Screening Soybean Genotypes for High-Temperature Tolerance by Maximin-Minimax Method Based on Yield Potential and Loss

Abstract

Temperature rise between 2.6 and 4.8 °C will impact the productivity of soybean at the turn of the twenty-first century. To predict differences in soybean genotypes to high temperatures, twelve soybean genotypes were grown in greenhouses maintained at a mean temperature of 26, 29, 32, and 35 °C, respectively, with one set in natural conditions. The leaf area, total biomass, photosynthesis, Fv/Fm, pollen germination, and reproductive efficiency were significantly high under natural conditions, and a further increase in temperature to 26, 29, 32, and 35 °C resulted in a decline in these parameters. The average seed yield of 12 soybean genotypes was 13.2 g/plant under ambient temperature and there was mild reduction of 8% and 14% when genotypes were grown at 26 and 29 °C, respectively. Severe decline by 51% and 65% in yield was observed at 32 and 35 °C, respectively. The total stress response index in twelve genotypes ranged from −1068 (JS 95-60) to −333 (EC 538828). NRC7 and EC 538828 performed comparatively better than other genotypes. Screening for high-temperature tolerance in soybean is very constrained in breeding programs. This genetic variability among soybean genotypes to elevated temperature reveals that heat tolerance can be improved through plant breeding programs. Additionally, it emphasizes the significance of identifying efficient selection strategies in improving the productivity of soybean in future climate scenarios.

1. Introduction

Climate change is a major abiotic stressor that would adversely affect agriculture worldwide, especially in tropical regions [1]. Global mean temperature is predicted to rise by 0.3 °C per ten years [2,3]. Soybean is the most widely cultivated seed legume, as they provides food, edible oil, protein concentrate for livestock feeding, constituent in formulated diet of fish and poultry, and a variety of industrial products [4,5,6]. At present, the area of soybean in India is 11.9 million ha and productivity is 13.5 million tons [7]. It is contributing a share of 42% to total oil seed and 22% to total oil [8]. Soybean is often exposed to heat stress, owing to which its growth and development are affected to a great extent [9,10,11,12]. In addition, under RCP 8.5 (representative concentration pathway), a yield reduction of 11.6% is predicted for soybean [13].

Among various environmental stresses, high temperature is considered a foremost disparaging stress [14]. A rise in temperature above a threshold level causes irreparable harm to plants [5,15,16,17]. In the coming years, the high-temperature effects are forecasted to amplify more with long-term effects [18,19]. High temperatures have a detrimental outcome on morphological, anatomical, physiological, and biochemical aspects, which sooner or later leads to a decline in crop productivity [20,21]. Several experiments have shown that elevated temperatures will cause a large decline in crop yields [22,23]. In India, soybean is frequently exposed to heat stress, and further rise in temperature above optimum may viciously impede the productivity of soybean, as the existing temperatures are already on the verge of the upper limit [5,8].

A reduction in soybean yield has been observed under high-temperature stress that has grown plants under short-term and long-term heat stress [24,25,26,27]. The soybean growth [28], photosynthesis [29,30], reproductive efficiency [31,32,33], crop yield [34], and seed quality [35,36,37,38] are greatly affected by temperature. A decline in yield will be very severe if high temperatures occur during the flowering or seed-fill stage of crops because of impaired pollen vitality, abortion of flowers, and abscission of young pods [39,40,41,42,43]. Therefore, efficient approaches to search for soybean genetic material to adapt to changing climate require a deep understanding of responses to elevated temperature stress. Although temperature is a major limiting factor for soybean productivity, efficient breeding programs for identifying heat tolerant genotypes are deficient due to a lack of appropriate screening methods. Therefore, only a few thermo-tolerant genotypes were identified. The aspiration of this study was to screen, identify, and classify thermo-tolerant soybean genotypes to promote breeding programs. The major objectives of this study were (I) to understand the impact of elevated temperature on growth, pollen germination, reproductive efficiency, photosynthesis, chlorophyll fluorescence, crop yield and its attributes, (II) to classify soybean genotypes for heat tolerance using maximin-minimax approach, (III) to determine the cumulative stress response index (CSRI) in soybean genotypes to elevated temperature, and (IV) to find the correlation of seed yield with photosynthesis, Fv/Fm (PS II efficiency), and pollen germination in twelve soybean genotypes.

2. Materials and Methods

2.1. Temperature Treatments

A pot experiment was performed under four polyhouses (6 m × 3 m) maintained at day/night temperatures of 30/22, 34/24, 38/26, and 42/28 °C with a mean temperature of 26, 29, 32, and 35 °C, respectively, and one set of plants in pots was also grown under natural conditions. The temperature in the greenhouses was maintained by an automatic temperature controller connected with a thermostat, cooling, and heating system.

2.2. Plant Material and Growing Conditions

In each greenhouse, twelve soybean genotypes, viz. NRC 7, JS 71-05, JS 93-05, EC 456548, NRC 37, JS 95-60, JS 97-52, Hardee, JS 335, EC 538828, EC 602288, and Punjab 1(details of genotypes has been attached in Supplementary File, Table S1), were grown in cement pots of 45 × 18 cm filled with 20 kg soil and manure in the ratio of 2:1. The seeds were treated with Rhizobium japonicum, Bavistin, and Dithane M fungicides before sowing. Overall, there were 72 pots (12 genotypes × 6 pots) for each temperature condition. The plants were irrigated as per requirement in each temperature treatment to avoid water deficit, and leaf water potential was measured using a psychrometer (Wescor Inc., Logan, UT, USA).

