in rainfall and irrigation between the Aman and Boro seasons have a significant impact on plant growth and yield. In the Aman season, rice depends mainly on monsoon rainfall, which can be unpredictable. Excess rainfall may cause waterlogging, while drought periods can stress the plants, leading to reduced growth and yield stability. In contrast, the Boro season is irrigated, ensuring a consistent water supply. This controlled irrigation helps promote uniform growth and reduces drought stress, generally resulting in higher yields compared to the Aman season. Thus, irrigation in the Boro season creates more stable conditions, leading to better plant growth and higher yields.

Data collection

Observations were recorded in all seasons for twelve quantitative traits viz. days to first flowering, plant height (cm), number of tillers hill⁻¹, number of effective tillers hill⁻¹, flagleaf length (cm), panicle length (cm), grains panicle⁻¹, filled grains panicle⁻¹, grain yield panicle⁻¹ (g), grain yield plant⁻¹ (g), straw yield plant⁻¹ (g), and weight of thousand seeds (g).

In this case, individual plant was used as a replication as data were collected at the individual plant level (see Supplementary Appendix II).

Statistical analysis

Statistical analysis such as analysis of variance (ANOVA), principal component analysis (PCA), and Pearson correlation analysis of two F₁ hybrids and their parents was done with MINITAB 20 (Minitab Inc., State College, Pennsylvania, USA) statistical software packages. One-way analysis of variance (ANOVA) was executed for different morphological traits responsible for yield following the general linear model (GLM). Tukey’s pairwise comparison was deployed as a posthoc analysis to distinguish any significant differences among genotypes (Parents, hybrids, and F₂ segregants). Two separate PCAs were conducted for two different data sets, one for parents and hybrids and another for parents and F₂ segregants. The eleven F₂ segregants used in this study were denoted by S1-S11. The phenotypic and genotypic coefficients of variation (PCV and GCV) were measured using the formula given by Burton [52] and Singh and Chaudhary [53].

Phenotypic coefficient of variation, PCV × 100.

Genotypic coefficient of variation, GCV × 100.

Here,

σ²p = Phenotypic variance.

σ²g = Genotypic variance.

= Population means.

Heritability in broad sense and genetic advance were estimated using the formula given by Johnson et al. [54] and heterosis by Briggle [55] and Fonseca and Patterson [56].

To check the maternal effect, a t-test of significance was conducted to see the variation between the average performances of two hybrids obtained from the cross (Kataribhog × BRRI dhan50) and reciprocal cross (BRRI dhan50 × Kataribhog). Heterosis was estimated as the percent change in F₁ over the mid parent (relative heterosis) and better parent (heterobeltiosis) following the formula:

Relative Heterosis × 100.

Here,

= Mean value of F₁ generation.

= Mean value of mid parent.

The test of significance of heterosis was accomplished by the ‘t’ test, as given below:

t .

Here,

S.E. of heterosis over mid parent = .

= Error variance obtained by using F₁s and parents together.

r = Number of replications.

The calculated ‘t’ value was compared with the table value of ‘t’ at error degrees of freedom at P = 0.05 and P = 0.01.

Heterobeltosis × 100.

Here,

= Mean value of F₁ generation.

= Mean value of better parent.

The test of significance of heterosis was accomplished by the ‘t’ test, as given below:

t .

Here,

S.E. of heterosis over better parent = .

= Error variance obtained by using F₁s and parents together.

r = Number of replications.

The calculated ‘t’ value was compared with the table value of ‘t’ at error degrees of freedom at P = 0.05 and P = 0.01.

Inbreeding depression was estimated as the percent change in F₂ over F₁.

Inbreeding Depression (%) X 100.

Here,

= Mean value of F₁ generation.

= Mean value of F₂ generation.

Significance of inbreeding depression was tested by ‘t’ test in terms of

t .

Here,

S.E. of inbreeding depression = .

= Error variance obtained by using F₁s and F₂s together.

r₁ = Number of replications of F₁.

r₂ = Number of replications of F₂.

The calculated ‘t’ value was compared with the table value of ‘t’ at error degrees of freedom at P = 0.05 and P = 0.01.

A t-test was conducted to investigate the significance difference between two different crosses to see any maternal effect.

Evidence of the existence of any transgressive segregation (TI) was tested by comparing the means of each of parents with the mean of each of the 11 F₂ segregant families. The value of TI was calculated as a ratio between variation in F₂ segregants (R) and that of two parents (D).

Results

Analysis of variance

For all morphological features, there were notable variations between parents and hybrids and parents and F₂ segregants according to the analysis of variance indicating a broad genetic base and the presence of sufficient variability for selection. (Tables 1 and 2). The analysis of variance in this study revealed that the genotypic mean sum of squares was significant at P ≤ 0.01 and P ≤ 0.001 for all the morphological traits.

Table 1.

