Всеобъемлющая оценка взаимодействия капсикум генотипа-окружающей среды с использованием моделей Biplot AMMI и GПЭ в различных поэтапных средах трансплантации.

Introduction

Capsicum ( Capsicum annuum L.) is a high-value horticultural crop of global significance, cultivated across a wide spectrum of agro-ecological niches characterized by pronounced spatial and temporal environmental heterogeneity 1 . Yield realization in capsicum is inherently complex, as it arises from the integrated expression of multiple morpho-phenological and fruit-specific component traits whose expression is strongly modulated by environmental conditions 2 – 4 . Consequently, phenotypic performance often reflects the combined influence of genetic potential and environmental modulation, resulting in genotype × environment interaction (GEI), which complicates the identification of consistently superior genotypes in breeding programs 5 . GEI can reduce selection efficiency by altering genotype rankings across environments and by inflating experimental variance, thereby obscuring stable genetic performance 6 . In capsicum, yield is governed by several environmentally responsive and genetically inter-related traits such as fruit number, fruit size, fruit weight, pericarp thickness, and harvest duration. The collective response of these traits to environmental variation often generates complex performance patterns that cannot be adequately interpreted using simple mean-based approaches. Therefore, robust multivariate analytical models are required to partition genotype and interaction effects and to interpret stability across variable environments 7 8 .

The Additive Main Effects and Multiplicative Interaction (AMMI) model integrates analysis of variance with principal component analysis to partition GEI into orthogonal interaction principal component axes, thereby improving interpretation of genotype responses across environments 9 , 10 . Similarly, the Genotype plus Genotype × Environment (GGE) biplot approach focuses on genotype evaluation by combining genotype main effects with GEI, enabling graphical visualization of genotype adaptability, stability, and test environment relationships 11 . The combined application of AMMI and GGE biplot analyses has proven useful for interpreting GEI patterns and identifying stable and high-performing genotypes in multi-environment trials 12 , 13 . In North-western Himalayan region, capsicum production is frequently constrained by fluctuating temperature regimes, erratic rainfall, and relatively short growing seasons. To cope with these constraints, farmers often adopt staggered transplanting schedules to optimize crop establishment and extend the harvesting window. Such temporal variation in transplanting time can create substantial environmental heterogeneity within a single location, thereby intensifying GEI and influencing genotype performance. However, systematic evaluation of GEI in capsicum under staggered transplanting environments remains limited 14 . Importantly, the environments considered in the present study represent temporal environmental variation generated by combinations of years and transplanting dates (year × transplanting date), rather than geographically distinct test locations. These temporally structured environments provide an opportunity to examine genotype responses to seasonal variability under hill production conditions.

Therefore, the present investigation hypothesizes that significant genotype × environment interaction exists among capsicum genotypes across staggered transplanting environments, and that multi-variate GEI models such as AMMI and GGE biplot can effectively identify stable and high-performing genotypes along with informative test environments. Accordingly, AMMI and GGE biplot models are employed to quantitatively characterize genotype × environment interaction for yield and its contributing traits in Capsicum annuum . The specific objectives were to evaluate genotype stability across temporally defined environments (year × transplanting date combinations), identify superior genotypes with consistent performance, and assess the discriminative ability and representativeness of the tested environments. By integrating multivariate GEI modelling with trait-level resolution, this study aims to provide analytical insights for genotype evaluation and selection under variable hill production conditions 15 .

Materials and methods

Plant materials

The experimental material consisted of 96 capsicum ( Capsicum annuum L.) genotypes (Table 1 ) sourced from diverse national and international institutions to capture broad genetic variability representing exotic introductions, indigenous germplasm, local cultivars, released varieties, advanced breeding lines, and MAGIC (Multi-parent Advanced Generation Inter-Cross) lines. All accessions were open-pollinated or stabilized breeding materials. These included accessions from the World Vegetable Center, Taiwan (2); Kyungpook National University, Daegu, Republic of Korea (1); National Bureau of Plant Genetic Resources (NBPGR), New Delhi, India (5); ICAR–Indian Institute of Horticultural Research (IIHR), Hesaraghatta, India (2); ICAR–Indian Institute of Vegetable Research (IIVR), Varanasi, India (2); ICAR–Indian Agricultural Research Institute (IARI) Regional Station, Katrain, India (3); Punjab Agricultural University (PAU), Ludhiana, India (1); Dr. Y.S. Parmar University of Horticulture and Forestry (YSP UHF), Nauni, India (1); YSP UHF–Krishi Vigyan Kendra, Kandaghat, India (1); and Chaudhary Sarwan Kumar Himachal Pradesh Krishi Vishvavidyalaya (CSK HPKV), Palampur, India (75). In addition, three genotypes were obtained through local collections from Jind (Haryana), Bhiwani (Haryana), and Chikkaballapur (Karnataka), India.

For ease of reference and analysis, the genotypes were coded sequentially from G1 to G96. Five widely cultivated cultivars viz. California Wonder (G1), Yolo Wonder (G3), Solan Bharpur (G69), KtPL-19 (G83), and Surajmukhi (G96)were included as standard checks.

