Genetic divergence in Capsicum annuum L. under protected conditions based on new DUS descriptors

Abstract

A comprehensive three-year pioneer study to develop DUS descriptors of Capsicum annuum under protected conditions was undertaken that defined 56 morphological descriptors and were utilized for genetic divergence analysis. The study revealed substantial phenotypic variability across plant, leaf, floral, fruit and seed traits, with the Shannon Diversity Index (H′) ranging from 0.09 to 1.57. High variability was recorded for length of first internode (H′ = 1.57), pericarp thickness (H′ = 1.43), fruit shape (H′ = 1.38), plant height (H′ = 1.24), fruit length (H′ = 1.21) and fruit stalk length (H′ = 1.10), indicating broad allelic diversity and strong potential for breeding interventions. Bell pepper genotypes demonstrated high variability in locule number (H′ = 1.10). The population was dominated by key traits, including thin pericarp (41.38%), long stalk length (50.00%), dentate calyx margins (86.21%), constricted calyx (91.38%), strong pedicel attachment (62.07%), presence of blossom-end appendage (93.10%), medium maturity (50.00%), heavy seed weight (> 6.0 g) and generally low seed recovery efficiency (65.52%). The genotypes were classified into ten distinct clusters using Tocher’s method, where clusters clusters I, II, IV, VII, VIII and X comprised hot pepper and paprika genotypes, while clusters III, V, VI and IX comprised bell pepper genotypes. Principal component analysis revealed that the first four principal components explained most of the existing variation (> 70%) in the germplasm. Overall, the substantial diversity observed across traits highlights the genetic potential of the evaluated Capsicum genotypes as a valuable resource for crop improvement, conservation and targeted breeding of desirable morphological and agronomic traits.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-54456-6.

Introduction

The genus Capsicum annuum L. (2n = 2x = 24) is one of the most diverse groups that encompasses a broad range of economically important vegetable and spice crops worldwide namely, chilli, paprika and bell pepper¹. These species display wide range of genetic and phenotypic variability for plant/growth architecture, flower and fruit morphology and behaviour, pungency and quality traits that make them perfect to study genetic diversity and varietal differentiation vis-a vis exploiting them for crop improvement programs^2,3. Further, the extensive genetic diversity helped their adaptation to diverse agro-climatic conditions and production systems that include field and protected conditions, helped to meet the diverse market preference and industrial requirements^4,5.

Recently, protected cultivation under net and polyhouses has emerged as an important production system for capsicum that has gained significant reputation and attracted growers particularly for bell pepper and speciality chilli types due to their ability to ensure year-round supply, high productivity, better fruit quality, efficient resource utilization, reduced biotic and abiotic stresses and more profitability than field conditions⁶. These crops exhibit a distinctly more vigorous plant growth, leaf size and prolonged duration of flowering and fruiting, fruit size and shape, colour intensity, better fruit set and uniform fruit development than field conditions due to regulated microclimate and nutrient supply⁷. Keeping these aspects in view, Protection of Plant Varieties and Farmers’ Rights Authority of India provided funds to precisely determine the environment specific expressions on morphological traits by formulating DUS guidelines for these crops under protected conditions to harmonize descriptors across chilli, paprika and bell pepper that reflect the wide range of variation within the species. Based on three years studies (2021–2023), 56 morphological descriptors were formulated in accordance with the provisions of the PPV&FRA, 2001 that will enable precise discrimination among genotypes belonging to differential market classes. Of these, 49 are established descriptors originally developed for open conditions and were adapted for evaluation under protected environment. Notably, no dedicated descriptor framework exists for protected cultivation in Capsicum annuum L. To address this gap, seven additional descriptors were developed for protected conditions [anthocyanin colouration of nodes (at 50% flowering), intensity of anthocyanin colouration of nodes (at 50% flowering), stem anthocyanin, petiole colour, petiole length, style colour and fruit purple colour intensity at the mature unripe stage] and five descriptors for field conditions were deleted after deliberations due to confusion in scoring. Furthermore, modified scoring scales were adopted for several quantitative traits on account of better manifestation of plant and fruit characteristics under protected conditions that result in a broader range of phenotypic variation.

The assessment of genetic diversity is prerequisite that forms the basis of crop improvement, germplasm conservation and varietal identification. The extent of morphological variation within core germplasm is particularly important, as it determines the potential for genetic enhancement of a crop⁸. The DUS descriptors under protected conditions provide a robust framework for varietal characterization, identification of unique and elite genotypes and establishment of reference collections. Additionally, DUS-based characterization enables the systematic grouping of genotypes according to agro-morphological traits thereby optimizing breeding strategy and accelerating the development of varieties tailored to specific agro-ecological zones and consumer requirements. DUS testing confirms that the new variety is clearly distinguishable from the existing ones, uniform in expression for essential traits and stable across environments⁹.

Evaluation of genetic divergence under protected conditions using these updated DUS descriptors is particularly relevant, as protected cultivation is increasingly becoming the target environment for capsicum improvement programmes. Mahalanobis D² statistics is an effective multivariate approach for quantifying genetic divergence and estimating relationships among genotypes based on their distance measures¹⁰. Such studies aid in identifying stable and distinct genotypes suitable for protected environments, understanding trait expression and plasticity under controlled conditions and selecting parents for hybrid breeding. Furthermore, characterization based on the new DUS descriptors strengthens varietal identification, supports effective implementation of plant variety protection laws and contributes to sustainable utilization and conservation of Capsicum annuum genetic resources. Beyond their regulatory function in variety registration, DUS testing provides a robust phenotypic framework that complements molecular diversity analyses by linking genotype-specific traits to molecular markers¹¹. This integration facilitates marker-assisted selection (MAS) and genomic selection enhancing precision in breeding programs. Therefore, comprehensive assessment of genetic divergence in chilli, paprika and bell pepper genotypes of Capsicum annuum under protected conditions, in accordance with the revised DUS descriptors, is essential for advancing breeding strategies, varietal protection and commercial exploitation of capsicum in modern production systems.

Materials and methods

Plant material and experimental site

The present investigation was carried out to characterize 58 genotypes of Capsicum annuum L. (Table S1), which were evaluated during the summer of 2021, 2022 and 2023 respectively, under naturally ventilated polyhouse conditions at the Research Farm of the Department of Vegetable Science and Floriculture, College of Agriculture, Chaudhary Sarwan Kumar Himachal Pradesh Krishi Vishvavidyalaya, Palampur. The Research Farm is situated at an elevation of 1290.8 m above mean sea level with 32˚6′ N latitude and 76˚3′ E longitude in the North-Western Himalayas region of India with humid, sub-temperate climate with an annual rainfall of 2,500 mm, of which 80% is received during June to September that represents the mid-hill zone of Himachal Pradesh (Zone-II). The soil is classified as Alfisoltypic Hapludalf clay with pH 5.7.