Analysis of variance for morphological traits of two hybrids and their parents

Sources of variation	df	Mean Squares
		DFF	PH	TH	ETH	FL	PL	GP	FGP	YPA	YP	SYP	WTS
Genotype	3	407.60 ***	334.11 **	42.867 ***	39.733 ***	182.98 ***	53.031 ***	31,784 ***	28,459 ***	8.172 ***	1522.3 ***	353.67 **	19.502 ***
Error	16	10.87	34.27	3.6	3.35	15.35	2.045	1583	1445	0.345	47.63	51.95	0.874

df degrees of freedom, DFF Days to first flowering, PH Plant height (cm), TH Number of tillers hill⁻¹, ETH Number of effective tillers hill⁻¹, FL Flagleaf length (cm), PL Panicle length (cm), GP Grains panicle⁻¹, FGP Filled grains panicle⁻¹, YPA Grain yield panicle⁻¹(g), YP Grain yield plant⁻¹ (g), SYP Straw yield plant⁻¹ (g), and WTS Weight of thousand seeds (g)

** indicate distinctly significant at P ≤ 0.01 and *** indicate highly significant at P ≤ 0.001

Table 2.

Analysis of variance for morphological traits of F₂ segregants and their parents

Sources of variation	df	Mean Squares
		DFF	PH	TH	ETH	FL	PL	GP	FGP	YPA	YP	SYP	WTS
Genotype	12	2484.93 ***	2958.69 ***	25.047 ***	26.21 ***	175.79 ***	82.095 ***	33411.1 ***	34304.1 ***	8.84 ***	442.69 ***	596.10 ***	27.197 ***
Error	117	0.69	16.01	2.649	2.745	6.897	1.118	446.7	405.6	0.093	23.55	45.93	0.73

df degrees of freedom, DFF Days to first flowering, PH Plant height (cm), TH Number of tillers hill⁻¹, ETH Number of effective tillers hill⁻¹, FL Flagleaf length (cm), PL Panicle length (cm), GP Grains panicle⁻¹, FGP Filled grains panicle⁻¹, YPA Grain yield panicle⁻¹(g), YP Grain yield plant⁻¹ (g), SYP Straw yield plant⁻¹ (g), and WTS Weight of thousand seeds (g)

*** indicate highly significant at P ≤ 0.001

Trait association

Correlation analysis

The correlation coefficients among different morphological traits are displayed in Table 3. The correlation analysis indicated that out of 66 associations, thirty-two associations were significant at P ≤ 0.001, nineteen associations were significant at P ≤ 0.01 and three associations were significant at P ≤ 0.05, and the rest twelve associations were non-significant. Among them, fifty-four associations were positively correlated, and twelve associations were negatively correlated.

Table 3.

Correlation coefficients among morphological traits of two hybrids and their parents

Traits	DFF	PH	TH	ETH	FL	PL	GP	FGP	YPA	YP	SYP
PH	−0.387
TH	−0.558 *	0.662 **
ETH	−0.582 **	0.592 **	0.990 ***
FL	−0.415	0.833 ***	0.650 **	0.607 **
PL	−0.876 ***	0.587 **	0.671 **	0.678 **	0.593 **
GP	−0.670 **	0.751 ***	0.804 ***	0.798 ***	0.612 **	0.843 ***
FGP	−0.662 **	0.738 ***	0.799 ***	0.794 ***	0.599 **	0.838 ***	0.997 ***
YPA	−0.748 ***	0.709 ***	0.787 ***	0.785 ***	0.576**	0.871 ***	0.985 ***	0.983 ***
YP	−0.719 ***	0.683**	0.864 ***	0.867 ***	0.643**	0.867 ***	0.944 ***	0.945 ***	0.944 ***
SYP	−0.430	0.582**	0.807 ***	0.779 ***	0.544 *	0.585 **	0.743 ***	0.745 ***	0.754 ***	0.833 ***
WTS	−0.627 **	−0.124	0.060	0.107	0.010	0.472 *	0.223	0.210	0.346	0.312	0.286

DFF Days to first flowering, PH Plant height (cm), TH Number of tillers hill⁻¹, ETH Number of effective tillers hill⁻¹, FL Flagleaf length (cm), PL Panicle length (cm), GP Grains panicle⁻¹, FGP Filled grains panicle⁻¹, YPA Grain yield panicle⁻¹(g), YP: Grain yield plant⁻¹ (g), SYP Straw yield plant⁻¹ (g), and WTS Weight of thousand seeds (g)

* indicates significant at P ≤ 0.05, ** indicate distinctly significant at P ≤ 0.01, and *** indicate highly significant at P ≤ 0.001

PCA between F₁ and parents

Principal component analysis revealed the most apposite association among the traits and genotypes. The first four principal components (PCs) described 94.1% of the total data variation for four genotypes (parents and two hybrids) on twelve important morphological traits (Table 4). PC1 and PC2 explained 70.7%, and 13.1%, data variation, respectively.

Table 4.

Coefficients of principal components for morphological traits of two hybrids and their parents

Variables	PC1	PC2
Days to first flowering	−0.257	0.423
Plant height (cm)	0.263	0.358
Number of tillers hill−1	0.306	0.182
Number of effective tillers hill−1	0.303	0.129
Flagleaf length (cm)	0.247	0.284
Panicle length (cm)	0.304	−0.239
Grains panicle−1	0.329	0.020
Filled grains panicle−1	0.327	0.023
Grain yield panicle−1 (g)	0.330	−0.080
Gain yield plant−1 (g)	0.335	−0.031
Straw yield plant−1 (g)	0.281	0.060
Weight of thousand seeds (g)	0.102	−0.702
% Variation explained	70.7	13.1
P-value	< 0.001	< 0.001

The first principal component (PC1) explained the highest variation (70.7%) of the data with strong positive coefficients for all the morphological traits viz. plant height, number of tillers hill⁻¹, number of effective tillers hill⁻¹, flagleaf length, panicle length, grains panicle⁻¹, filled grains panicle⁻¹, grain yield panicle⁻¹, grain yield plant⁻¹, straw yield plant⁻¹, and weight of thousand seeds; except days to first flowering, which had negative coefficient. PC1 showed a highly significant difference among genotypes (Table 4). The PC1 clearly separated two hybrids (Kataribhog × BRRI dhan50 and BRRI dhan50 × Kataribhog) from their parents (Kataribhog and BRRI dhan50) in terms of morphological traits indicating that hybridization successfully captured diverse trait combinations (Fig. 2).