Table 1 Catalogue of 96 capsicum genotypes used in the present study. S. no. Code Genotypes Source S. no. Code Genotypes Source 1 G1 California Wonder (C) ICAR-IARI, Katrain 49 G49 HPC-10 CSK HPKV, Palampur 2 G2 Kandaghat Selection YSP UHF, Kandaghat 50 G50 HPC-11 CSK HPKV, Palampur 3 G3 Yolo Wonder (C) ICAR-IARI, Katrain 51 G51 HPC-13 CSK HPKV, Palampur 4 G4 HPC-39 CSK HPKV, Palampur 52 G52 HPC-15 CSK HPKV, Palampur 5 G5 IC-653,105 NBPGR, New Delhi 53 G53 HPC-16 CSK HPKV, Palampur 6 G6 HPC-39-BR CSK HPKV, Palampur 54 G54 HPC-17 CSK HPKV, Palampur 7 G7 HPC-3–10 CSK HPKV, Palampur 55 G55 HPC-20 CSK HPKV, Palampur 8 G8 IC-653,106 CSK HPKV, Palampur 56 G56 HPC-22 CSK HPKV, Palampur 9 G9 HPC-3–3 L CSK HPKV, Palampur 57 G57 HPC-24 CSK HPKV, Palampur 10 G10 HPC-2–16 CSK HPKV, Palampur 58 G58 HPC-24-1 CSK HPKV, Palampur 11 G11 HPC-2-A CSK HPKV, Palampur 59 G59 HPC-25 CSK HPKV, Palampur 12 G12 HPC-2-B CSK HPKV, Palampur 60 G60 HPC-29 CSK HPKV, Palampur 13 G13 HPC-5-A CSK HPKV, Palampur 61 G61 HPC-30 CSK HPKV, Palampur 14 G14 HPC-5-B CSK HPKV, Palampur 62 G62 HPC-32 CSK HPKV, Palampur 15 G15 HPC-AC-48 CSK HPKV, Palampur 63 G63 HPC-35 CSK HPKV, Palampur 16 G16 HPC-S-22 CSK HPKV, Palampur 64 G64 HPC-N-13 CSK HPKV, Palampur 17 G17 EC-464,107 World Vegetable Center, Taiwan 65 G65 HPC-PC-1 CSK HPKV, Palampur 18 G18 EC-464,115 World Vegetable Center, Taiwan 66 G66 HPC-66 CSK HPKV, Palampur 19 G19 HPC-O-1 CSK HPKV, Palampur 67 G67 HPC-N-2 CSK HPKV, Palampur 20 G20 HPC-O-2 CSK HPKV, Palampur 68 G68 HPC-N-1 CSK HPKV, Palampur 21 G21 HPC-223 CSK HPKV, Palampur 69 G69 Solan Bharpur (C) YSP UHF, Nauni 22 G22 HPC-AC-48-DG CSK HPKV, Palampur 70 G70 HPC-N-20 CSK HPKV, Palampur 23 G23 HPC-A-3 CSK HPKV, Palampur 71 G71 HPC-N-23 CSK HPKV, Palampur 24 G24 HPC-A-1 CSK HPKV, Palampur 72 G72 HPC-12 CSK HPKV, Palampur 25 G25 HPC-A-2 CSK HPKV, Palampur 73 G73 HPC-23 CSK HPKV, Palampur 26 G26 HPC-A-5 CSK HPKV, Palampur 74 G74 HPC-26 CSK HPKV, Palampur 27 G27 HPC-A-6 CSK HPKV, Palampur 75 G75 HPC-28 CSK HPKV, Palampur 28 G28 HPC-123-13 CSK HPKV, Palampur 76 G76 HPC-40 CSK HPKV, Palampur 29 G29 HPC-122-2 CSK HPKV, Palampur 77 G77 HPC-231 CSK HPKV, Palampur 30 G30 HPC-123-3 CSK HPKV, Palampur 78 G78 HPC-219 CSK HPKV, Palampur 31 G31 HPC-S-31 CSK HPKV, Palampur 79 G79 HPC-79 CSK HPKV, Palampur 32 G32 HPC-S-32 CSK HPKV, Palampur 80 G80 HPC-80 CSK HPKV, Palampur 33 G33 HPC-S-33 CSK HPKV, Palampur 81 G81 IC-653,107 NBPGR, New Delhi 34 G34 HPC-S-34 CSK HPKV, Palampur 82 G82 EC-497,637 Kyungpook National University, Korea 35 G35 HPC-S-35 CSK HPKV, Palampur 83 G83 KtPL-19 (C) ICAR-IARI, Katrain 36 G36 HPC-S-36 CSK HPKV, Palampur 84 G84 HPC-P-84 CSK HPKV, Palampur 37 G37 HPC-S-37 CSK HPKV, Palampur 85 G85 Arka Lohit ICAR-IIHR, Bengaluru, Karnataka 38 G38 HPC-S-38 CSK HPKV, Palampur 86 G86 LC-C-1 Local collection from Bhiwani, Haryana 39 G39 HPC-1 CSK HPKV, Palampur 87 G87 Arka Meghna ICAR-IIHR, Hesaraghatta 40 G40 HPC-2 CSK HPKV, Palampur 88 G88 Punjab Tej PAU, Ludhiana 41 G41 HPC-3 CSK HPKV, Palampur 89 G89 LC-C-2 Local collection from Bhiwani, Haryana 42 G42 HPC-5-1 CSK HPKV, Palampur 90 G90 Kashi Ratna ICAR-IIVR, Varanasi 43 G43 HPC-6 CSK HPKV, Palampur 91 G91 LC-C-3 Local collection from Chikkaballapur, Karnataka 44 G44 HPC-6-1 CSK HPKV, Palampur 92 G92 LC-C-4 Local collection from Jind, Haryana 45 G45 HPC-7 CSK HPKV, Palampur 93 G93 Kashi Gaurav ICAR-IIVR, Varanasi 46 G46 HPC-7-1 CSK HPKV, Palampur 94 G94 IC-635,240 NBPGR, New Delhi 47 G47 HPC-8-1 CSK HPKV, Palampur 95 G95 IC-635,241 NBPGR, New Delhi 48 G48 HPC-9 CSK HPKV, Palampur 96 G96 Surajmukhi (C) CSK HPKV, Palampur Note - Throughout the study, genotypes were referred by codes instead of names; (C) = Check variety.

Experimental site

The Experimental Farm is situated in the mid-hill zone of Himachal Pradesh (Zone-II) at an elevation of 1290.8 m above mean sea level, lying between 32 o 6′ N latitude and 76 o 3′ E longitude. The soil is silty clay loam, having an acidic pH (5.0 to 5.6). The climate is predominantly humid sub-temperate, with an average annual rainfall of about 2500 mm, nearly 80 per cent of which occurs between June to September. The mean weekly meteorological data recorded at the Meteorological Observatory of the Department of Agronomy, College of Agriculture, CSK HPKV, Palampur, during the crop growing period of the location are given in Figs. 1 and 2 .

Fig. 1 Mean weekly meteorological data at Palampur (February-July 2022).

Fig. 2 Mean weekly meteorological data at Palampur (February-July 2023).

Experimental layout

Field evaluation trials were conducted during the summer–rainy seasons of 2022 and 2023 at the Experimental Farm of CSK Himachal Pradesh Krishi Vishvavidyalaya, Palampur. Seedlings were raised in plastic pro-trays under protected conditions using staggered sowing dates (5 January and 10 February in both years) following seed treatment with Bavistin and Dithane M-45.Transplanting was carried out on four dates to create distinct test environments: E1 (15 March 2022), E2 (15 April 2022), E3 (15 March 2023), and E4 (15 April 2023). Staggered transplanting dates were employed to deliberately expose genotypes to contrasting thermal and moisture regimes, thereby enhancing the detection and interpretation of genotype × environment interactions.

The experiment was laid out in an alpha-lattice design (12 × 8) with three replications in each environment. Each genotype was planted in a single row of ten plants at a spacing of 60 × 45 cm. The field was prepared by disc ploughing followed by rotavating, and fertilization was applied at 20 t ha⁻¹ farmyard manure and 90:75:50 kg N: P₂O₅:K₂O ha⁻¹ as per recommended practices. One-third of nitrogen and full doses of phosphorus and potassium were applied basally, while the remaining nitrogen was top-dressed in two equal splits at 30 and 45 days after transplanting.

Data observations

Observations were recorded on five randomly selected plants per genotype per replication, excluding border plants. Days to 50% flowering and days to first picking were noted as days from transplanting, while plant height (cm) was measured at final harvest. Harvest duration (days) was calculated as the interval between the first and last harvest. Fruit traits, including number of lobes per fruit, fruit length (cm), fruit width (cm), pericarp thickness (mm), average fruit weight (g), number of marketable fruits per plant, and marketable fruit yield per plant (g), were recorded from cumulative harvests; fruit shape index was calculated as the ratio of fruit length to fruit width. Total soluble solids (°Brix) were measured using a digital refractometer, and ascorbic acid content (mg 100 g⁻¹ fresh weight) was estimated by the 2,6-dichlorophenol indophenol titration method.

Statistical analysis

All statistical analyses were performed using R 16 software (version 4.2.1). Fruit yield and related traits were subjected to combined analysis of variance (ANOVA) to test the significance of genotype (G), environment (E), and genotype × environment interaction (GEI). In the ANOVA model, genotypes were considered as fixed effects, whereas environments (year × transplanting date combinations) were treated as random effects, following the standard framework for multi-environment trials aimed at genotype evaluation. The AMMI model was applied to partition additive main effects and multiplicative GEI components according to.

Where, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{Y}_{ij}$$\end{document} = observed mean of genotype i in environment j , \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mu\:$$\end{document} = grand mean, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{G}_{i}$$\end{document} = genotype effect, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{E}_{j}$$\end{document} = environment effect, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\lambda\:}_{k}$$\end{document} = singular value (eigenvalue) of IPC axis k , \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\alpha\:}_{ik}$$\end{document} = genotype score for axis k , \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{jk}$$\end{document} = environment score for axis k , N = number of significant interaction principal components (IPCs) retained, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\rho\:}_{ij}$$\end{document} = residual (noise). For GGE biplot analysis, genotype main effects and GEI were jointly analyzed using facilitating visualization graphical visualization of genotype performance and test environment discrimination.

