Assessment of growth variation among indigenous germplasm of turmeric (Curcuma longa L.)

Introduction

Turmeric (Curcuma longa L.) is an economically important rhizomatous species of the family Zingiberaceae and is widely cultivated in tropical and subtropical production systems. Its value is associated with culinary use, natural colour, medicinal applications, and the accumulation of curcuminoids and volatile constituents. For a vegetatively propagated crop, germplasm collections are particularly important because useful variability has to be identified and conserved within clonally maintained material. Consequently, systematic phenotypic characterization remains an essential first step before advanced breeding, molecular characterization, or targeted multilocation evaluation. Indian turmeric germplasm has repeatedly shown substantial morphological and genetic diversity. Roy et al. (2011) [10] , while evaluating 83 accessions from north-eastern India, reported useful variation in quantitative characters and demonstrated that multivariate methods could separate accessions on the basis of morphological traits. Mishra et al. (2015) [8] likewise documented considerable natural and genetic variability among 65 turmeric germplasms and reported high heritability for plant height. Gupta et al. (2015) [5] further demonstrated considerable genetic divergence among 65 genotypes using 13 agro-morphological characters, with plant height Among the major contributors to divergence. These findings establish that growth architecture is not merely descriptive: it can serve as an informative component of germplasm differentiation. More recent studies continue to support the value of growth traits in turmeric characterization. Anindita et al. (2020) [1] . reported broad phenotypic diversity in Indonesian local turmeric accessions and specifically identified plant height, number of shoots, number of leaves on the main shoot, and petiole length among the traits that distinguished accessions. This combination is especially relevant to the present study because the same four structural traits are available in the supplied Raigarh dataset. Similarly, Deve and Sharma (2022) [3] reported wide variability among turmeric genotypes, while Patel et al. (2023) [9] showed significant variability for growth and maturity-related characters in Indian germplasm and highlighted the usefulness of conventional phenotypic characterization alongside molecular diversity analysis. Current evidence also indicates that phenotypic characterization is increasingly being integrated with multivariate and molecular approaches. Verma et al. (2015) [12] found substantial molecular diversity among indigenous Indian turmeric genotypes, with high polymorphism and broad genetic distances. Dudekula et al. (2022) [4] used rhizome yield traits and curcuminoid profiles to resolve genetic diversity among Indian turmeric germplasm, reinforcing the concept that diverse phenotypes can provide a valuable entry point for more detailed selection. Most recently, Das et al. (2026) [2] demonstrated substantial variability for plant height and related traits across turmeric genotypes and emphasized the role of growth characters in integrated selection. Thus, a well-documented regional collection can contribute to breeding not only by identifying immediately vigorous accessions but also by preserving contrasting phenotypes for future traitspecific selection. The present dataset originates from the AICRP on Spices, CARS, Raigarh, and Chhattisgarh and comprises 54 indigenous accessions evaluated for days to maturity and four vegetative growth attributes over two seasons. Although the source document contains only these growth-related observations, it offers an unusually coherent opportunity to examine the magnitude, direction, and stability of phenotypic variation. The study therefore focuses specifically on (i) quantifying accessionwise variation in maturity and vegetative growth, (ii) identifying contrasting accessions based on pooled performance, (iii) comparing the consistency of trait expression between the two seasons, and (iv) examining accession-level interrelationships among the recorded traits. Importantly, the study does not infer rhizome yield or quality performance because those variables are not part of the supplied dataset.

Materials and Methods

Experimental site and genetic material

The experiment was conducted at the Research Farm of AICRP on Spices, College of Agriculture and Research Station (CARS), Raigarh, Chhattisgarh, India. The crop was turmeric (Curcuma longa L.). The material comprised 54 indigenous accessions, including the checks Roma, Suranjana, and BSR-2, together with the named Chhattisgarh materials Chhattisgarh Haldi-1 (IT-2), Chhattisgarh Haldi-2 (IT-10), and CG Raigarh Haldi-3 (IT-36). The source document does not provide soil analytical values, coordinates, rainfall, temperature, or other site-specific environmental measurements; these were therefore not added to the present manuscript. The experimental structure therefore provided replicated evaluation across two consecutive Kharif seasons. Each accession was represented in the RBD with two replications, and the same spacing and plot dimensions were used. The use of two seasons is important for distinguishing persistent accession differences from season-specific shifts in mean expression.