Fig. 2.

Biplot for morphological traits of two hybrids: F₁ (Kataribhog × BRRI dhan50), F₁′ (BRRI dhan50 × Kataribhog) and their parents. Here, DFF Days to first flowering, PH Plant height (cm), TH Number of tillers hill⁻¹, ETH Number of effective tillers hill⁻¹, FL Flagleaf length (cm), PL Panicle length (cm), GP Grains panicle⁻¹, FGP Filled grains panicle⁻¹, YPA Grain yield panicle⁻¹ (g), YP Grain yield plant⁻¹ (g), SYP Straw yield plant⁻¹(g), and WTS Weight of thousand seeds (g)

PC2 explained 13.1% of the total variation, which is mostly dominated by the positive coefficients of the majority of the characters except for panicle length, grain yield panicle⁻¹, grain yield plant⁻¹, and weight of thousand seeds. PC2 separated Kataribhog and Kataribhog × BRRI dhan50 from other genotypes as evident by their differential location in biplot (Fig. 2).

PCA between F₂ and parents

Considering F₂ segregants, the first four principal components (PCs) explained 90.8% of the total data variation for the genotypes on twelve yield and yield-contributing morphological traits (Table 5). PC1, PC2, PC3, and PC4 explained 54.1%, 20.4%, 8.6%, and 7.8% data variation, respectively.

Table 5.

Coefficients of principal components for morphological traits of F₂ segregants and their parents

Variables	PC1	PC2	PC3	PC4
Days to first flowering	−0.328	−0.122	0.259	0.334
Plant height (cm)	0.373	0.052	0.060	−0.112
Number of tillers hill−1	0.030	−0.613	−0.172	−0.013
Number of effective tillers hill−1	0.011	−0.617	−0.174	−0.000
Flagleaf length (cm)	0.276	−0.032	−0.069	−0.627
Panicle length (cm)	0.350	0.076	−0.156	−0.258
Grains panicle−1	0.356	0.054	0.300	0.153
Filled grains panicle−1	0.353	0.029	0.312	0.244
Grain yield panicle−1 (g)	0.352	0.120	0.028	0.331
Gain yield plant−1 (g)	0.280	−0.299	0.017	0.227
Straw yield plant−1 (g)	0.311	−0.183	−0.130	0.239
Weight of thousand seeds (g)	0.060	0.279	−0.797	0.338
% Variation explained	54.1%	20.4%	8.6%	7.8%
P-value	< 0.001	< 0.001	< 0.001	< 0.001

The first principal component (PC1) explained the highest variation (54.1%) of the data with strong positive coefficients for all the morphological traits, excluding days to first flowering, which had a negative coefficient. PC1 showed a highly significant difference among the genotypes (Table 5). The PC1 clearly separated S1, S2, S3, S4, S5, S6, and S10 from the rest F₂ segregants and their parents in terms of morphological traits as evident by their differential location in biplot (Fig. 3). This separation suggests successful recombination and emergence of novel genotypes, potentially due to transgressive segregation. PC2 explained 20.4% of total variation, which is governed by the positive coefficients of the traits viz. plant height, grains panicle⁻¹, filled grains panicle⁻¹, grain yield panicle⁻¹, and weight of thousand seeds.

Fig. 3.

Biplot for morphological traits of F₂ segregants and their parents. Here, S1-S11 F₂ segregants, DFF Days to first flowering, PH Plant height (cm), TH Number of tillers hill⁻¹, ETH Number of effective tillers hill⁻¹, FL Flagleaf length (cm), PL Panicle length (cm), GP Grains panicle⁻¹, FGP Filled grains panicle⁻¹, YPA Grain yield panicle⁻¹ (g), YP Grain yield plant⁻¹ (g), SYP Straw yield plant⁻¹ (g), and WTS Weight of thousand seeds (g)

Variability parameters

Comparison between the relative amount of phenotypic and genotypic coefficient of variation provides an estimate of the degree of variation. Grain yield plant⁻¹ showed the highest GCV both in hybrids (49.2%) and F₂ segregants (18.1%), indicating substantial heritable variation with relatively lower environmental influence (Table 6). Days to first flowering (7.6%) showed the lowest GCV in hybrid, while panicle length (3.9%) showed the lowest GCV in F₂, suggesting limited scope for genetic improvement through conventional selection (Table 6).

Table 6.

