1. Introduction
Soybean [
Glycine max (L.) Merr.] seeds are an important source of protein, folic acid, vitamins, and minerals [
1,
2,
3,
4,
5,
6]. The wide adaptability of this crop makes it popular worldwide and its cultivation is increasing gradually. Annually, the majority of the world’s soybeans are produced in the USA, while a large percentage of soybean is also produced in Sub-Saharan Africa, Brazil, and Nigeria [
7].
This crop has been cultivated since the early 1970s in Bangladesh by the Mennonite Central Committee; recently, the cultivation of soybean has been extended dramatically from only 5000 ha in 2005 [
8] to 62,508.50 ha in 2018–2019 [
9]. The consciousness about the high protein and nutrient content of soybean is increasing day by day [
10,
11]. Bangladesh has achieved almost self-sufficiency in cereal production, though the levels of malnutrition among children, adolescents, and women are amongst the highest in the world. Plants are natural sources of biochemicals with numerous phenolics, antioxidants, vitamins, flavonoids, minerals, numerous pigments, dietary fiber, protein, and carbohydrates [
12,
13,
14,
15,
16,
17]. Due to the high protein, oil, carbohydrate, sugar, dietary fiber, vitamin, and mineral content of soybean, it can be a good candidate crop for improving the nutrition of Bangladeshi people. Moreover, the isoflavones present in soybean seeds are beneficial for decreasing certain cancers, osteoporosis, cardiovascular diseases, and menopausal symptoms.
Three planting times may be recommended for soybean cultivation in Bangladesh, namely Rabi (started in mid-October and ended in mid-March), Kharif 1 (started in mid-March and ended in mid-July), and Kharif 2 (started in mid-July and ended in mid-October). However, most of the lands remain occupied with Aman rice in Kharif 2, while during Rabi, the land is used for growing winter crops. Therefore, Kharif 1 season may be a good option for growing soybean when only a few field crops are grown. However, it is difficult to harvest good crops in Kharif 1 season due to the shortage of water and prevailing high temperature. Another option to increase soybean production is to intensify its cultivation in char lands (land of a riverbank or any accretion in a river course or estuary), which comprised one million hectares. The soils of char lands are sandy or silty loam with low moisture-holding capacity. There is minimum crop diversity in chars compared to that of the mainland [
18]. Moreover, there is no good variety of soybeans developed so far for the Kharif 1 season, and no significant attempt has been taken to include soybean in Kharif 1 for fitting into a rice-cropping pattern in the drought-prone areas.
Drought, a shortage of water in the plant root zone, is the most significant abiotic stress affecting food production and security worldwide [
19]. It hampers farming, and changes the morphology of the plants, reducing seed quality and quantity [
20,
21]. Drought affects the physiological processes of plants that are related to crop growth, development, and economic yield [
22,
23,
24]. Drought stress reduces the production of crops [
25] by creating osmotic stress [
26,
27] and reactive oxygen species (ROS) [
28], which eventually generate oxidative damage and change numerous physiological and biochemical activities such as membrane, DNA, and protein damages, nutrient imbalance [
29,
30], and diminution in photosynthetic rates and changes color pigments [
31,
32,
33]. Plant cells lose their turgidity, which hampers cell enlargement and plant growth under drought conditions [
34]. The most important physiological process of photosynthesis (Pn) reduces under drought conditions resulting in decreased productivity of plants [
35]. Further, it also decreases the leaf area index (LAI) in various crops [
36]. Reduction in LAI causes lower Pn in plants leading to less dry matter (DM) production. Reduction in plant growth, leaf size, root, and stem DM is a common phenomenon when plants are exposed to drought at any growth stage. Moreover, plants are severely affected when water stress occurs at the reproductive stages rather than at the vegetative stage [
37]. Moradi et al. [
38] stated that drought during reproductive stages significantly reduced flower and pod numbers, and consequently, crop yield. To mitigate stresses, the plant has enhanced both enzymatic and non-enzymatic antioxidants, such as tocopherols, betalain, ascorbic acids, carotenoids, betacyanin, betaxanthin, chlorophyll
a (Chl
a), Chl
b, beta-carotene, phenolic and flavonoids [
39,
40,
41,
42,
43,
44,
45] and detoxify the ROS. Under drought conditions, grain yield could be considered as suitable criteria for the selection of drought tolerant variety. The varieties that perform better in terms of yield loss in drought conditions could be considered drought tolerant.