Experimental design and crop management

The experimental material was sown in nursery beds that were 3 m × 1 m × 15 cm in dimension consecutively for three years on the 27th, 25th and 27th of February 2021, 2022 and 2023, respectively. Seedlings were ready for transplanting to field plots within eight weeks after seed sowing. Each genotype was planted in 1.8 m × 1.8 m plots consisting of 16 plants in each replication with a spacing of 45 cm × 45 cm, respectively, between and within rows in a Randomized Complete Block Design (RCBD) during each of the years between 2021 and 2023, and replicated thrice. The crop was provided with recommended dosage of farmyard manure (20 tons ha^− 1) and fertilizer applications of 75: 60: 60 kg N, P₂O₅ and K₂O ha^− 1. The entire dose of P and K together with half dose of N was mixed in soil at the time of field preparation. The remaining dose of N was top-dressed in two equal splits at one month interval after transplanting. Besides, water soluble NPK (15:15:15) was also provided at 2% from flowering onwards at weekly intervals through drip. Drip irrigation system was used to apply water under polyhouse conditions at every other day.

DUS characterization and statistical analysis

The genotypes were evaluated for distinctness (differences between genotypes), uniformity (variation among individuals of the same genotype) and stability (consistency across replications) according to the guidelines provided by PPV&FRA (2001) and UPOV (International Union for the Protection of New Varieties of Plants) on Capsicum annuum L. under protected conditions. The observations were recorded on ten competitive plants selected randomly from each entry over the replications for 56 DUS characteristics (Table S2-S5) (as prescribed in the official PPV&FRA DUS Test Guidelines on plant habit, flower/fruit orientation, fruit bearing habit, fruit colour (at mature green/ripe fruit stage on plants), fruit shape in longitudinal section, fruit shape at the base and fruit shape of apex on a standard scale (e.g., 1–9 or UPOV descriptors)¹². Descriptive statistics were calculated for three years of data using Microsoft Excel. The mode was determined using the built-in formula functions in Excel to identify the most frequently occurring values in the dataset. The Shannon Diversity Index (H′) was computed using PAST 4.03 software of Hammer¹³. Mahalanobis D² analysis¹⁴ of 11 quantitative characters (Table S6) recorded during summer 2021–2023 was employed, followed by clustering of genotypes using Tocher’s method¹⁵ to elucidate genetic relationships. For conducting D² and principal component analysis, the computer program WINDOSTAT 8.0 developed by Indostat Services was used. The phenotypic frequencies calculated were further used to estimate Shannon’s Diversity Index (H) according to Negassa M¹⁶ to assess the present diversity.

Where, pi is the portion of the total number of entries belonging to the i^th class.

Results

Phenotypic characterization based on DUS descriptors

The frequency distribution (Table 1; Fig. 1) and mode values (Tables S2–S5) of 56 DUS characteristics of Capsicum annuum L. genotypes evaluated over three consecutive years (2021–2023) revealed a broad pattern of variation for all the traits (Fig. 2) with Shannon Diversity Index (H) ranging from 0.09 to 1.57 and a mean of 0.75 (Table 1). Most of the evaluated traits exhibited two or three phenotypic classes (26 and 20 traits, respectively), while seven traits showed four classes. The traits plant length of the first internode (on primary branches), plant height (in hot pepper and paprika), number of locules/lobes (in bell pepper) and fruit pericarp thickness (at physiological maturity stage) displayed the highest number of classes (five). The trait fruit shape in longitudinal section exhibited the maximum variation, with six distinct classes.

Table 1.

Morphological characterization depicting the predominate state, frequency percentage and Shannon diversity indices (H’) collection for each descriptor scored (2021-23).