2.3. Growth Analysis

To examine the growth in terms of dry matter accumulation, plants from three pots were sampled at R5 stage from each treatment and genotype. Plant parts were divided into leaves, stem, pod, and root, were oven dried at 65–70 °C for 3 days, and then the dry weights were recorded. The leaf area of the samples was measured using automatic leaf area meter (Model LI3100, LICOR, Nebaraska, Lincoln, NE, USA).

2.4. Seed Yield Attributes

At harvest, the plants from the remaining three pots of each temperature were harvested, and total biomass, pods/plant, seeds/plant, seeds/pod, seed weight, harvest index, and 100 seed weight were recorded.

2.5. Reproductive Efficiency

At the flowering stage, individual flower buds were tagged with twine and daily flowers were counted from five plants in all the treatments of each genotype. At harvest, the number of pods formed in each tagged plants was noted and reproductive efficiency was calculated as the ratio of total number of pods formed to number of flowers.

2.6. Gas Exchange and Chlorophyll Fluorescence Measurements

To measure photosynthetic rate and chlorophyll fluorescence, five plants were selected from each temperature treatment. The rate of photosynthesis was measured on the fully expanded third leaf from the top using IRGA photosynthesis system (LI-6400 XT, LI-COR, Inc., Lincoln, NE, USA). Measurements were made between 09:00 to 12:00 a.m. at 1000 µmol m⁻²s⁻¹ of PAR. Simultaneously, chlorophyll fluorescence was measured in dark-adapted leaves (30 min), using a 6400-40 leaf fluorometer system attached with LI-6400 XT.

2.7. Pollen Germination

At anthesis, little opened flowers were randomly sampled in the early morning. Firstly, the germinating medium was prepared in 100 mL of distilled water by mixing 15 g of sucrose, 0.03 g of calcium nitrate, and 0.01 g of boric acid [33]. The pollen germination percentage was determined by placing a drop of media on a glass cover slip. After that, pollen grains were powdered on drops, and then the cover slip was upturned and kept on the concavity slide, and incubated at 25 °C for 60 min dark [40,41]. On each slide, about 50–100 pollen grains were counted using the microscope (LEADZ optics model LOS 400, London, UK), and it was considered germinated only when the length of the pollen tube was equal to or greater than the diameter of the pollen grain [31].

2.8. Maximin-Minimax Approach

In order to categorize soybean genotypes into different classes of thermo-tolerance, a maximin-minimax approach was used [44]. In this approach, genotypes were distributed into different classes depending on their real yield potential under temperature stress. The genotype was categorized and ranked based on the criteria of having maximum seed yield and minimum yield loss percent under temperature stress conditions. The main rationale behind this approach is to maximize the minimum projected seed yield (maximin) and minimize the maximum estimated yield loss percent (minimax).

2.9. Cumulative Stress Response Index (CSRI)

The CSRI was measured as the adding together of individual elements from each temperature treatment [39,45]. CSRI was measured to assess the physiological and reproductive responses of soybean genotypes to different temperature treatments. CSRI was calculated as follows:

$$\begin{array}{l}\mathrm{CSRI}=(\frac{\left(\mathrm{SYt} - \mathrm{SYc}\right)}{\mathrm{SYc}}+\frac{\left(\mathrm{TBMt} - \mathrm{TBMc}\right)}{\mathrm{TBMc}}+\frac{\left(\mathrm{HIt} - \mathrm{HIc}\right)}{\mathrm{HIc}}+\frac{\left(100\mathrm{SWt} - 100\mathrm{SWc}\right)}{100\mathrm{SWc}}+\frac{\left(\mathrm{Pt} - \mathrm{Pc}\right)}{\mathrm{Pc}}\\ \begin{array}{ccc}& & \end{array}\hspace{1em} +\frac{\left(\mathrm{PNt} - \mathrm{PNc}\right)}{\mathrm{PNc}}+\frac{\left(\mathrm{REt} - \mathrm{REc}\right)}{\mathrm{REc}}+\frac{\left(\mathrm{PGt} - \mathrm{PGc}\right)}{\mathrm{PGc}})\times 100\end{array}$$

(1)

where SY—the seed yield, TBM—the total biomass, HI—the harvest index, 100SW—the 100 seed weight, P—the pod number, PN—the rate of photosynthesis, RE—the reproductive efficiency, PG—the pollen germination under t (different temperature), and c (ambient temperature control). The genotypes were categorized on the basis of total stress response index (TSRI).

2.10. Statistical Analysis

Analysis of variance (ANOVA) was calculated for all the parameters using SAS statistical software (ver.9.2; SAS Institute, Cary, NC, USA). The means were compared based on least significant differences (LSD) at p ≤ 0.05 using Duncan multiple range test.

3. Results

3.1. Weather Conditions

Under ambient conditions (control) the maximum temperature was 28.0 °C, minimum temperature was 22.4 °C, and mean temperature was 25.2 °C.

3.2. Effect of Temperature on Growth Parameters

3.2.1. Leaf Area

The maximum leaf area/plant in plants at ambient temperature was 2206 cm² and it was not statically significant with plants grown at 30/22 °C (2196 cm²) (

Table 1. Leaf area, above- and below-ground biomass in soybean genotypes grown at different temperature.