The crop damage due to environmental stress, i.e., drought and high temperature, differs among crop species and within the genotypes of a single species. Several physiological changes occur during the prevailing stress; notably, changes in water relations, biochemical and enzymatic activities, etc. [
46,
47]. According to Zlatev and Lindon [
48], the effect of drought is perceived in the decrease in growth and photosynthetic carbon assimilation. The changes in plant water content, physiological process, and biochemical attributes of the cell are the common changes during drought. However, changes in membrane structure and ultrastructure of subcellular organelles are also fundamental changes that occur under drought stress [
49]. Moreover, those changes are situation-specific and thus, it is necessary to analyze those changes that occur in plants using a particular situation of environmental stress to understand the mechanisms of stress tolerance of the particular crop. In this study, the popular soybean varieties, including new lines, were cultivated under favorable as well as drought conditions. The findings of the study will help to identify some mechanisms of drought tolerance in soybean and identify some phenotypes tolerant under water-scarce conditions. Understanding drought tolerance potential in the plant is crucial to facilitate the genetic improvement of crop plants, especially for developing climate-smart crop varieties [
50,
51]. Given the importance of soybean production under changing climatic conditions, the specific objectives of this present study were (i.) to select drought-tolerant soybean genotypes which can be grown under field conditions with less water, and (ii.) to determine the effects of drought stress on various morpho-physiological and biochemical properties, as well as the yield of soybean genotypes.
2. Materials and Methods
2.1. Experimental Site
Two experiments were carried out in the research field of Bangabandhu Sheikh Mujibur Rahman Agricultural University (24°09′ N and 90°26′ E), Gazipur, Bangladesh. The soil has low organic matter (1.5–2.0%) and the metrological data are shown in
Table 1. The air temperature and relative humidity were low in the early crop growth stage and increase gradually from January to June (
Table 1). Total monthly precipitation was minimal up to March, but dramatically increases from April.
2.2. Experimental Treatment and Design
The land of the experimental field was prepared by plowing with a tractor and then harrowing. At the final land preparation, nine soil samples were collected from 0–15 cm depth of the experimental plot. A composite soil sample was prepared by mixing the collected samples. The sample was air-dried, crushed, and passed through a 2 mm sieve. The pH of the experimental soil was 6.1, soil organic matter was 1.20%, and total N was 0.12%. The status of available P, exchangeable K, and available S were 6.33 μg g
−1, 0.18 meq/100 g, and 12 μg g
−1, respectively. The plot was fertilized with 55, 160, 110, 95, and 10 kg ha
−1 urea, triple super phosphate, muriate of potash, gypsum, and zinc sulfate, respectively. All the fertilizers except half of the urea were applied as basal and the remaining half of the urea was applied 30 days after sowing (DAS). The fertilizers were uniformly incorporated into the plot before sowing seeds. The first experiment was conducted in 2015, while the second one was in 2018. In the first experiment, 50 soybean genotypes were used as planting material (
Table 2) from which two genotypes (AGS383 and BD2336) were selected as drought tolerant and used in the second experiment. Seeds of soybean were sown by hand in mid-January maintaining 30 cm from line to line and 5 cm from plant to plant spacing, and the crop was harvested in mid-April each year. In both years, crops were grown in control (80% of field capacity, FC) and drought (40% FC) conditions. The findings of our previous study showed that soybean cultivars used in this experiment were tolerant up to 40% FC. BARI Soybean6 was used as a drought susceptible check variety. Each experiment was laid out in a randomized complete block design with three replications.