Sr. No.	Trait	Scale of descriptor	Distribution by classes of descriptor (%)	Shannon Diversity Index (DI)	Sr. No.	Trait	Scale of descriptor	Distribution by classes of descriptor (%)	Shannon Diversity Index (DI)
1.	Seedling: Anthocyanin colouration of hypocotyl	Absent	7 (12.07%)	0.37	30.	Fruit: Intensity of green colour (at mature unripe stage)	Light	13 (22.41%)	1.05
		Present	51 (87.93%)				Medium	28 (48.27%)
							Dark	17 (29.31%)
2.	Plant: Habit	Spreading	5 (8.62%)	0.78	31a .	Fruit: Length (cm) (hot pepper and paprika)	Short (< 8.0)	4 (9.09%)	1.21
		Semi-upright	41 (70.69%)				Medium (8.01-10.0)	20 (45.45%)
		Upright	12 (20.69%)				Long (10.01-12.0)	14 (31.82%)
							Very Long (> 12.0)	6 (13.64%)
3.	Plant: Length of main stem (cm)	Short (up to 5.0)	37 (63.79%)	0.90	31b .	Fruit: Length (cm) (bell pepper)	Short (< 3.0)	1 (7.14%)	0.51
		Medium (5.01-10.0)	9 (15.52%)				Medium (3.01-6.0)	1 (7.14%)
		Long (> 10.0)	12 (20.69%)				Long (> 6.0)	12 (85.71%)
4.	Plant: Length of first internode (on primary branches in cm)	Very short (up to 6.0)	7 (12.07%)	1.57	32a .	Fruit: Diameter (fruits measured at the widest point in cm) [hot pepper and paprika]	Narrow (< 0.8)	4 (9.09%)	0.73
		Short (6.01-7.0)	9 (15.52%)				Medium (0.81–1.5)	33 (75.00%)
		Medium (7.01-8.0)	13 (22.41%)				Broad (> 1.5)	7 (15.91%)
		Long (8.01-9.0)	15 (25.86%)
		Very long (> 9.0)	14 (24.14%)
5.	Plant: Anthocyanin colouration of nodes (at active vegetative stage)	Absent	6 (10.34%)	0.33	32b .	Fruit: Diameter (fruits measured at the widest point in cm) [bell pepper]	Narrow (< 6.0)	4 (28.57%)	1.09
		Present	52 (89.66%)				Medium (6.01-8.0)	5 (35.71%)
							Broad (> 8.0)	5 (35.71%)
6.	Plant: Intensity of anthocyanin colouration of nodes (at active vegetative stage)	Weak	9 (15.52%)	1.03	33.	Fruit: Shape in longitudinal section	Square	4 (6.90%)	1.38
		Medium	21 (36.21%)
		Strong	22 (37.93%)				Rectangular	5 (8.62%)
7.	Plant: Anthocyanin colouration of nodes (at 50% flowering)	Absent	2 (3.45%)	0.15			Trapezoidal	3 (5.17%)
		Present	56 (96.55%)				Moderately triangular	9 (15.52%)
8.	Plant: Intensity of anthocyanin colouration of nodes at 50% flowering	Weak	5 (8.62%)	0.92			Narrowly triangular	32 (55.17%)
		Medium	22 (37.93%)				Horn shaped	5 (8.62%)
		Strong	29 (50.00%)
9a.	Plant: Height (cm) (hot pepper and paprika)	Very Short (< 90.0)	1 (2.27%)	1.24	34.	Fruit: Curvature	Absent	47 (81.03%)	0.49
							Present	11 (18.96%)
		Short (90.01–110.0)	3 (6.82%)		35.	Fruit: Neck at basal end	Absent	35 (60.34%)	0.67
		Medium (110.01–130.0)	5 (11.36%)				Present	23 (39.66%)
		Tall (130.01–150.0)	20 (45.45%)
		Very tall (> 150.01)	15 (34.09%)
9b.	Plant: Height (cm) (bell pepper)	Short (< 80.0)	3 (21.43%)	0.89	36.	Fruit: Cross sectional corrugation (at level of placenta)	Round	44 (75.86%)	0.77
		Medium (80.01–100.0)	2 (14.29%)				Slightly corrugated	9 (15.52%)
		Tall (> 100.0)	9 (64.29%)				Strongly corrugated	2 (3.45%)
							Oval	3 (5.17%)
10	Plant: Spread (cm)	Narrow (< 50.0)	21 (36.21%)	0.73	37.	Fruit: Sinuation of pericarp	Weak	38 (65.52%)	0.87
							Medium	13 (22.41%)
		Medium (50.01–100.0)	36 (62.07%)				Strong	7 (12.07%)
		Broad (> 100.0)	1 (1.72%)		38.	Fruit: Texture of surface	Smooth	27 (46.55%)	0.69
							Rough	31 (53.45%)
11	Stem: anthocyanin	Absent	20 (34.48%)	0.64	39.	Fruit: Colour (at ripe maturity)	Yellow (12 A)*	4 (6.90%)	0.60
		Present	38 (65.52%)
12.	Stem: Pubescence (hairiness)	Absent	13 (22.41%)	0.53			Orange (N25A)*	2 (3.45%)
		Present	45 (77.59%)				Orange red (N30A)*	3 (5.17%)
							Red (42 A)*	49 (84.48%)
13.	Stem: Shape	Round	18 (31.03%)	0.62	40.	Fruit: Intensity of red colour (at ripe maturity)	Light (42 A)*	3 (5.17%)	0.49
		Angular	40 (69.00%)
14.	Leaf: Length of blade (measured from leaf base to leaf tip in cm)	Short (< 10.0)	6 (10.34%)	0.84			Medium	50 (86.21%)
		Medium (10.01-12.0)	13 (22.41%)				Dark (46 A)*	5 (8.62%)
		Long (> 12.0)	39 (67.24%)
15.	Leaf: Width of blade (measured on the widest part of the leaf in cm taken from the middle part of the plant)	Narrow (< 5.0)	7 (12.07%)	0.98	41.	Fruit: Color transition	One stage	40 (68.97%)	0.69
							Two stages	17 (29.31%)
		Medium (5.01-8.0)	25 (43.10%)				> Two stages	1 (1.72%)
		Broad (> 8.0)	26 (44.83%)		42.	Fruit: Glossiness	Weak	9 (15.52%)	0.43
							Strong	49 (84.48%)
16.	Leaf: Colour	Green	57 (98.27%)	0.09	43.	Fruit: Shape at the base	Acute	32 (55.17%)	1.00
		Purple (N77A) *	1 (1.72%)				Round	13 (22.41%)
							Sunken	13 (22.41%)
17.	Leaf: Intensity of colour	Light green (143 C)*	5 (8.62%)	0.92	44.	Fruit: Shape of apex	Acute	36 (62.07%)	0.92
		Medium green (N137B)*	37 (63.79%)				Blunt	9 (15.52%)
		Dark green (N137A)*	15 (25.86%)				Depressed	13 (22.41%)
		Purple (N77A)*	1 (1.72%)
18.	Leaf: Shape	Lanceolate	39 (67.24%)	0.78	45.	Fruit: Number of locules/lobes (bell pepper)	2	1 (7.14%)	1.10
		Ovate	16 (27.59%)
		Broad elliptic	3 (5.17%)				2–3	6 (42.86%)
19.	Leaf: Undulation of margin	Weak	42 (72.41%)	0.77			3	6 (42.86%)
		Medium	10 (17.24%)				3–4	1 (7.14%)
		Strong	6 (10.34%)				> 4	-
20.	Leaf: pubescence (hairiness)	Absent	10 (17.24%)	0.46	46.	Fruit: Pericarp thickness in mm (at physiological maturity stage)	Very thin (< 1.0)	15 (25.86%)	1.43
		Present	48 (82.76%)				Thin (1.01-2.0)	24 (41.38%)
21.	Petiole color	Green (129 B)*	45 (77.59%)	0.53			Medium (2.01-3.0)	4 (6.90%)
		Purple (77 C)*	13 (22.41%)				Thick (3.01-4.0)	7 (12.07%)
							Very thick (> 4.0)	8 (13.79%)
22.	Petiole length	Short (< 6.0 cm)	18 (31.03%)	1.07	47.	Fruit: Stalk Length (cm)	Short (1.6–2.5)	2 (3.45%)	1.10
		Medium (6.01–9.0 cm)	23 (39.65%)				Medium (2.6–3.5)	8 (13.79%)
		Long (> 9.0 cm)	17 (29.31%)				Long (3.6–4.5)	29 (50.00%)
							Very long (> 4.5)	19 (32.76%)
23.	Flower: Petal colour	White (NN 155D)*	36 (62.07%)	0.73	48.	Fruit: Calyx Cover	Non-enveloping	28 (48.27%)	0.69
		Yellowish white (158 D)*	21 (36.21%)				Enveloping	30 (51.72%)
		Purple (77 C)*	1 (1.72%)
24.	Flower: Anther colour	Yellow (9 C)*	3 (5.17%)	1.03	49.	Fruit: Calyx Margin	Smooth	8 (13.79%)	0.40
		Pale blue (107 C)*	26 (44.83%)				Dentate	50 (86.21%)
		Blue (99 C)*	3 (5.17%)		50.	Fruit: Calyx Constriction	Absent	5 (8.62%)	0.29
		Purple (83 A)*	26 (44.83%)				Present	53 (91.38%)
25.	Flower: Style color	Yellowish white (2 C)*	50 (86.21%)	0.40	51.	Fruit: Pedicel attachment	Weak	22 (37.93%)	0.66
		Purple (83 A)*	8 (13.79%)				Strong	36 (62.07%)
26.	Flower: Days to 50% flowering (from the date of sowing)	Early (≤ 70.0 days)	4 (6.90%)	0.86	52.	Fruit: Blossom end appendage	Absent	4 (6.90%)	0.25
		Medium (71–90.0)	20 (34.48%)
		Late (> 90.0)	34 (58.62%)				Present	54 (93.10%)
27(a)	Flower Orientation	Drooping	39 (67.24%)	0.85	53.	Fruit: Days to 50% ripening (from the date of sowing)	Early (≤ 130.0 days)	7 (12.07%)	0.97
		Semi drooping	12 (20.69%)				Medium (131–150 days)	29 (50.00%)
		Erect	7 (12.07%)				Late (> 150.0 days)	22 (37.93%)
27(b)	Fruit: Orientation	Drooping	49 (84.48%)	0.51	54.	Seed: 1000 seed weight (g)	Low (< 4.0)	5 (8.62%)	0.88
		Semi drooping	2 (3.45%)				Medium (4.01-6.0)	18 (31.03%)
		Erect	7 (12.07%)				High (> 6.0)	35 (60.34%)
28.	Fruit: Bearing habit (No. of fruits/ node)	1 (Solitary)	54 (93.10%)	0.29	55.	Seed: Recovery (%)	Low (< 30.0)	38 (65.52%)	0.82
		2–3	1 (1.72%)				Medium (30.1–40.0)	16 (27.59%)
		> 3 (Cluster)	3 (5.17%)				High (> 40.0)	4 (6.90%)
29.	Fruit: Colour (at mature unripe stage)	White	1 (1.72%)	0.29	56.	Seed: Colour	Light yellow	37 (63.79%)	0.65
		Cream	3 (5.17%)				Yellow	21 (36.21%)
		Green	54 (93.10%)
		Purple	-
Mean		0.75
Maximum		1.57
Minimum		0.09

Fig. 1.

