Introduction

Increasing greenhouse gas emissions are driving a rise in global temperatures, posing a significant threat to food and nutrition security due to their adverse effects on crop yields, including mung bean [Vigna radiata (L.) R. Wilczek]1,2,3,4,5,6. Mung bean is rich in protein (20.9–31.3%) and various micronutrients, vitamins, antioxidants, and health-benefiting bioactive compounds7,8,9,10, playing a vital role in nutritional security and helping combat micronutrient-related malnutrition, particularly in Southeast Asia and other parts of the world11. Mung bean thrives in tropical regions with dry or semiarid climates, typically grown during the post-rainy and summer seasons5. The crop grows best within an optimal mean temperature range of 27–30°C12. However, as a warm-season crop, mung bean is often exposed to extreme heat13, impairing vegetative and reproductive growth5,14.

In Southeast Asian mung bean-producing, especially in India, daytime maximum temperatures frequently exceed 40 °C, surpassing the crop’s optimal temperature range14. Moreover, due to climate change, heat spells are becoming more frequent during critical stages such as reproductive and seed-filling periods5,14,15. Heat stress during these stages has particularly severe effects on reproductive processes, including anthesis, pollen viability and germination, pistil development, stigma receptivity, fertilization, pod development, seed set, and seed filling, leading to substantial yield losses in mung bean3,5,15,16.

While the negative impact of heat stress on crop growth, biomass, yield, and yield attributes has been studied widely3,17,18, its effects on seed nutritional composition have received less attention, particularly in crops like mung bean5. Besides impacting crop growth and yield, heat stress negatively impacts seed quality traits critical for nutritional value18,19,20,21,22,23. Deleterious effects of heat stress on seed quality, including carbohydrates, proteins, and fats, have been documented in wheat (Triticumaestivum L.)20, rice (Oryza sativa L.)24, soybean (Glycine max. L.)25, and chickpea (Cicer arietinum L.)21,23. In lentil (Lens culinaris Medik.), heat stress led to reductions in starch (25–43%), protein (26–41%), amino acids, and fat (39–57%)19,26. Similar reductions in seed micronutrients such as iron (Fe), zinc (Zn), manganese (Mn), and copper (Cu) were reported in wheat and rice20,24, chickpea21, and lentil27 under heat stress.

However, the dynamic response of macro- and micronutrients—including seed protein, carbon (C), phosphorus (P), potassium (K), calcium (Ca), magnesium (Mg), Cu, Zn, Fe, Mn, and sulfur (S)—to heat stress in mung bean remain poorly understood5,28. To address this gap, we assessed the impact of heat stress on various seed nutrient traits and seed yield in ten diverse mung bean genotypes,growing them under optimal (34/25 °C) and heat stress (42/30 °C) conditions in growth chambers.We hypothesized that heat stress (42/30 °C) significantly reduces the concentration of key macro- and micronutrients in mung bean seeds compared to optimal conditions (34/25 °C), with genotypes displaying varying responses to heat stress.

Materials and methods

Ten diverse mung bean accessions (PI 425238, PI 425243, PI 425425, PI 379667, PI 363361, PI 363443, PI 364044, PI 237669, PI 223280, and PI 223002) (Table S1) were evaluated for their seed yield and nutrient quality responses to heat stress under controlled conditions at Kansas State University, Manhattan, from 2023 to 2024. The seeds were sourced from the USDA Plant Genetic Resources Conservation Unit in Griffin, Georgia, USA.

Experimental condition

The experiment was conducted in a randomized control block design with three replications, each comprising one pot with two plants under non-stress (control) and heat stress conditions. Seeds were sown in pots in 17.8 cm-diameter pots containing potting soil (Fafard®3B Mix/Metro-Mix®830, SUNGRO Horticulture, Agawam, MA, USA)29. In the growth chamber, photosynthetically active radiation (400–700 nm) was provided by cool fluorescent lamps at an intensity of 600 μmol m–2 s–1,with a 12 h photoperiod.The average relative humidity was maintained at 60%29. The plants were watered regularly (at field capacity i.e. the soil was kept saturated enough to hold water without excess runoff) to prevent drought stress, with nutrient supplementation applied every 7–14 days, mixing 1/2 teaspoon of Miracle-Gro water-soluble all-purpose plant food 24–8-16 in 4.5 L water (see Table S2 for details)29. Upon flowering onset, one set of genotypes was subjected to heat stress (42/30 °C), while another set was maintained at optimal conditions (34/25 °C). Temperature data were recorded using a HOBO®temperature logger (Onset Computer Corporation, Bourne, Massachusetts, USA), and the recorded data are provided in Fig.S1. Plants from non-stress and heat stress conditions were harvested at maturity to record yield and seed quality traits. The maturity dates (days after sowing) for each genotype under non-stress and heat stress conditions were as follows: PI363443 (75 and 63 days), PI237669 (74 and 67 days), PI425425 (73 and 67 days), PI363361 (77 and 68 days), PI364044 (79 and 69 days), PI425238 (78 and 67 days), PI379667 (74 and 67 days), PI425243 (73 and 69 days), PI223280 (77 and 69 days), and PI223002 (78 and 68 days).

