Alleviation of drought stress in soybean by applying of biostimulant based on amino acids and macro-and micronutrients

Drought stress is one of the most predominant environmental factors hindering the soybean productivity. This study investigated the effects of an exogenous application of biostimulants, consisting of nitrogen, phosphorous, iron, copper, boron, manganese, zinc and amino acids, in the physiological, biochemical and productive responses of soybean cultivated under drought stress. Findings showed that applying 0.5 kg ha − 1 of the biostimulant improved soybean tolerance to drought. The biostimulant application maintained the leaf photosynthetic rate (A), stomatal conductance (g s ), transpiration rate (E), leaf temperature, water use e�ciency (WUE) and carboxylation e�ciency (CE), in addition to increasing the SPAD index. Moreover, the biostimulant heightened the activity of antioxidant enzymes superoxide dismutase (SOD), catalase (CAT) and ascorbate peroxidase (APX) and maintained the activity of the nitrate reductase enzyme. Regarding osmoprotectant, the biostimulant application enhanced proline accumulation, which could improve the soybean’s osmotic adjustment under drought conditions. In addition, foliar application of the biostimulant maintained the biometric and production characteristics, stem diameters, number of branches, number of pods with 1, 2 and 3 grains, and signi�cantly increased leaf area, number of primary stem nodes, dry matter mass in the aerial part and roots, ultimately increasing yield. Based on the aforementioned bene�cial properties, the biostimulant based on macro-and micronutrients and amino acids, particularly in the dose 0.5 kg ha − 1 , has proven to effectively relieve the adverse effects of drought stress in soybean.

meet the growing demand for food, it is necessary to increase soybean yield, even in environments with low water availability (Rosa et al., 2021).
In the photosynthetic process, the lack of water leads to deleterious effects on important enzymes, such as ribulose-1 5-bisphosphate carboxylase/oxygenase (RUBISCO), phosphoenolpyruvate carboxylase, phosphoenolpyruvate carboxylase, pyruvate phosphate dikinase, NADP-malate dehydrogenase and NADP-malic enzyme (Farooq et al. 2009a; Taiz and Zeiger 2017; Shukla et al. 2018). This is due to imbalances of molecules or ions in cells, particularly in oxygen reactive oxygen species (ROS) (Schmidt et al. 2018), since these molecules are highly unstable and have a high reaction capacity, mainly damaging lipids, proteins, nucleic acids, and affecting the cell physiology (Podgórska et al. 2017).
Under drought stress conditions, there is an increase in ROS levels in the apoplast, due to the activity of the NADPH oxidase enzymes in the respiratory burst of plants (Podgórska et al. 2017). For such, plants have a machinery of antioxidant molecules and enzymes that helps to mitigate the deleterious effects, particularly the oxidative stress (Das and Roychoudhury 2014). Among these, superoxide dismutase (SOD), catalase (CAT), peroxidase (POX), ascorbate peroxidase (APX) stand out, which act in the maintenance of homeostasis redox (Skukla et al. 2018).
An alternative to improve crop yield under water de ciency conditions is the application of biostimulants that act to protect plants, minimizing the adverse effects caused by environmental stresses (Calvo et al. 2014). Biostimulants are classi ed as products containing active ingredients capable of directly or It is important to understand how biostimulants affect plants under water stress in order to understand the speci c mechanisms of action. Therefore, the hypothesis of this study is that the use of biostimulants mitigates the effect of water de cit through the maintenance of plant metabolism, which is mainly re ected in productivity. To answer the question, biometric, physiological, biochemical and productive parameters were analyzed to investigate the capacity of a biostimulant based on essential nutrients and amino acids to mitigate the effects of water stress in soybean. Yellow Latosol (RYL), consisting of 61% clay, 18% silt, and 21% sand, and its nutritional characteristics were corrected and the physicochemical characteristics are shown in Table 1.  Fertilization occurred according to the chemical analysis for fertility purposes (Table 1) and recommendation for the cultivation of soybean (Raij et al. 1996). All tested treatments received a standard seed treatment with the recommended dose of Bradyrhizobium-based inoculant. In the sowing, 50 kg ha − 1 of simple super phosphate and 20 kg ha − 1 of potassium chloride were applied.
The adopted experimental design was casualized blocks, with six treatments and four repetitions. During the V4 growth stage of soybean cultivation, all treatments -except T1 -were submitted to a continuous water de cit of 50% of eld capacity until the moment of analysis.
The maintenance of the water requirements of the treatments was performed daily by the method of the soil water retention curve and weighing of the pots. Thus, water de cit was imposed by weighing the pots, saturating the sampling of pots with water, draining for 12 hours to reach the eld capacity (FC) and weighing again to determine the mass of water in this situation. From then on, and with the aid of a table of maximum soil retention capacity and of the equation:

W = Wfc − Wd
Where: W = water to be added to the pot (mL); Wfc = initial pot weight with soil moisture at eld capacity or 50% (g); Wd = daily pot weight (g).
The pots were watered according to the treatment, that is, 100% of the FC for treatments without water de ciency and 50% of the FC with water de ciency. As a result, daily weighing and rehydration of the pots were carried out so that they reached the desired levels again.
Treatments consisted of ve biostimulant doses. Foliar applications occurred during the R1 growth stage. Applications were carried out using a high-pressure backpack sprayer (CO 2 ) equipped with a spraying boom with two nozzles 0.5 m apart, with a spray volume of 200 L ha − 1 , constant pressure of 1.5 bar.
The biostimulant is formulated with mineral components and organic like macronutrients and micronutrients chelated with EDTA, and the amino acid glycine betaine according shown in the Table 2.