Treatments	Leaf Area (cm2)	Above-Ground Biomass (g/Plant)	Below-Ground Biomass (g/Plant)
Day/Night Temperatures (°C)
Ambient	2206 a	23.4 a	4.94 a
30/22 °C	2196 a	23.0 b	3.63 b
34/24 °C	1960 b	20.5 c	2.72 c
38/26 °C	1613 c	16.3 d	2.13 d
42/28 °C	1302 d	14.5 e	1.62 e
LSD (T)	66.9	0.34	0.147
Genotypes
JS 97-52	2566 a	24.2 b	4.86 a
EC 602288	2301 b	26.3 a	4.60 a
JS 95-60	1047 f	10.0 i	1.25 g
JS 93-05	1104 f	12.8 h	2.43 e
EC 456548	2099 c	20.4 e	2.91 cd
Hardee	2375 b	22.4 c	3.01 c
NRC 37	2316 b	22.4 c	2.65 de
JS 335	2001 c	22.1 c	3.51 b
JS 71-05	1773 d	21.3 d	2.88 cd
EC 538828	1320 e	17.4 f	1.91 f
NRC 7	1619 d	20.7 de	2.66 de
Punjab-1	1742 d	14.8 g	3.42 b
Mean	1855.1	19.6	3.01
LSD (G)	78.3	0.65	0.274
LSD (T × G)	NS	1.44	0.612
ANOVA
T	<0.0001	<0.0001	<0.0001
G	<0.0001	<0.0001	<0.0001
T × G	0.1409	<0.0001	<0.0001

Means within a column followed by the same letters are not significantly different (p ≤ 0.05).

). A further increase in growing temperatures resulted in a significant decline in leaf area by 11, 27, and 41% as the temperatures rose to 34/24, 38/26, and 42/28 °C, respectively. Soybean genotypes also differed significantly for the leaf area/plant at R5 stage. The highest leaf area/plant was recorded in JS 97-52 (2566 cm²) while the lowest was observed in JS 95-60 (1047 cm²) (Table 1). The temperature and genotypes interactions were not significant for the leaf area.

3.2.2. Above-Ground Biomass

Among the temperatures, the maximum average above-ground biomass was in plants at ambient temperature (23.4 g/plant) and at 30/22 °C (23.0 g/plant) (Table 1). As the temperature increased, a significant decline in above-ground biomass by 13, 30, and 38% and at 34/24, 38/26 and 42/28 °C as compared to ambient temperature conditions, respectively, was observed. Average above-ground biomass has the highest value in EC 602288 (26.3 g/plant), while the lowest above-ground biomass was observed in JS 95-60 (10.0 g/plant). The temperature and genotype interaction was significant for the above-ground biomass.

3.2.3. Below-Ground Biomass

Maximum average root dry weight (4.94 g/plant) was observed under ambient temperature, which declined by 27, 45, 57, and 67% at 30/22, 34/24, 38/26, and 42/28 °C, respectively (Table 1). Unlike above-ground biomass, the reduction in below-ground biomass was of very high magnitude, indicating that the partitioning of dry matter to roots was more affected by increasing temperatures. The average root dry weight was the highest in JS 97-52 (4.86 g/plant) and was at par with root weight observed in EC 602288 (4.60 g/plant), whereas the lowest root dry weight was observed in JS 95-60 (1.25 g/plant) (Table 1). The interaction of temperature and genotype for the root growth was significant.

3.3. Effect of Temperature Photosynthetic Rate

The photosynthetic rate was highest under natural conditions (24.9 µmol CO₂ m⁻²s⁻¹) and a marginal decline was observed as the temperature increased to 24.2 µmol CO₂ m⁻²s⁻¹ (30/22 °C) and 22.9 µmol CO₂ m⁻²s⁻¹ (34/24 °C) (

Table 2. Rate of photosynthesis, Fv/Fm, pollen germination, and reproductive efficiency in soybean genotypes grown at different temperatures.

Treatments	Photosynthetic Rate	Fv/Fm	Pollen Germination (%)	Reproductive Efficiency (%)
Day/Night Temperatures (°C)
Ambient	24.9 a	0.797 a	87.3 a	42 a
30/22 °C	24.2 b	0.795 a	82.0 b	40 b
34/24 °C	22.9 c	0.785 b	72.2 c	37 c
38/26 °C	19.9 d	0.770 c	58.7 d	32 d
42/28 °C	16.0 e	0.745 d	48.0 e	28 e
LSD (T)	0.33	0.0037	1.28	1.2
Genotypes
JS 97-52	22.2 bc	0.785 b	71.9 d	24 e
EC 602288	21.2 efg	0.778 c	70.5 e	25 e
JS 95-60	22.0 bcd	0.766 e	60.1 j	39 c
JS 93-05	21.6 cde	0.770 e	64.6 i	40 c
EC 456548	21.2 efg	0.777 c	75.0 c	39 c
Hardee	20.6 gh	0.793 a	68.5 gf	28 d
NRC 37	20.3 h	0.775 cd	64.4 i	21 f
JS 335	21.7 bcde	0.770 e	65.7 hi	39 c
JS 71-05	22.4 b	0.786 b	68.9 f	41 c
EC 538828	20.9 fgh	0.786 b	78.4 b	54 b
NRC 7	23.3 a	0.785 b	81.3 a	58 a
Punjab-1	21.4 def	0.772 de	66.7 gh	21 f
Mean	21.6	0.779	69.7	36
LSD (G)	0.66	0.0036	1.31	2.0
LSD (T × G)	1.47	0.0081	2.93	4.5
ANOVA
T	<0.0001	<0.0001	<0.0001	<0.0001
G	<0.0001	<0.0001	<0.0001	<0.0001
T × G	<0.0001	<0.0001	<0.0001	<0.0001

Means within a column followed by the same letters are not significantly different (p ≤ 0.05).

