Идентификация генотипов теплотерпимых томатов на основе комплексного фенотипного, биохимического, молекулярного анализа и анализа бипло ГПЭ.

Introduction

Tomato ( Solanum lycopersicum L.) is among the most widely cultivated vegetable crops worldwide, contributing substantially to nutritional security, agro-industry, and farmers’ livelihoods. As a thermosensitive crop of tropical origin, tomato exhibits optimal growth and reproductive performance within a relatively narrow temperature range (21–27 °C) 1 , 2 . However, in recent decades, the accelerated pace of global warming has resulted in more frequent, prolonged, and intense heat waves, particularly in tomato-growing regions of Asia, Africa, and southern Europe. High temperature episodes exceeding 35–40 °C have been reported to cause severe physiological disruptions, leading to flower abortion, impaired pollen development, reduced fruit set, and significant yield penalties 3 , 4 . Consequently, identifying and developing heat-resilient tomato cultivars has become a critical component of climate-smart agriculture.

Heat stress triggers an intricate array of morphological, physiological, biochemical and molecular changes in tomato plants. Phenotypically, heat-tolerant genotypes typically maintain acceptable levels of pollen viability, fruit set, and yield under high-temperature episodes 5 – 7 . These traits are particularly important as reproductive processes are the most heat-sensitive phase in tomato. In contrast, susceptible genotypes exhibit floral drop, malformed fruits, reduced seed set, and premature senescence 8 . Understanding these phenotypic responses helps establish baseline screening criteria for differentiating tolerant and sensitive genotypes.

At the biochemical level, high temperatures lead to excessive accumulation of reactive oxygen species (ROS), causing lipid peroxidation, membrane disintegration, protein denaturation, and metabolic dysfunction 6 , 7 . To counteract these effects, heat-tolerant genotypes activate a suite of antioxidative defense mechanisms. Osmolytes such as proline play a crucial role in osmoprotection and stabilizing proteins and membranes. Lower levels of malondialdehyde (MDA) reflect restricted lipid peroxidation, indicating better membrane stability under stress. Enhanced activities of antioxidant enzymes—superoxide dismutase (SOD), catalase (CAT), and ascorbate peroxidase (APX)—help mitigate oxidative damage by detoxifying ROS 6 , 9 . Additionally, maintenance of total chlorophyll content, relative water content (RWC), and lower electrolyte leakage (EL) are reliable biochemical markers of stress resilience 7 , 10 . Integrating these parameters provides a mechanistic understanding of heat-stress adaptation and improves precision in selecting tolerant lines.

At the molecular level, heat tolerance is regulated by a complex transcriptional network involving heat-shock transcription factors and downstream heat-shock proteins, which protect cellular structures by acting as molecular chaperones 11 . Other stress-responsive genes associated with ROS scavenging, cell wall remodeling, signal transduction, and reproductive development also contribute to thermotolerance 11 . Differential gene expression profiling and molecular marker-based characterization enable the identification of allelic variants and candidate genes associated with heat tolerance, strengthening the foundation for molecular-assisted breeding 12 . Such molecular insights are indispensable for understanding genotype-specific tolerance mechanisms and for advancing precision breeding techniques 10 .

One of the major challenges in heat tolerance breeding is the pronounced genotype × environment (G × E) interaction, which complicates the consistent identification of stable genotypes across diverse environments 13 . Traditional selection methods often fail to capture the multidimensional nature of this interaction. In recent years, GGE biplot analysis has emerged as a powerful statistical tool for deciphering G and GE components simultaneously 13 . The GGE biplot facilitates visualization of discriminative and representative testing environments, identifies high-performing and stable genotypes, and allows “which-won-where” interpretation 14 . Its integration into heat-tolerance studies enhances the reliability of genotype selection by enabling breeders to identify both specifically adapted and broadly stable genotypes across multi-year or multi-season trials.

Considering the complexity of heat-stress responses, there is a critical need for integrated approaches that combine phenotypic and biochemical attributes along with robust multivariate analysis. Such comprehensive evaluations not only deepen our understanding of heat tolerance mechanisms in tomato but also accelerate the identification of elite donor lines for breeding programs targeting high-temperature environments 15 , 16 . In this context, the present study aims to assess: i ) phenotypic variability among tomato genotypes for key reproductive and yield-related traits under high-temperature conditions, ii ) elucidate biochemical and physiological mechanisms contributing to heat tolerance, and iii ) identify stable and superior heat-tolerant genotypes using GGE biplot analysis. Overall, this integrated approach is expected to generate valuable insights into the biological basis of heat tolerance and provide a scientific foundation for developing climate-resilient tomato varieties suited to warming agro-ecosystems.

Materials and methods

Plant materials and experimental conditions

A total of 35 tomato genotypes were used in this study (Supplementary File S1). The experiment was conducted during the summer seasons (February–May) of 2025 and 2026 under controlled environmental conditions at the ICAR–Indian Institute of Vegetable Research (ICAR-IIVR), Varanasi, India (25.18° N latitude, 82.83° E longitude). Two independent polycarbonate-sheet-covered polyhouse facilities were used to establish contrasting temperature regimes. The heat stress environment was maintained in a temperature-controlled polyhouse equipped with electric heaters, where the day time and night time temperatures were regulated at 40 ± 2 °C and 25 ± 2 °C, respectively. The optimal (control) environment was maintained in a separate air-conditioned polyhouse with temperatures ranged from 26 ± 2 °C during the day and 20 ± 2 °C at night. The relative humidity during the full crop cycle (Feb-May) under the heat stress treatment ranged from 25–30% during the day and 65–80% at night, whereas under the optimal environment it ranged from 25–40% during the day and 55–75% at night. Both environments were maintained under a 16 h light/8 h dark photoperiod with a photosynthetic photon flux density (PPFD) of 400–600 μmol m⁻ 2 s⁻ 1 . Other environmental parameters like soil parameters were not actively controlled.

Seedlings were raised under standard nursery conditions and transplanted into 12-inch diameter earthen pots filled with a potting mixture of soil and well-decomposed farmyard manure (FYM) in a 3:1 (v/v) ratio. Plants were allowed to establish under optimal conditions for 20 days after transplanting (DAT), after which one set of plants was transferred to the heat stress polyhouse, while the other set remained under optimal conditions. Both treatments were maintained until crop maturity and final harvest (120 DAT). The experiment was laid out in a completely randomized design (CRD) with three biological replications. Each replication consisted of five plants per genotype, resulting in 15 plants per genotype under each temperature regime. All recommended agronomic practices, including irrigation, nutrient management, and plant protection measures, were uniformly followed throughout the experimental period.

Phenotypic and biochemical characterization

Morphological data were collected from all five plants per genotype per replication as described earlier 17 for traits such as i) fruit weight (g), ii) yield (g/plant), iii) fruit length (cm), iv) fruit width (cm), v) pericarp thickness (cm), vi) total soluble solids (TSS) (°Brix), vii) Pollen viability (%), and viii) Fruit set (%). Fruit weight was calculated from an average of ten fruit weight. Pericarp thickness was measured with a Vernier caliper. TSS was determined using a hand refractometer. Pollen viability (%) was estimated using the acetocarmine staining, whereas fruit set (%) was determined by direct counting and calculated as Fruit set (%) = (Number of fruits set/Total number of flowers) × 100.

Leaf samples were harvested at flowering/fruiting stages for biochemical analyses under heat stress. Biochemical assays were performed following the original methods described by Thimmaiah 18 . Proline content (μg/g FW) was measured following the original protocols of Bates et al. 19 . Lipid peroxidation (MDA, malondialdehyde content) (μmol/g FW) was quantified using the Heath and Packer 20 method. Hydrogen peroxide (H 2 O 2 ) (μmol/g FW) levels were determined by the approach of Velikova et al. 21 . Antioxidant enzyme activities were assessed: superoxide dismutase (SOD) (units/mg protein) 22 , catalase (CAT) (μmol H 2 O 2 oxidized/min/mg protein) 23 , and ascorbate peroxidase (APX) 24 . Electrolyte leakage (EL) (%) was measured following Blum & Ebercon 25 to evaluate membrane stability. Pigment analyses included total chlorophyll (mg/g FW) 26 and carotenoids (mg/g FW) 27 . Relative water content (RWC) (%) was determined according to Barrs and Weatherley 28 .