2.3. Water Stress Imposition
Light irrigation was given after the sowing of soybean seeds. Most of the seedlings emerged within 3–4 DAS. Excess seedlings were thinned out after one week of emergence. Regular irrigation was applied with a hosepipe attached to a water tape both in control and drought plots up to the trifoliate stage (15 DAS) of soybean for seed germination and establishment of the young seedling. Drought treatments were imposed after the trifoliate stage of the crop. One day before treatment imposition, irrigation was applied to each plot to maintain the soil moisture content of all plots equally. Water stress condition was induced by withholding water until the wilting symptom was observed in plants. The wilting symptom in plants was visually observed every day.
To maintain 40% of FC, water was applied in each plot at the first appearance of wilting symptoms in plants. In general, water was applied after 3–5 days of the previous application. Before applying water, soil moisture content was measured using a soil moisture meter. During soil moisture content, 15 cm soil depth was considered. The soil of the experimental plot contains 30% soil moisture at FC. Thus, about 12% of soil moisture content was ensured through irrigation for maintaining 40% FC of the experimental soil. In the control treatment, water was applied to ensure 80% of FC by maintaining 24% soil moisture in the experimental plot.
2.4. Intercultural Operation and Harvesting Crops
Weeding and other cultural operations were done uniformly for the proper growth of the crop. Plant protection measures were taken by spraying admire @ 0.5 mL L−1 (Syngenta, Dhaka, Bangladesh). The crop was harvested when the plants attained full maturity.
2.5. Sampling and Data Collection
Data were collected on plant height, stem DM, and yield attributes from ten plants, and means were determined. Data were collected from the center of each plot to maintain data accuracy. For growth and DM, estimation sampling was done at 30 and 60 DAS. Five plants from each plot were sampled at the base. The plant parts were segmented into different components, such as leaf, root, nodule, stem, pod, and seed. The plant height was measured by a measuring scale (100 cm). The height was measured from the base of the cut plants to the tip of the shoot and the height of five plants was averaged. The total leaves of the collected five plants were counted and averaged for leaves plant−1. The leaf area was measured by an automatic leaf area meter (model: AAM-8, Hayashi Denko, Tokyo, Japan). To record DM of leaf, nodule, stem, and root, the plant parts were dried at 70 °C for 72 h.
2.6. Estimation of Proline and Malondialdehyde
Proline and malondialdehyde (MDA) content in the leaf of all soybean varieties grown in two water regimes was estimated at 60 DAS [
26]. Leaves were collected from each plot and immediately kept in an ice bag and brought to the laboratory. The 0.5 g of fresh weight (FW) of the leaf was taken for proline estimation and subsequently, proline was estimated. At first, a 0.5 g leaf sample was homogenized in 5 mL of 6% aqueous sulfosalicylic acid and centrifuged for 20 min at 4000 rpm. Two (2) mL supernatant was taken in a test tube with 2 mL of acid ninhydrin and 2 mL of glacial acetic acid and covered tightly with aluminum foil. The test tube was heated at 100 °C for 30 min and the reaction was terminated in an ice bath for 15 min. The reaction mixture was added with 4 mL toluene and mixed vigorously for 15–20 s. Keeping at room temperature for 10 min, the toluene layer was separated and absorbance was measured at 520 nm using a toluene blank. The proline concentration was determined from the standard curve and calculated on an FW basis as follows:
For MDA estimation, 0.5 g of fresh leaves are homogenized in 3 mL 5% trichloroacetic acid solution. The homogenate was centrifuged for 15 min at 15,500×
g at 4 °C. Then, 1 mL supernatant was added with 4 mL reaction mixture to the test tube and heated at 95 °C for 30 min in a water bath. After cooling down, the solution was centrifuged again at 15,500×
g for 10 min. Finally, the absorbance of the colored supernatant was measured at 532 nm and 600 nm. The MDA content was calculated on an FW basis as follows:
where A
532 = Absorbance reading at 532 nm, A
600 = Absorbance reading at 600 nm. The MDA concentration is calculated using the Lambert-Beer law with an extinction coefficient εM = 155 mM
−1 cm
−1.