). However, a greater decline of 20% (19.9 µmol CO₂ m⁻²s⁻¹) and 36% (16.0 µmol CO₂ m⁻²s⁻¹) was observed in plants grown at 38/26 °C and 42/28 °C. Among soybean genotypes, the highest photosynthetic rate was observed in genotype NRC 7 (23.3 µmol CO₂ m⁻²s⁻¹), while the lowest rate of photosynthesis was in NRC 37 (20.3 µmol CO₂ m⁻²s⁻¹) (Table 2). The temperature and genotype interaction was significant for the rate of photosynthesis.

3.4. Effect of Temperature on Fv/Fm

Fv/Fm was found to be stable across growing temperatures (Table 2). The difference in Fv/Fm value was not significant in ambient temperature and at 30/22 °C. There was a noteworthy reduction in Fv/Fm from 0.785 to 0.745 at 34/24, 38/26 and 42/28 °C, respectively (Table 2). The Fv/Fm values ranged from 0.770 to 0.793 among the 12 genotypes that were not significant among genotypes. The interaction of temperature and genotypes was significant for Fv/Fm.

3.5. Effect of Temperature on Pollen Germination

The decline in pollen germination percent was significant with an increase in temperature. Pollen germination was maximum (87.3%) under ambient temperature, which significantly declined to 82.1, 72.2, 58.7, and 48.1% as the growing temperatures were increased to 30/22, 34/24, 38/26, and 42/28 °C, respectively (Table 2). The pollen germination was lowest in JS 95-60 (60.1%) while the highest pollen germination was observed in NRC 7 (81.3%) (Table 2). The interaction of temperature with genotype was significant for pollen germination.

3.6. Effect of Temperature on Reproductive Efficiency

The reproductive efficiency was highest (42%) under ambient temperature, which declined to 40, 37, 32, and 28% as the temperatures increased to 30/22, 34/24, 38/26, and 42/28 °C, respectively (Table 2). Reproductive efficiency was maximum in variety NRC 7 (58%), while it was minimum in NRC 37 and Punjab 1 (21%). The interaction of temperature and genotype was significant for reproductive efficiency.

3.7. Effect of Temperature on Seed Yield and Its Attributes

3.7.1. Seed Yield

The seed yield of 12 soybean genotypes was 13.2 g/plant under natural temperature conditions and there was a mild reduction of 8% (12.2 g/plant) and 14% (11.4 g/plant) at 30/22 °C and 34/24 °C, respectively. A severe decline of 51% (6.4 g/plant) and 65% (4.7 g/plant) in yield was observed at 38/26 °Cand 42/28 °C, respectively (

Table 3. Effect of temperature on seed yield (g/plant) in soybean genotypes.

Genotypes	Ambient	30/22 °C	34/24 °C	38/26 °C	42/28 °C	Mean
JS 97-52	17.1	15.5	15.1	9.0	6.7	12.7 a
EC 602288	16.5	15.1	14.5	8.2	5.0	11.9 b
JS 95-60	10.5	9.0	8.2	2.5	1.5	6.3 f
EC 538828	13.5	13.6	13.0	9.5	7.9	11.5 b
JS 71-05	13.5	12.1	11.5	7.2	5.2	9.9 c
JS 93-05	11.6	10.3	8.9	3.5	2.0	7.3 e
EC 456548	15.2	14.3	13.7	7.8	6.2	11.4 b
Hardee	14.0	12.6	11.5	7.5	5.9	10.3 c
NRC 37	11.1	10.8	9.7	4.0	2.8	7.7 e
JS 335	12.1	10.7	10.2	5.4	3.5	8.4 d
NRC 7	14.0	13.4	13.4	9.6	7.4	11.6 b
Punjab 1	9.5	9.3	6.9	3.0	2.1	6.2 f
Mean	13.2 a	12.2 b	11.4 c	6.4 d	4.7 e	9.6
LSD
Temperature (T)			0.33
Genotype (G)			0.58
T × G			1.29
ANOVA
T			<0.0001
G			<0.0001
T × G			<0.0001

Means within a column followed by the same letters are not significantly different (p ≤ 0.05).

). The seed yield was maximum in JS 97-52 variety (12.7 g/plant), while minimum seed yield was in JS 95-60 variety (6.3 g/plant) and Punjab 1 (6.2 g/plant) (Table 3). The interaction of temperature and genotypes for seed yield was significant.

3.7.2. Total Biomass

Above-ground total biomass recorded at maturity also varied significantly with temperature. Highest total biomass (30.3 g/plant) was observed at ambient temperature, while it declined by 5, 12, 33, and 42% as the temperature increased to 30/22 °C (28.7 g/plant), 34/24 °C (26.8 g/plant), 38/26 °C (20.2 g/plant), and 42/28 °C (17.7 g/plant), respectively (

Table 4. Total biomass, pods/plant, harvest index, 100 seed weight, and seeds/pods in soybean genotypes grown at different temperatures.

