Agronomic performance, seed chemical composition, and bioactive components of selected Indonesian soybean genotypes

: The use of improved soybean varieties is crucial when it comes to the progress of soybean cultivation in Indonesia. This study presents the agronomic performance, seed chemical composition, and bioactive components of 12 soybean genotypes grown in Malang, Indonesia. Parameters included the agronomic characteristics, ash, protein, fat, total ﬂ avonoid content (TFC), total phenolic content (TPC), genistein and daidzein contents as well as antioxidant activity. The study found that Dena 1, Dering 1, and Deja 1 had the highest seed yield (2.76 – 2.84 t/ha), while Dega 1 had the largest seed size (24.69 g/100 seeds). The black-seeded genotype of Detam 1 had the highest protein content (39.79% dw), while GH 63 had the largest amount of total daidzein and genistein. Detam 4 (a black-seeded genotype) exhibited the highest values of TFC, TPC, and antioxidant activity. The agronomic characteristics and TPC signi ﬁ cantly correlated, suggesting that both traits are applicable as criteria for soybean breeding selection. Detam 4 and GH 63 are likely promising to be used as gene sources for improving the nutritional and health bene ﬁ ts of soybean genotypes. GH 63 and GH 73 also have the potential for further release as new high-yielding varieties with early maturity and large seed size. Overall, the study provides valuable information on the agronomic superiority and nutritional aspects of improved soybean varieties in Indonesia.


Introduction
Soybean is a rich protein source and plays an important role in the Indonesian diet with the majority as tempe and tofu that constitute about 83.7% of the total soybean available for consumption, while the rest is used as an ingredient for soy sauce, soy milk, sprouts, and flour [1].The consumption level of soybean in 2019 was reported about 6.43 kg/capita/year, resulting in a large amount of national demand for soybean c.a. 3.09 million tons per year.However, about 2.67 million ton (86.4%) of such demand is being imported [2].Therefore, intensive efforts have been conducted to increase the national soybean production, particularly through the introduction of high-yielding improved varieties and adapted to different agro-ecological conditions in order to expand the planting area to all Indonesian regions.In fact, soybean yield increases may appear modest compared to the significant yield increases seen in cereals, which mainly store starch.Soybean seeds primarily store protein (38-42%) and oil (18-22%) with more intricate pathways than those in cereals [3], suggesting more efforts are needed for increased production of soybean.
A number of soybean-improved varieties have been released for the last 20 years with high yielding (>2 t/ha) and particular agro-ecological characteristics, such as adapted to low land, water logging, acid soil, tidal land, shading (intercropping), and drought as well as specific characteristics, such as resistant to major pests and pod shattering [4].However, in addition to such agronomic characteristics, information on the nutritional aspects as well as the quality of foods prepared from improved varieties has not been sufficiently available.Large yellow-seeded is likely favored for tempe preparation as it would give good-looking and high-volume development of the seeds during fermentation.Meanwhile, tofu and soy milk need soybeans with high protein content (>40%) for the ingredient rather than the seed sizes as it linearly affects the yield recovery [5].However, small-seeded soybeans are less desired nowadays; thus, most of the newly released varieties belong to medium-and large-seeded soybeans that are around 10-14 g/100˗seed and >14 g/100-seed, respectively [4].Soybean with black seed color and a high amount of protein is especially suitable for soy sauce preparation [5].Therefore, next to seed size criteria, protein content is also an important trait regarding the quality and nutritional aspects of the products.
Soybean seeds contain numerous bioactive components that are beneficial for human health, with phenolic compounds such as flavonoids and tannins being the most abundant.Phenolic compounds, including flavonoids, play numerous molecular and biochemical roles in plants, such as non-enzymatic antioxidants (free radical scavenging), signaling, mediating auxin transport, and plant defense [6].Soybean seeds also contain several isoflavones, with genistin, daidzin, and glycitin glycones being the predominant components while genistein, daidzein, and glycitein are present in smaller amounts [7].The isoflavone contents may vary from 745 µg to 5253.98 µg/g [8] and 835.9 µg to 2130.2 µg/g [9].Isoflavones belong to phytoestrogens that have biological activity as estrogen and antioxidant capacities that preserve cells from aging related to oxidative damage, avoid osteoporosis, coronary heart disease, cancer, and health problems related to menopause, improve fat metabolism, and enhance cognitive function [10,11].Similarly, flavonoid and phenolic compounds also exhibit antioxidant activity [12].A substantial amount of total flavonoid and total phenolic was measured in 63 Indonesian soybean genotypes, ranging 0.22-5.30mg CE/g and 3.48-14.72mg GAE/g, respectively [13], while a lower total phenolic (1.15-1.77mg GAE/g) was noted in 24 soybean genotypes originated from nine countries [14].
Soybean variety, cultivation, growing environments (climate, latitude, and longitude), maturity, processing, and storage are greatly attributed to the agronomic characteristics of the crop, physicochemical properties of the seeds as well as the bioactive components [14][15][16][17].This would consequently contribute to the crop yield, nutritional, health benefits, and quality of the soybean products.Soybean having superior characteristics of agronomic performance, nutritional values and bioactive components is a distinct advantage.However, information regarding such superiority in Indonesian soybeans is yet limited.Therefore, this study aimed to evaluate both the agronomical and nutritional aspects of selected Indonesian soybean genotypes grown in the same location.This finding would be valuable as well in assessing the relation between the traits of yield component responses and seed chemical components for further breeding selection purposes.