Estimation of seed nutrition components

Mung bean seed samples (0.5 g dry seed per sample) from three replications were collected to quantify various nutritional components, including protein, C, P, K, Mg, Ca, S, Mn, Cu, Fe, and Zn concentrations. The samples were analyzed at the Kansas State Soil Testing Lab (Manhattan, KS, USA). The total levels of C and N were determined using a LECO TruSpec CN Carbon/Nitrogen combustion analyzer (LECO Corporation, St. Joseph, Michigan, USA), expressed as a percentage byweight. For other elements, a nitric perchloric digestion was performed, with the concentrations of Ca, Mg, Zn, Fe, Cu, Mn, S, K, and P measured using an Inductively Coupled Plasma (ICP) Spectrometer, Model 5800 ICP OES Agilent Technologies (Agilent Technologies, Santa Clara, California, USA)30. Seed protein was calculated by multiplying the N content by 6.25 as a conversion factor31.

At maturity, plants were harvested, and seed yield per plant (SYP) was measured by threshing both plants from each replication and averaging the results.

Statistical analysis

Analysis of variance (ANOVA) was performed using OPSTAT, with the least significant difference calculated at the 5% and 1% significance levels. Correlation analysis of seed nutrient traits and SYP under non-stress and heat stress conditions was conducted using PAST software. Principal component analysis was also performed using PAST software, and genotype clustering was carried out using Heatmapper32.

Results

Genetic variability for seed nutrition components

The ANOVA revealed significant genetic variation for all assessed nutrient parameters, except for seed Cu concentration, in the control environment. Under heat stress, all traits exhibited highly significant genetic variation. A two-way ANOVA was conducted to explore the genotype (G) × temperature (T) interaction, which showed a significant G × T effect for all traits except for seed P, Fe, and Cu concentrations (Table 4). The mung bean genotypes significantly differed for various nutrients traits, the individual effect and interaction effects are given in (Table S5 and Table S6).Under non-stress conditions, significant genetic variability was observed for protein (23.7–27.1%), C(42.2–42.3%), P (0.50–0.62%), K (1.68–3.04%), Ca (0.12–0.18%), Mg (0.22–0.27%), Mn (16.9–23.5 ppm), Zn (36.7–51.3 ppm), S (0.23–0.31%), Cu (1.1–2.9 ppm), Fe (41.5–76.2 ppm), and SYP (2.7–6.9 g) (see Tables 1, 2, S3). Similarly, under heat stress, considerable genetic variability was found for nutrient content, including protein (25.7–34.1%), C(42.1–42.3%), P (0.40–0.56%), K (1.37–2.43%), Ca (0.097–0.154%), Mg (0.197–0.252%), Mn (14.3–19.5 ppm), Zn (32.9–46.6 ppm), S (0.21–0.26%), Cu (0.7–3.07 ppm), Fe (38.3–58.8 ppm), and SYP (1.23–3.22 g) (Tables 1, 3, S4).

Table 1 Genetic variability for various seed nutritional components in ten mung bean genotypes under control (non-stress, NS) and heat stress (HS) conditions.
Full size table
Table 2 Seed nutritional and seed yield/plant parameters of 10 mung bean genotypes under non-stress conditions.
Full size table
Table 3 Seed nutritional and seed yield/plant parameters of 10 mung bean genotypes under heat-stress conditions.
Full size table

Seed nutrition components

Protein and carbon concentrations

Average seed protein concentration significantly increased from 27.1% to 30.2% under heat stress (Fig. 1). In contrast, the average seed C concentration decreased slightly from 42.5% to 42.2% under heat stress. Seed protein concentration showed significant treatment and interaction effects (Table 4). All genotypes experienced increased protein percentage under heat stress,with PI 425,425 showing the smallest increase compared to the other genotypes. Conversely, seed C concentration decreased significantly under heat stress, with PI 425,425 showing the least reduction in C concentration among all tested genotypes.

