Experimental Site and Soil Preparation
To test our hypothesis, two independent experiments were carried out: one conducted in soil and the other in a nutrient solution. The cultivations were carried out in a greenhouse at the College of Sciences and Engineering of the São Paulo State University, Tupã municipality, Brazil following a completely randomized experimental design.
For analysis of chemical characteristics, 20 soil subsamples from the first 20 cm depth were collected in the experimental area, the material was homogenized, and the following chemical characteristics were determined following Silva et al. (2019) showing chemical characteristics: P mg dm-3 = 4.0; organic matter mg dm-3 = 13.0; pH CaCl2 = 4.6; K mmolc dm-3 = 1.4; Ca mmolc dm-3 = 4.0; Mg mmolc dm-3 = 4.0; H + Al mmolc dm-3 = 18.0; Al mmolc dm-3 = 3.0; SB mmolc dm-3 = 9.4; CTC mmolc dm-3 = 27.4; V % = 34.0. For the soil-based experiments, the soil was first collected and prepared using 2.9 g of limestone per each 5 kg soil per pot, 30 days before the sowing to reach 70% base saturation. In addition, each pot received 0.124 g simple superphosphate, 0.25 g potassium chloride, 0.058 g boric acid, 0.083 g copper, and 0.25 g zinc
Readily available soil Se concentration was 3.6 µg kg-1. To determine Se concentration, 4 g of air-dried soil was added to a 50 mL centrifuge tube along with 20 mL of 0.01 mol L-1 KNO3. The solution was shaken in a rotary shaker for 2 h, then centrifuged for 30 min at 1650×g. After centrifugation, 9 mL of the supernatant was pipetted into a < 0.22 µm syringe filter. The supernatant was filtered into a tube containing 10% tetramethylammonium hydroxide (TMAH) and 0.1 mol L-1 KH2PO4 for determination of the readily available Se using inductively coupled plasma-mass spectrometry (ICP-MS; Thermo Fisher Scientific iCAPQ, Thermo Fisher Scientific, Bremen, Germany). Irrigation was performed daily keeping at 80% of soil water retention capacity.
Cultivation
Soybean cultivar BRS 1003 PRO was herein employed. The seeds were treated with pyraclostrobin, thiophanate-methyl, and fipronil, and inoculated with Bradyrhizobium japonicum SEMIA 5079. For the soil-based experiment, eight seeds were sown per pot. After the emergence of seedlings, thinning was performed, and two plants per pot were left. Weed control was carried out manually. Pyrethroid insecticide was used for insect control.
For the hydroponics experiment setup, the seeds were germinated in a sand substrate for 14 days, then transferred to a nutrient solution with a low-N (107 µM) containing 30 µM KNO3, 24 µM Ca(NO3)2·4H2O, 8 µM NH4NO3, 6 µM (NH4)2SO4, 1.2 mM CaCl2·2H2O, 300 µM K2SO4, 50 µM KH2PO4, 1.5 µM MnSO4·H2O, 1.5 µM ZnSO4·7H2O, 0.5 µM CuSO4·5H2O, 0.15 µM (NH4)6Mo7O24·4H2O, 2.5 µM NaB4O7·10H2O, 40 µM EDTA-Fe, and 0.1 µM CoCl2·6H2O, and 50 µM MgCl2·6H2O (Peng et al. 2018). After 1 week, plants were transferred into a 1/10 N (531 µM) nutrient solution as described by Peng et al. (2018). The solution was renewed every 5 days, and the pH was adjusted to 6.5 using hydrochloric acid. At each renewal of the solution, 2 mL of inoculant containing Bradyrhizobium japonicum SEMIA 5079 was added for each liter of solution.
Se application
When the plants reached the V2 growth stage, five Se doses (0, 7.5, 15, 30, and 45 µg kg-1) in pot experiments and four Se concentrations (0, 5, 10, and 15 µmol L-1) in hydroponics were applied using sodium selenate (Sigma-Aldrich) as Se source.