Plant materials
The research materials consisted of 12 soybean genotypes (10 improved varieties and two promising lines) as presented in Figure 1, while the variety description is listed in Table 1.They were grown in an irrigated lowland belonging to Iletri Jambegede Research Station in Malang, East Java, Indonesia (8°10′30″S, 112°33′32.4″E,altitude 335 m).The planting time was the 14th of July 2020 and harvested in the middle of October 2020 following the local crop rotation pattern of soybeans in the irrigated lowland (paddypaddy-soybean).The recorded weather conditions (monthly temperature, rainfall, and relative humidity) are presented in Figure 2. Before planting, the land was cleaned of the weeds and previous crop residues.The soil was taken at planting before the fertilizers were applied and analyzed in the Iletri Laboratory of Soil.
The soil tillage was optimally performed, followed by direct planting of the seeds.Each genotype was grown in a plot of 13.5 m 2 (4.5 m × 3 m) with a distance of 40 cm between rows and 15 cm in rows, and two seeds per hill, suggesting about 450 plants were available for each genotype and each plot/replicate.The drainage canals were made before planting and herbicides were applied.The application of fertilizer included 15.75 kg of N, 36 kg of P 2 O 5 , and 33.75 kg of K 2 O per ha as well as 5 t/ha of manure and 1 t/ha of organic fertilizer that were performed once at  the time of planting.Harvesting was done when the genotype reached the physiological maturity stage, which was evident by the presence of yellow or brown pods and fallen leaves.Postharvest handling included drying the soybean trunks, threshing, drying the seeds for 2-3 days, cleaning, and sortation.The seeds were then weighed for yield data collection.About 100 g of stock sample was randomly taken from the seed lot of each genotype and each replicate and divided into working and archive samples.The working sample was finely milled and used for the analysis of chemical composition in the Iletri Laboratory of Food Chemistry and Technology in Malang, Indonesia, and bioactive components in the Iletri Central Laboratory.

Experimental design
A randomized block design was used with a single factor/ treatment (genotype) and three replicates.A similar field replicate was also used during the analysis in the laboratory.Data collected were statistically calculated using an Analysis of Variance (ANOVA) and the differences between soybean genotypes were analyzed using an LSD test at a significant level of <0.05.