Traits recorded

Statistical analysis

The source tables report SEm (±), CD at the 5% probability level, and CV for each season and for pooled data. These supplied inferential statistics were retained without modification. For additional descriptive analysis, minimum, maximum, range, arithmetic mean, sample standard deviation, and descriptive coefficient of variation were calculated from the 54 reported pooled accession means for each trait. Seasonal means were calculated across the 54 accession means for each year, and Pearson correlation coefficients were calculated among pooled accession means for the five recorded traits. Pearson correlations were also calculated between the two seasonal accession means for each trait as a descriptive indicator of cross-season consistency. Because the source file does not report genotype × environment interaction statistics, F-values, variance components, heritability, genetic advance, or replication-level residual data, none of these parameters was reconstructed or inferred. Likewise, no artificial observations were generated to fill the missing analytical levels. This conservative approach preserves the evidentiary boundary of the supplied dataset.

Results

Overall magnitude of variation

The 54-accession collection expressed a wide phenotypic spectrum across all five recorded traits (Table 3). Among the pooled means, days to maturity ranged from 197.75 to 243.50 days, a spread of 45.75 days. Plant height varied from 74.71 to 134.79 cm, a 60.08 cm difference between the minimum and maximum accessions. Number of shoots ranged from 2.04 to 6.16, while leaves on main shoots ranged from 4.01 to 9.08. Petiole length showed the broadest proportional spread, from 11.39 to 35.18 cm. The pattern was biologically coherent across the four structural growth traits. IT-1 occupied the maximum position for plant height, shoot number, leaf number, and petiole length, whereas IT-43 occupied the minimum position for all four. Conversely, IT-1 was the earliest accession for maturity and IT-43 was the latest. Thus, the extremes of the dataset were not randomly distributed across traits; they represented strongly contrasting plant architectures. The supplied pooled SEm was 2.563 cm and the pooled CD at 5% was 7.27 cm. The extreme difference therefore exceeded the reported CD by more than eightfold. Such a large contrast is consistent with substantial phenotypic differentiation in plant stature within the evaluated germplasm. 6). The top group comprised IT-1, IT-28 (5.94), IT-3 (5.87), . The lowest group consisted of IT-43, BSR-2 (2.10), , . The maximum-to-minimum difference was 4.12 shoots, or approximately 202% of the minimum value. The pooled SEm was 0.128 and pooled CD (5%) was 0.36 shoots, while the pooled CV was 4.54%. The difference between IT-1 and IT-43 was therefore far larger than the supplied CD. The close correspondence between shoot number and the ranking of plant height suggests that shoot proliferation and overall vegetative stature were expressed as a coordinated phenotype in this collection.

Fig 4:

Pooled number of shoots for all 54 turmeric accessions.

Variation in leaves on main shoots

Leaves on main shoots ranged from 4.01 in IT-43 to 9.08 in IT-1 (Table 7). The highest values were recorded for IT-1, IT-28 (8.92), , . The lowest values were observed in IT-43, BSR-2 (4.37), , . The 5.07-leaf contrast between the extremes represents a 126.4% difference relative to the lowest accession. The source tables report pooled SEm of 0.215, pooled CD (5%) of 0.61 leaves, and pooled CV of 4.64%. The extreme difference was more than eight times the pooled CD. The parallel ranking of leaf number with plant height and shoot number indicates that the collection contains distinct high-leaf and low-leaf plant architectures rather than isolated differences in a single character.