Heat tolerance index

A modified Heat Tolerance Index (HTI) was developed to quantify the relative performance of tomato genotypes under heat stress compared with normal conditions. The index integrates pollen viability, fruit set, and yield retention, which are recognized as key indicators of thermotolerance in tomato 29 – 31 . Genotypes were classified based on the HTI value as i) Highly heat-tolerant (HHT: ≥ 70 HTI with ≥ 70% pollen viability and ≥ 40% fruit set), ii) Moderately heat-tolerant (MHT: 50–69 HTI with 50–69% pollen viability and 25–39% fruit set, and iii) Heat sensitive (HS: < 50 HTI with < 50% pollen viability and < 25% fruit set). In our study, HTI was calculated as: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$HTI = \frac{\left[\frac{PVs}{PVn}+\frac{FSs}{FSn}+ \frac{Ys}{Yn}\right]}{3}\times 100$$\end{document}

Symbols. HTI : Heat tolerant index PVs = Pollen viability under heat stress (%) PVn = Pollen viability under normal condition (%) FSs = Fruit set under heat stress (%) FSn = Fruit set under normal condition (%) Ys = Yield under heat stress (g) Yn = Yield under normal condition (g)

Gene expression analysis by quantitative real-time PCR (qRT-PCR)

The expression profiles of five genes including three heat-responsive genes- HSP70 (Heat shock protein 70), HSP90 (Heat shock protein 90) and HSFA2 (Heat shock transcription factor A2), and two antioxidant-related genes- CAT (Catalase), and SOD1 (Superoxide dismutase 1)- were analysed by qRT-PCR. The leaf tissues of heat-tolerant tomato genotypes along with the heat-susceptible control Punjab Chhuhara grown under heat stress conditions were used. Total RNA was extracted from fresh leaf tissues using the RNeasy Plant Mini Kit (QIAGEN, Hilden, Germany) following the manufacturer’s protocol. The quality and quantity of the isolated RNA were assessed prior to complementary DNA (cDNA) synthesis. First-strand cDNA was synthesized from purified RNA using the TaqMan Reverse Transcription Reagents Kit (Applied Biosystems, Foster City, CA, USA). Quantitative real-time PCR (qRT-PCR) was performed using SYBR Green chemistry on a qTOWER 3 Real-Time PCR System (Analytik Jena GmbH + Co. KG, Jena, Germany) with gene-specific primers targeting heat stress-responsive and antioxidant genes (Supplementary Table S2) 32 . Each reaction was performed using three biological replicates, with three technical replicates per biological sample. The thermal cycling conditions consisted of 50 °C/2 min, 95 °C/10 min, followed by 40 cycles of 95 °C/15 s, 60 °C/1 min, and 72 °C/30 s with melting curve analysis to confirm the specificity of the PCR products. The tomato Actin-2 gene ( SlACT2; Solyc11g005330 ) was used as the endogenous reference gene for normalization of transcript abundance. Relative gene expression levels were calculated using the comparative Ct (2⁻ ΔΔCt ) method 33 . Relative gene expression values (fold change) were normalized against the internal reference gene in treatment versus control (Punjab Chhuhara) under stress conditions.

Statistical analysis

The mean data recorded over two years were used for statistical analyses. Phenotypic, physiological, and biochemical data were subjected to pooled analysis of variance (ANOVA) using the OPSTAT statistical software 34 . The homogeneity of error variances across years was tested using Bartlett’s Chi-square test, and only traits showing homogeneous variances were included in the combined analysis. The pooled data were subsequently analyzed using a completely randomized design (CRD) model. Mean comparisons among genotypes were performed using the Critical Difference (CD) test ( P ≤ 0.05) to determine significant differences among genotypes, treatments, years, and their interactions. To identify genotypes with consistent performance across years, pooled mean values were used to estimate stable and favorable responses under contrasting temperature conditions. Furthermore, data from the optimal and heat stress environments were subjected to Genotype plus Genotype × Environment (GGE) biplot analysis to evaluate genotype performance, stability, and genotype × environment interactions. The Multi-trait Genotype–Ideotype Distance Index (MGIDI) analysis was performed to identify superior genotypes using the default algorithm in the ‘metan’ package in R (version 4.6.1).

Result

Variation in horticultural traits

The evaluation of 35 tomato genotypes under high temperature stress and optimal temperature conditions revealed significant genotypic variation for all growth, yield, and quality-related traits (Tables 1 and 2 ). In general, exposure to high temperature stress resulted in a reduction in fruit weight, fruit dimensions, pericarp thickness, total soluble solids (TSS), pollen viability, fruit set, and yield per plant as compared to optimal conditions. However, the extent of reduction varied considerably among the genotypes, indicating differential levels of heat tolerance. Table 1 Evaluation of tomato genotypes for horticultural traits under high temperature stress (40 ± 2 °C) and optimal (26 ± 2 °C) controlled conditions. Code Genotype Fruit weight (g) Fruit length (cm) Fruit width (cm) Pericarp thickness (cm) TSS (°Brix) Stress Optimal Stress Optimal Stress Optimal Stress Optimal Stress Optimal G1 Arka Vikash 72.50 72.33 5.78 4.78 4.43 4.62 0.62 0.44 4.16 5.40 G2 H-88-78-1 27.83 35.87 3.62 3.61 4.40 3.61 0.36 0.38 4.69 5.91 G3 Indam-2013-1 50.83 61.33 3.65 4.65 4.68 6.58 0.46 0.51 3.35 6.25 G4 Indam-2013-4-1 59.83 67.67 3.92 4.10 4.22 4.99 0.51 0.53 3.17 5.63 G5 Indam-2013-6-1 63.50 73.50 4.85 4.80 5.63 5.29 0.52 0.55 3.37 3.80 G6 Indam-2013-6-4 31.67 47.97 3.48 3.60 4.13 4.40 0.44 0.46 3.10 3.47 G7 Kashi Aman 61.00 74.18 4.97 5.00 5.38 4.66 0.60 0.50 4.57 5.22 G8 Kashi Chayan 62.17 99.72 4.53 4.76 5.32 5.83 0.46 0.54 4.50 5.57 G9 Nandi 51.50 48.12 4.98 3.86 4.33 4.10 0.41 0.47 4.03 6.80 G10 Prestige 42.00 53.00 4.12 4.25 4.53 4.73 0.55 0.56 3.01 3.60 G11 Punjab Varkha Bahar-1 57.83 80.62 4.53 4.79 5.58 5.92 0.47 0.58 3.76 6.23 G12 Punjab Varkha Bahar-2 28.17 45.44 4.73 4.65 5.63 5.93 0.43 0.48 4.10 4.53 G13 Punjab Chuhhara 26.83 52.00 4.90 6.75 5.20 4.03 0.55 0.49 3.25 6.63 G14 Sankranti 31.33 88.86 4.58 4.82 4.08 5.03 0.44 0.52 3.76 5.27 G15 Solan Vajra 27.50 59.94 4.00 4.41 4.45 4.57 0.32 0.49 3.40 4.30 G16 Superbug 32.50 43.66 4.68 4.99 3.65 3.49 0.51 0.34 3.10 6.07 G17 Swarn Gola 51.00 63.96 4.63 4.90 4.07 4.26 0.44 0.56 4.66 5.02 G18 ToDVAR-5 76.33 60.67 5.25 5.37 5.35 5.60 0.43 0.48 3.49 4.02 G19 ToLCV-32 45.17 51.00 5.17 4.93 4.65 4.13 0.55 0.46 4.78 5.57 G20 Vaibhav 33.00 50.67 4.28 5.46 4.15 4.72 0.31 0.53 3.29 5.13 G21 VRT-06 87.17 109.67 4.37 6.23 4.90 6.05 0.55 0.66 5.57 6.27 G22 VRT-101A 21.33 42.33 2.48 3.93 3.42 4.08 0.25 0.38 4.10 6.50 G23 VRT-11–21 32.00 62.57 4.32 5.10 4.42 5.61 0.40 0.50 3.09 5.42 G24 VRT-13 36.83 82.07 4.00 4.83 4.32 4.84 0.45 0.44 3.70 6.63 G25 VRT-19 50.17 98.05 4.45 4.56 4.48 6.24 0.42 0.54 4.64 6.26 G26 VRT-1923 38.17 104.67 4.17 4.90 4.52 5.75 0.44 0.57 2.89 5.77 G27 VRT-3122 45.67 74.85 4.52 4.43 4.30 5.15 0.53 0.52 4.16 6.27 G28 VRT-34 37.83 65.33 5.38 5.40 4.45 4.65 0.56 0.52 4.66 6.02 G29 VRT-50 43.33 92.67 4.52 4.70 4.93 5.38 0.54 0.58 3.19 5.73 G30 VRT-67 25.67 55.87 3.75 4.70 4.70 4.35 0.41 0.48 5.20 5.41 G31 VRT-69 31.67 47.50 5.85 4.54 4.50 4.05 0.57 0.60 3.18 5.90 G32 VRT-458 73.67 91.07 5.80 5.36 5.82 5.00 0.64 0.58 3.20 5.67 G33 VRT-494 52.50 48.66 5.93 4.24 4.77 3.72 0.66 0.60 5.12 5.50 G34 VRT-500 55.33 85.93 5.22 5.18 4.05 5.00 0.47 0.58 5.39 6.47 G35 VRT-613 44.00 27.89 4.87 3.02 5.15 5.18 0.50 0.35 3.47 6.30 Mean 45.94 66.27 4.58 4.73 4.65 4.90 0.48 0.51 3.92 5.56 C.D 2.64 5.401 0.214 0.218 0.19 0.191 0.02 0.045 0.174 0.389 SE(m) 0.934 1.911 0.076 0.077 0.067 0.067 0.007 0.016 0.062 0.137 SE(d) 1.321 2.702 0.107 0.109 0.095 0.095 0.01 0.023 0.087 0.194 C.V 3.522 4.994 2.863 2.825 2.51 2.384 2.546 5.438 2.73 4.284 Table 2 Evaluation of tomato genotypes for heat tolerant index (HTI) variables under high temperature stress (40 ± 2 °C) and optimal (26 ± 2 °C) controlled conditions. Sample code Genotype Pollen viability (%) Fruit setting (%) Yield/plant (g) Heat tolerant index (HTI) Class Stress Optimal Stress Optimal Stress Optimal G1 Arka Vikash 55.42 86.78 34.69 48.55 320.06 495.48 66.64 MHT G2 H-88-78-1 77.93 91.41 57.17 59.41 470.11 1489.76 71.01 HHT G3 Indam-2013-1 59.72 90.09 33.39 41.95 291.94 713.31 62.27 MHT G4 Indam-2013-4-1 42.46 77.77 21.86 35.09 283.42 850.11 50.07 HS G5 Indam-2013-6-1 45.74 78.10 22.42 34.42 212.78 980.11 48.47 HS G6 Indam-2013-6-4 45.38 77.74 19.42 33.73 306.86 1116.74 47.80 HS G7 Kashi Aman 38.41 92.42 22.28 28.13 294.78 2517.39 44.15 HS G8 Kashi Chayan 40.13 91.42 20.19 23.88 296.58 3251.09 45.86 HS G9 Nandi 43.11 74.46 21.17 33.43 237.69 2087.75 44.20 HS G10 Prestige 47.97 72.76 21.53 40.41 251.86 1099.07 47.37 HS G11 Punjab Varkha Bahar-1 42.46 84.50 23.11 32.08 201.72 4255.31 42.34 HS G12 Punjab Varkha Bahar-2 48.03 85.09 22.78 33.43 195.47 1694.16 45.37 HS G13 Punjab Chuhhara 35.50 76.23 19.42 31.78 106.22 1031.19 39.32 HS G14 Sankranti 44.19 88.47 26.78 40.45 110.72 507.76 45.98 HS G15 Solan Vajra 44.83 75.81 20.69 36.48 288.31 911.08 49.17 HS G16 Superbug 71.17 90.28 43.25 37.40 583.42 2629.78 72.21 HHT G17 Swarn Gola 48.47 82.46 22.28 42.13 283.14 770.12 49.48 HS G18 ToDVAR-5 74.50 87.95 63.17 53.14 326.27 1261.43 76.48 HHT G19 ToLCV-32 47.64 92.09 23.78 35.48 315.22 1085.75 49.26 HS G20 Vaibhav 43.16 74.85 24.75 34.47 296.53 1901.12 48.35 HS G21 VRT-06 82.35 94.40 62.43 68.42 846.19 1451.75 78.93 HHT G22 VRT-101A 44.58 75.43 23.33 39.73 245.53 1051.22 47.06 HS G23 VRT-11–21 40.52 78.15 23.17 40.09 179.22 990.12 42.58 HS G24 VRT-13 42.79 80.09 22.39 40.46 347.61 1013.19 47.69 HS G25 VRT-19 62.39 88.75 36.39 45.09 121.22 1583.07 52.89 MHT G26 VRT-1923 62.97 88.27 39.72 46.92 200.50 1261.41 57.30 MHT G27 VRT-3122 40.76 82.45 20.28 36.40 394.14 1051.09 47.55 HS G28 VRT-34 74.10 88.78 46.58 43.75 551.84 2303.44 71.30 HHT G29 VRT-50 43.43 88.15 23.67 42.37 155.42 1985.25 37.65 HS G30 VRT-67 55.77 90.11 64.42 56.41 596.33 1471.26 72.21 HHT G31 VRT-69 43.13 91.09 22.78 38.29 161.72 2261.02 37.99 HS G32 VRT-458 83.08 88.41 55.10 57.46 480.64 1093.11 77.95 HHT G33 VRT-494 81.10 87.71 62.67 63.09 548.25 1503.56 76.09 HHT G34 VRT-500 79.17 87.42 60.43 59.75 400.31 1227.57 74.77 HHT G35 VRT-613 74.67 92.78 63.58 57.74 255.08 1003.41 72.01 HHT Mean 54.49 84.93 34.03 43.62 318.77 1482.83 CD ( P < 0.05) 4.38 4.16 3.25 3.46 14.88 51.94 SE(m) 1.55 1.47 1.15 1.22 5.264 18.37 SE(d) 2.19 2.08 1.62 1.73 7.445 25.98 CV (%) 6.84 7.06 5.7 7.85 7.92 8.74 Heat tolerant index (HTI): Highly heat tolerant (HHT) = ≥ 70; Moderately heat tolerant (MHT) = 50–69; and Heat sensitive (HS) = < 50.