Where, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{Y}_{ij}$$\end{document} = observed value of genotype i in environment j , \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mu\:$$\end{document} = grand mean, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{E}_{j}$$\end{document} = environment main effect, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\lambda\:}_{k}$$\end{document} = singular value for PC axis k , \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\xi\:}_{ik}$$\end{document} = genotype eigenvector score, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\eta\:}_{jk}$$\end{document} = environment eigenvector score, N = number of principal components retained (usually 2), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\epsilon\:}_{ij}$$\end{document} = residual. Each year-transplanting combination was treated as an independent environment.

AMMI analysis was performed using the agricolae package 17 . The significance of interaction principal component axes (IPCAs) was evaluated using Gollob’s F-test at the 5% probability level, and IPCAs were sequentially tested to determine the number of statistically significant interaction components retained in the AMMI model. Only significant IPCAs representing meaningful interaction patterns were retained for interpretation, while non-significant components were considered as noise and excluded from further analysis. This approach enabled effective separation of signal from noise in the GEI structure, thereby improving the reliability of genotype stability assessment.

GGE biplots were generated using the GGEBiplotGUI package 18 – 20 . Environment-centered data (centering = 2) were used to remove environmental main effects, while scaling was set to 0 (no scaling) to preserve the original variance structure of the dataset. Singular value partitioning was genotype-focused (SVP = 1) to facilitate genotype comparison across environments. The “which-won-where” polygon view of the GGE biplot was used to identify winning genotypes and delineate possible mega-environment patterns, while the average environment coordination (AEC) view was employed to evaluate genotype mean performance and stability across environments. Additionally, the discriminative ability and representativeness of test environments were assessed using the environment vector view of the GGE biplot.

Results

Combined analysis of variance

Pooled ANOVA revealed highly significant ( p < 0.01) effects of genotype (G), environment (E), and genotype × environment interaction (G×E) for all traits studied (Table 2 ). The presence of significant genotypic variation provided a robust basis for subsequent GEI analysis and the identification of stable, high-performing genotypes for yield and related traits in capsicum. Partitioning of total variation revealed that genotypic effects contributed the largest proportion of variance for most of the traits studied, particularly pericarp thickness (99.01%), fruit width (97.84%), average fruit weight (96.98%), fruit shape index (96.88%), fruit length (95.30%), harvest duration (91.60%), marketable fruits per plant (98.93%), marketable fruit yield per plant (81.37%), plant height (69.84%), days to first picking (61.22%), and days to 50% flowering (63.53%), indicating strong genetic control and substantial variability among the evaluated genotypes. Genotype × environment interaction (GEI) accounted for a relatively greater proportion of variation for ascorbic acid (18.91%), days to 50% flowering (13.23%), marketable fruit yield per plant (10.78%), days to first picking (10.56%), plant height (7.38%), harvest duration (3.58%), fruit length (1.79%), average fruit weight (1.51%), fruit shape index (1.15%), fruit width (1.04%), marketable fruits per plant (0.83%), pericarp thickness (0.29%), lobes per fruit (4.04%), and total soluble solids (0.59%), suggesting differential genotype responses across the evaluated environments. Environmental effects were comparatively more pronounced for days to first picking (13.15%) and plant height (7.64%), reflecting the influence of temporal environmental conditions on these traits.

Table 2 Pooled analysis of variance for fruit yield traits in 96 capsicum genotypes. Source of variation Days to 50 per cent flowering Days to first picking df SS MS %SS SS MS %SS Environments 3 1012.90 337.64* 5.61 2649.10 883.04* 13.15 Genotypes 95 11473.90 120.78 63.53 12334.00 129.83 61.22 Replications 8 97.10 12.14 0.54 139.30 17.41 0.69 G×E interaction 285 2389.70 8.39* 13.23 2127.60 7.47* 10.56 Blocks within replications 132 458.40 3.47* 2.54 524.90 3.98* 2.61 Pooled error 628 2629.20 4.19 14.56 2371.10 3.78 11.77 Source of variation Plant height (cm) Harvest duration (days) df SS MS %SS SS MS %SS Environments 3 4224.00 1407.97* 7.64 2872.00 957.40* 1.60 Genotypes 95 38613.00 406.45 69.84 164177.00 1728.18 91.60 Replications 8 842.00 105.28 1.52 77.00 9.60 0.04 G×E interaction 285 4083.00 14.33* 7.38 6424.00 22.54* 3.58 Blocks within replications 132 1422.00 10.77* 2.57 859.00 6.51* 0.48 Pooled error 628 6104.00 9.72 11.04 4815.00 7.67 2.69 Source of variation Lobes per fruit Fruit length (cm) df SS MS %SS SS MS %SS Environments 3 0.00 0.001* 0.00 2.61 0.87* 0.21 Genotypes 95 368.80 3.88 368.80 1182.36 12.45 95.30 Replications 8 0.09 0.01 0.09 0.34 0.04 0.03 G×E interaction 285 4.04 0.01* 4.04 22.22 0.08* 1.79 Blocks within replications 132 1.17 0.01* 1.17 5.33 0.04* 0.43 Pooled error 628 4.88 0.01 4.88 27.77 0.04 2.24 Source of variation Fruit width (cm) Fruit shape index df SS MS %SS SS MS %SS Environments 3 3.02 1.01* 0.15 1.60 0.55* 0.03 Genotypes 95 2011.81 21.18 97.84 4745.70 49.96 96.88 Replications 8 0.03 0.00 0.00 1.20 0.16 0.02 G×E interaction 285 21.36 0.07* 1.04 56.40 0.20* 1.15 Blocks within replications 132 3.15 0.02* 0.15 16.20 0.12* 0.33 Pooled error 628 16.95 0.03 0.82 77.50 0.12 1.58 Source of variation Pericarp thickness (mm) Average fruit weight (g) df SS MS %SS SS MS %SS Environments 3 0.88 0.29* 0.07 1334.00 444.66* 0.67 Genotypes 95 1247.48 13.13 99.01 193298.00 2034.72 96.98 Replications 8 0.07 0.01 0.01 48.00 5.95 0.02 G×E interaction 285 3.61 0.01* 0.29 3009.00 10.56* 1.51 Blocks within replications 132 1.58 0.01* 0.13 285.00 2.16* 0.14 Pooled error 628 6.37 0.01 0.51 1339.00 2.13 0.67 Source of variation Marketable fruits per plant Marketable fruit yield per plant(g) df SS MS %SS SS MS %SS Environments 3 927.00 309.10* 0.08 732263.00 244088.00* 4.37 Genotypes 95 1202232.00 12655.10 98.93 13649012.00 143674.00 81.37 Replications 8 27.00 3.30 0.00 13612.00 1701.00 0.08 G×E interaction 285 10115.00 35.50* 0.83 1808733.00 6346.00* 10.78 Blocks within replications 132 332.00 2.50* 0.03 89866.00 681.00* 0.54 Pooled error 628 1621.00 2.60 0.13 481434.00 767.00 2.87 Source of variation Total soluble solids (°Brix) Ascorbic acid (mg/100 g) df SS MS %SS SS MS %SS Environments 3 0.01 0.00* 0.00 2931.00 977.11* 0.93 Genotypes 95 873.99 9.20 94.74 247377.00 2603.97 78.13 Replications 8 0.22 0.03 0.02 46.00 5.80 0.01 G×E interaction 285 5.43 0.02* 0.59 59882.00 210.11* 18.91 Blocks within replications 132 7.86 0.06* 0.85 1081.00 8.19* 0.34 Pooled error 628 35.03 0.06 3.80 5303.00 8.44 1.67 *, indicates significance at P ≤ 0.05, SS- sum of squares, MS- mean sum of squares, %SS- per cent sum of squares.