Evaluated parameters
Parameters covered the seed yield and yield performances, such as days to flowering, days to maturity (DTM), plant height, number of branches, number of productive nodes, number of filled and unfilled pods, 100-seed weight, and seed yield.Days to flowering, DTM, and seed yield were observed for all plants of each genotype in each plot/replicate, while other agronomic characteristics were collected from 10 random plants.The seed chemical composition included the parameters of ash content (gravimetry method using a muffle furnace) and fat content (direct Soxhlet extraction method) following Badan Standarisasi Nasional [18].About 1-2 g of sample was used for fat analysis.The sample was put into a dry thimble and subsequently attached to the holder in the Soxhlet extraction unit.The known weight of the extraction tube was filled with 25 mL of solvent (petroleum benzene) and coupled with the thimble holder; then, it was boiled or extracted for 15 min followed by rinsing for 30-45 min and distillation of the solvent for 15 min.Finally, the extraction tube was removed from the Soxhlet unit, dried in the oven at 105°C for about 30 min, transferred to a desiccator, and weighed.The fat content (%) was calculated using the differences in initial and final weights of the extraction tube.
Protein content was analyzed using a micro-Kjeldhal method referred to as AOAC [19]: About 15-20 mg of sample was added with 2.0 mL of H 2 SO4 and 200 mg K 2 SO4 (Kjeldhal tablet) and let be digested starting with the temperature of 200°C until it raised about 450°C, and the solution was finally clear or colorless and the fuming ended.A blank sample consisting of 2.0 mL of water was also undertaken digestion along with the sample.About 15 mL of 40% NaOH was added to both sample and blank, and subsequently steam distilled.Ammonia that was liberated from the distillation unit was collected in the Erlenmeyer containing solution of 4% boric acid, and the nitrogen contents were determined by titration with 20 mN HCl after blank correction.Protein content (%) was calculated using a conversion factor of 5.75 for soybeans.
Determination of bioactive components and antioxidant activity were performed as follows: Estimation of the total flavonoid was carried out following the previous study [20].Briefly, 0.5 g of soybean seed sample was extracted in 5 mL of 50% acetone, and the extraction was done twice.The extract was added with distilled water (1:5 v/v), followed by the addition of 150 µL of 5% NaNO 2 solution, 300 µL of aluminum chloride, 1,000 µL of 1 M sodium hydroxide, and subsequently adjusted to the final volume of 5,000 µL with distilled water.A spectrophotometer (Genesys 10 s, USA) was employed to measure the absorbance against the blank at a wavelength of 510 nm.The concentration was recorded as catechin equivalents per gram of sample (mg CE/g), which were calculated using catechin calibration curve.

Total phenolic content (TPC)
Measurement of the total phenolic in soybean seed had referred to Yusnawan [13].Similarly, about 0.5 g of soybean seed sample was extracted in 5 mL of 50% acetone, and the extraction was done twice.Distilled water was added to the extract (1:60 v/v), and then undergone a reaction with 250 µL of Folin-Ciocalteu's reagent and 750 µL of sodium carbonate.After incubation of the mixture, the extract was subjected to a spectrophotometer (Genesys 10 s, USA) and detected the absorbance at 765 nm.The TPC was presented as gallic acid equivalents per gram of sample (mg GAE/g) through a calculation using the gallic acid calibration curve.

Daidzein and genistein content
In this study, only two isoflavone components, namely daidzein and genistein, were analyzed due to the limited availability of the standards.The sample extraction process was accomplished using a method employed by Ginting et al. [16].The seed samples were finely ground, followed by maceration of 1 g sample with 20 mL of 50% methanol and shaking for 24 h at room temperature, and then filtered using a Whatmann paper No. 42.One milliliter of the sample extract was taken using a syringe and purified through a microfilter (0.45 μm, 33 mm i.d., SLHV033RS/Millex, Millipore, Merck) and gently put into a vial.The extract was then subjected to a UHPLC unit for separation and detection.The instrument belonged to Agilent Technologies 1290 Infinity II series, USA, which consisted of a quaternary pump (Agilent G7104A), a degasser (Agilent G1322A), an auto multi sampler (Agilent G7167B), and a Diode Array Detector (Agilent G1315D).About 10 μL of the analyte was injected onto a reversed-phase InfinityLab Poroshell 120EC-C18 column (75 mm × 4.6 mm i.d., 2.7 μm, Agilent, USA) equipped with a guard column of Zorbax Eclipse Plus-C-18 (12.5 mm × 4.6 mm i.d., 5 μm, Agilent, USA).The gradient elution was applied for daidzein and genistein separation using the mobile phase of acetonitrile containing 0.1% acetic acid (A) and water containing acetic acid 0.1% (B) as follows: 0 min 20% A, 3.3 min 25% A, 7.5 min 35% A, 10 min 40% A, and went back to 20% of A after 11 min.The flow rate of the mobile phase was 1.0 mL/min, and the time between injections was 1 min.Detection of the analyte was performed at 260 nm of wavelength and quantified using the calibration curves of both daidzein and genistein standards (Sigma, Aldrich, St. Louis, MO, USA).The individual content of daidzein and genistein was expressed in µg/g, and the total content was calculated as the sum of both isoflavone components.