A pot experiment was carried out using eight independent biological replications, of which four were used for destructive analysis at flowering, and four were carried out until grains maturation to estimate soybean yield. Hydroponic-based experiments contained four biological replications that were conducted until flowering for nodule investigation and flavonoid analysis.
Quantification of antioxidant metabolism and photosynthetic pigments
At flowering for both experiments, the diagnosis leaves were collected and macerated in liquid nitrogen and stored in a freezer at -80 ºC until processed. Lipid peroxidation was based on malondialdehyde (MDA) concentration as described by Heath and Packer (1968). The concentration of H2O2 was performed according to Alexieva et al. (2001). Besides, the material preparation for total proteins determination, as well as SOD (EC:1.15.1.1), CAT (EC:1.11.1.6), and APX (EC:1.11.1.11) activity were performed as described by Gomes-Junior et al. (2007). The quantification of total proteins, SOD, CAT, and APX activity were performed as described by Bradford (1976), Giannopolitis and Ries (1977), Azevedo et al. (1998) and Moldes et al. (2008), respectively. Finally, the photosynthetic pigments quantification was performed at the flowering stage and according to Lichtenthaler (1987).
Determination of N-compounds in leaves and nodules
Leaf and nodule concentration of total sugars, amino acids, nitrate, ammonia, and ureides (allantoin and allantoic acid) was determined. For extraction and determination of total sugar, amino acids, nitrate, ammonia, and ureides, 1 g of fresh tissue was added into 10 mL of MCW solution (60% v/v methanol, 25% v/v chloroform, and 15% v/v H2O). After 48 h extraction in the refrigerator, 4 mL of the supernatant was collected and 1 mL of chloroform and 1.5 mL of deionized water were added, and the mixture was diverted to another tube. After phase separation, total sugars, amino acids, nitrate, ammonia, and ureides were determined according to DuBois et al. (1956), Yemm and Cocking (1955), Cataldo et al. (1975), McCullough (1967) and Vogels and Van Der Drift (1970), respectively. The total concentrations of sugar, amino acids, nitrate, ammonia, and ureides were based on the standard curve of sucrose, methionine, calcium nitrate tetrahydrate, ammonium sulfate, and allantoin, respectively, and the results were expressed in µmol g− 1 FW (fresh weight).
Nitrate reductase activity was performed as described by Reis et al. (2009). All measurements herein described were carried out using an ultraviolet-visible spectrophotometer (SP-220, BioespectroTM, Brazil). All quantifications were performed using external standard-based calibration curves.
Determination of flavonoids in roots
To determine the flavonoids, i.e., rutin, genistein, daidzein, kaempferol, and naringenin, roots tissues were splatted from the shoots and dried in an oven with forced air circulation at 60 ºC until constant weight. Thereafter, the roots were macerated, and 0.1 g of material was weighed for each replication and mixed with 2 mL of 70% v/v methanol. Afterward, the samples were exposed for 30 minutes to sonication. Finally, the samples were centrifuged at 5000 rpm for 20 minutes and filtered in vials. Flavonoid readings were performed in a high-performance liquid chromatography (HPLC) system. To determine the results, a standard solution containing known concentrations of daidzein, genistein, kaempferol, rutin, and naringenin was used.
Determination of biometrics and yield components
At grain maturation, the plants were collected, divided into shoots and roots, and placed in an oven with forced air circulation at 60 ºC until constant weight to determine the dry weight of shoots and roots. Furthermore, the plant height, the number of pods, grains, the weight of 100 seeds, and soybean yield per pot were determined.
Statistical analysis
The results were first subjected to the Shapiro-Wilk normality test and after analysis of variance (ANOVA) followed by Tukey’s mean comparison test at p < 0.05. To verify the correlation between the variables, Pearson's correlation was performed. The R software version 4.0.3 was used.