Variation in petiole length

Petiole length displayed the widest proportional variation among the measured structural traits, ranging from 11.39 cm in IT-43 to 35.18 cm in IT-1 (Table 8). The high group included IT-1, IT-28 (34.84 cm), , . The low group included , , . The absolute difference between IT-1 and IT-43 was 23.79 cm, equivalent to 208.9% of the IT-43 value. The pooled SEm for petiole length was 1.153 cm, the pooled CD was 3.27 cm, and the pooled CV was 7.03%. Again, the extreme contrast greatly exceeded the supplied CD. Because petiole length contributes to the spatial arrangement of the leaf canopy, its close association with plant height and leaf number strengthens the interpretation that the collection contains distinct architectural types.

Fig 6:

Pooled petiole length for all 54 turmeric accessions.

Seasonal comparison and consistency

Across all 54 accession means, the seasonal averages were remarkably stable (Table 9). Mean days to maturity changed by only 0.03 days between seasons, plant height increased by 0.99 cm, shoot number increased by 0.09, leaves on main shoots increased by 0.02, and petiole length increased by 0.33 cm. These small shifts indicate that the collection-level central tendency was broadly maintained across the two evaluation seasons. The strongest cross-season correspondence occurred for petiole length (r = 0.991) and plant height (r = 0.988), followed by leaves (r = 0.967) and shoot number (r = 0.965). Days to maturity also showed a strong association (r = 0.907). These coefficients indicate that accession ranking was generally stable across years, although they are correlations of accession means and should not be interpreted as formal estimates of broad-sense heritability or genotype × environment stability. Among the three named checks, Roma showed 86.98 cm height and 14.99 cm petiole length, Suranjana showed 97.78 cm and 20.48 cm, and BSR-2 showed 77.09 cm and 11.96 cm, respectively. Thus, the local Chhattisgarh materials, particularly IT-10, expressed stronger vegetative growth than the named checks in this dataset. Such comparisons are useful for identifying regional materials that merit further testing, but superiority for growth alone should not be equated with superiority for yield, curcumin content, disease resistance, or market quality. (2011) [10] reported substantial agro-morphological diversity among turmeric accessions and identified plant height and leaf number among traits associated with rhizome yield. Mishra et al.

Performance of named checks and Chhattisgarh materials

(2015) [8] also found considerable variability among 65 germplasms and reported very high broad-sense heritability for plant height. Gupta et al. (2016) [6] identified substantial variability and interrelationships among 13 agro-morphological traits, reinforcing the value of vegetative characters as components of turmeric phenotypic characterization. An especially close methodological parallel is provided by Anindita et al. (2020) [1] , whose turmeric dataset included plant height, number of shoots, number of leaves on the main shoot, and petiole length as distinguishing phenotypic traits. The present Raigarh dataset reproduces this same core architectural set and demonstrates that these traits can jointly discriminate accessions in a geographically different germplasm context. Although the two collections cannot be directly compared numerically because their environments and genetic materials differ, the recurrence of the same informative trait classes supports their continued use in germplasm characterization.

Early maturity versus vegetative vigor

One of the most striking patterns was the inverse relationship between maturity duration and vegetative expression. The pooled correlation between days to maturity and plant height was -0.989, and correlations with shoots, leaves, and petiole length were similarly negative (-0.994, -0.992, and -0.984, respectively). At the phenotypic level, this indicates that accessions requiring more days to maturity tended, within this particular collection, to have lower values for the measured structural growth traits. The extreme contrast between IT-1 and IT-43 illustrates this pattern: IT-1 matured 45.75 days earlier while simultaneously exceeding IT-43 by 60.08 cm in height, 4.12 shoots, 5.07 leaves, and 23.79 cm in petiole length.Such a relationship should not automatically be interpreted as a universal physiological rule for turmeric. Days to maturity integrates developmental progression and environmental response, while plant height, shoot number, leaf number, and petiole length describe structural expression. The strong inverse association may therefore reflect the particular genetic composition of this collection, correlated developmental strategies, or shared responses to the environmental conditions of the evaluation site. Patel et al. (2023) [9] also emphasized variation in days to maturity and other agronomic traits among Indian turmeric genotypes, illustrating that phenology can be a useful discriminating character in germplasm evaluation. From a breeding perspective, the pattern creates an important selection question: whether early maturity and vigorous vegetative growth can be retained simultaneously in a genotype. IT-1 is notable precisely because it combines both features in the supplied data. However, because rhizome yield and quality are absent from the current dataset, it would be scientifically inappropriate to conclude that IT-1 is a superior production genotype. Instead, it should be regarded as a high-priority candidate for subsequent yield and quality testing.