Frequency of 58 Capsicum genotypes on the basis of morphological description.

Fig. 2.

Images depicting morphological traits across different populations.

Plant characteristics

The anthocyanin coloration of the hypocotyl in the evaluated Capsicum genotypes was predominantly present in 51 genotypes (87.93%) whereas only 7 genotypes (12.07%) lacked anthocyanin coloration with a low Shannon Diversity Index of 0.37. Analysis of plant growth habit revealed that the majority of genotypes (41; 70.69%) exhibited a semi-upright habit, followed by upright (12; 20.69%) and spreading (5; 8.62%) forms. A majority of genotypes (37; 63.79%) were classified as short (≤ 5.0 cm) length of the main stem, followed by long (> 10.0 cm; 12; 20.69%) and medium (5.01–10.0 cm; 9; 15.52%) stem lengths, with moderate Shannon Diversity Index of 0.90, indicating a predominance of compact plant architecture. The length of the first internode on primary branches also varied widely among the genotypes. About 15.52% of genotypes exhibited short internodes (6.01–7.0 cm), 22.41% were medium (7.01–8.0 cm), 25.86% were long (8.01–9.0 cm), and 24.14% showed very long internodes (> 9.0 cm), while 12.07% were very short (≤ 6.0 cm) with high Shannon Diversity Index of 1.57. The anthocyanin coloration of nodes at the active vegetative stage was present in 52 genotypes (89.66%) exhibiting weak pigmentation in 15.52% of plants and at 50% flowering, with 96.55% of plants retaining anthocyanin deposition.

Plant height also varied among species. Hot pepper and paprika accessions predominantly exhibited tall growth (130.01–150.0 cm, 45.45%) and very tall stature (> 150.01 cm, 34.09%), whereas medium (110.01–130.0 cm), short (90.01–110.0 cm), and very short (< 90.0 cm) plants represented smaller fractions of 11.36%, 6.82%, and 2.27%, respectively. Bell pepper accessions showed a similar trend, with the majority (64.29%) classified as tall (> 100.0 cm), while short (< 80.0 cm) and medium (80.01–100.0 cm) plants comprised 21.43% and 14.29%, respectively. The spread of the plants was largely medium (50.01–100.0 cm, 62.07%), with 36.21% of accessions exhibiting narrow spread (< 50.0 cm) and only 1.72% showing broad growth (> 100.0 cm). The evaluated accessions exhibited considerable variation in stem and leaf morphological traits. Stem anthocyanin pigmentation was present in 38 of 58 genotypes (65.52%), while 20 genotypes (34.48%) lacked pigmentation, resulting in a moderate Shannon diversity index (H’ = 0.64), indicating variability in stem color expression. Stem pubescence was observed in 45 genotypes (77.59%) corresponding to a Shannon index of 0.53, suggesting moderate diversity in hairiness. Stem shape was predominantly angular, recorded in 40 genotypes (69.00%), whereas 18 genotypes (31.03%) had round stems, yielding a Shannon index of 0.62. Leaf blade length considerably, with long leaves (> 12.0 cm) observed in 39 genotypes (67.24%), medium-length leaves (10.01–12.0 cm) in 13 genotypes (22.41%), and short leaves (< 10.0 cm) in 6 genotypes (10.34%). Leaf width also showed high variability, with 26 genotypes (44.83%) exhibiting broad leaves (> 8.0 cm), 25 genotypes (43.10%) medium-width leaves (5.01–8.0 cm) and 7 genotypes (12.07%) narrow leaves (< 5.0 cm), producing a high Shannon index (H’ = 0.98).

Leaf colour was overwhelmingly green in 57 genotypes (98.27%), with only a single genotype with purple-leaves (1.72%), corresponding to a low Shannon index (H’ = 0.09). Leaf colour intensity varied, with medium green leaves in 37 genotypes (63.79%), dark green in 15 genotypes (25.86%), light green in 5 genotypes (8.62%) and purple in one genotype (1.72%), resulting in a high Shannon index (H’ = 0.92). Leaf shape was predominantly lanceolate (39 genotypes, 67.24%), followed by ovate (16 genotypes, 27.59%) and broad elliptic (3 genotypes, 5.17%), with H’ = 0.78. Leaf margin undulation was mostly weak (42 genotypes, 72.41%), medium in 10 genotypes (17.24%) and strong in 6 genotypes (10.34%), yielding a Shannon index of 0.77. Leaf pubescence was present in 48 genotypes (82.76%) and absent in 10 genotypes (17.24%), with a Shannon index of 0.46, indicating moderate variation in leaf hairiness.

Flower and fruit characteristics

Analysis of flowering and fruit traits among the studied genotypes revealed considerable variation in several characteristics. Green petiole color was most common, observed in 45 genotypes (77.59%), whereas purple petioles occurred in 13 genotypes (22.41%) with medium-length petioles (6.01–9.0 cm) being most frequent (23, 39.65%). White petals predominated (36 genotypes, 62.07%), followed by yellowish white (21, 36.21%), while purple petals were rare (1, 1.72%). Anther coloration showed considerable diversity, with pale blue and purple anthers each recorded in 26 genotypes (44.83%) and yellow and blue anthers being less frequent (3 genotypes each, 5.17%), corresponding to a high diversity index of 1.03. Style color was largely conserved, with yellowish white styles in 50 genotypes (86.21%) and purple styles in 8 genotypes (13.79%), indicating low diversity index of 0.40). Flowering time varied markedly, with late-flowering genotypes (> 90 days) being most frequent (34, 58.62%), medium-flowering genotypes (71–90 days) in 20 (34.48%), and early-flowering genotypes (≤ 70 days) being rare (4, 6.90%). Flower orientation was predominantly drooping (39, 67.24%), with semi-drooping (12, 20.69%) and erect forms (7, 12.07%) less frequent. Fruit orientation was largely drooping (49, 84.48%), while semi-drooping (2, 3.45%) and erect fruits (7, 12.07%) were uncommon, reflecting low to moderate diversity (H′ = 0.51). The majority of genotypes produced solitary fruits per node (54, 93.10%), whereas 2–3 fruits per node and clustered fruits (> 3) were rare (1, 1.72% and 3, 5.17%, respectively), indicating very low variation in fruit-bearing habit (H′ = 0.29). Fruit colour at the mature unripe stage was largely green (54, 93.10%), with medium green being most common (28, 48.27%), followed by dark green (17, 29.31%) and light green (13, 22.41%), corresponding to a high diversity index (H′ = 1.05).