Under stress conditions, fruit weight ranged from 21.33 g in VRT-101A to 87.17 g in VRT-06, with a population mean of 45.94 g, whereas under optimal conditions the mean fruit weight increased to 66.27 g. Similarly, fruit length and fruit width showed overall reductions under stress, although some genotypes maintained comparatively stable fruit morphology. Genotypes such as VRT-458, VRT-494, ToDVAR-5, and VRT-500 exhibited comparatively higher fruit size and pericarp thickness under heat stress, suggesting better reproductive stability and sink strength under elevated temperature conditions. Total soluble solids (TSS) also declined under heat stress, with mean values decreasing from 5.56°Brix under optimal conditions to 3.92°Brix under stress. Nevertheless, certain genotypes including VRT-06, VRT-500, VRT-494, and ToLCV-32 maintained relatively higher TSS values under stress, indicating their ability to preserve fruit quality traits under adverse thermal environments (Table 1 ).

High temperature stress severely affected reproductive traits, particularly pollen viability and fruit set percentage. Mean pollen viability declined from 84.93% under optimal conditions to 54.49% under stress, while mean fruit set decreased from 43.62% to 34.03%. Among the evaluated genotypes, VRT-06 (82.35%), VRT-458 (83.08%), VRT-494 (81.10%), VRT-500 (79.17%), H-88–78-1 (77.93%), ToDVAR-5 (74.50%), and VRT-613 (74.67%) recorded comparatively higher pollen viability under stress conditions. Likewise, higher fruit set percentages under stress were observed in VRT-67 (64.42%), VRT-613 (63.58%), VRT-494 (62.67%), ToDVAR-5 (63.17%), and VRT-06 (62.43%), demonstrating superior reproductive resilience under elevated temperature conditions. Yield per plant was drastically reduced under stress conditions, with the population mean decreasing from 1482.83 g under optimal conditions to 318.77 g under heat stress. Despite this reduction, several genotypes maintained relatively higher yield under stress, including VRT-06 (846.19 g), VRT-67 (596.33 g), Superbug (583.42 g), VRT-34 (551.84 g), and VRT-494 (548.25 g), indicating their superior adaptation to high temperature stress (Table 2 ).

Heat tolerance classification

Significant variation in heat-tolerance traits was observed among the 35 tomato genotypes evaluated under high-temperature stress and optimal conditions (Tables 1 and 2 ). Based on Heat Tolerant Index (HTI), the genotypes were classified into highly heat-tolerant (HHT), moderately heat-tolerant (MHT), and heat-sensitive (HS) categories. Ten genotypes, namely VRT-06, VRT-67, Superbug, VRT-494, H-88–78-1, VRT-458, VRT-500, ToDVAR-5, VRT-34, and VRT-613, recorded HTI values greater than 70 and were categorized as highly heat-tolerant. Four genotypes, namely Arka Vikash, Indam-2013–1, VRT-19, and VRT-1923, were identified as moderately heat-tolerant (MHT) with HTI values ranging from 50 to 69, whereas the remaining 21 genotypes were classified as heat-sensitive with HTI values below 50 (Table 3 ). Among all the genotypes, VRT-06 exhibited the highest HTI value (78.93), followed by VRT-458 (77.95), ToDVAR-5 (76.48), and VRT-494 (76.09), indicating their superior performance and stability under high temperature stress conditions. Table 3 Identification of representative heat-tolerant tomato lines under high temperature stress. Category No. of genotypes Genotypes Heat tolerance index (HTI) Highly heat-tolerant (HHT) 10 VRT-06, VRT-67, Superbug, VRT-494, H-88-78-1, VRT-458, VRT-500, ToDVAR-5, VRT-34, and VRT-613 > 70 Moderately heat-tolerant (MHT) 4 Arka Vikash, Indam-2013-1, VRT-19, and VRT-1923 50–69 Heat-sensitive (HS) 21 Indam-2013-4-1, Indam-2013-6-1, Indam-2013-6-4, Kashi Aman, Kashi Chayan, Nandi, Prestige, Punjab Varkha Bahar-1, Punjab Varkha Bahar-2, Punjab Chuhhara, Sankranti, Solan Vajra, Swarn Gola, ToLCV-32, Vaibhav, VRT-101A, VRT-11–21, VRT-13, VRT-3122, VRT-50, and VRT-69 < 50