AMM I biplot analysis for marketable fruit yield per plant

AMMI 1 biplot analysis

The AMMI1 biplot revealed a pronounced genotype × environment (G×E) interaction for marketable fruit yield per plant, with the first interaction principal component axis (IPCA1) accounting for 45.4% of the total interaction variance (Fig. 3 ). Genotypes G4, G5, G6, G8, and G9, positioned towards the positive end of the x-axis, exhibited higher mean yield across environments, indicating superior yield potential. In contrast, genotypes G35, G46, G57, G68, and G69, clustered close to the origin, displayed minimal interaction effects, reflecting greater yield stability across environments. Environmental differentiation was evident, with E3 showing a strong positive IPCA1 score, indicative of favourable conditions for yield expression and enhanced genotype discrimination. Conversely, E2, located in the negative IPCA1 direction, exerted a restrictive influence on yield performance for most genotypes. Genotypic proximity to the IPCA1 zero line indicated stability, whereas increasing deviation reflected greater environmental sensitivity.

AMMI 2 biplot analysis

The AMMI 2 biplot, integrating IPCA1 (45.4%) and IPCA2 (34.3%), cumulatively explained 79.7% of the total G×E interaction for marketable fruit yield per plant. Genotypes G12, G39, G41, G56, G78, G79, and G89, located near the origin, exhibited limited interaction with environments, confirming their broad adaptability and consistent yield performance. The polygon view of the biplot enabled clear identification of environment-specific winners, with G86 in E1, G48 and G52 in E2, G23 and G34 in E3, and G88 in E4, demonstrating specific adaptation to respective environments. The relative length of environmental vectors reflected differential interaction intensities, highlighting the heterogeneity among test environments. Collectively, the AMMI2 biplot complemented the AMMI1 results by distinguishing genotypes with wide adaptation from those with pronounced environment-specific yield advantages.

Fig. 3 AMMI1 and AMMI2 biplots for marketable fruit yield per plant.

The letters ‘E’ represents the transplanting dates of environments where E1- March 15, 2022; E2- April 15, 2022; E3- March 15, 2023; and E4- April 15, 2023 at CSK HPKV, Palampur; 1–96 represents genotypes codes from G1 to G96.

AMMI biplot interpretation of phenological, morphological, yield-component and quality traits

AMMI1 biplot analysis across phenological traits revealed a moderate to high magnitude of genotype × environment (G×E) interaction, with IPCA1 explaining 59.5% and 64.9% of the interaction variance for days to 50% flowering and days to first picking, respectively. Early flowering and early maturity were consistently associated with negative IPCA1 scores, with genotypes G4, G5, G6, G20, G29, and G37 exhibiting early flowering, while G4, G5, G6, G17, G18, and G62 were early for first picking. Genotypes G14, G24, G59, G61, G75 (flowering) and G41, G44, G66, and G76 (first picking), positioned close to the IPCA1 origin, demonstrated greater phenological stability. AMMI2 biplots cumulatively explained 89.2% and 93.0% of the interaction variance, respectively, and identified environment-specific winners, notably G52 (E1), G25 and G56 (E2), G55 and G92 (E3), and G4 (E4) for flowering, and G71 (E1), G40 (E2), G75 (E3), and G30 and G59 (E4) for first picking, confirming the presence of crossover interaction (Fig S1 ). The identification of “winning” genotypes in the AMMI2 polygon view is based on graphical inference from the AMMI Model.

For vegetative growth and harvest traits, AMMI1 biplots indicated substantial interaction effects, with IPCA1 accounting for 52.3% of the variance for plant height and 51.0% for harvest duration. Taller plant stature was associated with genotypes G18, G42, G85, G88, and G91, whereas G12, G15, G43, G44, and G45 showed shorter stature. Genotypes G8, G9, G40, and G60 combined moderate height with stability. Extended harvest duration was observed in G85, G87, G91, G93, G94, and G95, while G2, G31, G33, G40, and G51 displayed stable, moderate harvest duration. AMMI2 biplots explaining 83.6% (plant height) and 82.6% (harvest duration) of the interaction variance identified G87 (E1), G47 (E2), G18 and G78 (E3), and G50 (E4) as winning genotypes for plant height, and G94 (E1), G48 (E2), G85 and G96 (E3), and G36 and G57 (E4) for harvest duration, demonstrating strong environment-specific adaptation.

Fruit morphology and yield-component traits exhibited comparatively stronger interaction effects. IPCA1 explained 78.7% of the interaction for lobes per fruit, 55.5% for fruit length, 59.3% for fruit width, 82.6% for fruit shape index, 80.2% for pericarp thickness, 46.3% for average fruit weight, and 47.4% for marketable fruits per plant. High lobe number was associated with G17, G20, G28, G29, G38, G39, G52, and G93, whereas G46, G53, G70, G79, and G90 were stable across environments. Longer fruits were observed in G65, G81, G82, G83, and G85, while G13, G14, G27, G42, and G57 showed stability. Wider fruits were linked to G7 and G10, with G41, G85, G90, and G93 being stable. Elongated fruit shape was most pronounced in G76, G77, G78, G81, G94, and G95, while G10, G26, G42, and G72 exhibited stable expression. Thicker pericarp was recorded in G3, G15, G69, G70, G89, and G93, whereas G47, G56, G60, G86, and G91 showed minimal interaction. AMMI2 polygon views (cumulative variance 79.1–95.8%) consistently identified environment-specific winners, including G49, G31, G47, G17 (lobes), G54, G65, G12, G83 (fruit length), G36, G66, G26, G5 (fruit width), G81, G26, G77, G94 (fruit shape index), G16, G92, G3, G19 (pericarp thickness), G7, G1 and G11, G69, G14 and G17 (average fruit weight), and G95, G7 and G51, G94, and G84 (marketable fruits per plant) Fig S2 .

Quality traits displayed the strongest interaction structure. For total soluble solids, IPCA1 alone explained 97.6% of the interaction, while ascorbic acid showed 79.7% contribution from IPCA1. High TSS expression was observed in G40, G79, and G81, whereas G12, G48, G50, G54, and G90 were stable across environments. For ascorbic acid, G21, G22, G61, and G78 recorded higher values, while G51, G55, G79, and G85 exhibited stability. AMMI2 biplots (cumulative variance 95.0–99.4%) identified G1 (E1), G39 and G63 (E2), G88 and G85 (E3), and G81 (E4) as winners for TSS, and G6 (E1), G32 (E2), G15 and G57 (E3), and G66 (E4) for ascorbic acid, highlighting pronounced environment-specific expression of biochemical traits Fig S3.

GGE biplot of marketable fruit yield per plant (g)

GGE biplot analysis for marketable fruit yield per plant across four environments revealed pronounced genotype × environment interaction (Fig. 4 ). The first two principal components explained 99.62% of the total variation, with PC1 accounting for 99.04% and PC2 for 0.58%.

The environment vector view showed wide dispersion of environments along PC1, reflecting substantial differences in their discriminating ability. Environments clustered into two groups, with E1 and E4 forming one group and E2 and E3 forming another. Acute angles (< 90°) among all environment vectors indicated positive correlations, suggesting that the environments were comparable while still effective in differentiating genotypic performance. The close proximity of E2 and E3 vectors further indicated a high degree of similarity between these environments.

In the discriminativeness versus representativeness view, E2 exhibited the longest vector length and was identified as the most discriminative environment. In contrast, E3 showed minimal deviation from the average environment axis, indicating high representativeness and suitability as an ideal test environment for selecting stable genotypes.

The mean performance versus stability biplot identified G5 as the highest-yielding genotype, followed by G6, G4, G8, and G9, all positioned far along the positive PC1 axis. Genotypes G1, G41, G43, G56, and G83 were located close to the biplot origin, indicating minimal interaction effects and superior yield stability across environments.

The which-won-where view of the GGE biplot partitioned the polygon into ten sectors defined by vertex genotypes. However, only four sectors contained test environments, indicating four environment-specific winning genotypes. Genotype G5 was the winning genotype in E1, G9 in E2, and G6 in both E3 and E4, demonstrating distinct patterns of specific adaptation across environments.