Treatments	TBM (g/Plant)	HI (%)	Pods/Plant	100 Seed Weight (g)	Seeds/Pods
Day/Night Temperatures (°C)
Ambient	30.3 a	43.8 a	61 a	12.5 a	1.94 a
30/22 °C	28.7 b	43.0 a	59 b	12.3 a	1.90 a
34/24 °C	26.8 c	42.7 a	56 c	12.0 a	1.91 a
38/26 °C	20.2 d	32.2 b	40 d	10.1 b	1.70 b
42/28 °C	17.7 e	26.4 c	35 e	8.8 c	1.61 c
LSD (T)	0.60	1.09	1.4	0.64	0.075
Genotypes
JS 97-52	32.0 a	38.6 cd	80 a	7.9 g	1.97 cd
EC 602288	32.3 a	35.3 e	79 a	8.3 fg	1.72 ef
JS 95-60	15.7 g	37.0 de	20 h	12.1 c	2.29 a
JS 93-05	17.7 f	37.8 cde	32 f	9.5 e	2.11 b
EC 456548	26.6 c	42.6 b	47 d	14.9 b	1.59 fg
Hardee	26.6 c	37.9 cde	62 b	9.6 e	1.68 ef
NRC 37	25.1 d	29.4 g	58 c	8.0 g	1.53 g
JS 335	26.4 c	30.6 fg	48 d	9.1 ef	1.84 de
JS 71-05	26.3 c	36.6 de	49 d	11.5 d	1.72 ef
EC 538828	21.5 e	53.3 a	24 g	23.7 a	2.03 bc
NRC 7	28.4 b	40.2 c	63 b	11.9 cd	1.52 g
Punjab-1	18.3 f	31.9 f	43 e	7.4 g	1.75 ef
Mean	24.7	37.6	50	11.2	1.81
LSD (G)	0.83	2.57	2.2	1.00	0.139
LSD (T × G)	1.85	5.75	5.0	NS	NS
ANOVA
T	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
G	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
T × G	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001

Means within a column followed by the same letters are not significantly different (p ≤ 0.05).

). On the other hand, the overall extent of decline in total biomass was relatively lower as compared to the reduction observed for seed yield. The average TBM was highest for EC 602288 (32.3 g/plant), which was not significantly different with JS 97-52 (32.0 g/plant), while the lowest above-ground total biomass was in JS 95-60 (15.7 g/plant) (Table 4). The temperature and genotype interaction for TBM was significant.

3.7.3. Harvest Index

The maximum harvest index was observed under ambient temperature (43.8%) and was not affected significantly at 30/22 °C (43.0%) and 34/24 °C (42.7%) (Table 4). The harvest index sharply declined to 32.2 and 26.4% when plants were grown at 38/26 °C and 42/28 °C, respectively. Harvest index was highest (53.3%) in EC 538828, while it was lowest in NRC 37 (29.4%). The temperature and genotype interaction for harvest index was significant.

3.7.4. Number of Pods/Plant

Pods/plant was highest (61) under ambient temperature condition and declined by only 4 and 9% at 30/22 °C (59) and 34/24 °C (56), respectively (Table 4). However, a further increase in growing temperatures lead to a considerable decline in pods/plant to 40 (34%) at 38/26 °C and 35 (43%) at 42/28 °C (Table 4). The pods/plant was very high for soybean genotype JS 97-52 (80) and EC 602288 (79), while a lesser number was observed in JS 95-60 (20) and EC 538828 (24). The temperature and genotype interaction was significant for the number of pods.

3.7.5. Seeds/Pod

Seeds/pod were not affected in plants grown under ambient temperature (1.94), 30/22 °C (1.90), and 34/24 °C (1.91). On the other hand, a rise in temperature to 38/26 °C and 42/28 °C significantly declined the seeds/pod to 1.70 and 1.61, respectively (Table 4). Seeds/pod was significantly higher in JS 95-60 (2.29) and JS 93-05 (2.11), while it was low in NRC 7 (1.52) and NRC 37 (1.53) (Table 4). The temperature and genotypes interaction for seeds/pod was non-significant.

3.7.6. 100 Seed Weight

The average seed size, as reflected by 100 seed weight, was 12.5 g at ambient temperature conditions and was not greatly influenced at 30/22 °C (12.3 g) and 34/24 °C (12.0 g). As the temperatures were increased to 38/26 °C and 42/28 °C, a significant decline in 100 seed weight was observed to10.1 and 8.8 g, respectively (Table 4). Among genotypes, EC 538828 with an average 100 seed weight of 23.7 g had the largest seed size, while genotypes NRC 37, JS 97-52, and Punjab 1 with 8.0, 7.9, and 7.4 g of 100 seed weight possessed small seeds, respectively (Table 4). The temperature and genotype interaction was not significant, representing that a reduction in seed size in response to increasing growing temperatures was similar among the genotypes.