Determination of physiological variables
Physiological evaluations were carried out during the R6 phenological growth stage (plant stress peak) and consisted of the following variables: leaf gas exchanges, based on the net CO 2 assimilation rate (A), stomatal conductance (g s ), transpiration rate (E), leaf temperature (Tl) and intercellular CO 2 concentration (C i ), using an Infrared Gas Analyzer (IRGA) (LI-COR Biosciences Inc., Li-6400xt, Lincoln, Nebraska, USA), with measurements taken between 9:00 a.m. and 11:30 a.m., using the atmospheric CO 2 concentration, with room temperature and humidity, and constant photosynthetically active radiation (PAR) (1500 µmol photons m − 2 s − 1 ). Water-use e ciency (WUE) was calculated based on the A/E ratio, and carboxylation e ciency (CE) was calculated based on the A/C ratio. The SPAD index was measured through a portable chlorophyll meter (SPAD-502®, Minolta, Konica Minolta Sensing, Inc., Osaka, Japan).

Determination of antioxidant compound and enzymes
To analyze the antioxidant enzymes SOD, APX, CAT, POX, reductase nitrate (RN), and the non-enzymatic compound proline (Prol), samples were collected during the R6 phenological growth stage (plant stress peak).
For the activity of enzymes SOD, CAT, POX and APX, 300-mg samples of expanded leaves were milled in liquid nitrogen and added to a homogenization medium. The medium consists of a potassium phosphate buffer 0.1 M, pH 6.8, ethylenediaminetetraacetic acid (EDTA) 0.1 mM, phenylmethylsulfonyl uoride (PMSF) 1 mMe polyvinylpyrrolidone (PVPP) 1% (p/v). Next, homogenized samples were centrifuged in a refrigerated centrifuge (Hettich, Universal 320R, Tuttlingen, Germany) at 12,000 g at 4ºC for 15 minutes and the supernatant was used as crude enzyme extract. For Prol determination, 100 mg of leaf tissue were homogenized in 2 mL of sulfosalicylic acid 3% (p/v) and placed in the refrigerated centrifuge (Hettich, Universal 320R, Tuttlingen, Germany) at 6300 g for 10 min. Samples of 100 µL of the extract were added to 200 µL of acid ninhydrin solution (1.25 g ninhydrin, 30 mL glacial acetic acid, and 20 mL of phosphoric acid 6M) and the mixture was incubated at 100 ºC for 1 hour. The reaction was paralyzed in ice bath and supernatant absorbance was measured in a spectrophotometer (Shimadzu, UV-2700, Kyoto, Japan) at the wavelength of 520 nm. Absorbance results were compared to the standard curve of proline (0 to 100 µg mL − 1 ) (Bates 1973) and results were expressed in µmol proline g − 1 fresh matter (FM) −1 .
To determine RN activity, 200 mg of leaf sample was placed in a tube with penicillin and added 10 mL of the extraction solution; subsequently, plants were vacuum-incubated for 3 cycles of 2 minutes each. After incubation, samples were placed in water bath for another 30°C for 1 hour. Next, 1 mL of the extracted solution was collected and transferred to tubes, where 1 ml of the sulfanilamide solution and 1 mL of the N-Naphthyl solution were added; readings were made through spectrometry at 540 nm, in accordance with the methodology proposed by Jaworski (1971). NN was counted after washing the roots with water.

Determination of production components
The evaluations of production components were collected during harvesting stage, when grains had a humidity of approximately 13% and consisted in counting the average number of pods per plant (NPP), average number of pods with 1 grain (NP1), average number of pods with 2 grains (NP2), and average number of pods with 3 grains (NP3), and productivity (P). P (g plant − ¹) was obtained through the mass of grains measured in a precision analytical scale (Shimadzu, BL-3200H, Kyoto, Japan), adjusting humidity to 13%.

Statistical analysis
Results were submitted to variance analysis, polynomial reduction to assess product doses under water de cit and mean test to compare doses with the control without water de cit and without biostimulant application, at a level of 0.05 of probability. The non-signi cance of the regression deviation and/or higher value of the determination coe cient (R²) express the signi cance of parameters of the statistical model, using the statistics software SISVAR® (Ferreira 2014). Pearson's correlation analysis was performed with normalized data from the treatments adopted to verify the relationship among analyzed variables. Pearson's correlation heatmap was generated with software RStudio® (R Software (R Development Core Team)).

