Molecular analysis of BC4F1 plants
Molecular screening using the Bradbury primer set identified 20 heterozygous BC4F1 plants out of 42 individuals analysed (Fig 1A). These plants showed both the 355 bp fragment associated with the non-aromatic allele and the 257 bp fragment associated with the aromatic allele. Chi-square analysis indicated that the segregation pattern followed the expected 1:1 ratio for a backcross population (χ
2= 3.0<3.841), confirming the predicted distribution of genotypes. Additionally, ten representative BC4F1 plants carrying the aromatic allele were documented alongside both parental varieties for morphological comparison (Fig 1B).
Evaluation of heading date, plant height and panicle length
Heading date is a key phenological trait in rice. The BC4F1 plants had a mean heading date of 61 days after transplanting (DAT), which was later than Inpari 33 (57.2 DAT) and comparable to Merah Wangi (61.2 DAT) (Fig 2A, Table 1). The low coefficient of variation (2.49%) indicates that heading date was relatively uniform within the BC4F1 population.
The average plant height of the BC4F1 plants was 100.38 cm, slightly taller than Inpari 33 (98.4 cm) but shorter than Merah Wangi (128.6 cm) (Fig 2B, Table 1) and the narrow variability observed for this trait (6.13%) indicates limited phenotypic dispersion within the population. Panicle length averaged 25.27 cm, which was longer than Inpari 33 (23.66 cm) and shorter than Merah Wangi (28.17 cm) (Fig 2C, Table 1) and this trait exhibited broad variability (5.75%), reflecting considerable differences in panicle size among BC4F1 individuals.
Evaluation of productive tillering and yield component
Tillering is an important trait in rice because each tiller has the potential to develop into a panicle and contribute directly to yield. The BC4F1 plants produced an average of 44.7 tillers per plant, an intermediate value between Inpari 33(51.6) and Merah Wangi (26.6), with broad variability (15.53%) (Fig 3A, Table 2). On average, BC4F1 plants formed 32.65 productive tillers per plant, fewer than Inpari 33 (44.2) but more than Merah Wangi (21) and this trait also displayed broad variability (24.63%) (Fig 3B, Table 2).
The number of filled grains per panicle is a vital yield component because it directly contributes to total grain production. The BC4F1 plants averaged 105.55 filled grains per panicle, a value higher than Inpari 33 (77.33) but lower than Merah Wangi (175.85) (Fig 3C, Table 2). The proportion of empty grains was 25.81%, which was slightly lower than Inpari 33 (28.72%) and higher than Merah Wangi (23.79%) and this trait exhibited broad variability (23.1%) (Fig 3D, Table 2). For grain yield per plant, the BC4F1 lines produced an average of 43.41 g, exceeding Inpari 33 (33.40 g) but remaining below Merah Wangi (50.31 g), with yield showing broad variability (9.15%) across the population (Fig 3E, Table 2).
The 1000-grain weight and kernel dimensions are important indicators of grain size and density, which influence both yield and market value. The BC4F1 plants had an average 1000-grain weight of 22.76 g, lower than both parents, Inpari 33 (25.51 g) and Merah Wangi (25.75 g) (Fig 4B, Table 3), with broad variability (5.27%). The average kernel length of the BC4F1 plants was 7.05 mm, an intermediate value between Merah Wangi (6.01 mm) and Inpari 33 (7.16 mm) (Fig 4C, Table 3). Kernel width averaged 2.22 mm, slightly smaller than Inpari 33(2.23 mm) and markedly smaller than Merah Wangi (2.90 mm) (Fig 4D, Table 3). Kernel length and width exhibited broad variability (3.66% and 3.12%, respectively), reflecting considerable phenotypic diversity within the BC4F1 population.
This study demonstrates that Marker-Assisted Backcrossing (MAB) effectively introgressed the aromatic trait into Inpari 33, while maintaining desirable agronomic performance in the BC4F1 generation. Molecular screening using the Bradbury marker identified 20 heterozygous BC4F1 plants, confirming the transfer and stable inheritance of the
BADH2 aromatic allele. As a recessive loss-of-function mutation governs rice fragrance, aroma develops only when both
BADH2 alleles are inactivated, leading to the accumulation of 2-acetyl-1-pyrroline (
Behera and Panda, 2023;
Bradbury et al., 2005). The BC4F1 plants exhibited an intermediate heading date of 61 days after transplanting, with low variation indicating a uniform flowering response. Rice as a short-day crop typically requires 7 to 10 hours of light to induce flowering (
Anggraeni et al., 2021) and synchronized heading is beneficial for crop management
(Khotasena et al., 2022). A flowering time of around 60 days is advantageous in tropical environments such as Indonesia, where early- to medium-maturing varieties support higher cropping intensity
(Adhi et al., 2024; Liang et al., 2024). Panicle length averaged 25.27 cm, longer than both parents, a pattern that may reflect transgressive segregation
(Rieseberg et al., 1999). Longer panicles are often associated with increased grain number per panicle
(Fang et al., 2024), although genetic and environmental factors jointly influence this trait
(Yadav et al., 2023) and optimal assimilate distribution is required to avoid increases in poorly filled grains
(Li et al., 2023; Won et al., 2022).
The BC4F1 plants also exhibited enhanced tillering, producing an average of 45 tillers per plant, representing an intermediate expression level between Inpari 33 and Merah Wangi. Tillering is governed by genetic, hormonal and environmental regulations, involving the initiation of axillary buds and their subsequent elongation
(Yuan et al., 2024). Although the total number of tillers is one component of yield potential, productive tillers are more critical. The BC4F1 plants produced an average of 33 productive tillers, indicating promising yield attributes, while the observed variability suggests the need for further selection to improve trait stability. Improved grain filling was also observed, as the BC4F1 plants produced more filled grains per panicle than both parents. Grain filling is strongly influenced by environmental conditions, including temperature, water supply and light availability
(Hu et al., 2021). A relatively high proportion of empty grains in the BC4F1 plants indicates heterogeneity in grain development, which may be associated with stress conditions such as heat, drought, or pest incidence during flowering
(Shrestha et al., 2022). Because the present evaluation was conducted under non-replicated, single-season field conditions, part of this variability may also reflect uncontrolled weather fluctuations rather than genotype alone, as comparable weather-driven variation in panicle number, grains per panicle and 1000-grain weight has been documented in other aromatic rice cultivars
(Hanglem et al., 2025).
A comprehensive analysis of grain quality traits showed that the BC4F1 plants performed intermediately between Inpari 33 and Merah Wangi. The lower 1000 grain weight in the BC4F1 population suggests partial inheritance of grain mass from both parents. The broad variation observed indicates considerable genetic diversity, which can be utilized to select lines with improved grain size and market value
(Li et al., 2018; Xuedan et al., 2023). Kernel morphology also showed intermediate expression, with average kernel length of 7.05 mm and width of 2.22 mm. The coefficients of variation for these traits highlight meaningful phenotypic diversity, providing breeders with opportunities to refine grain shape and size characteristics that influence consumer preferences and market acceptance
(Saidon et al., 2020).