3.7.7. 0-,1-,2-,3-, and 4-Seeded Pods

Soybean plants could have one, two, three, or four seeds in a single pod. The proportion of these pods on a plant is under genetic control, but environment also influences this trait. Under reduced availability of source, soybean plants may tend to produce one- or two-seeded pods and under acute limitations of source could result in empty pods (zero-seeded pods). In our study, the highest proportion of two-seeded pods was observed across genotypes and temperatures (

Table 5. Effect of temperature on zero-, one-, two-, three-,and four-seeded pods in soybean genotypes.

Treatments	0 Seeded Pods (%)	1 Seeded Pods (%)	2 Seeded Pods (%)	3 Seeded Pods (%)	4 Seeded Pods (%)
Day/Night Temperatures (°C)
Ambient	1.3 e	6.4 e	63.8 a	26.8 a	1.8 a
30/22 °C	2.1 d	9.8 d	64.1 a	22.4 b	1.7 a
34/24 °C	2.8 c	12.7 c	63.1 b	20.1 c	1.2 b
38/26 °C	5.6 b	22.8 b	57.5 c	13.7 d	0.4 c
42/28 °C	7.4 a	26.6 a	55.0 d	10.8 e	0.2 d
LSD (T)	0.17	0.29	0.38	0.53	0.09
Genotypes
JS 97-52	3.0 e	15.8 fg	69.3 d	12.0 g	0.0 c
EC 602288	2.9 e	12.7 i	70.6 c	13.8 f	0.0 c
JS 95-60	7.2 a	18.5 c	31.8 i	37.5 b	4.9 b
JS 93-05	5.0 b	11.6 j	32.2 i	43.5 a	7.9 a
EC 456548	4.0 d	19.1 b	58.8 g	18.0 e	0.0 c
Hardee	2.5 f	17.2 d	72.2 b	8.1 i	0.0 c
NRC 37	4.4 c	20.7 a	68.2 e	6.7 j	0.0 c
JS 335	5.1 b	16.7 e	58.3 g	19.9 d	0.0 c
JS 71-05	3.1 e	13.2 h	54.1 h	29.7 c	0.0 c
EC 538828	1.8 g	11.0 k	77.7 a	9.6 h	0.0 c
NRC 7	3.1 e	16.1 f	63.0 f	17.8 e	0.0 c
Punjab-1	3.9 d	15.4 g	72.2 b	8.5 i	0.0 c
Mean	3.8	15.7	60.7	18.8	1.0
LSD (G)	0.17	0.49	0.66	0.62	0.14
LSD (T × G)	0.39	1.09	1.48	1.39	0.31
ANOVA
T	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
G	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
T × G	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001

Means within a column followed by the same letters are not significantly different (p ≤ 0.05).

). A significant increase in the percent of zero-and one-seeded pods was observed with an increase in growing temperatures. In contrast, the percent of two-and three-seeded pods considerably declined, particularly at 38/26 °C and 42/28 °C. Four-seeded pods were observed in a very small proportion in two genotypes (JS 95-60 and JS 93-05), and also reduced as the growing temperatures increased. Genotypes also varied significantly for the proportion of seeds/pod, and the maximum percentage of zero-seeded pods was observed in JS 95-60 (7.2%) and JS 335 (5.1%), while the lowest was in EC 538828 (1.8%). The average proportion of one-, two-, and three-seeded pods varied from 11.0 (EC 538828) to 20.7% (NRC 37), 31.8 (JS 95-60) to 77.7% (EC 538828), and 6.7 (NRC 37) to 43.5% (JS 93-05), respectively (Table 5). The interaction of temperature and genotype was also significant, demonstrating that variation in number of seeds in each pod was differentially influenced by temperature in these genotypes.

3.8. Maximin-Minimax Approach

Using the maximin-minimax approach, taking into consideration the potential yield and the relative loss due to high temperature, soybean genotypes were classified into four classes:(1) Heat-resistant and high yielding (HR), (2) Heat-tolerant and high yielding (HT), (3) Heat-resistant and low yielding (LR), and (4) Heat-susceptible and low yielding (LS). Loss due to high temperature was very high at 42/28 °C, therefore the seed yield observed at 38/26 °C was used for calculating yield potential in soybean genotypes as compared to ambient temperature. Soybean genotypes with a loss in yield lower than 50% were considered as heat-resistant and those with a yield higher than 75% were considered high yielding. Based on evaluation under ambient and at 38/26 °C temperature, only two genotype were recognized as heat-resistant and high yielding (NRC 7and EC 538828), five as heat-tolerant and high yielding (EC 456548, JS 97-52,EC 602288, Hardee and JS 71-05), none as heat-resistant and low yielding, while five genotypes (JS 95-60, JS 93-05, JS 335, NRC 37, and Punjab 1) were considered as low yielding and heat-susceptible (Figure 1).

3.9. Correlation of Seed Yield with Photosynthesis, Fv/Fm, and Pollen Germination

In order to understand the role of different physiological traits such as rate of photosynthesis, Fv/Fm, and pollen germination in imparting heat tolerance, their association with seed yield under all temperature regimes in all soybean genotypes was worked out. A significant and positive relationship between seed yield and rate of photosynthesis (R² = 0.608) was found among these genotypes (Figure 2). The correlation of seed yield and Fv/Fm was also significant (R² = 0.728) (Figure 3). The strongest positive association with seed yield was observed with pollen germination (R² = 0.818) (Figure 4).These relations clearly indicate that rate of photosynthesis, Fv/Fm, and pollen germination had a pivotal role in influencing the yield in soybean under elevated temperature conditions. Therefore, these traits can be used as simple criteria for the screening and selecting of soybean genotypes for improved yield under high-temperature conditions.