Fig. 1
figure 1

Seed (a) protein (genotype × treatment: 3.02), (b) carbon (genotype × treatment: 0.039), (c) phosphorus (genotype × treatment: 0.002), and (d) potassium (genotype × treatment: 0.02) concentrations (%) of mung bean genotypes under control (non-stress, NS) and heat stress (HS) conditions. LSD values (P < 0.05); values are means + SE. (n = 3).

Full size image
Table 4 ANOVA values for mean sum of squares for 12 different traits across the ten mung bean genotypes.
Full size table

Primary and secondary macronutrients

Total seed P concentration was influenced by treatment and genotype but not by their interaction. Seed K, Ca, Mg, and S concentrations showed significant treatment, genotype, and G × T interaction effects, with all nutrients decreasing under heat stress. Notably, PI 425,238 and PI 237,669 exhibited the least reduction in K and Ca concentrations under heat stress (Fig. 2).

Fig. 2
figure 2

Seed (a) calcium, (b) magnesium, (c) sulfur, and (d) manganese concentrations (ppm) of mung bean genotypes under control (non-stress, NS) and heat stress (HS) conditions. LSD values (P < 0.05); genotype × treatment: 0.00 for (Ca%), 0.00 for (Mg%), 0.001 for (S%), and 3.82 for (Mn). Values are means + SE. (n = 3).

Full size image

Micronutrients

Seed Mn and Zn concentrations showed significant genotype, temperature, and G × T interaction effects (Table 4). However, seed Fe and Cu concentrations did not exhibit significant G × T interaction effects. Among the ten genotypes, PI425423 and PI364044 demonstrated the least reduction in seed Mn and Zn concentrations under heat stress (Fig. 3).

Fig. 3
figure 3

Seed (a) copper (Cu), (b) zinc (Zn) and (c) iron (Fe) concentrations (ppm) of mung bean genotypes under control (non-stress, NS) and heat stress (HS) conditions. LSD values (P < 0.05); genotype × treatment: 0.35 for (Cu), 17.34 for (Zn), and 152.9 for (Fe). Values are means + SE (n = 3).

Full size image

Seed yield components

Seed yield/plant decreased significantly under heat stress, ranging from 49.8% to 78.9%. The SYP reduction showed significant genotype, temperature, and G × T interaction effects. Among the tested genotypes, PI363361 exhibited the smallest reduction in SYP under heat stress. This genotype may, therefore, have the potential to improve heat tolerance in mung bean breeding programs.

Correlation analysis

Association analysis of seed nutrient components revealed a significant positive correlation between seed P and protein concentrations (0.62*) (Fig. 4). High and significant associations were also observed between seed Ca and S (0.58*), S and Mg (0.67*), Fe and protein (0.74*), and S and P (0.67*). In contrast, under non-stress conditions, significant negative correlations occurred between seed C and protein (–0.83*), Fe and C (–0.68*), and SYP and Zn (–0.68*).

Fig. 4
figure 4

Correlation analysis of various nutritional traits under (A) non-stress and (B) heat stress conditions. Traits: Protein % = Protein (%); C = Carbon (%), P = Phosphorous (%), K = Potassium (%), Ca = Calcium (%), Mg = Magnesium (%), S = Sulfur (%), Mn = Manganese (ppm), Zn = Zinc (ppm), Cu = Copper (ppm), Fe = Iron (ppm), and SYP = Seed yield per plant (g).

Full size image

Under heat stress, notable positive and highly significant associations were found between seed protein and P (0.78*), P and Mg (0.63*), Fe and P (0.67*), and S and Mg (0.64*) (Fig. 4). Conversely, significant negative associations occurred between seed Cu and Ca (–0.65*), Cu and Mg (–0.70*), and SYP with all studied nutritional components.