Fig. 4 GGE biplots for marketable fruit yield per plant. (a). GGE biplots for four environments. (b). Discriminativeness vs. representativeness biplot. (c). Mean vs. stability biplots for 96 capsicum genotypes. (d). Which-won-where biplot with polygon view for winning genotypes.

The letters ‘E’ represents the transplanting dates of environments where E1- March 15, 2022; E2- April 15, 2022; E3- March 15, 2023; and E4- April 15, 2023 at CSK HPKV, Palampur; 1–96 represents genotypes codes from G1 to G96.

GGE biplot analysis of phenological, morphological, yield components, and quality traits

GGE biplot analyses across phenological, morphological, yield-component, and quality traits revealed that the first two principal components (PC1 and PC2) together accounted for more than 90% of the total variation for all traits, confirming the robustness of the model and the reliability of graphical interpretations (Fig S4). Across traits, environments exhibited predominantly acute angles among their vectors, indicating positive correlations and consistent ranking of genotypes, though their discriminating ability and representativeness varied substantially. Environment E2 emerged as highly discriminative for several phenological and quality traits, while E3 was generally the most representative, making it a reliable test environment for identifying broadly adapted and stable genotypes.

For flowering and earliness traits, environments clustered into two major groups, with E3 and E4 showing strong similarity. Genotypes G1, G5, G6, G20, and G80 combined early flowering with high stability, whereas genotypes such as G24, G51, G61, G78, and G79 exhibited minimal G×E interaction. For days to first picking, genotypes G5, G17, G18, and G93 consistently showed early maturity, while G2, G25, G45, G62, and G76 were among the most stable across environments. Plant height was strongly influenced by environmental effects, with E4 being the most discriminative environment; genotypes G42, G48, G85, G88, and G91 were superior for this trait, whereas G8, G9, G40, G60, and G94 exhibited high stability.

Harvest duration displayed pronounced G×E interaction, with E3 showing the greatest discriminating power and E1 being most representative. Genotypes G85, G87, G93, G95, and G96 recorded extended harvest duration, while G16, G31, G33, G40, and G82 remained stable across environments. For fruit morphology traits, including lobes per fruit, fruit length, fruit width, fruit shape index, and pericarp thickness, PC1 and PC2 jointly explained more than 99% of total variation. E2 and E4 were consistently the most discriminative environments, whereas E3 was generally the most representative. Genotypes G7, G65, G81, G82, G83, and G85 showed superior performance for fruit size and shape attributes, while genotypes such as G12, G13, G23, G29, G39, G41, G43, G66, and G72 demonstrated stable expression with low interaction effects.

For average fruit weight and marketable fruits per plant, E3 exhibited the highest discriminating ability, while E4 and E1 were more representative. Genotypes G1, G5, G8, G23, and G69 recorded higher fruit weight, whereas G25, G29, G35, G53, and G57 were highly stable. Genotypes G86, G87, G93, G94, and G96 were superior for marketable fruits per plant, with several showing specific adaptation to individual environments.

Quality traits showed contrasting patterns of environmental sensitivity. Total soluble solids exhibited a relatively higher G×E interaction effect, with E4 being the most discriminative and E3 the most representative environment. Genotypes G21 and G22 consistently recorded higher TSS, while G14, G22, G60, G64, and G78 showed stable performance. For ascorbic acid content, E2 emerged as the most discriminative environment and E1 as the most representative. Genotypes G23, G48, G79, G81, and G84 exhibited higher ascorbic acid concentration, whereas G7, G11, G14, G38, and G73 showed greater stability across environments.

Overall, the GGE biplot analyses effectively distinguished between broadly adapted, stable genotypes and those with specific environmental adaptation across yield, yield-contributing, and quality traits, highlighting the potential of staggered transplanting environments for dissecting genotype × environment interactions in capsicum.

Discussion

The highly significant effects of genotype, environment, and their interaction for all studied traits clearly demonstrate the complex and non-additive nature of yield expression and associated characters in capsicum across staggered transplanting environments 21 , 22 . Although a high proportion of genotypic variance (> 90%) was observed for several traits, these estimates should be interpreted with caution, as they may be influenced by the specific experimental conditions and relatively limited environmental variation. Such high values do not necessarily indicate strong genetic control in the absence of supporting genetic parameters such as heritability or repeatability. Therefore, these results should be considered indicative rather than conclusive for genetic determination of the traits. The magnitude of the G×E interaction observed in the pooled ANOVA underscores the necessity of employing multivariate stability models such as AMMI and GGE biplot analysis, as conventional mean-based comparisons alone would inadequately capture genotype performance under variable agro-temporal conditions 23 . Such pronounced interaction effects are typical for capsicum, a crop known for its sensitivity to temperature, photoperiod, and moisture fluctuations during flowering and fruit development 24 , 25 .

The relative contribution of genotypic effects exceeding environmental and G×E components for most quality and yield-contributing traits 26 . This indicates strong prospects for selection and genetic improvement. Conversely, the predominance of environmental effects for traits such as days to first picking and plant height highlights their plastic response to transplanting time-induced climatic variation, emphasizing the importance of environment-specific management rather than genetic selection for these attributes 27 – 30 . The high proportion of interaction variance captured by IPCA1 for total soluble solids should be interpreted with caution when the overall genotype × environment interaction is very small. In such situations, the percentage explained by the first interaction component is calculated relative to a very small interaction sum of squares, which can lead to apparently inflated values. Therefore, a large contribution of IPCA1 does not necessarily indicate strong or complex interaction effects. Instead, it suggests that the limited interaction variance present can be adequately summarized by a single interaction axis in the AMMI model, reflecting a relatively simple interaction structure rather than pronounced genotype-specific environmental responses 31 . Although PC2 was used to describe genotype stability, its contribution to the interaction variance was relatively small for certain traits (e.g., marketable fruit yield). Therefore, interpretations based on PC2 should be considered primarily descriptive and interpreted with caution. In such cases, greater emphasis may be placed on PC1, which accounts for the major proportion of the interaction variation.

The GGE biplot analysis for marketable fruit yield per plant partitioned genotype main effects and G×E interaction, revealing distinct adaptation patterns among the tested genotypes 32 – 35 . The first two principal components explained a very high proportion of the variation and PC2 for only 0.58%. While this indicates that most of the variation is captured by the first principal component, the extremely small contribution of PC2 suggests that interpretations of stability based on PC2 should be made cautiously. In such situations, the biplot largely reflects differences in mean genotype performance rather than strong differential stability patterns among genotypes. Accordingly, the graphical interpretation should be considered primarily descriptive of genotype performance across transplanting environments rather than definitive evidence of complex stability relationships. The dispersion of environments along the PC1 axis indicates differences in their discriminating ability, suggesting that staggered transplanting dates generated varying test conditions capable of eliciting differential genotypic responses 36 , 37 . However, it should be noted that all environments represent different transplanting dates within the same experimental location, and therefore the observed variation primarily reflects temporal climatic differences rather than spatial environmental heterogeneity.

The positive correlations among environments, as evidenced by acute angles between environment vectors, suggest a degree of consistency in genotype ranking despite differences in discriminative power 38 . This indicates that while the environments varied in their ability to expose genetic differences, they were not antagonistic, thereby enabling reliable identification of broadly adapted genotypes. The close alignment of E2 and E3 further implies similarity in prevailing climatic conditions during critical crop growth phases, resulting in comparable genotype responses 39 – 41 . The discriminativeness versus representativeness view provided insight into the utility of individual environments for selection. The high discriminating ability of E2 suggests its effectiveness in revealing genotypic differences, making it a useful environment for identifying high-performing genotypes. In contrast, the strong representativeness of E3, as indicated by its proximity to the average environment axis, suggests that it may serve as a suitable environment for selecting genotypes with relatively stable performance across transplanting windows. These interpretations should be considered cautiously because the environments differ only temporally within a single research location rather than representing independent agro-climatic zones 42 , 43 .