Estimation of genetic parameters for morphological traits of parents, hybrids, and F₂ segregants

Traits	F1							F2
	GV	PV	GCV (%)	PCV (%)	h2b	GA	GAM (%)	GV	PV	GCV (%)	PCV (%)	h2b	GA	GAM (%)
DFF	79.3	90.2	7.59	8.09	87.9	17.2	14.7	50.8	57.7	8.25	8.79	88.0	13.8	15.9
PH	59.9	94.2	8.54	10.7	63.6	12.7	14.0	61.9	79.0	6.22	7.03	78.2	14.3	11.3
TH	7.85	11.4	17.8	21.6	68.6	4.78	30.4	2.71	5.01	11.4	15.6	54.1	2.49	17.3
ETH	7.28	10.6	17.3	20.9	68.5	4.60	29.5	2.87	5.29	12.0	16.3	54.3	2.57	18.3
FL	33.5	48.9	19.3	23.2	68.6	9.88	32.8	5.98	12.8	6.92	10.1	46.7	3.44	9.74
PL	10.2	12.2	12.9	14.2	83.3	6.00	24.4	1.30	2.35	3.95	5.31	55.3	1.75	6.06
GP	6040	7623	35.5	39.9	79.2	142.5	65.2	1541	1993	14.1	16.0	77.3	71.1	25.5
FGP	5402	6847	36.6	41.2	78.9	134.5	67.0	1924	2343	17.1	18.9	82.1	81.9	31.9
YPA	1.57	1.91	38.8	42.9	81.9	2.33	72.4	0.51	0.61	18.0	19.6	84.1	1.35	33.9
YP	294.9	342.6	49.2	53.1	86.1	32.8	94.1	30.5	55.7	18.1	23.2	54.8	8.42	26.1
SYP	60.3	112.3	22.6	30.9	53.7	11.7	34.2	27.4	75.8	11.9	19.8	36.1	6.48	14.7
WTS	3.73	4.60	11.5	12.8	81.0	3.58	21.4	2.21	3.02	9.07	10.6	73.3	2.62	16.0

GV Genotypic variance, PV Phenotypic variance, GCV Genotypic coefficient of variation, PCV Phenotypic coefficient of variation, h²b  Heritability in broad sense, GA Genetic advance, GAM (%) Genetic advance as percentage of mean, DFF Days to first flowering, PH Plant height (cm), TH Number of tillers hill⁻¹, ETH Number of effective tillers hill⁻¹, FL Flagleaf length (cm), PL Panicle length (cm), GP Grains panicle⁻¹, FGP Filled grains panicle⁻¹, YPA Grain yield panicle⁻¹(g), YP Grain yield plant⁻¹ (g), SYP Straw yield plant⁻¹ (g), and WTS Weight of thousand seeds (g)

Heritability and genetic gain in yield-related traits

The heritability along with genetic advance is more meaningful and helps in predicting the resultant effect of selection on phenotypic expression. The magnitude of the estimated broad sense heritability (%) in this study ranged from 53.7 to 87.9% for the morphological traits which were responsible for yield. The estimation of GAM ranged from 14.0 to 94.1% for morphological traits. High heritability (h²b > 60%) with high genetic advance (GAM > 20%) showed in number of tillers hill⁻¹ (h²b = 68.6, GAM = 30.4), number of effective tillers hill⁻¹ (h²b = 68.5, GAM = 29.5), flagleaf length (h²b = 68.6, GAM = 32.8), panicle length (h²b = 83.3, GAM = 24.4), grains panicle⁻¹ (h²b = 79.2, GAM = 65.2), filled grains panicle⁻¹ (h²b = 78.9, GAM = 67.0), grain yield panicle⁻¹ (h²b = 81.9, GAM = 72.4), grain yield plant⁻¹ (h²b = 86.1, GAM = 94.1), straw yield plant⁻¹ (h²b = 53.7, GAM = 34.2), and weight of thousand seeds (h²b = 81.0, GAM = 21.4). High heritability (h²b > 60%) with moderate GAM was found in days to first flowering (h²b = 87.9, GAM = 14.7), and plant height (h²b = 63.6, GAM = 14.0) (Table 6).

In case of F₂ generation, the magnitude of the estimated broad sense heritability in this study ranged from 36.1 to 88.0% for the morphological traits which were responsible for yield. The estimation of GAM ranged from 6.06 to 33.9% for morphological traits. High heritability (h²b > 60%) with high genetic advance (GAM > 20%) showed in grains panicle⁻¹ (h²b = 77.3, GAM = 25.5), filled grains panicle⁻¹ (h²b = 82.1, GAM = 31.9), grain yield panicle⁻¹ (h²b = 84.0, GAM = 33.9), and grain yield plant⁻¹ (h²b = 54.8, GAM = 26.1). High heritability (h²b > 60%) coupled with moderate GAM was found in days to first flowering, plant height, and weight of thousand seeds (Table 6).

Estimation of heterosis and inbreeding depression

According to t-test, there was no significant difference between the two reciprocal crosses indicating that there was no notable maternal effect between Kataribhog × BRRI dhan50 and BRRI dhan50 × Kataribhog for yield and other morphological traits (Table S3). In case of F₁ (Kataribhog × BRRI dhan50), heterobeltiosis (heterosis over better parent) ranged from − 15.7 to 125.2, while it ranged from − 13.4 to 156.9 in F₁′ (BRRI dhan50 × Kataribhog) (Table 7). Days to first flowering had the lowest negative heterosis and grain yield plant⁻¹ showed the highest positive heterosis for both cross and reciprocal cross. F₁ generation estimated the highest relative heterosis (heterosis over mid-parent) of 133.1% for grain yield plant⁻¹. Meanwhile F₁′ generation showed the highest relative heterosis of 165.9% for grain yield plant⁻¹. However, the grain yield plant⁻¹ accounted for 25.9% and 44.9% inbreeding depression in F₂ generation for both cross and reciprocal cross, respectively (Table 7). It confirms the loss of heterozygosity in F₂ and the involvement of non-additive gene action. This reinforces the need for careful selection in segregating generations to recover superior homozygous lines.