2.7. Determination of Chlorophyll Content
At 60 DAS, Chl was determined on an FW basis extracted with 80% acetone using a double-beam spectrophotometer [
30]. The formulae for computing Chl
a,
b and total Chl were—
where D (663, 645) = Optical density of the Chl extract at a wavelength of 663 and 645 nm, respectively. V = Final volume (mL) of the 80% acetone with Chl extract and W = Weight of fresh leaf sample in g.
2.8. Measurement of Photosynthetic Traits
Photosynthetic traits such as Pn, transpiration rate (Tr), stomatal conductance (Gs), and leaf temperature were measured on young, fully expanded leaves in the same position on 60 DAS at full sunshine. The measurement was taken using a portable Pn system (Li-COR-LI-6400) assembled with an infra-red gas analyzer (Li-COR-LI-6250).
2.9. Estimation of Water-Related Parameters
The FW, turgid weight (TW), dry weight (DW), and area of leaves were recorded at each sampling time. The segmented plant parts were then dried in an oven at 70 °C for 72 h and weighed. Relative water content (RWC) was measured using fully expanded leaves of each variety under both control and water deficit condition. Immediately after cutting at the base of the lamina, leaves were sealed within plastic bags and kept in the ice box and quickly transferred to the laboratory. The FW of the leaf from each treatment was recorded just after removal. The TW was obtained after soaking leaves in distilled water in beakers for 24 h at room temperature and under low light conditions in the laboratory. After soaking, leaves were quickly and carefully blotted and dried with tissue paper for determining TW. The DW of the leaf was obtained after oven drying the leaf samples for 72 h at 70 °C. The RWC was calculated in the following equation.
Water saturation deficit (WSD), water retention capacity (WRC), and water uptake capacity (WUC) were calculated as follows according to Sangakkara et al. [
52].
The plant growth rate (PGR) was calculated by the following formula as
where W
1 = DW at time T
1 (30 DAS) and W
2 = DW at time T
2 (60 DAS).
Ten plants from each plot were sampled randomly for collecting yield component data. All pods of collected plants were counted and averaged. Pods having at least one seed were considered in this measurement. The seeds of the collected pods were separated and counted to determine seeds plant−1. After counting the seeds, weight was taken to determine the 100-seed weight. Plants from a 4.5 m−2 area were harvested for taking yield data at harvest. The plant was cut at the soil surface level. Threshing, cleaning, and drying of seeds were done separately and the weight of seeds was recorded plot-wise and adjusted at 12% moisture content.
2.10. Data Analysis
Comparison of genotypes based on a single morphological character is often inaccurate, artificial, and cumbersome, especially when a large number of genotypes and multiple characters for each genotype must be screened. However, using cluster analysis, genotypes can be scored on multiple parameters simultaneously. In the first experiment, all the collected data were converted to relative values, i.e., drought tolerance indexes before cluster analysis. The drought tolerance index was defined as the observations under drought divided by the means of the controls. Cluster analysis followed the methods described by Khrais et al. [
53]. Cluster group rankings were obtained based on Ward’s minimum variance cluster analysis on the means of the drought tolerance indexes for four parameters, i.e., stem DM per plant, pods per plant, seeds per plant, and seed yield. The cluster groups were identified in dendrograms. The cluster group rankings were obtained from the averages of means over multiple parameters in each cluster group, i.e., cluster mean, in order from highest to lowest averages. A sum was obtained by adding the numbers of cluster group ranking at each drought level in each genotype. The genotypes were finally ranked based on the sums so that those with the smallest sums were ranked as the most tolerant and those with the largest sums were ranked as the least tolerant in terms of relative drought tolerance. The collected data were analyzed location-wise each year. The ANOVA of different responses within the location was performed with the computer software package Crop Stat, version 7.2 [
54]. The pairwise treatments mean was compared with
t statistics at
p < 0.05. Graphical analyses were done using Excel software (Microsoft Corporation, Redmond, WA, USA).