Genotypes positioned far along the positive PC1 axis with minimal projection onto PC2 combined high mean yield with acceptable stability, indicating their potential suitability for cultivation across the tested transplanting environments. In contrast, genotypes clustered near the origin exhibited minimal interaction effects, reflecting consistent performance but relatively moderate yield potential 44 . This distinction is important from a breeding perspective, as it enables breeders to select either broadly adapted stable genotypes or specifically adapted high-yielding ones depending on breeding objectives. The which-won-where pattern revealed the presence of crossover interactions, with different genotypes emerging as winners in specific transplanting environments 45 . However, although the polygon structure delineates several geometric sectors, only a subset of these sectors contains test environments. Therefore, the interpretation should be limited to identifying environment-specific winning genotypes. As all environments represent staggered transplanting dates at a single location, the results reflect temporal adaptation patterns rather than true mega-environment delineation 46 .

Overall, the integration of AMMI and GGE biplot analyses provided a comprehensive framework for examining genotype performance, stability, and adaptation under staggered transplanting environments 47 , 48 . Nevertheless, the present study was conducted at a single research farm, and therefore the findings should be interpreted as location-specific responses to temporal environmental variation. Additional multi-location testing across the northwestern Himalayan region would be required before making broader regional recommendations. These findings nevertheless highlight the usefulness of multivariate GEI analysis in identifying promising genotypes and understanding their responses to varying transplanting schedules under changing agro-climatic conditions 49 , 50 .

Conclusion

The present investigation highlights the pronounced influence of genotype × environment interaction on yield and its attributing traits in capsicum evaluated under staggered transplanting environments at Palampur. Highly significant effects of genotype, environment, and their interaction across all traits confirm the complex and non-additive nature of trait expression, emphasizing the limitations of selection based solely on mean performance. The combined use of the AMMI Model and GGE Biplot Analysis effectively dissected GEI patterns, enabling the identification of both widely adapted and environment-specific genotypes. Genotypes such as G5 and G6 exhibited consistently high yield and broad adaptability, while several others showed specific superiority under transplanting environments, supporting targeted genotype deployment under staggered transplanting conditions. Among environments, E3 was identified as the most representative and suitable for selecting stable genotypes, whereas E2 was highly discriminative for differentiating genotypic performance. Overall, this study demonstrates that integrating AMMI and GGE analyses provides a robust framework for genotype evaluation and environment characterization under staggered transplanting conditions at Palampur. These findings may guide future capsicum breeding programmes and multilocation trials aimed at validating genotype stability and refining selection strategies for staggered transplanting systems.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Jasdeep Kaur, Sonia Sood and Nikhil Thakur contributed equally to this work.

Acknowledgements

The authors sincerely acknowledge the Chaudhary Sarwan Kumar Himachal Pradesh Krishi Vishvavidyalaya, Palampur, Himachal Pradesh, for providing the necessary research facilities, experimental land, and infrastructural support to conduct the present investigation. The authors are grateful to the Department of Vegetable Science and Floriculture for continuous academic guidance and technical assistance during the study. The authors also acknowledge the field and technical staff of the Experimental Farm, CSK HPKV, Palampur, for their assistance in crop management, data recording, and trial maintenance.

Author contributions

J.K. and S.S conceived and designed the study. S.S. contributed to data acquisition and experimental work. N.T. and J.K performed data analysis and wrote the first draft of the manuscript. J.K. and N.T. critically revised the manuscript for important intellectual content. All authors reviewed and approved the final version of the manuscript.

Funding

This research received no external funding.

Data availability

The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

References

1. Jokarfard, V., Rabiei, B., Laki, E. S. & Börner, A. Stability and adaptability of grain yield in quinoa genotypes in four locations of Iran. Front Plant. Sci 15 , 1-16. 10.3389/fpls.2024.1487106 (2024). PMC11637842 39678005

2. Singh V Sharma A Kumar N Rana N Srishti. Stability models regulate the adaptation of male sterility-based chilli hybrids for agro-ecologically diverse regions of north-western Himalayas Sci. Rep. 2025 15 34742 10.1038/s41598-025-18393-0 41053325 PMC12500983 Singh, V., Sharma, A., Kumar, N. & Rana, N. Srishti. Stability models regulate the adaptation of male sterility-based chilli hybrids for agro-ecologically diverse regions of north-western Himalayas. Sci. Rep. 15 , 34742. 10.1038/s41598-025-18393-0 (2025). 41053325 PMC12500983

3. Thakur N Sharma D Kaur J Sharma V Characterizing tomato genotypes in the varied climates of north-western Himalayas and implications for environmental resilience using GGE biplot analyses Sci. Rep. 2025 15 28524 10.1038/s41598-024-83454-9 40764500 PMC12325667 Thakur, N., Sharma, D., Kaur, J. & Sharma, V. Characterizing tomato genotypes in the varied climates of north-western Himalayas and implications for environmental resilience using GGE biplot analyses. Sci. Rep. 15 , 28524. 10.1038/s41598-024-83454-9 (2025). 40764500 PMC12325667

4. Kona P AMMI and GGE biplot analysis of genotype by environment interaction for yield and yield contributing traits in confectionery groundnut Sci. Rep. 2024 14 2943 10.1038/s41598-024-52938-z 38316821 PMC10844612 Kona, P. et al. AMMI and GGE biplot analysis of genotype by environment interaction for yield and yield contributing traits in confectionery groundnut. Sci. Rep. 14 , 2943. 10.1038/s41598-024-52938-z (2024). 38316821 PMC10844612

5. Sood R AMMI and GGE biplot analysis for seed yield and stability performance of selected buckwheat genotypes under multi-environmental trials Sci. Rep. 2025 15 36830 10.1038/s41598-025-10939-6 41125639 PMC12546826 Sood, R. et al. AMMI and GGE biplot analysis for seed yield and stability performance of selected buckwheat genotypes under multi-environmental trials. Sci. Rep. 15 , 36830. 10.1038/s41598-025-10939-6 (2025). 41125639 PMC12546826

6. Sultan A Stability analysis for yield and maturity traits in coloured sweet pepper ( Capsicum annuum L. var. grossum Sendt) J. Community Mobil. Sustain. Dev. 2021 16 633 638 Sultan, A. et al. Stability analysis for yield and maturity traits in coloured sweet pepper ( Capsicum annuum L. var. grossum Sendt). J. Community Mobil. Sustain. Dev. 16 , 633–638 (2021).

7. Noman VV Bhuiyan MSR Islam MS G × E interaction of yield and yield contributing characters in chilli ( Capsicum frutescens L.) using AMMI model Bangladesh J. Plant. Breed. Genet. 2015 28 1 7 10.3329/bjpbg.v28i2.29959 Noman, V. V., Bhuiyan, M. S. R. & Islam, M. S. G × E interaction of yield and yield contributing characters in chilli ( Capsicum frutescens L.) using AMMI model. Bangladesh J. Plant. Breed. Genet. 28 , 1–7 (2015).

8. Sahmat SS Unravelling the dynamics of genotype and environment interactions on chilli ( Capsicum annuum L.) yield-related attributes in soilless planting systems Sci. Rep. 2024 14 1698 10.1038/s41598-023-50381-0 38242885 PMC10799017 Sahmat, S. S. et al. Unravelling the dynamics of genotype and environment interactions on chilli ( Capsicum annuum L.) yield-related attributes in soilless planting systems. Sci. Rep. 14 , 1698 (2024). 38242885 10.1038/s41598-023-50381-0 PMC10799017

9. Senapati BK Sarkar G Genotype-environment interaction and stability for yield and yield components in chilli ( Capsicum annuum L) Veg. Sci. 2002 29 146 148 Senapati, B. K. & Sarkar, G. Genotype-environment interaction and stability for yield and yield components in chilli ( Capsicum annuum L). Veg. Sci. 29 , 146–148 (2002).