Table 7.

Heterosis over better parent (BP), mid parent (MP), and inbreeding depression at F₂ for yield and yield contributing traits

Traits	Kataribhog × BRRI dhan50			BRRI dhan50 × Kataribhog
	Heterosis		Inbreeding Depression	Heterosis		Inbreeding Depression
	BP	MP		BP	MP
DFF	−15.7**	−12.0**	19.2**	−13.4**	−9.59*	24.4**
PH	15.6**	19.2**	−31.2**	7.61*	10.9*	−29.5**
TH	31.9**	37.9**	23.6**	31.9**	37.9**	20.8**
ETH	28.9**	34.8**	23.7**	31.9**	37.9**	21.9**
FL	35.4**	45.9**	7.13**	22.3*	31.8**	−8.23**
PL	16.1**	24.7**	−5.12**	19.8**	25.2**	−4.39**
GP	42.3**	80.6**	−12.9*	92.1**	100.3**	19.4**
FGP	41.9**	82.1**	−16.1**	94.2**	105.6**	22.3**
YPA	78.0**	89.1**	−11.9**	101.8**	114.4**	23.9**
YP	125.2**	133.1**	25.9**	156.9**	165.9**	44.9**
SYP	28.9*	32.1**	−29.1**	59.1**	63.2**	12.2**
WTS	−9.74**	2.30*	0.90	−1.49	11.6**	10.4**

BP Better parent, MP Mid parent, DFF Days to first flowering, PH Plant height (cm), TH Number of tillers hill⁻¹, ETH Number of effective tillers hill⁻¹, FL Flagleaf length (cm), PL Panicle length (cm), GP Grains panicle⁻¹, FGP Filled grains panicle⁻¹, YPA Grain yield panicle⁻¹(g), YP Grain yield plant⁻¹ (g), SYP Straw yield plant⁻¹ (g), and WTS Weight of thousand seeds (g)

 *- indicate significant at P ≤ 0.05, **- indicate distinctly significant at P ≤ 0.01

Evidence of transgressive segregation

The F₂ generation exhibited transgressive segregation, with all F₂ segregants displaying higher positive transgressive values than the parents for key agronomic traits including plant height, flagleaf length, panicle length, grains panicle⁻¹, filled grains panicle⁻¹, grain yield panicle⁻¹, grain yield plant⁻¹, and straw yield plant⁻¹ (Table 8). Conversely, for days to first flowering, all F₂ segregants showed negative transgressive segregation, flowering earlier than both parents (Table 8). The segregant family S10 resulted in approximately 97.24% higher grain yield plant⁻¹ compared to the better parent BRRI dhan50. The value of transgressive segregation (TI) was estimated the highest for grain yield plant⁻¹, 321.7 (Table 8). A value of TI greater than 1.0 indicated that transgressive segregation is evident for 12 morphological traits measured (Table 8).

Table 8.

Comparison of means between two parents and 11 F₂ segregants for 12 morphological traits

Genotype	DFF	PH	TH	ETH	FL	PL	GP	EGP	YPA	YP	SYP	WTS
P1	129.7 a	88.32 e	14.2 b-d	14.2 b-e	27.6 de	20.98 g	163.7 e	150.1 e	2.002 h	21.28 f	27.2 e	13.98 d
P2	118.4 b	81.83 f	13.5 cd	13.5 c-e	24.89 e	22.68 f	154.8 e	137.7 e	2.411 gh	21.34 f	30.4 de	18.37 b
S1	81 h	137.41 a	12.2 d	11.9 e	36.41 b	30.55 a	337.1 a	307.9 a	4.76 ab	29.7 c-e	46.4 ab	16.23 c
S2	81 h	132.17 ab	14.6 b-d	14 b-e	34.12 bc	28.6 b-d	299.6 bc	279.7 ab	4.47 a-c	34.06 bc	49.2 a	15.6 c
S3	98 d	124.39 c	13.5 cd	13.3 c-e	30.92 cd	27.68 de	330.1 ab	306.5 a	4.53 a-c	33.7 bc	42.0 a-c	15.8 c
S4	80 h	134.24 ab	12.7 cd	12.3 de	34.84 bc	29.4 a-c	282.5 c	268.1 b	4.92 a	35.2 a-c	49.4 a	20.03 a
S5	101 c	131.37 ab	16.5 ab	16.4 ab	34.41 bc	27.03 e	286.2 c	282.5 ab	3.91 d	39.71 ab	49.2 a	14.25 d
S6	81 h	132.24 ab	13.5 cd	13.3 c-e	35.06 b	30.26 ab	297 c	278.9 ab	4.35 b-d	32.4 bc	50.4 a	18.19 b
S7	85 g	117.67 d	12.2 d	11.9 e	37.6 ab	27.61 de	222.8 d	183.5 d	2.78 fg	22.8 ef	34.8 cde	15.99 c
S8	86 fg	118.15 d	17.2 a	17 a	34.08 bc	28.3 c-e	245.5 d	202.1 cd	3.42 e	29.6 c-e	38 bcd	16.4 c
S9	87 f	116.53 d	14.7 bc	14.5 a-d	34.61 bc	29.1 a-d	238.3 d	219.1 c	3.41 e	24.7 d-f	40.2 a-d	16.16 c
S10	81 h	130.18 bc	14.9 a-c	14.5 a-d	41.34 a	30.2 ab	300.4 bc	283.4 ab	4.16 cd	42.1 a	46.6 ab	15.6 c
S11	89 e	116.39 d	16 ab	15.8 a-c	35.58 b	28.3 c-e	227.3 d	209 cd	3.01 ef	30.25 cd	37.2 b-e	16.12 c
P-value	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001
Transgressive segregation (TI)	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Value of TI	1.77	3.24	7.14	6.43	3.85	2.07	12.84	10.03	5.23	321.67	4.88	1.32