DPPH assay for free radical scavenging activity
Antioxidant activity was estimated according to Yusnawan [13].About 0.5 g of soybean seed sample was extracted in 5 mL of 70% ethanol, and the extraction was done twice.The soybean extract was reacted with ethanolic DPPH solution (0.1 mM; 1:19 v/v) and incubated for 30 min, followed by reading the absorbance at 515 nm using a spectrophotometer (Genesys 10 s, USA).The antioxidant activity was calculated as micromoles of Trolox equivalent per gram of sample (µmol TE/g).

Weather conditions and soil chemical properties
Figure 2 shows that the amount of rainfall was considerably low from the beginning of soybean planting in July up to September.Therefore, watering the plant was performed once a week during this period as such condition was less favorable for soybean growth and development.The soil chemical properties obtained from the three replicates were almost similar (Table 2), except for the P 2 O 5 content, which was lower in replicate 2 relative to the other replicates.All the P 2 O 5 contents were much higher than the recommended dosage as well as for the K contents.This suggests that P 2 O 5 and K were sufficiently available for plant growth and development.In contrast, the N contents were considerably low which may cause low absorption of other elements, such as P and K.The CEC value in replicate 1 was lower than in other replicates; however, the Fe content was higher.Soils with high CEC retain more nutrients than low-CEC soils, reflecting that the nutrient availability is high.

Agronomic characteristics
The agronomic characteristics of soybean genotypes, particularly the days to flowering, maturity, and plant height, were significantly affected by replicate (Table 3), while other characteristics showed no effect of replicate.Conversely, the soybean genotype showed a very significant effect on all observed agronomic characteristics, with the exception of the number of unfilled pods (Table 3).This illustrates that the 12 genotypes studied were obviously different with respect to their agronomic characteristics.Days to flowering is the initial change from the vegetative to generative phase; thus, this is an important characteristic for plant growth and development.The days to flowering of 12 soybean genotypes varied from 30 to 43 days (Table 4).The earliest days of flowering were shown in Dega 1 and GH 73, while the longest one was seen in Deja 1. None of the genotypes had early days to flowering (less than 30 days).
Table 4 shows a significant variation on the DTM.Detam 1 had the longest day of maturity, while the earliest maturity was found in Dega 1.The genotypes with early days to flowering are normally followed by early DTM as well (Table 4).However, the genotypes with relatively longer days to flowering tended to have a more diverse range of maturity than those with early days to flowering.This can be observed in Detam 1, which had the longest days of maturity, even though the days to flowering were earlier than the other five varieties.In Indonesia, the soybean maturity is classified as early maturity genotypes (<80 days), medium maturity (80-85 days), and late maturity (>85 days).In this study, the shortest plant height was seen in Dega 1, followed by GH 73 and GH 63, which belong to early maturity.Dering 1, Deja 1, and Dena 1 which had medium maturity showed the highest plant height (Table 4).
The highest number of plant branches was obtained in Dering 1, followed by Detam 4. Dega 1, an early maturity genotype showed the least number of branches (1.93) due to a short vegetative phase of the plant.However, GH 63 and GH 73, which are also grouped as early maturity genotypes, could develop branches up to 3.5 and 3.4 per plant, respectively.Similarly, Anjasmoro that belongs to medium maturity did not have many branches compared to other medium maturity genotypes.Meanwhile, the average number of productive nodes was 19 per plant (Table 4).However, two genotypes showed more than 20 nodes per plant, namely Detam 1 and Dering 1.Meanwhile, the smallest number of productive nodes (9.7) was noted in Dega 1.In particular, the promising lines of GH 63 and GH 73 had 14 and 15 nodes per plant, respectively.
Detam 1, which had the highest productive node numbers, showed the highest number of filled pods.Likewise, Dega 1 which had the least productive node numbers also had the least filled pod numbers (Table 4).A similar finding occurred in two early maturity genotypes (GH 63 and GH 73), which had a smaller number of filled pods than those of medium and late maturity genotypes.The number of pods per plant is dictated by the interaction between the variety and the environment.The number of filled pods was not consistently followed by the number of unfilled pods, except for Anjasmoro, Detam 1, and Devon 1.Meanwhile, Dega 1 exhibited the smallest number of unfilled pods.
Seed yield per plot (t/ha) (YLD) varied significantly among the genotypes with the highest value achieved by Dena 1, Dering 1, and Deja 1, while the lowest yield was shown by Dega 1.These three genotypes belonged to medium maturity, while Dega 1 was categorized as early maturity.GH 63 and GH 73 genotypes, which also belong to early maturity, had higher seed yields than those of varieties with medium maturity, including Detam 1 and Devon 1. Growing environmental conditions may be attributed to such differences in seed yield.
The 100-seed weight represents the seed size of soybean, which is large (if >14 g), medium (10-14 g), and small (<10 g).This characteristic obviously had a different pattern compared to other agronomic characteristics (Table 4).Based on such criteria, eight genotypes belonged to largeseeded, and four genotypes were medium-seeded.The largest seed size was shown by Dega 1 (24.69 g/100 seeds), followed by GH 73, GH 63, and Dena 1, while Dering 1 and Detam 4 had the smallest sizes (Table 4).The large-seeded genotypes normally have early maturity, as seen in Dega 1, GH 73, and GH 63.