Biochemical traits analysis

Significant variation was observed among the highly heat-tolerant (HTT) tomato genotypes for all biochemical traits under high-temperature stress condition (Table 4 ). These parameters reflected distinct physiological adjustments that differentiated highly heat-tolerant genotypes from sensitive ones. Proline concentration varied widely from 17.12 µg/g FW in Punjab Chhuhara to 29.28 µg/g FW in VRT-613, with an overall mean of 22.6 µg/g FW. The highly heat-tolerant (HHT) genotypes— VRT-06, VRT-67, Superbug, VRT-494, H-88–78-1, VRT-458, VRT-500, ToDVAR-5, VRT-34, and VRT-613 —accumulated markedly higher proline levels, whereas the heat-sensitive genotype Punjab Chhuhara recorded the lowest. Enhanced proline accumulation in tolerant lines contributes to osmotic balance and ROS detoxification, aiding membrane stability. Malondialdehyde (MDA), a marker of lipid peroxidation, ranged from 18.70 nmol g⁻ 1 FW (VRT-06) to 33.38 nmol g⁻ 1 (Punjab Chhuhara). HHT genotypes maintained 30–45% lower MDA levels, confirming less oxidative membrane damage under stress. VRT-06 and VRT-458 exhibited the lowest MDA, supporting their superior membrane thermostability. Antioxidant enzyme activity varied significantly among genotypes. SOD activity ranged from 15.9 U mg⁻ 1 protein (Punjab Chhuhara) to 40.56 U mg⁻ 1 (VRT-613), with tolerant genotypes showing 40–60% higher activity than sensitive ones. CAT activity varied from 32.32 U mg⁻ 1 (Punjab Chhuhara) to 91.74 U mg⁻ 1 (VRT-494). APX ranged from 15.98 U mg⁻ 1 (Punjab Chhuhara) to 31.71 U mg⁻ 1 (VRT-06) (Table 4 ). Table 4 Biochemical profiling of highly heat-tolerant (HHT) tomato lines exposed to high temperature stress (40 ± 2 °C) under controlled temperature gradient tunnel (TGT) conditions. Genotype Proline (µg/g FW) MDA (µmol/g FW) H 2 O 2 (µmol/ g FW) SOD (units/ Mg protein) CAT (μmol H 2 O 2 oxidized/min/mg protein) APX (μmol ascorbate oxidized/min/mg protein) Carotenoids (mg/g FW) Total chlorophyll (mg/g FW) Relative water content (%) Electrolyte leakage (%) VRT-06 23.00 ± 1.41 bc 18.70 ± 0.99f. 41.01 ± 1.30 d 27.07 ± 1.30 c 64.69 ± 0.43f. 31.71 ± 0.41 a 4.94 ± 0.08 b 12.28 ± 0.39 ab 69.74 ± 0.36 bcd 20.87 ± 1.59 g VRT-67 20.98 ± 0.02 cde 21.10 ± 1.56 cde 34.38 ± 0.42 e 24.70 ± 0.42 cd 77.06 ± 2.91 d 26.68 ± 0.45 bc 6.08 ± 0.11 a 12.73 ± 0.37 ab 61.34 ± 1.90f. 31.37 ± 0.89 b Superbug 21.84 ± 0.21 cde 22.62 ± 1.95 bcd 33.57 ± 0.21 e 29.84 ± 0.21 b 80.56 ± 0.79 cd 25.42 ± 0.82 c 4.94 ± 0.08 b 11.83 ± 0.24 abc 65.37 ± 2.29 ef 23.39 ± 0.86 efg VRT-494 24.93 ± 0.09 b 24.96 ± 1.35 b 56.49 ± 1.91 ab 30.64 ± 1.91 b 91.74 ± 0.36 a 26.98 ± 1.44 bc 5.85 ± 0.21 a 12.48 ± 0.72 ab 69.27 ± 1.79 cde 27.47 ± 0.75 bcd H-88–78-1 21.40 ± 1.98 cde 22.28 ± 1.01 cde 38.79 ± 0.82 d 22.58 ± 0.82 de 83.77 ± 0.32 bc 24.87 ± 0.17 cd 3.81 ± 0.26 c 11.35 ± 0.91 abc 67.01 ± 2.85 de 22.35 ± 2.74 efg VRT-458 20.19 ± 1.13 de 19.95 ± 0.06 ef 40.43 ± 0.87 d 30.61 ± 0.87 b 71.80 ± 0.28 e 22.88 ± 0.16 d 5.78 ± 0.02 a 13.37 ± 0.89 a 71.44 ± 2.03 abc 20.93 ± 1.51 fg VRT-500 21.99 ± 1.42 cde 21.36 ± 0.90 cde 52.58 ± 1.20 b 25.85 ± 1.20 c 85.47 ± 2.08 b 24.86 ± 1.22 cd 4.85 ± 0.21 b 12.30 ± 0.99 ab 68.59 ± 0.83 cde 25.59 ± 0.57 cde ToDVAR-5 22.32 ± 0.95 cde 21.64 ± 0.50 cde 41.25 ± 1.63 d 21.84 ± 1.63 e 77.64 ± 0.50 d 28.55 ± 2.04 b 5.06 ± 0.08 b 13.12 ± 1.23 a 73.82 ± 2.57 a 29.21 ± 1.11 bc VRT-34 19.60 ± 0.55 ef 20.84 ± 0.22 def 47.31 ± 0.21 c 30.84 ± 0.21 b 84.88 ± 2.65 b 30.89 ± 1.56 a 3.85 ± 0.21 c 10.46 ± 0.65 cd 68.52 ± 0.74 cde 24.90 ± 1.28 def VRT-613 29.28 ± 1.01 a 23.34 ± 0.93 bc 58.57 ± 2.02 a 40.56 ± 2.02 a 64.74 ± 2.06f. 31.48 ± 0.72 a 5.15 ± 0.21 b 12.11 ± 0.15 abc 73.43 ± 2.02 ab 24.45 ± 2.05 defg Punjab Chhuhara (Control) 17.12 ± 1.58f. 33.38 ± 0.87 a 20.25 ± 0.13f. 15.90 ± 0.13f. 32.32 ± 1.78 g 15.98 ± 0.20 e 2.89 ± 0.14 d 8.77 ± 1.09 d 43.34 ± 0.93 g 50.96 ± 2.18 a Mean 22.06 22.74 42.24 27.31 74.06 26.39 4.83 11.89 66.53 27.41 CD ( P < 0.05) 2.48 2.38 4.12 2.60 4.10 2.28 0.37 1.71 4.05 4.01 SE(m) 0.79 0.76 1.32 0.83 1.31 0.73 0.12 0.55 1.30 1.29 SE(d) 1.13 1.08 1.87 1.18 1.86 1.03 0.16 0.78 1.84 1.82 CV (%) 5.12 4.74 4.42 4.32 2.51 3.92 3.46 6.55 2.94 6.66 Relative water content (%), Electrolyte leakage (%), Total chlorophyll (mg/g FW), carotenoids (mg/g FW), MDA (malondialdehyde)/LPO (Lipid peroxidation) (μmol/g FW), proline (μg/g FW), H 2 O 2 (hydrogen peroxide) (μmol/g FW), CAT (catalase) (μmol H 2 O 2 oxidized/min/mg protein), SOD (superoxide dismutase) (units/mg protein), APX (ascorbate peroxidase) (μmol ascorbate oxidized/min/mg protein); Std. (Std. deviation) value is shown after (±).

These results indicate that the HHT group exhibited a more efficient antioxidant defense system, consistent with their high pollen viability and yield stability. Total chlorophyll content ranged between 8.77 mg g⁻ 1 (VRT-34) and 13.37 mg g⁻ 1 (VRT-458). The tolerant genotypes retained higher chlorophyll levels, suggesting better photosynthetic activity and delayed senescence compared to Punjab Chhuhara, which showed marked pigment loss under stress. Relative water content (RWC) ranged from 43.34% (Punjab Chhuhara) to 73.82% (ToDVAR-5). Electrolyte leakage (EL) varied inversely, from 20.87% (VRT-06) to 50.96% (Punjab Chhuhara). The HHT genotypes maintained significantly higher RWC (65–74%) and lower EL (< 26%), indicating superior cellular water retention and membrane stability during stress (Table 4 ).

Overall, the biochemical analysis corroborates the morphological and physiological findings. The highly heat-tolerant genotypes (VRT-06, VRT-67, Superbug, VRT-494, H-88-78-1, VRT-458, VRT-500, ToDVAR-5, VRT-34, and VRT-613) —maintained elevated antioxidant enzyme activity, osmolyte content, and water balance while minimizing oxidative damage and electrolyte leakage. In contrast, Punjab Chhuhara displayed severe oxidative injury and membrane destabilization. These biochemical adaptations provide strong mechanistic evidence for the superior thermotolerance and yield stability of elite tomato genotypes under extreme heat stress.

Stability analysis using GGE biplot

Joint ANOVA revealed highly significant genotypic differences ( P < 0.001) for fruit weight, yield, pollen viability, and fruit set, indicating substantial genetic variability among the evaluated tomato genotypes (Supplementary Table S3). Genotype × environment interaction was significant for yield and fruit set, suggesting differential genotype responses across environments and highlighting the need for stability analysis through GGE biplot. In contrast, fruit weight and pollen viability exhibited non-significant GEI effects indicating greater stability of these traits across environments.

GGE biplot analysis revealed significant variations in the traits analyzed like pollen viability, fruit set and yield/plant. Highly significant ( P < 0.01) variation was observed for genotype, environment, replication and genotype × environment interaction (E1: stress and E2: optimal conditions). In the GGE biplot analysis, the first two principal components (PCs) recorded variation for pollen viability (PC1: 91.05% and PC2: 8.95%), fruit set (PC1: 97.03% and PC2: 2.97%) and yield/plant (PC1: 96.45% and PC2: 3.55%) (Fig. 1 ). This reflects two PCs accounted for a wide variation in the study. Fig. 1 Basic GGE biplot views of 35 tomato genotypes (G1–G35) evaluated across two environments i.e. E1 (heat stress) and E2 (optimal conditions) for ( a ) pollen viability (%), ( b ) fruit set (%), and ( c ) yield per plant (g plant⁻ 1 ) generated using the default settings of the metan package in R (Scaling = 0, Centering = 2, Singular Value Partitioning (SVP) = 2). The corresponding Mean vs. Stability views are presented for ( d ) pollen viability (%), ( e ) fruit set (%), and ( f ) yield per plant (g plant⁻ 1 ) using the default parameters of the metan package (Scaling = 0, Centering = 2, SVP = 1). Genotype details and their corresponding codes (G1–G35) are provided in Table 1 .