P1 Kataribhog, P2 BRRI dhan50, S1-S11 F₂ segregants, DFF Days to first flowering, PH Plant height (cm), TH Number of tillers hill⁻¹, ETH Number of effective tillers hill⁻¹, FL Flagleaf length (cm), PL Panicle length (cm), GP Grains panicle⁻¹, FGP Filled grains panicle⁻¹, YPA Grain yield panicle⁻¹ (g), YP Grain yield plant⁻¹(g), SYP Straw yield plant⁻¹ (g), and WTS Weight of thousand seeds (g). TI R/D, R range of variation in F₂ segregants, D parental difference

Discussion

The experiment was conducted to estimate variability, correlation, and genetic parameters of hybrids and F₂ generation based on morphological traits. Additionally, heterosis for F₁ hybrids was estimated to determine their superiority compared to their parents. However, to estimate the reduced fitness, inbreeding depression of the F₂ generation was done.

The estimation of correlation among the studied morphological characters bears great importance in identifying the key characters that can be exploited for increased yield. Grain yield plant⁻¹ was significantly and positively correlated with the number of tillers hill⁻¹, number of effective tillers hill⁻¹, flagleaf length, panicle length, grains panicle⁻¹, filled grains panicle⁻¹, and grain yield panicle⁻¹ (Table 3). These findings suggest that improving tiller production and panicle-related traits can effectively enhance grain yield (Table 3) [57]. On the contrary, a significant negative correlation was found for days to first flowering with grain yield plant⁻¹ (Table 3) [58]. It suggests that selecting for earlier flowering genotypes may improve yield by enabling plants to avoid adverse late-season conditions and maximize grain-filling duration.

The majority of characters has complex inheritance pattern and are significantly influenced by multiple genes acting with several environmental factors; as a result, the study of phenotypic and genotypic coefficient of variation is very effective in determining the probability for selection-based improvement. There was an environmental effect on phenotype which is suggested by a relatively higher phenotypic coefficient of variation than the genotypic coefficient of variation (Table 6).

Grain yield plant⁻¹ had the highest genotypic coefficient of variation (GCV) and phenotypic coefficient of variation (PCV) values in hybrid, followed by grain yield panicle⁻¹ and filled grains panicle⁻¹, which indicates that these traits possess substantial genetic variability and are less influenced by the environment, making them ideal candidates for effective selection (Table 6) [59, 60]. On the other hand, days to first flowering recorded lower values of PCV and GCV, reflecting limited genetic variability and a narrow genetic base for these traits (Table 6) [61]. The F₂ segregants exhibited higher PCV and GCV values for grains panicle⁻¹, filled grains panicle⁻¹, grain yield panicle⁻¹, and grain yield plant⁻¹, recommending the scope for choosing genotypes with the aforementioned traits further supporting the potential to isolate superior genotypes in segregating generations (Table 6) [62, 63]. Contrarily, plant height showed lower PCV and GCV, suggesting less effect of this trait for improvement [61]. Overall, these findings underscore that effective selection for traits with high PCV and GCV, particularly those directly related to grain yield, could significantly enhance breeding efficiency and accelerate the development of high-yielding genotypes.

Knowledge of heritability is essential in plant breeding as it provides insights into the genetic control of traits and the potential for selection [64]. However, heritability alone does not guarantee effective selection outcomes. When considered alongside genetic advance, it offers a more reliable prediction of selection response [65]. In particular, traits showing high heritability (> 60%) coupled with high genetic advance (> 20%) are generally governed by additive gene action and thus are highly responsive to selection [54, 59].

In the present study, most traits exhibited high heritability with high genetic advance, except for days to first flowering and plant height in hybrids (Table 6) [22, 66]. These exceptions suggest greater influence from non-additive genetic effects or environmental interactions. For instance, days to first flowering is often regulated by multiple genes and affected by factors such as photoperiod and temperature [67], while plant height may be influenced by heterosis and complex parental interactions [68]. The low genetic advance associated with these traits indicates limited scope for improvement through conventional selection alone, highlighting the potential utility of molecular tools such as marker-assisted or genomic selection [69].