Fruit length among hot pepper and paprika genotypes displayed marked variability, ranging from short (< 8.0 cm) to very long (> 12.0 cm). The majority of genotypes (45.45%) exhibited medium fruit length (8.01–10.0 cm), followed by long fruits (31.82%) and very long fruits (13.64%), while only 9.09% were short. In contrast, fruit length variation among bell pepper genotypes was relatively narrow. Long fruits (> 6.0 cm) predominated (85.71%), whereas short (< 3.0 cm) and medium (3.01–6.0 cm) fruits each accounted for only 7.14% of the population. Fruit diameter ranged widely, though most genotypes (75.00%) were classified as medium (0.81–1.5 cm). Narrow (< 0.8 cm) and broad (> 1.5 cm) fruits represented 9.09% and 15.91%, respectively with moderate phenotypic heterogeneity for this trait. A more even distribution of diameter classes was observed in bell pepper genotypes: 28.57% narrow (< 6.0 cm), 35.71% medium (6.01–8.0 cm) and 35.71% broad (> 8.0 cm). The Shannon diversity index (H′ = 1.09) revealed high morphological diversity, suggesting that fruit width contributes significantly to the phenotypic distinction among bell pepper cultivars. Six distinct fruit shape categories were identified, with “narrowly triangular” fruits predominating (55.17%), followed by “moderately triangular” (15.52%), “rectangular” and “horn shaped” (each 8.62%), “square” (6.90%) and “trapezoidal” (5.17%). The Shannon diversity index (H′ = 1.38) indicated a high degree of morphological diversity for fruit shape. Most genotypes (81.03%) lacked fruit curvature, while 18.96% displayed a distinct curved shape. The presence of a neck at the basal end was observed in 39.66% of genotypes, while 60.34% lacked this feature. The majority of the evaluated genotypes (75.86%) exhibited a round cross-sectional shape, whereas 15.52% were slightly corrugated, 5.17% oval and only 3.45% strongly corrugated with moderate Shannon diversity index of H′ = 0.77.

Variation in pericarp sinuation was evident across the genotypes, with 65.52% expressing weak sinuation, 22.41% medium and 12.07% strong sinuation with high Shannon diversity index (H′ = 0.87). Fruit surface texture exhibited a nearly even distribution between smooth (46.55%) and rough (53.45%) fruit types. For fruit colour at ripe maturity, ripe fruit colour showed distinct categorical variation, with red fruits overwhelmingly predominant (84.48%), followed by yellow at 6.90%, orange-red at 5.17% and orange at 3.45%. The Shannon diversity index (H′ = 0.60) signified moderate but limited diversity, largely driven by the dominance of red-fruited genotypes. Among red-fruited genotypes, colour intensity varied from light (5.17%) and dark (8.62%) to predominantly medium red (86.21%) with low Shannon diversity index (H′ = 0.49). Distinct variation was observed in the number of fruit colour transitions during fruit ripening. Most genotypes (68.97%) exhibited a single-stage colour change, typically transitioning directly from green to red, whereas 29.31% showed two-stage transitions (e.g., green → yellow → red) and only 1.72% displayed more than two transitional stages. The Shannon diversity index (H′ = 0.69) indicated moderate diversity, suggesting that the ripening pattern varies among genotypes, potentially reflecting genetic differences in pigment biosynthesis pathways. Fruit glossiness displayed low phenotypic variation, with the majority of genotypes (84.48%) having a strong glossy surface, while 15.52% were weakly glossy. Fruit shape at the base revealed three distinct types, acute (55.17%), round (22.41%) and sunken (22.41%) with an overall Shannon diversity index (H′ = 1.00), demonstrating high morphological diversity. The shape of the fruit apex varied widely among the genotypes examined. A majority (62.07%) exhibited an acute apex, while depressed (22.41%) and blunt (15.52%) shapes were less common with high Shannon diversity index (H′ = 0.92). Bell pepper genotypes showed considerable variability in locule number, ranging from two to four. The most frequent categories were 2–3 locules (42.86%) and three locules (42.86%), while genotypes with two and three-to-four locules were rare (7.14% each). None exhibited more than four locules. The Shannon diversity index (H′ = 1.10) reflected a very high level of structural diversity.

Fruit pericarp thickness at physiological maturity exhibited substantial variation among the Capsicum genotypes evaluated. The majority (41.38%) possessed a thin pericarp (1.01–2.0 mm), while very thin fruits (< 1.0 mm) constituted 25.86%. A smaller proportion of genotypes exhibited medium (6.90%), thick (12.07%), and very thick (> 4.0 mm; 13.79%) pericarps. The Shannon diversity index (H′ = 1.43) indicated a high level of diversity, reflecting broad genetic differentiation for this trait. Fruit stalk length also varied widely, ranging from short (1.6–2.5 cm) to very long (> 4.5 cm) with high Shannon diversity index (H′ = 1.10). The long-stalked genotypes (3.6–4.5 cm) were predominant (50.00%), followed by very long (32.76%) and medium (13.79%), whereas short-stalked fruits were rare (3.45%). Fruit calyx cover exhibited a nearly equal distribution between non-enveloping (48.27%) and enveloping (51.72%) types, with a Shannon diversity index (H′ = 0.69), indicating moderate diversity. Calyx margin morphology displayed low phenotypic diversity (H′ = 0.40). A pronounced dominance of the dentate type (86.21%) was observed compared to smooth margins (13.79%). Most genotypes (91.38%) exhibited a constricted calyx, while only 8.62% lacked this feature. For fruit pedicel attachment, majority of the genotypes exhibited a strong pedicel attachment (62.07%), while a smaller proportion showed weak attachment (37.93%). The presence of a blossom-end appendage was dominant across genotypes, observed in 93.10%, whereas only 6.90% lacked the appendage. The Shannon diversity index for this characteristic was 0.25, suggesting low diversity and limited variation for this morphological trait. Substantial variability was recorded in ripening behaviour among the genotypes. Half of the genotypes (50.00%) exhibited medium maturity (131–150 days), followed by late types (> 150 days) constituting 37.93% and early maturing genotypes (≤ 130 days) representing 12.07%. The Shannon diversity index of 0.97 reflects high genetic diversity for earliness in ripening. Significant variation was noted in seed weight, with 60.34% of genotypes producing heavy seeds (> 6.0 g), 31.03% showing medium seed weight (4.01–6.0 g) and only 8.62% categorized as light (< 4.0 g). Genotypes differed markedly in seed recovery efficiency. A majority (65.52%) showed low recovery (< 30.0%), while 27.59% and 6.90% were classified as medium (30.1–40.0%) and high (> 40.0%) recovery, respectively. Variation in seed coat coloration was evident among genotypes, with 63.79% producing light yellow seeds and 36.21% exhibiting yellow coloration. The diversity index (H’ = 0.65) suggests moderate phenotypic diversity for seed colour within the evaluated population.

Genetic divergence based on quantitative characteristics

Dendrogram based on standardized squared euclidean distance was generated using Tocher’s method for 11 quantitative characters representing 58 genotypes (Fig. 3) during summer 2021–2023. The genotypes were classified into ten distinct clusters using Tocher’s method (Fig. 3 and Table S7), where clusters I, II, IV, VII, VIII and X comprised hot pepper and paprika genotypes, while clusters III, V, VI and IX comprised bell pepper genotypes. Cluster I categorized as the largest with 39 genotypes followed by 5, 5 and 3 genotypes in clusters V, VI and III, respectively while cluster II, IV, VII, VIII, IX and X were monogenotypic that revealed optimum genetic divergence in the germplasm. Inter- and intra-cluster distances and genotypes representing the respective clusters are presented in Table 2. The intra-cluster distance was highest (19.33) in cluster VI and lowest (0.00) in clusters II, IV, VII, VIII, IX and X. The inter-cluster distance ranged from 16.37 to 74.84 with maximum divergence in clusters II and IX (74.84) followed by clusters IV and IX (70.53), I and IX (69.25) and VIII and IX (66.13). Cluster means also showed differences for all the traits (Table 3). Cluster IX recorded high mean values for PLS (11.46 cm), FD (8.19 cm) and FPT (6.76 mm). Similarly, high means for PLI (10.08 cm) and PH (125.32 cm) was observed in cluster II and for LLB (20.15 cm) and LWB (11.32 cm) in Cluster VI. Cluster VII recorded high mean values for FL (15.35 cm) and cluster VIII for PL (12.38 cm). Least cluster mean value was calculated for FL (4.29), FD (0.69), FSL (2.65), PH (75.20) and LLB (10.91). In addition, the relative contribution of individual traits indicated the maximum contribution towards genetic divergence by fruit diameter (37.93%), fruit length (22.87%), petiole length (16.09%) and fruit pericarp thickness (10.77%) (Table 4).