3.10. Cumulative Stress Response Index (CSRI)

The CSRI is the summation of individual factors which exhibited that soybean issensitive to high temperatures. It is an integration of the effects of elevated temperature on number of pods, pollen germination, reproductive efficiency, rate of photosynthesis, seed yield, total biomass, 100 seed weight, and harvest index (

Table 6. Cumulative stress response index in twelvesoybean genotypes.

Genotypes	30/22 °C	34/24 °C	38/26 °C	42/28 °C	TSRI
EC 538828	+1.0	−25	−117	−191	−333
NRC 7	−15	−21	−138	−194	−367
EC 456548	−18	−62	−223	−273	−575
JS 97-52	−41	−58	−212	−269	−580
JS 71-05	−38	−100	−199	−250	−587
Hardee	−39	−85	−219	−264	−607
EC 602288	−43	−68	−232	−302	−644
JS 335	−62	−79	−265	−305	−711
NRC 37	−16	−79	−300	−328	−724
Punjab 1	−17	−102	−322	−329	−770
JS 93-05	−39	−115	−317	−378	−848
JS 95-60	−79	−151	−393	−445	−1068

). In all the treatments, the highest CSRI among all the genotypes was observed at 42/28 °C. At this temperature, all the parameters were drastically reduced as compared to lower temperatures. Additionally, relative sensitivities of the soybean genotypes varied across each temperature. The total stress response index (TSRI) over the treatments ranged from −1068 (JS 95-60) to −333 (EC 538828). Thus, EC 456548, EC 538828, NRC 7, and JS 97-52 were the least affected genotypes, while JS 95-60, Punjab-1, and JS 93-05 were the most affected soybean genotypes athigh temperatures (Table 6).

4. Discussions

Soybean is susceptible to high temperatures, as shown by the impact on growth, photosynthesis, pollen germination, reproductive efficiency, and ultimately seed yield was observed in the present study. The optimum temperature for soybean growth and yield in India is 28 °C [46], and existing temperatures in soybean growing regions are observed as close to the upper limit of this optimal temperature [5]. Due to climate change, additional enhancement in temperature will have a comprehensive impact on soybean productivity in India [47]. The adverse impact of high temperature on shoot dry mass, biomass accumulation, and leaf area development has also been reported in previous studies [12,15,16,28,36]. The detrimental effect of high-temperature conditions was observed on leaf area and above- and below-ground biomass. The reduction in leaf area and dry matter increased with raising temperatures. The leaf area at any given point of time depends on the rate of appearance and the senescence of old leaves [48]. In the present study, besides the reduction in the rate of expansion, a higher degree of senescence of leaves was observed at high temperatures, resulting in a considerable loss of leaf area (Table 1). Similar to leaf area, no significant difference in above-ground biomass was observed under ambient temperature and at temperatures of 30/22 °C, while a substantial reduction of 13, 30, and 38% in above-ground biomass was observed with the further rise in temperature to 34/24, 38/26, and 42/28 °C (Table 1). However, the degree of decline was very high for below-ground biomass, which was reduced by 27, 45, 57, and 67% as the temperatures increased to 30/22, 34/24, 38/26, and 42/28 °C, respectively (Table 1). Hence, the present results clearly indicated that below-ground biomass accumulation was more sensitive to temperature as compared to above-ground biomass accumulation in soybean genotypes. Other studies have also shown that with the rise in temperatures above 30 °C, the partitioning of photosynthates between shoots and roots and between vegetative and reproductive structures are strongly altered in soybean [28,49].

Photosynthesis is considered as an important indication of growth because of its direct association to crop productivity [5,15,50,51,52,53,54]. Elevated temperature can affect photosynthesis through stomatal/non-stomatal factors [55,56,57,58]. The rate of photosynthesis in genotypes was similar under ambient conditions and 30/22 °C, while a substantial reduction with increasing magnitude of 8, 20, and 36% was detected as the temperatures increased to 34/24, 38/26, and 42/28 °C, respectively (Table 2). Fv/Fm is usually used as a function of maximal photochemical efficiency of photosystem II (PSII) [5], and it is reported to range between 0.750–0.850 for healthy plants. A slight decrease in Fv/Fm shows no structural damage to the PSII site. In the present study, the Fv/Fm was found to be stable across growing temperatures (Table 2). A very low decline of 0.797 under ambient conditions to 0.745 at 42/28 °C was observed. It clearly showed that under high-temperature conditions PSII was not damaged, and its efficiency was not compromised (Table 2). In plants, reproductive processes are more sensitive to high temperatures as compared to vegetative processes, so reproductive organs are more susceptible to high temperatures [45]. Pollen grains are very sensitive to high-temperature conditions compared to ovules because they are exposed to an ambient environment [59,60]. At an ambient temperature pollen germination was 87.3%, which declined to 48.1% at 42/28 °C (Table 2). The reproductive efficiency was also severely reduced as the temperature increased to 42/28 °C (Table 2). Under ambient temperature the reproductive efficiency was 42%, which decreased to 28% at 42/28 °C.