In the F₂ generation, traits like grains panicle⁻¹, filled grains panicle⁻¹, grain yield panicle⁻¹, and grain yield plant⁻¹ showed both high heritability and genetic gain over the percentage of mean, indicating a predominance of additive genetic control (Table 6). These traits are thus ideal candidates for early-generation selection. Conversely, traits with low heritability and genetic advance are less promising for selection and may be deprioritized in breeding pipelines [70].

No significant differences were observed between the reciprocal F₁ crosses of Kataribhog × BRRI dhan50 and BRRI dhan50 × Kataribhog for the evaluated traits, indicating the absence of maternal effects [71]. This suggests that the heterotic effect was primarily governed by nuclear genes rather than cytoplasmic or maternal inheritance.

Since increasing yield is the main objective of rice breeding, highly significant positive heterosis estimates are preferred for yield and features that contribute to yield, such as the number of tillers hill⁻¹, panicle length, flagleaf length, grains panicle⁻¹, and filled grains panicle⁻¹ [72]. On the contrary, negative heterosis is preferable for flowering and maturation to produce short-duration variety because hybrids or crosses are more likely to achieve early maturity with early flowering than those with positive heterosis. Both crosses had significant and negative relative heterosis and heterobeltiosis in days to first flowering which indicates the scope of the extensive use of heterosis for earliness [Table 7, 73].

In contrast, plant height exhibited significant positive heterosis over both the mid-parent and better parent in both crosses (Table 7). While taller plants may be more prone to lodging, the observed increase in plant height reflects the expression of dominant or overdominant gene actions contributing to vegetative growth. These results suggest that the hybrids possessed enhanced biomass, although this may not be ideal if lodging resistance or shorter stature is a breeding priority. To develop dwarf plant types, Sari et al. [74] and Shukla et al. [72] emphasized the importance of negative heterosis for plant height.

The tillering ability of a genotype is an important yield determinant, and both crosses exhibited significant positive heterosis for the number of tillers hill⁻¹ and number of effective tillers hill⁻¹. This indicates a genetic advantage of the hybrids in tiller production, which could directly contribute to higher grain output (Table 7) [75].

Both crosses also recorded significant positive heterosis for panicle length, suggesting improved reproductive development. Longer panicles are often associated with more spikelets and potentially greater grain set, which was reflected by the results.

Similarly, the number of grains panicle⁻¹ and number of filled grains panicle⁻¹ exhibited significant positive heterosis in both crosses. These components are strongly correlated with final yield and their enhancement through hybridization highlights the effectiveness of the parental combinations used (Table 7) [74, 76]. The most notable result was the high magnitude of positive heterosis for grain yield plant⁻¹ in both hybrids, confirming the superiority of these combinations over their respective parents in terms of productivity (Table 7) [77].

In contrast, the weight of thousand seeds showed negative heterosis over the better parent. While this might appear unfavorable at first glance, it aligns with breeding objectives aimed at producing fine-grained rice. The reduction in seed weight likely reflects the inheritance of slender grain traits preferred by consumers and suitable for premium markets (Table 7) [72, 78–80].

In general, 20%−30% heterosis is desirable to develop a new hybrid or an inbred after selfing. However, we observed heterosis of > 120% in cross and heterosis of > 150% in reciprocal cross for grain yield plant⁻¹. This higher level of heterosis will be declined due to inbreeding depression in the subsequent generations.

However, significant negative inbreeding depression was recorded for plant height, which revealed that the F₂ segregants had larger plant heights than hybrids (Table 7) [81]. These types of segregants would be carefully observed in the next generation whether they lodge or not for their larger plant height. Positive inbreeding depression for plant height is desirable for obtaining dwarf plant types, as reduced stature enhances plant stability, particularly under adverse environmental conditions, by minimizing the risk of lodging. Dwarf plants also tend to allocate more energy to reproductive structures, potentially improving grain yield and harvest index [82]. Exploiting inbreeding depression in plant height can facilitate the development of compact, high-yielding varieties with enhanced resistance to lodging and better resource allocation, ultimately increasing the efficiency of crop production.

Panicle length, grains panicle⁻¹, filled grains panicle⁻¹, grain yield panicle⁻¹, and straw yield plant⁻¹ showed significant negative inbreeding depression, indicating improved performance of these traits in the segregating generations. This suggests that selection based on these traits could be effective. In contrast, traits such as days to first flowering, number of tillers hill⁻¹, number of effective tillers hill⁻¹, flagleaf length, and grain yield panicle⁻¹ exhibited significant positive inbreeding depression in F₁, reflecting reduced performance likely due to increased homozygosity and the expression of deleterious alleles (Table 7) [83]. Similarly, the F₁′ generation showed positive and significant inbreeding depression for most traits, including days to first flowering, number of tillers hill⁻¹, number of effective tillers hill⁻¹, grains panicle⁻¹, filled grains panicle⁻¹, grain yield panicle⁻¹, grain yield plant⁻¹, straw yield plant⁻¹, and weight of thousand seeds, indicating their decreased output comparing to their prior generation (Table 7) [72, 84]. These findings highlight the importance of maintaining hybrid vigor in breeding programs and suggest that incorporating diverse genetic backgrounds or population improvement strategies could help sustain desirable trait performance in later generations.