Principal component analysis (PCA)

Under control conditions, four principal components (PCs) collectively explained 82.5% of the total variability across all 12 traits. The individual contributions of each component were 33.6%, 20.6%, 16.8%, and 11.4% (Fig. 5). For PC1, the major contributors were seed C(0.84), K(0.80), SYP (0.72), and Cu(0.62), withseed P contributing the most negatively (–0.80). For PC2, the greatest positive contributors were seed Ca(0.74) and SYP (0.52), with seed Cu contributing the most negatively (–0.66) (Table 5). For PC3, seed S(0.82) and Mg(0.42) had the greatest positive contributions, while SYP had the largest negative impact(–0.28). For PC4, seed Fe contributed the most positively (0.77), while seed Ca had the largest negative contribution(–0.36).

Fig. 5
figure 5

Principal component analysis (PCA) of various seed nutritional traits under (A) non-stress and (B) heat stress conditions. Traits: P = Protein (%), C = Carbon (%), P = Phosphorous (%), K = Potassium (%), Ca = Calcium (%), Mg = Magnesium (%), S = Sulfur (%), Mn = Manganese (ppm), Zn = Zinc (ppm), Cu = Copper (ppm), Fe = Iron (ppm), and SYP = Seed yield per plant (g).

Full size image
Table 5 Correlation between seed variables and principal components (PCs) under non-stress conditions.
Full size table

Under heat stress, four PCs collectively accounted for 83.5% of the total variability across all ten traits. The individual contributions of each component were 36.7%, 18.7%, 17.2%, and 10.8% (Fig. 5). For PC1, the major positive contributors were seed K(0.95), C(0.86), Mn(0.77), and SYP (0.47), while seed Fe (–0.68) had the most substantial negative effect. For PC2, seed Zn (0.91) and C(0.60) had the greatest positive contributions, while seed P(–0.34) had the largest negative impact (Table 6). For PC3, seed S(0.88) and P(0.64) exhibited the greatest positive contributions, while SYP had the largest negative contribution(–0.41). For PC4, SYP contributed the most positively (0.62), while seed C had the largest negative contribution (–0.52).

Table 6 Correlation between seed variables and PCs under heat stress conditions.
Full size table

Clustering and heat map analysis

Heat map analysis revealed that, under non-stress conditions, the ten genotypes assessed for nutritional and yield parameters could be grouped into two major clusters. The first cluster included PI 425243, PI 379667, and PI 237669, while the second cluster contained PI 223002, PI 425238, PI 364044, PI 223280, PI 363443, PI 363361, and PI 425425 (Fig. 6). Similarly, under heat stress, the genotypes were grouped into two clusters. The first cluster contained PI 223002, PI 425238, PI 379667, PI 363361, PI 425425, and PI 237669, while the second cluster included PI 223280, PI 425243, PI 364044, and PI 363443 (Fig. 7). The rank summation index, based on the percentage reduction of evaluated traits under non-stress and heat stress conditions, indicated that PI 363443, PI 425425 and PI 379667 (Fig. 8) could serve as valuable donor parents for improving quality traits in mung bean under heat stress.

Fig. 6
figure 6

Heat map analysis of mung bean genotypes under control (non-stress) conditions, showing responses of various seed nutritional traits. Traits: P = Protein (%), C = Carbon (%), P = Phosphorous (%), K = Potassium (%), Ca = Calcium (%), Mg = Magnesium (%), S = Sulfur (%), Mn = Manganese (ppm), Zn = Zinc (ppm), Cu = Copper (ppm), Fe = Iron (ppm), and SYP = Seed yield per plant (g).

Full size image
Fig. 7
figure 7

Heat map analysis of mung bean genotypes under heat stress conditions,showing responses of various seed nutritional traits. Traits: Protein = Protein (%), C = Carbon (%), P = Phosphorous (%), K = Potassium (%), Ca = Calcium (%), Mg = Magnesium (%), S = Sulfur (%), Mn = Manganese (ppm), Zn = Zinc (ppm), Cu = Copper (ppm), Fe = Iron (ppm), and SYP = Seed yield per plant (g).

Full size image
Fig. 8
figure 8

Ranking of genotypes based on the percentage reduction of assessed quality traits under control (non-stress) and heat stress conditions. Traits: Protein = Protein (%), C = Carbon (%), P = Phosphorous (%), K = Potassium (%), Ca = Calcium (%), Mg = Magnesium (%), S = Sulfur (%), Mn = Manganese (ppm), Zn = Zinc (ppm), Cu = Copper (ppm), Fe = Iron (ppm), and SYP = Seed yield per plant (g).