10. Sran TS Jindal SK Chawla N Genotype by environment interaction for quality traits in chilli pepper ( Capsicum annuum L) Genetika 2021 53 23 49 10.2298/GENSR2101023S Sran, T. S., Jindal, S. K. & Chawla, N. Genotype by environment interaction for quality traits in chilli pepper ( Capsicum annuum L). Genetika 53 , 23–49 (2021).

11. Yan W Hunt LA Sheng Q Szlavnics Z Cultivar evaluation and mega-environment investigation based on the GGE biplot Crop Sci. 2000 40 596 605 10.2135/cropsci2000.403597x Yan, W., Hunt, L. A., Sheng, Q. & Szlavnics, Z. Cultivar evaluation and mega-environment investigation based on the GGE biplot. Crop Sci. 40 , 596–605 (2000).

12. Tembhurne BV Rao SK Stability analysis in chilli ( Capsicum annuum L) J. Spices Aromat. Crops 2013 22 154 164 Tembhurne, B. V. & Rao, S. K. Stability analysis in chilli ( Capsicum annuum L). J. Spices Aromat. Crops . 22 , 154–164 (2013).

13. Wondimu G Aklilu S Stability analysis of chilli ( Capsicum frutescens L.) genotypes in the rift valley areas of Ethiopia J. Agric. Sci. Food Res. 2018 9 1 5 Wondimu, G. & Aklilu, S. Stability analysis of chilli ( Capsicum frutescens L.) genotypes in the rift valley areas of Ethiopia. J. Agric. Sci. Food Res. 9 , 1–5 (2018).

14. Yan W Kang M Ma B Woods S Cornelius P GGE biplot vs. AMMI analysis of genotype-by-environment data Crop Sci. 2007 47 643 653 10.2135/cropsci2006.06.0374 Yan, W., Kang, M., Ma, B., Woods, S. & Cornelius, P. GGE biplot vs. AMMI analysis of genotype-by-environment data. Crop Sci. 47 , 643–653 (2007).

15. Hajjarpoor A Nelson WC Vadez V How process-based modelling can help plant breeding deal with G × E × M interactions Field Crops Res. 2022 283 108554 10.1016/j.fcr.2022.108554 Hajjarpoor, A., Nelson, W. C. & Vadez, V. How process-based modelling can help plant breeding deal with G × E × M interactions. Field Crops Res. 283 , 108554. 10.1016/j.fcr.2022.108554 (2022).

16. RStudio. RStudio: Integrated Development Environment for R. RStudio. (2023).

17. Olivoto T Lúcio ADC Rossoni DF metan: An R package for multi-environment trial analysis Methods Ecol. Evol. 2020 11 783 789 10.1111/2041-210X.13384 Olivoto, T., Lúcio, A. D. C. & Rossoni, D. F. metan: An R package for multi-environment trial analysis. Methods Ecol. Evol. 11 , 783–789. 10.1111/2041-210X.13384 (2020).

18. Gauch HG Model selection and validation for yield trials with interaction Biometrics 1988 44 705 715 10.2307/2531585 Gauch, H. G. Model selection and validation for yield trials with interaction. Biometrics 44 , 705–715. 10.2307/2531585 (1988).

19. Annicchiarico P Additive main effects and multiplicative interaction (AMMI) analysis of genotype-location interaction in variety trials repeated over years Theor. Appl. Genet. 1997 94 1072 1077 10.1007/s001220050517 Annicchiarico, P. Additive main effects and multiplicative interaction (AMMI) analysis of genotype-location interaction in variety trials repeated over years. Theor. Appl. Genet. 94 , 1072–1077. 10.1007/s001220050517 (1997).

20. Olivoto T Mean performance and stability in multi-environment trials I: Combining features of AMMI and BLUP techniques Agron. J. 2019 111 2949 2960 10.2134/agronj2019.03.0220 Olivoto, T. et al. Mean performance and stability in multi-environment trials I: Combining features of AMMI and BLUP techniques. Agron. J. 111 , 2949–2960. 10.2134/agronj2019.03.0220 (2019).

21. Crossa J Statistical analyses of multi-environment trials Adv. Agron. 1990 44 55 85 10.1016/S0065-2113(08)60818-4 Crossa, J. Statistical analyses of multi-environment trials. Adv. Agron. 44 , 55–85. 10.1016/S0065-2113(08)60818-4 (1990).

22. Wani KP Ahmed N Tanki MI Narayan R Stability analysis for yield and quality characters in hot pepper ( Capsicum annuum L) Capsicum Eggplant Newsl. 2003 22 75 78 Wani, K. P., Ahmed, N., Tanki, M. I. & Narayan, R. Stability analysis for yield and quality characters in hot pepper ( Capsicum annuum L). Capsicum Eggplant Newsl. 22 , 75–78 (2003).

23. Yan, W. K. & Kang, M. S. GGE Biplot Analysis: A Graphical Tool for Breeders, Geneticists and Agronomists (CRC, 2003).

24. Mohammadi R Amri A Analysis of genotype × environment interactions for grain yield in durum wheat Crop Sci. 2009 49 1177 1186 10.2135/cropsci2008.09.0537 Mohammadi, R. & Amri, A. Analysis of genotype × environment interactions for grain yield in durum wheat. Crop Sci. 49 , 1177–1186. 10.2135/cropsci2008.09.0537 (2009).

25. Datta S Jana JC Stability of chilli genotypes for different seasons at Terai zone of West Bengal J. Spices Aromat. Crops 2012 21 71 75 Datta, S. & Jana, J. C. Stability of chilli genotypes for different seasons at Terai zone of West Bengal. J. Spices Aromat. Crops . 21 , 71–75 (2012).

26. Yousefabadi VA Evaluation of yield and stability of sugar beet ( Beta vulgaris L.) genotypes using GGE biplot and AMMI analysis Sci. Rep. 2024 14 27384 10.1038/s41598-024-78659-x 39521894 PMC11550840 Yousefabadi, V. A. et al. Evaluation of yield and stability of sugar beet ( Beta vulgaris L.) genotypes using GGE biplot and AMMI analysis. Sci. Rep. 14 , 27384. 10.1038/s41598-024-78659-x (2024). 39521894 PMC11550840

27. Ortiz R Studying the effect of environmental variables on the genotype × environment interaction of tomato Euphytica 2007 153 119 134 10.1007/s10681-006-9248-7 Ortiz, R. et al. Studying the effect of environmental variables on the genotype × environment interaction of tomato. Euphytica 153 , 119–134. 10.1007/s10681-006-9248-7 (2007).

28. Basak J Identification of stable high-yielding hybrids in chilli ( Capsicum annuum L) J. Food Legumes 2019 32 139 144 Basak, J. Identification of stable high-yielding hybrids in chilli ( Capsicum annuum L). J. Food Legumes . 32 , 139–144 (2019).

29. Yan W Tinker NA An integrated biplot analysis system for displaying, interpreting and exploring genotype × environment interaction Crop Sci. 2005 45 1004 1016 10.2135/cropsci2004.0076 Yan, W. & Tinker, N. A. An integrated biplot analysis system for displaying, interpreting and exploring genotype × environment interaction. Crop Sci. 45 , 1004–1016. 10.2135/cropsci2004.0076 (2005).

30. Cabral NSS Genotype × environment interaction on experimental hybrids of chilli pepper Genet. Mol. Res. 2017 16 1 9 10.4238/gmr16029551 28437553 Cabral, N. S. S. et al. Genotype × environment interaction on experimental hybrids of chilli pepper. Genet. Mol. Res. 16 , 1–9 (2017). 10.4238/gmr16029551 28437553

31. Sood R Genotype × environment interaction studies under natural farming and inorganic production systems in maize ( Zea mays L) Electron. J. Plant. Breed. 2023 14 1446 1452 10.37992/2023.1404.175 Sood, R. et al. Genotype × environment interaction studies under natural farming and inorganic production systems in maize ( Zea mays L). Electron. J. Plant. Breed. 14 , 1446–1452. 10.37992/2023.1404.175 (2023).