Fig. 3.

Dendrogram depicting grouping of 58 genotypes of Capsicum on Mahalanobsis D² statistics using Tocher’s method. (1) LCA 305; (2) LCA 334; (3) LCA 206; (4) LCA 235; (5) LCA 620; (6) LCA 353; (7) LCA 625; (8) CA 960; (9) Arka Lohit; (10) Arka Uphal; 11) Arka Sweta; 12) Arka Harita; 13) Arka Meghana; 14) Kashi Tej; 15) Kashi Gaurav; 16) Kashi Surkh; 17) Kashi Anmol; 18) Kashi Ratna; 19) Kashi Abha; 20) Punjab Sindhuri; 21) Punjab Tej; 22) Punjab Gucchedar; 23) CH-52; 24) CH-27; 25) Pusa Jwala; 26) Eagle; 27) Hy-Veg 78; 28) Gagan; 29) Aastha; 30) Him Palam Mirch-1; 31) Him Palam Mirch-2; 32) Surajmukhi; 33) KBCH-1; 34) Byadgi Kaddi; 35) Gowri Bidanoor Local; 36) AVPP9824; 37) LCA 436; 38) LCA 424; 39) Arka Abhir; 40) Kashi Sindhuri; 41) Byadgi Dabbi; 42) Kashmiri Chilli; 43) KTPL-19; 44) Paprika; 45) Arka Basant; 46) HYB-1; 47) YSC-5; 48) Yolo Wonder; 49) California Wonder; 50) KTC-1; 51) VS-1; 52) Solan Bharpur; 53) Orobelle; 54) Indera; 55) Him Palam Yellow Capsicum-1; 56) NS-292; 57) NS-82; 58) Hybrid Bombay.

Table 2.

Average intra and inter-cluster distance of D² and √D² among clusters (2021-23).

Clusters	I	II	III	IV	V	VI	VII	VIII	IX	X
I	13.60	17.94	57.57	18.27	42.13	60.12	20.52	19.22	69.25	25.69
II		0.00	62.72	16.37	45.93	63.28	30.59	21.63	74.84	34.08
III			13.43	58.28	23.76	29.20	53.75	54.33	27.87	40.39
IV				0.00	43.23	60.77	28.85	27.10	70.53	29.28
V					13.96	25.97	40.13	39.96	34.60	30.29
VI						19.33	56.39	57.28	23.21	46.65
VII							0.00	24.77	61.70	18.03
VIII								0.00	66.13	24.98
IX									0.00	50.25
X										0.00

Bold values are intra-cluster distance.

Table 3.

Cluster means for different traits of chilli genotypes distributed in seven clusters (2021-23).

Traits	I	II	III	IV	V	VI	VII	VIII	IX	X	Mean	Maximum	Minimum
PLS	3.90	2.56	19.74	3.00	15.13	10.05	2.95	2.36	11.46	2.42	7.36	11.46	2.36
PLI	8.0 m 3	10.08	9.73	8.09	8.91	7.21	8.60	8.11	5.84	9.43	8.40	10.08	5.84
PH	114.08	125.32	98.44	87.42	101.16	93.32	92.02	109.58	103.60	75.20	100.01	125.32	75.20
PS	61.27	31.91	54.81	70.20	54.78	49.52	57.87	62.18	51.32	40.81	53.47	70.20	31.91
LLB	12.81	13.05	14.73	15.78	17.54	20.15	10.92	13.06	16.52	10.91	14.55	20.15	10.91
LWB	6.24	8.58	8.34	3.60	10.09	11.32	8.45	8.48	9.27	4.68	7.91	11.32	3.60
PL	7.42	9.53	7.54	4.96	9.45	11.02	3.40	12.38	6.78	4.65	7.71	12.38	3.40
FL	9.92	4.81	9.77	4.29	8.12	7.76	15.35	12.42	12.52	13.81	9.88	15.35	4.29
FD	1.11	0.69	8.17	1.23	5.61	6.83	1.52	2.06	8.19	3.48	3.89	8.19	0.69
FPT	1.09	0.83	3.73	1.13	3.76	6.08	2.04	0.79	6.76	2.27	2.85	6.76	0.79
FSL	4.03	2.95	4.66	3.14	4.58	4.27	3.98	2.65	3.47	3.22	3.70	4.66	2.65

PLS- Plant: Length of main stem (cm); PLI- Plant: Length of first internode; PH-Plant: Height (cm); PS- Plant: Spread (cm); LLB- Leaf: Length of blade (cm); LWB- Leaf: Width of blade (cm); PL- Petiole length (cm); FL- Fruit: Length (cm); FD- Fruit: Diameter (cm); FPT- Fruit: Pericarp thickness (mm); FSL- Fruit: Stalk Length (cm).

Table 4.

Relative contribution (%) of different traits to the genetic divergence (2021-23).

Source	Times Ranked 1st	Contribution (%)
Plant: Length of main stem (cm)	11	0.67
Plant: Length of first internode (cm)	20	1.21
Plant: Height (cm)	24	1.45
Plant: Spread (cm)	1	0.06
Leaf: Length of blade (cm)	2	0.12
Leaf: Width of blade (cm)	41	2.48
Petiole length (cm)	266	16.09
Fruit: Length (cm)	378	22.87
Fruit: Diameter (cm)	627	37.93
Fruit: Pericarp thickness (mm)	178	10.77
Fruit: Stalk Length (cm)	105	6.35

Principal components with Eigen values > 1.0 were considered significant. As a result, the first four significant principal components explained 74.21% of the variation (Table 5). PC1 explained 39.00% of the total variability that was mainly influenced by fruit pericarp thickness, fruit diameter, leaf length of blade, leaf width of blade and plant length of main stem. PC2 showed the association of stalk length of fruit, plant height and length of first internode with 13.94% contribution. In PC3 (11.94%), plant spread, plant height and petiole length had a major impact. The fourth component explained 9.33% of total variability which were due to fruit length. Principal components 1 and 2 had high positive loading for the characters. Further, it was observed from the scattered diagram (Fig. 4) that genotypes namely, Him Palam Yellow Capsicum-1, KTC-1, California Wonder and Yolo Wonder were the most diverse, being placed on the vertices of the polygon while the others were placed within the polygon representing variable performance. The other promising and diverse genotypes were YSC-5, Arka Basant, Hybrid Bombay and Solan Bharpur and may be suitable for Capsicum improvement programmes.

Table 5.

Eigen values, proportion of total variance, cumulative percentage variance and component loading of 11 characters of Capsicum (2021-23).