Genotype stability was evaluated based on the “mean vs stability” view of the GGE biplot using the average environment coordinate (AEC) method (Fig. 1 ). Genotypes located close to the AEC abscissa (horizontal axis) with the shortest vertical projection toward the AEC ordinate (vertical axis) were considered more stable. In contrast, genotypes with longer projections from the AEC abscissa were regarded as unstable. The arrow on the AEC abscissa indicates higher mean performance, whereas the AEC ordinate represents genotype stability. In this study, the stable genotypes with high mean values were G21 (VRT-06), Superbug (G16) and VRT-67 (G30) for pollen viability, fruit set and yield per plant.

The "which-won-where" view of the GGE biplot revealed superior genotypes in different environments. The vertex genotypes having longer vector, located at the polygon edge within the biplot, were the most responsive (Fig. 2 ). Out study observed, the superior or vortex genotypes for various traits such as G21 (VRT-06) for pollen viability and yield per plant and G30 (VRT-67) for fruit set. Overall, G21 (VRT-06) was found to be the superior and stable genotype under high temperature stress under controlled conditions. Further, in this study, ideal environments were identified based on the “discriminativeness vs. representativeness” view of GGE biplot (Fig. 2 ). The ideal environment is based on the longest vector, (discriminativeness), whereas representativeness reflects an acute angle (< 90º) with the AEC abscissa (horizontal axis). Overall, we confirmed that the most discriminative and representative environments such as E1 (stress condition) for high temperature stress tolerance in tomato. Fig. 2 Polygon (Which-won-where) views of 35 tomato genotypes (G1–G35) evaluated across two environments i.e. E1 (heat stress) and E2 (optimal conditions) for ( a ) pollen viability (%), ( b ) fruit set (%), and ( c ) yield per plant (g plant⁻ 1 ) generated using the default settings of the metan package in R (Scaling = 0, Centering = 2, Singular Value Partitioning (SVP) = 3). The corresponding Discriminativeness vs. Representativeness views are presented for ( d ) pollen viability (%), ( e ) fruit set (%), and ( f ) yield per plant (g plant⁻ 1 ) using the default parameters of the metan package (Scaling = 0, Centering = 2, SVP = 3). Genotype details and their corresponding codes (G1–G35) are provided in Table 1 .

The MGIDI tool was used to rank genotypes according to their distance from an ideal genotype (Fig. 3 ). Accordingly, we observed the stable genotypes as indicated by a red circle, including G21 (VRT-06), G16 (Superbug), G33 (VRT-494), G30 (VRT-67) and G2 (H-88–78-1). These selected and stable genotype can be successfully utilized in tomato breeding for high temperature tolerance. Fig. 3 Multi-trait Genotype–Ideotype distance index (MGIDI) plot based on the GGE biplot analysis of 35 tomato genotypes (G1–G35) evaluated across two environments E1 (heat stress) and E2 (optimal conditions) using three key traits: pollen viability (%), fruit set (%), and yield per plant (g plant⁻ 1 ). The analysis was performed using the default settings of the metan package in R. Genotypes enclosed within the red circle represent those selected by the MGIDI index based on their superior overall performance across the evaluated traits and environments. Genotype details and their corresponding codes (G1–G35) are provided in Table 1 .

Real-time quantitative gene expression analysis

Gene expression analysis of selected highly heat-tolerant tomato genotypes along with the heat-sensitive susceptible control (Punjab Chhuhara) revealed differential regulation of heat-responsive and antioxidant defense genes under stress condition (Fig. 4 ). The relative gene expression levels (fold change) of HSP70, HSP90, HSFA2, CAT , and SOD1 genes were induced in tolerant genotypes compared with control, indicating activation of cellular defense mechanisms against thermal stress. Among the evaluated genotypes, VRT-06 exhibited the highest upregulation of major heat shock-responsive genes such as HSP70 (5.87), HSP90 (5.01), and HSFA2 (5.17). This genotype also showed elevated expression of antioxidant genes CAT (3.85) and SOD1 (4.17), suggesting a strong coordinated heat stress response involving both molecular chaperones and reactive oxygen species (ROS) scavenging systems. Similarly, VRT-67 also demonstrated substantial induction of HSP70 (5.50), HSP90 (4.12), HSFA2 (4.50), and CAT (4.43), confirming its superior thermotolerance capacity. Fig. 4 Relative expression analysis of heat-responsive and antioxidant genes in tomato genotypes under heat stress conditions: ( a ) HSP70 ( Heat Shock Protein 70 ), ( b ) HSP9 0 ( Heat Shock Protein 90 ), ( c ) HSFA2 ( Heat Shock Transcription Factor A2 ), ( d ) CAT ( Catalase ), and ( e ) SOD1 ( Superoxide Dismutase 1 ). Relative gene expression fold change (FC) levels are presented as mean ± standard deviation (SD) of biological replicates. Error bars indicate ± SD. Statistical differences among genotypes were determined using the critical difference (CD) test at P ≤ 0.05. Bars annotated with different letters indicate significant differences among genotypes, whereas bars sharing the same letter are not significantly different.

The genotype Superbug exhibited moderate to high expression (fold change) of HSP70 (3.40), HSP90 (4.05), HSFA2 (3.41), and CAT (3.51), whereas VRT-494 showed balanced upregulation of HSP70 (3.21), HSFA2 (3.77), and SOD1 (4.09). Genotypes VRT-500 and VRT-613 recorded relatively higher SOD1 expression levels of 4.00 and 3.86 fold change, respectively, indicating enhanced antioxidant protection against oxidative damage induced by heat stress. Among antioxidant-related genes, CAT expression was highest in VRT-67 (4.43), followed by ToDVAR-5 (4.06) and VRT-06 (3.85), whereas SOD1 expression was highest in VRT-06 (4.17), followed by VRT-494 (4.09) and VRT-500 (4.00). These results suggest efficient detoxification of reactive oxygen species in highly tolerant genotypes under elevated temperature conditions.

In contrast, the heat-sensitive control Punjab Chhuhara exhibited comparatively lower gene expression levels ( HSP70: 1.42, HSP90: 1.95; HSFA2: 2.04, CAT: 2.08, and SOD1 : 1.88 fold change). The comparatively weak induction of stress-responsive and antioxidant genes in the susceptible genotype may account for its poor performance under heat stress conditions. Overall, the results demonstrated that highly heat-tolerant tomato genotypes possess stronger transcriptional activation of heat shock proteins, heat stress transcription factors, and antioxidant defense genes, which collectively contribute to enhanced thermotolerance and maintenance of reproductive and yield stability under high temperature stress.

Correlation analysis

Correlation analysis among phenotype and reproductive traits like pollen viability, fruit set, and yield per plant under high temperature stress revealed strong positive associations among these reproductive and yield-related traits (Fig. 5 ). Pollen viability exhibited a highly significant positive correlation with fruit set (r = 0.72) and yield per plant (r = 0.52) indicating that maintenance of viable pollen under high temperature conditions directly contributed to improved fruit set ability and higher fruit yields. Fruit set percentage also exhibited a strong positive correlation with yield per plant (r = 0.45), demonstrating yield stability in tomato. These findings collectively indicate that pollen viability, fruit set and yield are critical selection criteria for identifying heat-resilient tomato genotypes under high temperature stress conditions. Fig. 5 Hierarchically clustered Pearson correlation heatmap illustrating the relationships among fruit quality traits (FWT: Fruit weight, FL: Fruit length, FWD: Fruit width, PERI: Pericarp thickness, and TSS: Total soluble solids), reproductive traits (PV: Pollen viability, FS: Fruit set, and YLD: Yield (g/plant), and heat-responsive/antioxidant genes ( HSP70, HSP90, HSFA2, CAT, and SOD1 ) in tomato genotypes evaluated under high temperature stress. Correlation coefficients (r) are shown within each cell. Red shades indicate positive correlations, blue shades indicate negative correlations, and color intensity reflects the strength of association. Hierarchical clustering groups traits and genes with similar correlation patterns.

Further, correlation analysis revealed strong positive associations among reproductive traits (pollen viability, fruit set, and yield) and the expression levels of heat-responsive genes, particularly HSP70 (r = 0.81), HSP90 (r = 0.81), and HSFA2 (r = 0.78). Antioxidant gene CAT (r = 0.53) and SOD1 (r = 0.27) also exhibited positive relationships with yield and reproductive traits under heat stress. The clustering pattern grouped stress-responsive genes with pollen viability, fruit set, and yield, suggesting coordinated regulation of heat tolerance mechanisms. Conversely, traits showing weaker associations formed separate clusters, indicating differential contributions to thermotolerance. These findings highlight the importance of heat-shock proteins and antioxidant defense systems in maintaining reproductive traits and productivity under elevated temperature conditions.

Discussion

High temperature stress is among the most severe abiotic constraints limiting tomato productivity worldwide and is expected to intensify under future climate change scenarios. Temperatures exceeding the optimum range (25–30 °C) during the reproductive stage adversely affect pollen development, fertilization, fruit set, assimilate partitioning and ultimately marketable yield 31 , 32 . In the present study, substantial phenotypic variation was observed among the 35 tomato genotypes under contrasting temperature regimes, demonstrating the existence of considerable genetic variability for heat tolerance. The consistent performance of elite genotypes (VRT-06, VRT-67, Superbug, VRT-494, H-88–78-1, VRT-458, VRT-500, ToDVAR-5, VRT-34 and VRT-613) under heat stress indicates that these genotypes possess effective heat tolerance mechanism. Similar levels of genetic variation for thermotolerance have been reported in tomato germplasm 32 , 35 , 36 , highlighting the potential of exploiting natural genetic diversity for breeding climate-resilient cultivars.