32. Joshi BK Yield stability of Tartary buckwheat genotypes in Nepal Fagopyrum 2004 21 1 5 Joshi, B. K. Yield stability of Tartary buckwheat genotypes in Nepal. Fagopyrum 21 , 1–5 (2004).

33. Barchenger, D. W. et al. Stability of yield and yield components of pepper ( Capsicum annuum ), and evaluation of publicly available predictive meteorological data in East and Southeast Asia. HortScience 53, 1776–1783 (2018). 10.21273/HORTSCI13581-18

34. Raghavendra H Stability analysis of different chilli hybrids ( Capsicum annuum L.) for yield and yield-attributing traits J. Appl. Hortic. 2017 19 218 221 10.37855/jah.2017.v19i03.38 Raghavendra, H. et al. Stability analysis of different chilli hybrids ( Capsicum annuum L.) for yield and yield-attributing traits. J. Appl. Hortic. 19 , 218–221 (2017).

35. Khan MMH AMMI and GGE biplot analysis for yield performance and stability assessment of selected Bambara groundnut genotypes under multi-environment trials Sci. Rep. 2021 11 22791 10.1038/s41598-021-01411-2 34815427 PMC8611061 Khan, M. M. H. et al. AMMI and GGE biplot analysis for yield performance and stability assessment of selected Bambara groundnut genotypes under multi-environment trials. Sci. Rep. 11 , 22791. 10.1038/s41598-021-01411-2 (2021). 34815427 PMC8611061

36. Anputhas M Stability and adaptability analysis of rice cultivars using environment-centred yield in a two-way ANOVA model Commun. Biom Crop Sci. 2011 6 80 86 Anputhas, M. et al. Stability and adaptability analysis of rice cultivars using environment-centred yield in a two-way ANOVA model. Commun. Biom Crop Sci. 6 , 80–86 (2011).

37. Cooper M Predicting genotype × environment × management interactions for the design of crop improvement strategies Plant. Breed. Rev. 2022 46 467 485 10.1002/9781119874157.ch8 Cooper, M. et al. Predicting genotype × environment × management interactions for the design of crop improvement strategies. Plant. Breed. Rev. 46 , 467–485. 10.1002/9781119874157.ch8 (2022).

38. Ozukum, C. & Seyie, K. Stability analysis of Naga King Chilli ( Capsicum chinense Jacq). Indian J. Hill Farming, 5–9 (2019).

39. Singh TN Mean performances, character associations and multi-environment evaluation of chilli landraces in North Western Himalayas Sci. Rep. 2024 14 769 10.1038/s41598-024-51348-5 38191594 PMC10774388 Singh, T. N. et al. Mean performances, character associations and multi-environment evaluation of chilli landraces in North Western Himalayas. Sci. Rep. 14 , 769. 10.1038/s41598-024-51348-5 (2024). 38191594 PMC10774388

40. Farshadfar E GGE biplot analysis of genotype × environment interaction in wheat–barley disomic addition lines Aust J. Crop Sci. 2012 6 1074 1079 Farshadfar, E. et al. GGE biplot analysis of genotype × environment interaction in wheat–barley disomic addition lines. Aust J. Crop Sci. 6 , 1074–1079 (2012).

41. Roostaei M Mohammadi R Amri A Rank correlation among different statistical models in ranking of winter wheat genotypes Crop J. 2014 2 154 163 10.1016/j.cj.2014.02.002 Roostaei, M., Mohammadi, R. & Amri, A. Rank correlation among different statistical models in ranking of winter wheat genotypes. Crop J. 2 , 154–163. 10.1016/j.cj.2014.02.002 (2014).

42. Liu WC Stability and high-yield analysis of buckwheat varieties in Inner Mongolia based on the AMMI model J. North. Agric. 2021 49 18 24 10.12190/j.issn.2096-1197.2021.05.03 Liu, W. C. et al. Stability and high-yield analysis of buckwheat varieties in Inner Mongolia based on the AMMI model. J. North. Agric. 49 , 18–24. 10.12190/j.issn.2096-1197.2021.05.03 (2021).

43. Piepho HP Methods for comparing the yield stability of cropping systems J. Agron. Crop. Sci. 2001 186 135 141 10.1046/j.1439-037x.2001.00427.x Piepho, H. P. Methods for comparing the yield stability of cropping systems. J. Agron. Crop. Sci. 186 , 135–141. 10.1046/j.1439-037x.2001.00427.x (2001).

44. Pham HN Kang MS Interrelationship among and repeatability of several stability statistics estimated from international maize trials Crop Sci. 1998 28 925 928 10.2135/cropsci1988.0011183X002800060010x Pham, H. N. & Kang, M. S. Interrelationship among and repeatability of several stability statistics estimated from international maize trials. Crop Sci. 28 , 925–928. 10.2135/cropsci1988.0011183X002800060010x (1998).

45. Ndhlela T Genotype × environment interaction of maize grain yield using AMMI biplots Crop Sci. 2014 54 1992 1999 10.2135/cropsci2013.07.0448 Ndhlela, T. et al. Genotype × environment interaction of maize grain yield using AMMI biplots. Crop Sci. 54 , 1992–1999. 10.2135/cropsci2013.07.0448 (2014).

46. Parihar AK Targeting test environments and rust-resistant genotypes in lentil ( Lens culinaris ) using heritability-adjusted biplot analysis Crop Pasture Sci. 2018 69 1113 1125 10.1071/CP18259 Parihar, A. K. et al. Targeting test environments and rust-resistant genotypes in lentil ( Lens culinaris ) using heritability-adjusted biplot analysis. Crop Pasture Sci. 69 , 1113–1125. 10.1071/CP18259 (2018).

47. Thakur, N. et al. Study on genetic variability and heritability in cherry tomato ( Solanum lycopersicum var. cerasiforme ) for yield and quality traits. Agric. Res. J. 59 (6). 10.5958/2395-146X.2022.00144.2 (2022).

48. Kaur J Genetic potential and trait analysis of garlic (Allium sativum L.) landraces from the lower western Himalayas Bangladesh J. Bot. 2025 54 2 319 324 10.3329/bjb.v54i2.82026 Kaur, J. et al. Genetic potential and trait analysis of garlic (Allium sativum L.) landraces from the lower western Himalayas. Bangladesh J. Bot. 54 (2), 319–324. 10.3329/bjb.v54i2.82026 (2025).

49. Sood T Characterisation of bell pepper ( Capsicum annuum L. var. grossum Sendt.) accessions for genetic diversity and population structure based on agro-morphological and microsatellite markers Sci. Hort. 2023 321 112308 10.1016/j.scienta.2023.112308 Sood, T. et al. Characterisation of bell pepper ( Capsicum annuum L. var. grossum Sendt.) accessions for genetic diversity and population structure based on agro-morphological and microsatellite markers. Sci. Hort. 321 , 112308 (2023).

50. Kaur J Sood S Assessment of genetic association and Direct-Indirect Effects of Morphological and Quality Traits on Marketable Fruit Yield in Diverse Capsicum ( Capsicum annuum L.) Genotypes Under the North-Western Himalayas Himachal J. Agricultural Res. 2025 51 2 207 212 10.18311/HJAR/2025/53845 Kaur, J. & Sood, S. Assessment of genetic association and Direct-Indirect Effects of Morphological and Quality Traits on Marketable Fruit Yield in Diverse Capsicum ( Capsicum annuum L.) Genotypes Under the North-Western Himalayas. Himachal J. Agricultural Res. 51 (2), 207–212. 10.18311/HJAR/2025/53845 (2025).