Plant architecture and potential functional interpretation

Plant height, shoot number, leaf number, and petiole length were all positively correlated at values approaching unity. The plant architecture implied by these data is internally consistent: accessions with more shoots also tended to have taller plants, more leaves on the main shoot, and longer petioles. Such coordinated expression is biologically plausible because shoot initiation, leaf production, petiole elongation, and overall canopy extension are interconnected components of vegetative development. Nevertheless, the present dataset does not contain leaf area, biomass, photosynthetic rate, light interception, or rhizome accumulation measurements; therefore, mechanistic explanations should remain hypotheses rather than direct conclusions. The close association between vegetative traits and agronomic performance has been noted in turmeric research. Roy et al. (2011) [10] reported positive associations of plant height and leaf number with rhizome yield in north-eastern Indian accessions. Das et al. (2026) [2] similarly documented significant variability in plant height and other growth attributes and demonstrated that growth characters can contribute to multivariate differentiation among turmeric genotypes. These studies support the use of plant architecture as a useful intermediate phenotype, while also emphasizing that yield relationships must be validated independently. The present results also suggest that petiole length deserves greater attention in regional turmeric characterization. It had a pooled descriptive CV of 7.03%, somewhat higher than plant height (3.46%) and leaves on main shoots (4.64%) but still within a moderate experimental range, and it exhibited very high cross-season consistency (r = 0.991). This combination of measurable variability and strong temporal consistency makes petiole length potentially useful as a complementary descriptor, particularly when combined with other architecture traits.

Consistency across seasons

The two-season comparison provides an important robustness check. Across the collection, seasonal means changed only slightly, while the accession-level correlations between years were high for every trait. Plant height and petiole length showed the strongest cross-season correspondence, suggesting that accession ranking for these structural traits was highly conserved between 2021-22 and 2022-23. Days to maturity showed the lowest, though still strong, correlation (r = 0.907), indicating somewhat greater sensitivity of accession ranking for phenological duration. This consistency is valuable for germplasm selection because a phenotype that persists across seasons is more useful as a screening character than one that is strongly season-dependent. However, the present analysis cannot partition the causes of stability. No environmental covariates, replication-level observations, or genotype × environment interaction statistics were provided. Therefore, the results demonstrate repeatability of accession means within the two supplied seasons, not formal stability across environments. The low pooled CV values reported in the original tables further support the precision of the trial. Pooled CV ranged from 1.75% for days to maturity to 7.03% for petiole length. These values indicate that experimental variation relative to the trait means was comparatively limited. The combination of low CV, reported CD values, and strong cross-season correlations strengthens confidence that the major contrasts are not simply the result of random experimental noise.

Implications for Indian turmeric germplasm management

India possesses extensive turmeric diversity, and molecular studies have demonstrated that indigenous collections retain substantial genetic variation. Verma et al. (2015) [12] detected high polymorphism and broad genetic distances among Indian turmeric genotypes, while Dudekula et al. (2022) [4] demonstrated that Indian germplasm can be differentiated using rhizome yield traits and curcuminoid profiles. Against this background, the present regional collection should be viewed as a phenotypic layer of a larger conservation and pre-breeding strategy.