According to Abdel-Moneam et al. [85] the inheritance of yield and yield components is regulated by the non-additive type of gene action. Since increased yield is the ultimate aim in plant breeding, it is recommended to give more emphasis on yield-contributing traits, as there is no separate gene for yield.

The F₂ population exhibited significant variability for all agronomic traits but the average performance of several segregating progenies was better than both parents indicating the presence of transgressive segregation (Table 8). This suggests that the parents had distinct genes and alleles that affected yield and its constituent parts [86]. Earlier studies showed that multiple linked QTLs, complementary gene action etc. are often responsible for transgressive segregation of days to heading, grain characteristics and yield of rice [87–90]. Therefore, by targeted selection in subsequent generations, there is a great chance to combine these advantageous alleles into a single genotype in order to enhance yield and associated attributes (Table 8).

One of the primary limitations of this study is the relatively small sample size, comprising only 11 F₁ plants and 11 F₂ segregant families derived from them. While these numbers were sufficient to detect initial trends and assess basic genetic parameters, such sample sizes may not fully capture the genetic diversity or environmental interactions present in a larger breeding population. Some degree of inflation in these values may likely be attributable to limited number of samples although it is expected that an inflated estimate of heterosis, for example, will be gradually reduced with increasing homozygosity in the subsequent generation. Further studies incorporating a large number of replicates across multiple environments may refine these estimates and confirm their stability.

This experiment was carried out in two different rice growing seasons i.e. Boro and Aman. These two seasons differ notably in terms of environmental conditions, with Boro being an irrigated and relatively stable season, while Aman is dependent on monsoon rainfall, which introduces greater variability in moisture, temperature, and other climatic factors. Such differences could influence trait expression, particularly for traits like flowering time, plant height, and yield components. To minimize these effects, all experiments were conducted at the same geographical location using consistent management practices. Furthermore, the environmental influence was statistically accounted for through the estimation of phenotypic and genotypic coefficients of variation (PCV and GCV), where relatively small differences between the two indicated limited environmental interference were estimated for most of the traits.

Conclusions

The current study was conducted to determine heterosis, correlation, genetic variability, and inbreeding depression in F₁ and F₂ generations. Substantial variations were observed among all the genotypes for all morphological traits, highlighting the potential for genetic improvement in rice. Grain yield plant⁻¹ was positively correlated with all the studied traits, excluding days to first flowering indicating that selection for these traits could improve yield performance. Additionally, grain yield panicle⁻¹ and grain yield plant⁻¹ showed the highest PCV and GCV, suggesting a high potential for selection response. High heritability and genetic advance accounted for grains panicle⁻¹, filled grains panicle⁻¹, grain yield panicle⁻¹, and grain yield plant⁻¹. However, both the hybrids showed the highest positive heterosis for grain yield plant⁻¹, demonstrating that exploiting heterosis can significantly increase yield potential which can address the growing demand for food security, especially in regions where rice is a staple crop. The grain yield plant⁻¹ also accounted for a remarkably higher value of transgressive segregation. These findings underscore the importance of hybrid breeding as a strategy to enhance rice productivity. A total of 36 promising segregants from the F₂ generation, with superior yield characteristics, were identified and selected for further generation advancement. The results support the development of high-yielding commercial hybrids and the adoption of strategic breeding approaches, such as recurrent selection and hybridization, to meet the growing demand for rice and improve food security. Future research should evaluate a wider range of genotypes across diverse environments to assess the stability of heterosis and genetic potential. Additionally, exploring the molecular mechanisms of heterosis, inbreeding depression, and the role of genomic selection, along with enhancing stress tolerance, will be crucial for developing more resilient rice varieties for climate change adaptation.

Abbreviations

%

Percentage

ANOVA

Analysis of variance

BP

Better parent

DAT

Days after transplanting

Df

Degrees of freedom

DFF

Days to first flowering

ETH

Number of effective tillers hill^-1

et al.

And other people (Latin et alia)

FGP

Filled grains panicle^-1

FL

Flagleaf length

GA

Genetic advance

GAM

Genetic advance as percentage of mean

GCV

Genotypic coefficient of variation

GDP

Gross domestic product

GP

Grains panicle^-1

GV

Genotypic variance

h²b

Heritability in broad sense

i.e

That is

MP

Mid parent

P

Probability value

PCV

Phenotypic coefficient of variation

PCA

Principal component analysis

PH

Plant height

PL

Panicle length

PV

Phenotypic variance

SYP

Straw yield plant^-1

TH

Number of tillers hill^-1

Viz.

Namely

 

Mean

WTS

Weight of thousand seeds

YP

Grain yield plant^-1

YPA

Grain yield panicle^-1

σ²g

Genotypic variance

σ²p

Phenotypic variance

Authors’ contributions

A. H. K. R. and A. M. M. M. U. T. conceived and designed the study. A. M. M. M. U.T. conducted the experiment, formally analyzed the data, and wrote the original draft. L. H. provided technical advice on how to prepare tables and figures. M. A. H. and S. N. assisted in data analysis and edited the manuscript. A. H. K. R. supervised and monitored the experimental work, provided technical support, and extensively revised the manuscript. All authors read and approved the final manuscript.

Funding

This research received no funding.

Data availability

The data will be available from the corresponding author on reasonable requests.

Declarations

Ethics approval and consent to participate

This article does not include any clinical trials and/or studies involving human participants, animals, endangered, or protected species.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.