Trait	PC1	PC2	PC3	PC4
Eigen value	4.29	1.53	1.31	1.03
Variability (%)	39.00	13.94	11.94	9.33
Cumulative%	39.00	52.94	64.88	74.21
Plant: Length of main stem (cm)	0.32	0.31	-0.07	0.08
Plant: Length of first internode (cm)	0.02	0.42	-0.21	-0.71
Plant: Height (cm)	-0.16	0.44	0.52	-0.14
Plant: Spread (cm)	-0.22	0.27	0.53	0.25
Leaf: Length of blade (cm)	0.41	-0.10	0.23	-0.07
Leaf: Width of blade (cm)	0.41	-0.02	0.21	-0.16
Petiole length (cm)	0.25	-0.27	0.46	0.03
Fruit: Length (cm)	-0.14	0.24	-0.24	0.51
Fruit: Diameter (cm)	0.44	0.16	-0.16	0.13
Fruit: Pericarp thickness (mm)	0.45	0.03	-0.07	0.22
Fruit: Stalk Length (cm)	0.12	0.49	-0.04	0.22

Fig. 4.

Biplot graphical display of PCA showing the genetic diversity. 1)LCA 305; 2) LCA 334; 3) LCA 206; 4) LCA 235; 5) LCA 620; 6) LCA 353; 7) LCA 625; 8) CA 960; 9) Arka Lohit; 10) Arka Uphal; 11) Arka Sweta; 12) Arka Harita; 13) Arka Meghana; 14) Kashi Tej; 15) Kashi Gaurav; 16) Kashi Surkh; 17) Kashi Anmol; 18) Kashi Ratna; 19) Kashi Abha; 20) Punjab Sindhuri; 21) Punjab Tej; 22) Punjab Gucchedar; 23) CH-52; 24) CH-27; 25) Pusa Jwala; 26) Eagle; 27) Hy-Veg 78; 28) Gagan; 29) Aastha; 30) Him Palam Mirch-1; 31) Him Palam Mirch-2; 32) Surajmukhi; 33) KBCH-1; 34) Byadgi Kaddi; 35) Gowri Bidanoor Local; 36) AVPP9824; 37) LCA 436; 38) LCA 424; 39) Arka Abhir; 40) Kashi Sindhuri; 41) Byadgi Dabbi; 42) Kashmiri Chilli; 43) KTPL-19; 44) Paprika; 45) Arka Basant; 46) HYB-1; 47) YSC-5; 48) Yolo Wonder; 49) California Wonder; 50) KTC-1; 51) VS-1; 52) Solan Bharpur; 53) Orobelle; 54) Indera; 55) Him Palam Yellow Capsicum-1; 56) NS-292; 57) NS-82; 58) Hybrid Bombay.

Discussion

The use of morphological descriptors remains a practical and widely accepted approach for diversity assessment, particularly in the context of varietal characterization and DUS testing. However, it is important to recognize that phenotypic traits are influenced not only by genetic factors but also by environmental conditions and genotype × environment interactions. Consequently, estimates of diversity based on such descriptors may not fully correspond to underlying molecular variation, although they provide valuable insights into functional and adaptive diversity relevant to crop performance. In this context, the present study represents a comprehensive evaluation of genetic divergence of Capsicum annuum L. under protected conditions through newly developed DUS descriptors, addressing a critical gap in varietal characterization for polyhouse grown capsicum though it should be interpreted with appropriate caution. Diversity assessment remains a practical and cost-effective approach for variety differentiation, particularly when grown in controlled environments that minimize environmental variability and amplify genetic expression differences¹⁷. Under open field conditions, the expression of morphological traits is often influenced by fluctuating environmental factors such as temperature, rainfall and biotic stresses, which may mask the true potential of genotypes. In contrast, the protected environment used in the present study provides more uniform growing conditions, allowing clearer expression of inherent phenotypic differences by minimizing environmental fluctuation and enhance detection of variability, thereby justifying the need for specific descriptors under such conditions. Significant morphological variation observed across vegetative, floral and fruit traits in the current germplasm underscores the high genetic heterogeneity inherent in Capsicum annuum. (Table S2-S5), suggests a rich reservoir of allelic diversity that can be exploited in breeding programs. The protected cultivation does not mask genetic expression rather it enables clearer differentiation of genotypes by minimizing environmental noise¹⁸. This reinforces the suitability of DUS-based morphological descriptors for assessing phenotypic diversity and varietal distinctness in controlled environments. The Shannon Wiener diversity index (H′) values revealed differential levels of variability among traits, indicating trait-specific patterns of selection and conservation during domestication and breeding. The characters such as length of first internode, fruit pericarp thickness, fruit shape, fruit length and fruit stalk length exhibited high H′ values, suggesting broad allelic diversity and relatively low directional selection pressure for these characters. These traits are known to be strongly associated with plant architecture, fruit quality, market preference and yield stability and therefore offer significant scope for selection in breeding programmes targeting protected cultivation systems. Conversely, traits such as leaf colour, hypocotyl pigmentation, blossom-end appendage presence and fruit bearing habit showed low diversity indices, reflecting strong stabilizing selection and genetic fixation¹⁹.

The dominance of green leaf colour, solitary fruiting habit, drooping fruit orientation and red ripe fruit colour suggests that these characters have been conserved due to their adaptive and commercial advantages. Under protected conditions, where light interception, canopy management and uniform fruit development are critical, such conserved traits confer agronomic stability and market acceptability^20,21. The predominance of anthocyanin pigmentation in the hypocotyl across most Capsicum annuum genotypes reflects a strong expression of pigment-related genes, which are known to be regulated by MYB-bHLH-WD40 transcription complexes controlling flavonoid biosynthesis²². The predominance of semi-upright to upright plant architecture, tall stature, medium plant spread and larger leaf area reveals adaptation to protected environments where optimized light distribution and canopy structure improves photosynthetic efficiency and fruit set. Enhanced vegetative growth, longer internodes and increased plant height under protected conditions are expected responses to regulated temperature, humidity and nutrient availability²³. Tripodi and Greco²⁴ and Parisi M²⁵ et al. have also indicated the potential of structural traits as a key selection criteria for improving canopy efficiency and light interception in Capsicum. Leaf morphological traits, particularly leaf blade length and leaf width contributed moderately to phenotypic diversity. The high diversity for petiole length further indicates scope for modifying leaf orientation and canopy architecture, which is crucial for light interception and disease management in dense planting systems^26,27.

Flower and fruit morphological traits in Capsicum are of considerable interest in breeding programmes because they influence pollinator attraction, fruit set, agronomic performance and market acceptability. Moon S²⁸ et al. have shown that phenotypic variation for floral colouration, orientation of flowers, fruit-bearing habit and fruit colour at different maturity stages contribute significantly to the phenotypic diversity available in germplasm. Flower traits such as white petal colour and style colour were largely uniform, anther colour and flowering time showed considerable diversity. The predominance of late and medium flowering genotypes reflects selection for prolonged harvesting duration, a desirable trait in commercial polyhouse capsicum production. Variation in flowering phenology can be strategically used to synchronize flowering and harvest windows in protected cultivation, facilitating extended market supply and staggered production. High variability in fruit shape, size, diameter, pericarp thickness, surface texture and locule number (bell pepper) highlights the effectiveness of the new DUS descriptors in capturing commercially relevant diversity. Pericarp thickness which showed one of the highest diversity indices is a critical quality trait associated with fruit firmness, shelf life and transportability. Similarly, diversity in locule number among bell pepper genotypes reflects structural variation directly linked to fruit weight and consumer preference^29,30. The dominance of red fruit colour with moderate variation in colour intensity and ripening pattern suggested differential expression of pigment biosynthesis pathways across genotypes. The predominance of glossy fruits and strong pedicel attachment further reflects selection for traits that enhance visual appeal and reduce post-harvest losses³¹.