Heat stress markedly reduced fruit weight and yield in susceptible genotypes, reflecting the detrimental effects of elevated temperature on photosynthesis, carbohydrate metabolism and assimilate translocation towards developing fruits 3 . High temperatures reduce Rubisco activity, accelerate photorespiration and impair source–sink relationships, resulting in poor fruit development 37 . Nevertheless, highly heat-tolerant genotypes such as VRT-06, VRT-67, Superbug, VRT-494, H-88-78-1 and VRT-458 maintained significantly higher fruit weight and yield under stress, suggesting greater photosynthetic resilience and more efficient assimilate partitioning 7 . Maintenance of sink strength during reproductive development has been recognized as a major determinant of yield stability under heat stress 38 . The superior performance of these genotypes therefore indicates enhanced physiological capacity to sustain carbon assimilation and reproductive growth under adverse environmental conditions.

Among the reproductive traits evaluated, pollen viability and fruit set emerged as the most reliable indicators of thermotolerance 29 , 30 . Tomato pollen development is particularly sensitive to elevated temperatures because meiosis, tapetal development and pollen maturation are highly susceptible to heat-induced oxidative damage 31 . The tolerant genotypes maintained significantly higher pollen viability and fruit set compared with the susceptible genotype, resulting in superior yield under heat stress. Furthermore, the strong positive association observed among pollen viability, fruit set and yield demonstrates that preservation of male reproductive function is fundamental to heat tolerance. Similar conclusions have been reported by many researchers who identified pollen fertility as one of the strongest predictors of fruit yield under elevated temperatures 29 – 31 , 39 , 40 . These findings suggest the utility of pollen viability and fruit set as rapid and reliable screening criteria in tomato breeding programmes targeting heat tolerance.

The HTI effectively classified genotypes into highly heat-tolerant, moderately heat-tolerant and heat-sensitive categories by integrating reproductive and productivity traits under stress. The HTI genotypes exhibited superior pollen viability, fruit set and yield. Unlike conventional indices that rely solely on yield performance 29 , the HTI method employed in this study integrates reproductive and yield traits, thereby providing a more biologically meaningful assessment of thermotolerance. Similar integrated stress indices have been successfully used for identifying stable tomato genotypes under high-temperature environments 41 , 42 . Hence, this study demonstrate HTI method for screening of tomato genotypes for thermotolerance.

Physiological and biochemical analyses further elucidated the mechanisms underlying heat tolerance 39 , 41 . Heat stress induces excessive accumulation of reactive oxygen species (ROS), leading to oxidative damage, membrane destabilization and impaired cellular metabolism 39 . The heat-tolerant genotypes accumulated significantly higher concentrations of proline than the susceptible control (Punjab Chhuhara). Proline serves multiple protective functions by acting as an osmolyte, molecular chaperone, ROS scavenger and stabilizer of proteins and cellular membranes 43 . Concurrently, lower malondialdehyde (MDA) content and reduced electrolyte leakage observed in tolerant genotypes indicated diminished lipid peroxidation and superior membrane thermostability 43 . Similar responses have been documented in heat-tolerant tomato genotypes by Rajametov et al. 44 and Osei-Bonsu et al. 45 , confirming that maintenance of membrane integrity is a critical parameters of thermotolerance.

Enhanced antioxidant defence also contributed significantly to heat tolerance 11 . The higher antioxidant activity observed in tolerant genotypes suggests more efficient detoxification of ROS generated under heat stress. Antioxidant enzymes such as superoxide dismutase (SOD) and catalase (CAT) constitute the primary enzymatic defence system by converting superoxide radicals into hydrogen peroxide and subsequently detoxifying hydrogen peroxide into water and oxygen 10 , 11 . Efficient ROS scavenging minimizes oxidative damage, preserves chloroplast integrity and maintains metabolic homeostasis under stress. Consequently, tolerant genotypes observed in our study sustained greater reproductive success and yield under elevated temperatures. These observations agree with previous reports describing enhanced antioxidant capacity as a hallmark of thermotolerant tomatoes 9 , 44 .

The integration of multivariate analyses substantially improved genotype discrimination 46 . GGE biplot analysis revealed that VRT-06, Superbug and VRT-67 combined high mean performance with remarkable stability across optimal and heat-stress environments. The "which-won-where" pattern further confirmed VRT-06 and VRT-67 as winning genotypes for the target environments, indicating broad adaptation and stable performance. Complementarily, MGIDI analysis simultaneously integrated multiple agronomic, reproductive and physiological traits into a single selection index, identifying VRT-06, Superbug, VRT-494, VRT-67 and H-88-78-1 as the most desirable genotypes. The combined application of GGE biplot and MGIDI provides a powerful framework for multi-trait selection because it simultaneously considers genotype performance, stability and ideotype proximity, thereby enhancing selection efficiency in tomato breeding programmes 14 . Thus, our study identifies superior and stable performing genotypes for high temperature tolerance.

At the molecular level, highly tolerant genotypes exhibited coordinated activation of heat-responsive and antioxidant genes, providing genetic evidence for the observed phenotypic differences 9 . The up-regulation of genes (e.g. HSP70 and HSP90 ) indicates activation of molecular chaperone systems that stabilize proteins, prevent aggregation of denatured proteins and facilitate refolding during thermal stress 43 . Likewise, elevated expression of HSFA2 , a master regulator of acquired thermotolerance, suggests activation of downstream heat-responsive regulatory networks that sustain prolonged stress adaptation 10 . Simultaneously, increased expression of CAT and SOD1 reflects enhanced antioxidant defence capable of mitigating ROS accumulation and protecting cellular components from oxidative injury. Similar coordinated transcriptional responses have been documented by Karkute et al. 9 and Fernández-Crespo et al. 32 . The coordinated induction of heat shock proteins and antioxidant enzymes in elite genotypes such as VRT-06, VRT-67 and VRT-494 demonstrates that thermotolerance in tomato is governed by an integrated molecular network involving protein protection, ROS detoxification and maintenance of cellular homeostasis.

Overall, the integration of morphological, reproductive, physiological, biochemical and molecular analyses provides a comprehensive understanding of heat tolerance mechanisms in tomato. The superior performance of VRT-06, VRT-67, Superbug, VRT-494 and H-88–78-1 across multiple evaluation criteria demonstrates their considerable potential as donor parents for developing heat-resilient tomato cultivars. These genotypes also represent valuable genetic resources for dissecting the molecular basis of thermotolerance and for deployment in marker-assisted and genomics-assisted breeding programmes 47 aimed at sustaining tomato productivity under future climate change scenarios 48 .

Conclusion

The present study demonstrated considerable genetic variability among tomato genotypes for morphological, reproductive, biochemical, and molecular traits under high temperature stress conditions. Heat stress significantly reduced fruit yield, pollen viability, fruit set, and overall physiological performance. Based on HTI Multi-trait genotype-ideotype distance index, GGE biplot, MGIDI analysis, and biochemical and molecular characterization, the genotypes VRT-06, VRT-67, Superbug, VRT-494, H-88-78-1, VRT-458, VRT-500, ToDVAR-5, VRT-34, and VRT-613 were identified as highly heat-tolerant. The tolerant genotypes maintained higher pollen viability, fruit set, yield stability, chlorophyll content, relative water content, and antioxidant enzyme activities, along with lower lipid peroxidation and electrolyte leakage under stress conditions. Furthermore, enhanced expression of heat-responsive genes ( HSP70, HSP90 , and HSFA2 ) and antioxidant defense genes ( CAT and SOD ) indicated the presence of coordinated molecular defense mechanisms contributing to thermotolerance. The strong positive association among pollen viability, fruit set, and yield further highlighted the importance of reproductive traits as reliable selection criteria for heat tolerance breeding. The integration of morphological, biochemical, molecular, GGE biplot, and MGIDI analyses provided a robust and comprehensive framework for identifying stable heat-tolerant tomato genotypes. Among the evaluated lines, VRT-06 emerged as the most promising genotype due to its superior yield performance, reproductive stability, biochemical resilience, and enhanced expression of stress-responsive genes under high temperature stress. The elite genotypes identified in this study can serve as valuable genetic resources for developing heat-tolerant tomato cultivars and may be effectively utilized in the future breeding programs.

Supplementary Information

Supplementary Information.

Publisher’s note

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

Acknowledgements

Research was supported by the Indian Council of Agricultural Research, Department of Agricultural Research and Education, Government of India. We thank the Competent Authority for support under the ICAR-NICRA and the ICAR-CRP-HT (Tomato) projects at ICAR-IIVR, Varanasi. The ChatGPT 4o was used for language editing.

Author contributions

JKT designed the experiment, performed research work, analysed data and prepared the manuscript. NR provided resources. MKS, RS, TK and YS measured field data and performed biochemical analysis. AKS performed qPCR analysis. YSR performed data analysis. RK edited the manuscript. All authors confirmed the manuscript for further publication.

Funding

This work was supported under the ICAR-NICRA, ICAR-CRP Hybrid Technology (Tomato) & Solanaceous projects at the institute.

Data availability

All data supporting the findings of this study are available within the paper.

Declarations

Competing interests

The authors declare no competing interests.

References

1. Lokesha AN Response of contrasting tomato genotypes under high temperature stress Mysore J. Agric. Sci. 2019 53 9 14 Lokesha, A. N. et al. Response of contrasting tomato genotypes under high temperature stress. Mysore J. Agric. Sci. 53 , 9–14 (2019).