Physiologicfal variables
Highest A highest was observed without water de cit. However, under water de cit conditions, the dose of 0.5 kg ha − ¹ of the biostimulant reached similar A in compared to 0, 0.75 and 1.00 kg ha − ¹ doses. Under water de cit conditions and biostimulant application conditions, A results were adjusted to the quadratic model and increased by 34.75% up to a dose of 0.5 kg ha − ¹ compared to a dose of 0 kg ha − ¹, with subsequent reduction in larger doses (Fig. 2a).
There was no signi cant difference in g s between the different amounts of biostimulant tested, except between the dose of 0.25 kg ha − ¹ and 0 kg ha − ¹ without water de cit, with a reduction of 266.38% in g s (Fig. 2b).
The use of biostimulant under water de cit conditions increased C i by 197.96% with the dose of 0.5 kg ha − ¹ compared to the dose 0 kg ha − ¹ (Fig. 2c). However, there was no signi cant effect of biostimulant application on E (Fig. 2d), Tl (Fig. 2e) and WUE (Fig. 2f).  (Fig. 2g).
The relative chlorophyll content increased by 24% under the 0.5 kg ha − ¹ dose compared to the 0 kg ha − ¹ dose of the biostimulant in plants subjected to water de cit conditions (Fig. 2f).
3.2 Antioxidant compound and enzymes SOD activity was higher under application of 0.5 kg ha − ¹ under water de cit conditions, with an increase of 420% compared to plants that did not receive biostimulant and 86.57% compared to the control (Fig. 3a).
CAT activity increased 167.24% under application of 0.5 kg ha − ¹ of biostimulant and water de cit compared to 0 kg ha − ¹, but it did not differ statistically from the control and doses 0.25, 0.75 and 1.00 kg ha − ¹ (Fig. 3b).
Higher APX activity was observed under the application of 0.75 kg ha − ¹ of biostimulant, with an increase of 695.04% in relation to the 0 kg ha − ¹ dose, but with no signi cant difference with the 0.5 kg ha − ¹ dose (Fig. 3c). While the POX activity was not in uenced by the evaluated treatments (Fig. 3d).
Under water de cit, RN activity increased by 134.15% with application of 0.5 kg ha − ¹ compared to the dose of 0 kg ha − ¹, however, it did not differ from control and dose 0.25 kg ha − ¹ (Fig. 3e).
Higher accumulation of Prol was observed at the dose of 0.5 kg ha − ¹, with an increase of 105.79% in relation to the dose of 0 kg ha − ¹, but it was similar to the dose of 0.75 kg ha − ¹ (Fig. 3f).

Biometric components
The highest PH was seen in the control without water de ciency. Under water de cit conditions, the highest plant height was observed in the biostimulant dose of 0.5 kg ha − ¹, which did not differ from the doses of 0 kg ha − ¹ and 0.75 kg ha − ¹ (Fig. 4a).
SD was not affected by treatments (Fig. 4b). However, NB reduced as the biostimulant dose increased. Highest NB was observed in the sample with no water de cit, which did not differ from samples under water de cit and biostimulant application from dose 0 to dose 0.75 kg ha − ¹. Smallest NB was observed in the dose of 1.0 kg ha − ¹ (Fig. 4c).
The largest LA was observed under dose of 0.5 kg ha − ¹, with increase of 278.75% compared to the dose of 0 kg ha − ¹ (Fig. 4d). While NN was highest under application of 0.5 kg ha − ¹, with an increase of 90% compared to dose 0 kg ha − ¹ and 73% compared to the control (Fig. 4e).
SDM under dose of 0.5 kg ha − ¹ under water de cit conditions increase of 66.35% compared to plants that did not receive biostimulant, with subsequent reduction at higher doses, and an increase of 44.74% compared to the control (Fig. 4f). This performance was also observed in RDM, however the increase in this case was 26.33% in the dose of 0.5 kg ha − ¹ compared to plants that did not receive biostimulant, and of 12.00% compared to the control (Fig. 4g).

Production components
NPP was negatively impacted by water de cit conditions even with the application of the biostimulants, thus the control treatment presented higher NPP (Fig. 5a). The NP1 was not affected by treatments, resulting in an average of 2.11 pods plant − ¹ (Fig. 5b).
NP2 increased 105.18% with the dose of 0.5 kg ha − ¹ compared to the dose of 0 kg ha − ¹, however, it did not differ of the control and dose 0.25 (Fig. 5c). This tendency was also observed in NP3, with increase of 65.09% compared to the dose of 0 kg ha − ¹, but without signi cant difference in relation to control, doses 0.25 and 0.75 kg ha − ¹ (Fig. 5d).
P increased 22.15% under 0.5 kg ha − ¹ of biostimulant, compared to dose 0 kg ha − ¹, in addition to 19.55% compared to the control (Fig. 5e). There was a greater correlation among productivity and LA, NN, SOD, SDM, RDM, SPAD index, C i and CAT (Fig. 6).

Discussion
Photosynthesis is one of the processes most impacted by drought stress ( Our results demonstrate that the use of biostimulant similarly in uenced the activity of POX in all tested treatments, suggesting a greater effect of the product on the other enzymes of the antioxidant complex mentioned above. In fact, POX was more correlated to Prol (Fig. 6).