Multivariate analysis based on quantitative traits clearly separated the genotypes into ten distinct clusters (Table S7 and Fig. 3) broadly differentiating hot pepper/paprika from bell pepper types, validates the robustness of the DUS descriptors and confirms their ability to reflect underlying genetic relationships. Cluster I contained a large number of genotypes, indicating similarity for different traits that may reflect a relatively narrow genetic base due to shared ancestry or breeding history. Additionally, the 11 quantitative traits used in D² analysis, though important, may have limited resolution to distinguish closely related genotypes. Hence, the observed homogeneity likely the result of both genetic relatedness and restricted trait discrimination that warrants inclusion of additional descriptors, especially molecular markers for better differentiation. The presence of monogenotypic clusters indicates the existence of highly distinct genotypes, which can serve as valuable parents for hybridization to generate maximum heterosis^32,33. High inter-cluster distances, particularly between clusters II and IX, further emphasize the scope for exploiting genetic divergence in breeding programmes¹⁰. Principal component analysis (PCA) revealed that a small number of components accounted for the majority of total variation, with plant spread, plant height, fruit length, stalk length of fruit, pericarp thickness and fruit diameter contributing most strongly (Table 5). This underscores the central role of plant architecture and fruit traits in shaping overall diversity under protected conditions which is reflected from both D² and PCA. The identification of highly diverse genotypes viz., Him Palam Yellow Capsicum-1, KTC-1, California Wonder, Yolo Wonder, YSC-5, Arka Basant, Hybrid Bombay and Solan Bharpur positioned at the extremes of the PCA plot highlights promising candidates for use in varietal development and pre-breeding programmes. The use of standardized DUS descriptors under protected conditions enhances the reliability of phenotypic characterization by reducing cofounding environmental effects. A comparative assessment of genotypes clearly indicated differential trait expression under open and protected conditions (Table S8). For plant height, Him Palam Mirch-1, Him Palam Mirch-2 and Surajmukhi showed substantial increase from field (58.15, 61.25 and 65.00 cm) to polyhouse (122.00, 135.00 and 138.00 cm)³⁴ conditions, respectively that also reflected from the present study (78.22, 81.32 and 116.83 cm). Similarly, fruit length of Surajmukhi, Him Palam Mirch-1 and Him Palam Mirch-2 increased from 4.80, 5.64 and 6.49 cm under field³⁴ to 5.26, 6.96 and 8.85 cm, respectively under protected conditions. These consistent upward shifts and expanded ranges of quantitative traits under protected environments confirm enhanced phenotypic expression, thereby justifying the need for adapted descriptors and modified scoring scales specific to protected conditions. The current results indicated that the developed descriptors effectively capture relevant phenotypic diversity and can thereby support both legal protection mechanisms and practical breeding decisions. Integrating morphological descriptors with molecular tools such as SSR or SNP markers can further validate genetic relationships and confirms that phenotypic diversity often reflects underlying genomic variation.

Elucidations of diversity based on either of phenotypic or molecular analysis, are often indirectly outlined within the assumptions of the Neutral Theory of molecular evolution that has historically guided analyses of genetic variation. However, its adequacy in explaining observed diversity has been gradually questioned due to reported patterns inconsistent with key assumptions of strict neutrality^35–38. Furthermore, emerging evidence suggests that molecular variation previously considered neutral may in fact reflect adaptive processes, including adaptive tracking under antagonistic pleiotropy³⁹. In addition, genomic regions traditionally treated as neutral markers, such as mitochondrial DNA and short tandem repeats, have been shown to possess functional and regulatory roles^40,41. Therefore, these findings suggest that genetic diversity may not be fully explained by neutrality-based models alone and that both phenotypic and molecular variation can be shaped by selective as well as non-neutral processes. Therefore, the diversity patterns observed in the present study are best interpreted as reflecting functional and adaptive variability rather than strictly neutral genetic divergence. In spite of these considerations, the identified variability provides a useful foundation for selection and breeding. Genotypes exhibiting desirable trait combinations and greater divergence may be utilized in hybridization programmes to exploit heterosis and to broaden the genetic base. Overall, the study contibutes to the characterization of available germplasm and offers insights that are directly relevant for varietal improvement and sustainable crop production.

Conclusion

Morphological characterization of Capsicum germplasm is essential for developing passport data, trait cataloguing and classification and generating baseline information for advanced research. This study represents the first comprehensive characterization aimed at formulating novel DUS guidelines of Capsicum annuum under protected conditions in India. The study revealed a wide spectrum of morphological variations across vegetative, floral, fruit and seed traits, underscoring the substantial phenotypic diversity present within the evaluated germplasm. The application of the Shannon Wiener diversity index (H′) enabled a quantitative assessment of trait polymorphism, highlighting traits with limited variability such as hypocotyl pigmentation, leaf colour and certain floral characteristics as well as those displaying considerable heterogeneity, including internode length, petiole length, fruit shape, pericarp traits and stalk length. The predominance of specific phenotypes such as green leaf colour, white corolla, drooping fruit orientation, red ripe fruit colour and conserved anthocyanin expression suggests stabilizing selection under domestication, whereas the high variability in architectural and fruit-related traits reflects a rich reservoir of alleles suitable for targeted improvement. The documentation of distinct fruit shapes, pericarp textures, ripening behaviours and reproductive traits provides valuable descriptors for varietal differentiation under the PPV&FRA sui generis framework, thereby supporting both plant breeders’ and farmers’ rights. Clustering of genotypes grouped them into ten distinct clusters, clearly separating hot pepper and bell pepper types. Cluster I was the largest, while clusters II, IV, VII, VIII, IX and X were monogenotypic, indicating high genetic uniqueness. Maximum inter-cluster distance was observed between clusters II and IX, suggesting greater divergence and potential for heterotic crosses. Principal component analysis revealed that four principal components with eigen values > 1.0 explained 74.21% of total variation, with PC1 alone accounting for 39.00% due to fruit pericarp thickness and fruit diameter. Overall, the phenotypic divergence captured across genotypes not only strengthens the foundation for selection and hybrid development but also contributes to the conservation and sustainable utilization of Capsicum genetic resources. Future work integrating these morphological descriptors with molecular markers and genomic tools will further refine genotype discrimination, elucidate genetic control of key traits and accelerate the development of improved cultivars tailored to emerging agro-climatic and market demands.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Author contributions

A.S. contributed to conceptualization, methodology, investigation, data curation, visualization, supervision, and writing—review and editing; P.S. contributed to supervision and writing—review and editing; N.K., A.T. and M.K. executed the experiments and were involved in data curation; N.K. also performed data analysis. A.S., H.T., D.K.A., R.K. and P.K. contributed to writing—review and editing. S. analyzed the data, prepared tables and figures and wrote the original draft. The manuscript was revised with input from all authors. All authors reviewed and approved the final manuscript.

Funding

The researchers stated that they did not receive any financial assistance, including funding or grants for the development of this manuscript.

Data availability

All data generated or analyzed in this study are included in this article. For more information, please contact the corresponding author.

Declarations

Competing interests

The authors declare no competing interests.