Chemical composition
The ash contents which reflect the amounts of minerals in soybean seeds were slightly different between genotypes that ranged from 5.62 to 6.20% (dw) (Table 5).The protein contents varied from 35.08 to 39.79% (Tabel 5).All genotypes had protein contents ≥35% and Detam 1 had the highest value, followed by Detam 4. A significant difference in the amounts of fat was obtained in 12 soybean genotypes that varied from 12.19% (dw) in Devon 1 up to 18.61% (dw) in GH 63 as shown in Table 5.Among 12 genotypes, only Devon 1 had fat content <15%; however, no genotype contained a high amount of fat (>20%).

Bioactive component
In the present study, the isoflavones were detected in the aglycone forms as daidzein and genistein (Figure 3).The daidzein and genistein contents and total (sum) of both isoflavones were significantly different between genotypes (Table 5).GH 63 possessed the highest daidzein and genistein contents with a total value of 338.61 µg/g (dw), followed by Devon 1 and Dega 1.Meanwhile, the lowest value was seen in Detam1, Dering 1, and Anjasmoro.
Two black-seeded genotypes, namely Detam 4 and Detam 1, had higher TFC than those of ten yellow-seeded genotypes (Table 5).No variation in TFC was observed in the yellow-seeded genotypes.Meanwhile, variations in TFC were noted between black-and yellow-seeded soybeans (Table 5), which showed genotypic variations based on the seed coat color.The black-seeded genotypes contained higher TPC than those of the yellow-seeded.Detam 4 gave the highest value of TPC (7.18 mg GAE/g), followed by Detam 1.Similarly, Detam 4 had the highest antioxidant activity compared to the yellow-seeded genotypes as well as Detam 1 (Table 5).The antioxidant activities in yellowseeded genotypes varied from 4.25 to 5.42 mol TE/g.

Genetic variability, heritability, and correlation
The genetic variability values of all agronomic characteristics studied were classified as narrow, except for the number of unfilled pods and seed yield (Table 6).In this study, most of the soybean genotypes (10) belonged to released varieties; thus, they might have narrow genetic variability.Broad genetic variability was seen in the number of unfilled pods as normally this characteristic is less considered in releasing a new variety.High heritability was observed in all agronomic characteristics with an exception for the number of unfilled pods per plant (POU; Table 6).This reflects that all characteristics were controlled more by genetic factors relative to environment factors.Based on such reason, if the genotypes are crossed each other, expectedly the offsprings would have high heritability, and the characteristic used as a criterion for selection can be simply predicted.A significant phenotypic correlation was exhibited by days to flowering (DTF) with all agronomic characteristics, except for POU, DTM with HIG and branches per plant (BRC), HIG with BRC and number of productive nodes per plant (NOD), NOD with number of filled pods per plant (POF) and POU, POF with POU, POU with YLD, and SEW with YLD (Table 7).In addition, a significant genetic correlation was also obtained for DTF with TT, BRC with SEW, as well as NOD with POF (Table 7).In the present study, YLD had a positive phenotypic correlation with DTF; however, a negative correlation was seen for YLD with POU and SEW.Meanwhile, no genetic correlation between YLD with other characteristics was noted.
The genetic variability of all genotypes was narrow, while the heritability was high (Table 8).The category for genetic variability in this study was calculated based on genetic standard deviation.The broad sense of heritability of chemical composition and bioactive components varied from 0.82 to 0.96.The lowest heritability was reached by fat content, and the highest value was reached by TPC.Table 9 shows the relation between seed chemical composition and bioactive components.There was no correlation between protein and fat contents.Meanwhile, daidzein positively correlated with genistein.However, both isoflavones negatively correlated with TFC.The latter was noted to have a positive correlation with TPC.
No genetic correlation was observed between the agronomic and seed chemical characteristics (Table 10).However, a phenotypic correlation was noted for TFC and TPC.In particular, TPC showed phenotypic correlations with all agronomic characteristics studied, except for unfilled pods.Similar phenotypic correlations were also seen for TPC with all characteristics, with the exception of POU.Meanwhile, TFC showed significant phenotypic correlations with flowering day, number of productive nodes, and 100-seed weight.