2. Rai, N. et al. Climate resilient tomato: breeding and production technologies. Technical Bulletin No. 119. ICAR-Indian Institute of Vegetable Research, Varanasi. p116. (2025).

3. Alsamir M An overview of heat stress in tomato (Solanum lycopersicum L.) Saudi J. Biol. Sci. 2021 28 1654 1663 10.1016/j.sjbs.2020.11.088 33732051 PMC7938145 Alsamir, M. et al. An overview of heat stress in tomato (Solanum lycopersicum L.). Saudi J. Biol. Sci. 28 , 1654–1663. 10.1016/j.sjbs.2020.11.088 (2021). 33732051 PMC7938145

4. Ayenan MAT Genomic and phenotypic diversity of cultivated and wild tomatoes with varying levels of heat tolerance Genes 2021 12 503 10.3390/genes12040503 33805499 PMC8067180 Ayenan, M. A. T. et al. Genomic and phenotypic diversity of cultivated and wild tomatoes with varying levels of heat tolerance. Genes 12 , 503. 10.3390/genes12040503 (2021). 33805499 PMC8067180

5. Luo J Effect of nitrogen application on enhancing high-temperature stress tolerance of tomato plants during the flowering and fruiting stage Front. Plant Sci. 2023 14 1172078 10.3389/fpls.2023.1172078 37360700 PMC10285307 Luo, J. et al. Effect of nitrogen application on enhancing high-temperature stress tolerance of tomato plants during the flowering and fruiting stage. Front. Plant Sci. 14 , 1172078. 10.3389/fpls.2023.1172078 (2023). 37360700 PMC10285307

6. Raza A Role of plant peroxisomal catalase in temperature and drought stress: Physio-biochemical and molecular perspectives Plant Physiol. Biochem. 2025 229 Pt D 110730 10.1016/j.plaphy.2025.110730 41240811 Raza, A. et al. Role of plant peroxisomal catalase in temperature and drought stress: Physio-biochemical and molecular perspectives. Plant Physiol. Biochem. 229 (Pt D), 110730. 10.1016/j.plaphy.2025.110730 (2025). 41240811

7. Raza A Trehalose: A sugar molecule involved in temperature stress management in plants Crop J. 2024 12 1 16 10.1016/j.cj.2023.09.010 Raza, A. et al. Trehalose: A sugar molecule involved in temperature stress management in plants. Crop J. 12 , 1–16. 10.1016/j.cj.2023.09.010 (2024).

8. Tokic M Growth and molecular responses of tomato to prolonged and short-term heat exposure Int. J. Mol. Sci. 2023 24 4456 10.3390/ijms24054456 36901887 PMC10002527 Tokic, M. et al. Growth and molecular responses of tomato to prolonged and short-term heat exposure. Int. J. Mol. Sci. 24 , 4456. 10.3390/ijms24054456 (2023). 36901887 PMC10002527

9. Karkute SG Characterization of high-temperature stress-tolerant tomato (Solanum lycopersicum L.) genotypes by biochemical analysis and expression profiling of heat-responsive genes 3 Biotech 2021 11 45 10.1007/s13205-020-02587-6 33489667 PMC7801581 Karkute, S. G. et al. Characterization of high-temperature stress-tolerant tomato (Solanum lycopersicum L.) genotypes by biochemical analysis and expression profiling of heat-responsive genes. 3 Biotech 11 , 45. 10.1007/s13205-020-02587-6 (2021). 33489667 PMC7801581

10. Staacke T Mueller-Roeber B Balazadeh S Stress resilience in plants: The complex interplay between heat stress memory and resetting New Phytol. 2025 245 2402 2421 10.1111/nph.20377 39853503 PMC11840417 Staacke, T., Mueller-Roeber, B. & Balazadeh, S. Stress resilience in plants: The complex interplay between heat stress memory and resetting. New Phytol. 245 , 2402–2421. 10.1111/nph.20377 (2025). 39853503 PMC11840417

11. Liu J Jin W Molecular underpinnings of plant stress granule dynamics in response to high-temperature stress New Crops 2025 2 100069 10.1016/j.ncrops.2025.100069 Liu, J. & Jin, W. Molecular underpinnings of plant stress granule dynamics in response to high-temperature stress. New Crops 2 , 100069 (2025).

12. Delarue M Benhamed M Fragkostefanakis S The role of epigenetics in tomato stress adaptation New Crops 2025 2 100044 10.1016/j.ncrops.2024.100044 Delarue, M., Benhamed, M. & Fragkostefanakis, S. The role of epigenetics in tomato stress adaptation. New Crops 2 , 100044. 10.1016/j.ncrops.2024.100044 (2025).

13. Yan W GGEbiplot – a windows application for graphical analysis of multi-environment trial data and other types of two way data Agron. J. 2001 93 1111 1118 10.2134/agronj2001.9351111x Yan, W. GGEbiplot – a windows application for graphical analysis of multi-environment trial data and other types of two way data. Agron. J. 93 , 1111–1118. 10.2134/agronj2001.9351111x (2001).

14. Tiwari JK Delineating genotype × environment interaction for horticultural traits in tomato using GGE and AMMI biplot analysis Sci. Rep. 2025 15 23796 10.1038/s41598-025-09021-y 40610582 PMC12229600 Tiwari, J. K. et al. Delineating genotype × environment interaction for horticultural traits in tomato using GGE and AMMI biplot analysis. Sci. Rep. 15 , 23796. 10.1038/s41598-025-09021-y (2025). 40610582 PMC12229600

15. Singh U Screening of tomato genotypes under high temperature stress for reproductive traits Veg. Sci. 2015 42 52 55 Singh, U. et al. Screening of tomato genotypes under high temperature stress for reproductive traits. Veg. Sci. 42 , 52–55 (2015).

16. Vijayakumar A High temperature induced changes in quality and yield parameters of tomato (Solanum lycopersicum L.) and similarity coefficients among genotypes using SSR markers Heliyon 2021 7 e05988 10.1016/j.heliyon.2021.e05988 33644434 PMC7889828 Vijayakumar, A. et al. High temperature induced changes in quality and yield parameters of tomato (Solanum lycopersicum L.) and similarity coefficients among genotypes using SSR markers. Heliyon 7 , e05988. 10.1016/j.heliyon.2021.e05988 (2021). 33644434 PMC7889828

17. Tiwari JK Identification of tomato lines based on conventional and molecular tools for breeding Curr. Sci. 2024 1126 569 573 10.18520/cs/v126/i5/569-573 Tiwari, J. K. et al. Identification of tomato lines based on conventional and molecular tools for breeding. Curr. Sci. 1126 , 569–573 (2024).

18. Thimmaiah SK Standard Methods of Biochemical Analysis 2021 Kalyani Publishers 517 Thimmaiah, S. K. Standard Methods of Biochemical Analysis 517 (Kalyani Publishers, 2021).

19. Bates LS Rapid determination of free proline for water-stress studies Plant Soil 1973 39 205 207 10.1007/BF00018060 Bates, L. S. et al. Rapid determination of free proline for water-stress studies. Plant Soil 39 , 205–207. 10.1007/BF00018060 (1973).

20. Heath RL Packer L Photoperoxidation in isolated chloroplasts. I. Kinetics and stoichiometry of fatty acid peroxidation Arch. Biochem. Biophys. 1968 125 189 198 10.1016/0003-9861(68)90654-1 5655425 Heath, R. L. & Packer, L. Photoperoxidation in isolated chloroplasts. I. Kinetics and stoichiometry of fatty acid peroxidation. Arch. Biochem. Biophys. 125 , 189–198 (1968). 5655425 10.1016/0003-9861(68)90654-1

21. Velikova V Yordanov I Edreva A Oxidative stress and some antioxidant systems in acid rain-treated bean plants: protective role of exogenous polyamines Plant Sci. 2000 151 59 66 10.1016/S0168-9452(99)00197-1 Velikova, V., Yordanov, I. & Edreva, A. Oxidative stress and some antioxidant systems in acid rain-treated bean plants: protective role of exogenous polyamines. Plant Sci. 151 , 59–66 (2000).

22. Giannopolitis CN Ries SK Superoxide dismutases I Occurrence in higher plants. Plant Physiol. 1977 59 309 314 10.1104/pp.59.2.309 16659839 PMC542387 Giannopolitis, C. N. & Ries, S. K. Superoxide dismutases I. Occurrence in higher plants. Plant Physiol. 59 , 309–314. 10.1104/pp.59.2.309 (1977). 16659839 PMC542387

23. Aebi H Catalase in vitro Methods Enzymol. 1984 105 121 126 10.1016/S0076-6879(84)05016-3 6727660 Aebi, H. Catalase in vitro. Methods Enzymol. 105 , 121–126. 10.1016/S0076-6879(84)05016-3 (1984). 6727660

24. Nakano Y Asada K Hydrogen peroxide is scavenged by ascorbate-specific peroxidase in spinach chloroplasts Plant Cell Physiol. 1981 22 867 880 10.1093/oxfordjournals.pcp.a076232 Nakano, Y. & Asada, K. Hydrogen peroxide is scavenged by ascorbate-specific peroxidase in spinach chloroplasts. Plant Cell Physiol. 22 , 867–880 (1981).

25. Blum A Ebercon A Cell membrane stability as a measure of drought and heat tolerance in wheat Crop Sci. 1981 21 43 47 10.2135/cropsci1981.0011183X002100010013x Blum, A. & Ebercon, A. Cell membrane stability as a measure of drought and heat tolerance in wheat. Crop Sci. 21 , 43–47 (1981).