Full size image

Discussion

This study provides new insights into the negative impact of heat stress on yield and seed quality parameters—including macro- and micronutrients—in mung bean. The genotype-by-temperature interaction on yield and seed quality is a key finding consistent with previous studies in various crops, including grain legumes4,5,13,16.

In this investigation, exposing mung bean genotypes to heat stress resulted in contrasting effects on seed C and protein concentrations, with seed protein increasing. Similar findings have been observed in mung bean5, chickpea23, and lentil18, where seed carbohydrates decreased. The decline in seed C concentrationis likely due to impaired carbohydrate synthesis under heat stress33. Conversely, seed protein concentration increased under heat stress, as reported in wheat34, which experienced higher globulin grain protein levels35. An increase in vicilin seed protein due to heat stress has also been observed in pea (Pisum sativum L.)36, and a similar pattern was found in soybean37. However, some studies have reported reduced seed protein concentration in mung bean under heat stress33.

The increased seed protein concentration under heat stress may be due to reduced activity of the enzyme glutamine synthetase38,39. In maize (Zea mays L.), glutamine synthetase activity declined during the grain-filling stage, whereas enhanced glutamate synthase (GOGAT) and glutamate pyruvate transaminase activity increased grain protein concentration40. Moreover, plants may produce high defense-related proteins for survival, viz., heat shock proteins,with enhanced up-regulatory activity under heat stress41,42,43.

We observed a significant positive correlation between seed P and seed protein under non-stress conditions, suggesting that selecting genotypes with high seed protein content could enhance seed P. Similarly, Chen et al.44 reported a significant positive relationship between seed P and seed protein in lentil. The positive correlation between seed Ca and S suggests that higher seed S levels may promote better calcium uptake in plants, as suggested by Brodowska and Kaczor45. Additionally, the positive correlation between seed P and S observed in our study also mirrors findings in sesame (Sesamum indicum)46.

A strong negative correlation was found between seed Zn and SYP, similar to findings in common bean (Phaseolus vulgaris L.)47. Additionally, a significant negative correlation was observed between seed C and Fe,which echoes studies in soybean48, where a negative relationship between Fe and soluble sugar content was noted. The potential trade-off between seed C and protein concentrations suggests that reducing seed C could promote protein accumulation by enhancing its synthesis. Likewise, a significant negative association between grain starch and protein was reported in lentil44.

The positive correlation between seed Mg and P concentrations suggests that Mg may act as a carrier for P, facilitating appropriate plant growth. Similar findings have been reported in soybean by Webb et al.49, who found a positive relationship between seed Mg and P. Moreover, this study’s positive relationship between seed P and Fe concentrations under heat stress suggests that the fixation of bivalent Fe ions by P could improve seed germination and growth under heat stress. This finding aligns with the results of Teboul et al.46 in sesame.

Under heat stress, seed protein positively correlated with P, Ca, K, Mg, Fe, and S concentrations. This finding is consistent with Chen et al.44 in lentil and Samineni et al.21 in chickpea, where nutrient components were assessed under heat and drought stress conditions. However, SYP negatively correlated with most seed quality traits, indicating that improving SYP may simultaneously enhance seed protein, C, and other micronutrients under heat stress. These correlations suggest that improving seed protein content in mung bean under heat stress could result in concomitant improvements in seed P, Ca, Mg, Fe, and S.

The PCA analysis revealed that the prime traits contributing to the total variability in the studied genotypes were seed yield and various seed nutritional components, as supported by previous studies in mung bean5, lentil27, and chickpea23. The rank summation index-based clustering analysis identified PI 363,443, PI 425,425, and PI 379,667 as genotypes with improved quality traits, making them suitable candidates for inclusion in mung bean breeding programs to enhance seed nutrition quality under heat stress.

Conclusion

Heat stress at 42/30 °C significantly reduces mung bean yield and diminishes the quality of key attributes, including seed C and other essential nutrients such as P, Ca, K, Mg, S, Zn, Fe, Cu, and Mn concentrations. However, seed protein concentration remained unaffected by heat stress. The positive correlation observed between seed protein and other nutrients suggests that improving protein could enhance the nutritional profile of mung bean under heat stress. Therefore, maintaining a balanced level of seed C, protein, and other nutritional elements is essential for ensuring mung bean’scalorific and nutritional value as temperatures rise.