The Chhattisgarh materials included in the experiment are particularly relevant for regional breeding. IT-10 (Chhattisgarh Haldi-2) expressed a comparatively strong vegetative phenotype and relatively early pooled maturity, whereas IT-36 showed an intermediate architecture and IT-2 was more moderate in growth. These differences provide a rational basis for prioritizing materials for follow-up trials under local conditions. Yet a breeding programme should avoid selecting exclusively for vegetative vigor because rhizome yield, dry recovery, curcumin content, disease resistance, and market-preferred quality can respond differently. Recent molecular work also supports integrating phenotypic and molecular evidence. Patel et al. (2023) [9] combined morphological characterization with SSR-based molecular diversity, while Verma et al. (2015) [12] used DAMD and ISSR markers to reveal genetic relationships among indigenous Indian turmeric. A logical next step for the Raigarh material would therefore be to combine the present growth descriptors with rhizome yield and quality measurements and, where resources permit, molecular marker data. Such integration would help distinguish accessions that are merely vigorous from those that combine desirable architecture with commercially valuable rhizome traits.

Methodological significance and limitations

The present study deliberately follows a data-conservative analytical strategy. Only values explicitly provided in the source document were used. Consequently, the manuscript does not invent F-statistics, genotype × environment effects, heritability estimates, genetic advance, or replication-level distributions. This is important because the source file supplies treatment means and summary error statistics, but not the complete raw experimental matrix required to reconstruct every conventional quantitativegenetic parameter. The strongest limitation is therefore the absence of yield and quality variables in the supplied Paper 2 growth dataset. Growth superiority cannot be assumed to translate into rhizome yield superiority. A second limitation is that the Pearson correlation analysis uses pooled accession means, so the very high coefficients should be interpreted as descriptors of coordinated accession phenotype rather than as genetic correlations. A third limitation is that only two seasons at one research location are represented; broader adaptation requires multi-location and multiyear testing. Despite these limitations, the study has a clear practical value: it identifies a small set of contrasting accessions that can be targeted for the next stage of evaluation. IT-1, IT-28, IT-3, IT-39, and IT-11 repeatedly ranked highly for structural growth, while IT-43, BSR-2, IT-51, IT-25, and IT-45 represented the opposite end. Such contrasting material is valuable for parent selection, descriptor development, and studies of the biological basis of architecture and maturity.

Supplementary Tables

The following tables reproduce the complete numerical dataset.

Values are transcribed from the original document; the pooled column is the source-reported pooled mean. These tables are included to preserve transparency and allow verification of all accession-level analyses reported in the manuscript.

Supplementary Table S1 From a breeding and conservation perspective, IT-1, IT-28, IT-3, IT-39, and IT-11 merit priority for subsequent evaluation because of their consistently strong structural growth, while the contrasting low-growth group may be useful for diversity studies and trait-specific selection. Among the named Chhattisgarh materials, IT-10 showed particularly strong vegetative performance. However, no accession should be declared superior for overall productivity on the basis of growth alone. Future evaluation should integrate rhizome yield, dry recovery, curcuminoid concentration, disease response, and molecular diversity before recommending accessions for cultivar development or commercial deployment.

Practical recommendations for subsequent evaluation

Retain the complete 54-accession collection as a phenotypically diverse regional germplasm resource. Prioritize IT-1, IT-28, IT-3, IT-39, IT-11 and IT-10 for independent trials incorporating rhizome yield and quality traits. Include contrasting low-growth accessions such as IT-43 and BSR-2 in diversity and trait-association studies rather than discarding them solely because of low vegetative growth. Conduct multi-location evaluation to determine whether the observed maturity-growth relationship persists across environments.

Combine the present morphological descriptors with molecular markers and rhizome quality measurements for integrated germplasm characterization.

Data availability

All numerical observations used in this article were derived from the supplied document titled "Paper 2 Growth". The complete accession-wise values for the five traits are reproduced in the supplementary tables of this manuscript, together with the original season-wise and pooled values. No external dataset was merged with the experimental observations.

Conflict of interest

No conflict-of-interest statement was provided in the source document; therefore, no additional declaration is made in this data-based manuscript.

References