26. Aron D Copper enzymes isolated chloroplasts, polyphenoloxidase in Beta vulgaris Plant Physiol. 1949 24 1 15 10.1104/pp.24.1.1 16654194 PMC437905 Aron, D. Copper enzymes isolated chloroplasts, polyphenoloxidase in Beta vulgaris . Plant Physiol. 24 , 1–15 (1949). 16654194 10.1104/pp.24.1.1 PMC437905

27. Lichtenthaler H Wellburn A Determinations of total carotenoids and chlorophylls a and b of leaf extracts in different solvents Biochem. Soc. Trans. 1983 603 591 592 10.1042/bst0110591 Lichtenthaler, H. & Wellburn, A. Determinations of total carotenoids and chlorophylls a and b of leaf extracts in different solvents. Biochem. Soc. Trans. 603 , 591–592. 10.1042/bst0110591 (1983).

28. Barrs HD Weatherley PE A re-examination of the relative turgidity techniques for estimating water deficits in leaves Australian J. Biol. Sci. 1962 15 413 428 10.1071/BI9620413 Barrs, H. D. & Weatherley, P. E. A re-examination of the relative turgidity techniques for estimating water deficits in leaves. Australian J. Biol. Sci. 15 , 413–428. 10.1071/BI9620413 (1962).

29. Abdul-Baki AA Stommel JR Pollen viability and fruit set of tomato genotypes under optimum- and high-temperature regimes HortScience 1995 30 115 117 10.21273/HORTSCI.30.1.115 Abdul-Baki, A. A. & Stommel, J. R. Pollen viability and fruit set of tomato genotypes under optimum- and high-temperature regimes. HortScience 30 , 115–117. 10.21273/HORTSCI.30.1.115 (1995).

30. Firon N Pollen grains of heat-tolerant tomato cultivars retain higher carbohydrate concentration under heat stress conditions Sci. Hortic. 2006 109 212 217 10.1016/j.scienta.2006.03.007 Firon, N. et al. Pollen grains of heat-tolerant tomato cultivars retain higher carbohydrate concentration under heat stress conditions. Sci. Hortic. 109 , 212–217. 10.1016/j.scienta.2006.03.007 (2006).

31. Paupière MJ Screening for pollen tolerance to high temperatures in tomato Euphytica 2017 213 130 10.1007/s10681-017-1927-z Paupière, M. J. et al. Screening for pollen tolerance to high temperatures in tomato. Euphytica 213 , 130. 10.1007/s10681-017-1927-z (2017).

32. Fernández-Crespo E Exploiting tomato genotypes to understand heat stress tolerance Plants 2022 11 3170 10.3390/plants11223170 36432899 PMC9696584 Fernández-Crespo, E. et al. Exploiting tomato genotypes to understand heat stress tolerance. Plants 11 , 3170. 10.3390/plants11223170 (2022). 36432899 PMC9696584

33. Livak KJ Schmittgen TD Analysis of relative gene expression data using real-time quantitative PCR and the 2⁻ ΔΔCt method Methods 2001 25 402 408 10.1006/meth.2001.1262 11846609 Livak, K. J. & Schmittgen, T. D. Analysis of relative gene expression data using real-time quantitative PCR and the 2⁻ ΔΔCt method. Methods 25 , 402–408. 10.1006/meth.2001.1262 (2001). 11846609

34. Sheoran, O. P. et al. Statistical Software package for agricultural research workers (OPSTAT). CCS Haryana Agricultural University, Hisar, India. (accessed on 10 April 2026); http://opstat.somee.com/opstat/ (1998).

35. Aldubai AA Response of tomato (Solanum lycopersicum L.) genotypes to heat stress using morphological and expression study Plants 2022 11 615 10.3390/plants11050615 35270087 PMC8912326 Aldubai, A. A. et al. Response of tomato (Solanum lycopersicum L.) genotypes to heat stress using morphological and expression study. Plants 11 , 615. 10.3390/plants11050615 (2022). 35270087 PMC8912326

36. Gonzalo MJ Identification of tomato accessions as source of new genes for improving heat tolerance: from controlled experiments to field BMC Plant Biol. 2021 21 345 10.1186/s12870-021-03104-4 34294034 PMC8296629 Gonzalo, M. J. et al. Identification of tomato accessions as source of new genes for improving heat tolerance: from controlled experiments to field. BMC Plant Biol. 21 , 345. 10.1186/s12870-021-03104-4 (2021). 34294034 PMC8296629

37. Bhattarai S Exploring morpho-physiological variation for heat stress tolerance in tomato Plants 2021 10 347 10.3390/plants10020347 33673031 PMC7918821 Bhattarai, S. et al. Exploring morpho-physiological variation for heat stress tolerance in tomato. Plants 10 , 347. 10.3390/plants10020347 (2021). 33673031 PMC7918821

38. Graci S Barone A Tomato plant response to heat stress: a focus on candidate genes for yield-related traits Front. Plant Sci. 2024 14 1245661 10.3389/fpls.2023.1245661 38259925 PMC10800405 Graci, S. & Barone, A. Tomato plant response to heat stress: a focus on candidate genes for yield-related traits. Front. Plant Sci. 14 , 1245661. 10.3389/fpls.2023.1245661 (2024). 38259925 PMC10800405

39. Hoshikawa K Genetic and molecular mechanisms conferring heat stress tolerance in tomato plants Front. Plant Sci. 2021 12 786688 10.3389/fpls.2021.786688 35003175 PMC8739973 Hoshikawa, K. et al. Genetic and molecular mechanisms conferring heat stress tolerance in tomato plants. Front. Plant Sci. 12 , 786688. 10.3389/fpls.2021.786688 (2021). 35003175 PMC8739973

40. Miller G Contrasting processing tomato cultivars unlink yield and pollen viability under heat stress AoB PLANTS 2021 13 plab046 10.1093/aobpla/plab046 34394907 PMC8356174 Miller, G. et al. Contrasting processing tomato cultivars unlink yield and pollen viability under heat stress. AoB PLANTS 13 , plab046. 10.1093/aobpla/plab046 (2021). 34394907 PMC8356174

41. Ro S Response of tomato genotypes under different high temperatures in field and greenhouse conditions Plants 2021 10 449 10.3390/plants10030449 33673537 PMC7997173 Ro, S. et al. Response of tomato genotypes under different high temperatures in field and greenhouse conditions. Plants 10 , 449. 10.3390/plants10030449 (2021). 33673537 PMC7997173

42. Scarano A Selection of tomato landraces with high fruit yield and nutritional quality under elevated temperatures J. Sci. Food Agric. 2020 100 2791 2799 10.1002/jsfa.10312 32022274 PMC7187367 Scarano, A. et al. Selection of tomato landraces with high fruit yield and nutritional quality under elevated temperatures. J. Sci. Food Agric. 100 , 2791–2799. 10.1002/jsfa.10312 (2020). 32022274 PMC7187367

43. Mohammed SP Integrative trait analysis for enhancing heat stress resilience in tomato (Solanum lycopersicum L.): a focus on root, physiological, and yield adaptations Plants 2025 14 533 10.3390/plants14040533 40006792 PMC11858947 Mohammed, S. P. et al. Integrative trait analysis for enhancing heat stress resilience in tomato (Solanum lycopersicum L.): a focus on root, physiological, and yield adaptations. Plants 14 , 533. 10.3390/plants14040533 (2025). 40006792 PMC11858947

44. Rajametov SN Heat treatment in two tomato cultivars: A study of the effect on physiological and growth recovery Horticulturae 2021 7 119 10.3390/horticulturae7050119 Rajametov, S. N. et al. Heat treatment in two tomato cultivars: A study of the effect on physiological and growth recovery. Horticulturae 7 , 119. 10.3390/horticulturae7050119 (2021).

45. Osei-Bonsu MK Assessing the heat stress tolerance potential of tomato lines under poly-house and open field conditions Cogent Food Agric. 2022 8 2115665 10.1080/23311932.2022.2115665 Osei-Bonsu, M. K. et al. Assessing the heat stress tolerance potential of tomato lines under poly-house and open field conditions. Cogent Food Agric. 8 , 2115665. 10.1080/23311932.2022.2115665 (2022).

46. Olivoto T Nardino M MGIDI: Toward an effective multivariate selection in biological experiments Bioinformatics 2021 37 1383 1389 10.1093/bioinformatics/btaa981 33226063 Olivoto, T. & Nardino, M. MGIDI: Toward an effective multivariate selection in biological experiments. Bioinformatics 37 , 1383–1389. 10.1093/bioinformatics/btaa981 (2021). 33226063

47. Bashary N New population of Solanum pimpinellifolium backcross inbred lines as a resource for heat stress tolerance in tomato Front. Plant Sci. 2024 15 1386824 10.3389/fpls.2024.1386824 39011307 PMC11246914 Bashary, N. et al. New population of Solanum pimpinellifolium backcross inbred lines as a resource for heat stress tolerance in tomato. Front. Plant Sci. 15 , 1386824. 10.3389/fpls.2024.1386824 (2024). 39011307 PMC11246914

48. Gonzalo MJ Genetic control of reproductive traits in tomatoes under high temperature Front. Plant Sci. 2020 11 326 10.3389/fpls.2020.00326 32391023 PMC7193983 Gonzalo, M. J. et al. Genetic control of reproductive traits in tomatoes under high temperature. Front. Plant Sci. 11 , 326. 10.3389/fpls.2020.00326 (2020). 32391023 PMC7193983