Genotypic variation in responses to pollen germination to heat stress has already been well-documented in soybean [33,40,41]. On the other hand, evidence about genotype-dependent variability of seed yield under high temperature for seed yield is very limited. A slight decline of 8 and 14% in seed yield was observed under 30/22 °C and 34/24 °C, respectively (Table 3). However, the severe reduction of 51 and 65% was observed as the temperatures rose to 38/26 °C and 42/28 °C, respectively, as compared to ambient temperature. Various studies have shown that the increase in temperature from 26 to 30 °C reduced photosynthesis, seed yield, and number of pods and seeds [10,11].

Seed yield is linked with biomass accumulation, and as a result, any influence of temperature on biomass production will also affect seed yield [48]. There was a negative influence of high temperature on most of the physiological traits, which was proved by severe decline in seed yield. The harmful effect of these physiological processes manifested in reduced yield through their impact on yield attributing characters such as total biomass, pods/plant, seeds/plant, 100 seed weight, and harvest index. A detrimental effect of temperature beyond 34/24 °C was observed on all these yield-contributing factors. The total dry weight, pods/plant, seeds/pod, and 100 seed weight were reduced by 42, 43, 17, and 31% at 42/28 °C as compared to ambient temperature (Table 4). Pod number was the most exaggerated trait by high temperature, which was due to a reduction in pod to flower ratio (Table 4). It clearly indicates that, under high temperatures, reproductive processes like pollen germination and stigma receptivity are highly susceptible in soybean. Interestingly, the temperature influenced even the percent of zero-, one-, two-, three-, and four-seeded pods (Table 5). At high temperatures, the percentage of three-seeded pods was significantly reduced, along with the concomitant increase in empty and single-seeded pods. Harvest index was 43.8% at ambient temperature, which was reduced to 26.4% at 42/28 °C. A decline in harvest index was due to reduced partitioning to different organs of plants because of a shortage in reproductive sink (Table 4).

Based on percent decrease in seed yield, among 12 soybean genotypes, NRC 7 and EC 538828 were found to be less sensitive, JS 97-52, EC 456548, EC 602288, Hardee, and JS 71-05 were moderately sensitive, and JS 93-05, NRC 37, JS 335, Punjab 1, and JS 95-60 were highly sensitive to high-temperature stress. The maximin-minimax method was used to categorize the soybean genotypes into four different classes of high-temperature response (Figure 1). The soybean genotypes in class I (NRC 7 and EC 538828) and II (JS 71-05, EC 602288, JS 97-52, EC 456548, and Hardee) are good genetic sources for soybean breeding programs designed at improving heat tolerance, and can also be endorsed for cultivation under high-temperature conditions in India. CSRI showed that soybean is highly susceptible to temperature. The TSRI, over the temperature treatments, ranged from −1068 (JS 95-60) to −333 (EC 538828) (Table 6). Thus, EC 538828 was found to be the most tolerant and JS 95-60 was found to be the most sensitive to high temperature. JS 97-52, EC 538828, and EC 602288 have also been identified as tolerant to water stress and high-temperature conditions [61,62,63]. Thus, future studies on their combined effect of temperature and water stress are needed because both these stresses frequently occur in combination, and JS 97-52, EC 602288, and EC 538828 possess both drought and temperature tolerance characteristics.

A significant variation in seed yield and other physiological traits was observed in twelve soybean genotypes. Additionally, the interaction of temperature and genotype was significant for the majority of the physiological traits, which clearly highlighted that the reaction of these genotypes to higher temperatures differed substantially. The adverse effect of high temperature on physiological traits such as dry matter production, photosynthesis, pollen germination, and reproductive efficiency was less in those soybean genotypes, which has shown less reduction in seed yield as compared to sensitive genotypes. Attempts were made to correlate the seed yield with photosynthesis (Figure 2), Fv/Fm (Figure 3), and pollen germination (Figure 4). A linear, strong, positive, and significant relationship was found between seed yield and these parameters. Thus, pollen germination, photosynthesis, and Fv/Fm can be used as selection criterion for screening soybean genotypes for better yield under high-temperature conditions.

5. Conclusions

Although temperature is the major limiting factor for soybean productivity, efficient breeding programs for developing heat-tolerant genotypes are deficient due to a lack of appropriate screening techniques, and thus there are less genetic sources identified for temperature tolerance. Our dataset offers clear evidence for substantial variations in terms of yield and physiological parameters against high temperature among soybean genotypes. High temperatures during reproductive processes cause intense loss due to abscission of flowers and pod abortion, which ultimately reduces seed weight. Therefore, crop breeding programs need to be focused on improving photosynthesis, pollen germination, and reproductive efficiency for developing soybean genotypes to perform better under elevated temperature conditions. A great deal of genetic variability and the strong relationships of physiological traits helped in the identification of soybean genotypes, which might be used as trustworthy donors in plant breeding programs. Therefore, there is a vital need to evaluate a large number of soybean genotypes which have a superior range of tolerance to heat stress conditions. Such thermo-tolerance will help in improving the soybean productivity under future climatic scenarios. Thus, using these physiological parameters, soybean genotypes better grown for high-temperature tolerance can be identified. Additionally, for high-temperature tolerance, novel approaches to crop breeding are needed where suitable tolerant soybean varieties can be selected, which will play an important role. Here, we propose a rapid and reliable method to select and rank soybean genotypes in terms of thermo-tolerance, and we found that the genotypes EC 538828 and NRC 7 maintained the highest potential yield under high temperature, there by resulting in promising source material for further breeding programs.