Discussion
The environment's optimal temperature and relative humidity for soybean growth and development are about 22-27°C and 78.5-81.6%,respectively [21]. Figure 2 shows that the relative humidity met the crop requirement; however, the temperature was relatively high.Elevated temperature may speed up the pod filling and increase the total number of pods per plant [22].
The significant effects of replication on days to flowering, DTM, and plant height might be caused by the differences in soil chemical properties as replicate 1 had lower CEC and higher Fe values relative to other replicates.The plant height in replicate 1 (57.3 cm) was lower than in replicate 2 (64.1 cm) and 3 (63.1 cm).This may be related to the higher amount of Fe in the soil, which may decrease  category for GCV/GSD; Broadmore than 2 GSD; Narrowbelow 2 GSD; DTFdays to flowering; DTMdays to maturity; HGTplant height (cm); BRCnumber of branches per plant; NODnumber of productive nodes per plant; POFnumber of filled pods per plant; POUnumber of unfilled pods per plant; SEW -100 seeds weight (g); YLDseed yield per plot (t/ha); PCVphenotypic coefficient of variation; GCVgenetic coefficient of variation; MSmean square; Vpphenotypic variance; Vggenetic variance; Veenvironmental variance; Hbsbroad sense heritability.
the plant height [23].A lower CEC value in replicate 1 may decrease the nutrient availability [24], resulting in extended days to flowering and maturity.
In general, days to flowering and maturity expose high heritability [25]; thus, differences in both characteristics due to crossed location may not be significant.In contrast to our findings, Kuswantoro et al. reported that the days to flowering and maturity of a particular genotype would be consequently different when grown in locations with considerable differences in temperature and rainfall [26].The heritability of the number of branches belonged to moderate, suggesting that genetic and growing environments considerably dictated the number of branches [25].The number of pods per plant is dictated by the interaction between the variety and the environment.The heritability of this characteristic in soybeans varied from moderate [26] to high [27]; thus, it might give different results from different studies.
100-seed weight obtained in the present study was relatively greater than those listed in the variety description, which was 22.98 and 14.3 g for Dega 1 and Dena 1, respectively [4].The 100-seed weight was significantly affected by genetic, growing conditions or environments, and interaction of both factors [30].A study conducted under 20 different environments proved that the heritability of 100-seed weight was 79%, indicating that most of the variation is related to genetic factors [30].A similar finding was also recorded for 146 soybean accessions grown in three eco-regions in China with a broad-sense heritability ranging from 69.93 to 95.99% [31].
No single genotype showed protein content >40%, including Detam 1 which was previously recorded to have high protein content (45.36% dw) in the variety description [4].Growing environment and planting time may contribute to such variations in protein content [32].The values obtained in this study were within the range of protein contents of 49 soybean genotypes (yellow and blackseeded) grown previously at the same location c.a. 34.72-40.54%(dw) [17].In terms of utilization for food products, 12 genotypes that contained protein >35% (dw) were tailored for ingredients of tofu, soymilk, and soy sauce as well as with large seed sizes for tempe [5].
The fat values studied were slightly smaller than most of the improved soybean varieties released in Indonesia that ranging from 13.06% to 20.48% (dw) [4] as well as 20 soybean genotypes adapted to acid soil c.a. 16.85-20.58%(dw) [16].Besides genetic factors, growing conditions, particularly temperature would linearly increase the amount of fat [33].Maturity and planting season or date in a particular region were also essential in determining the soybean composition, including fat [34].In addition to genetic factors, the ash content is likely dictated by the amounts of Ca, P, Na, K, Fe, and Mg, which are closely related to soil fertility conditions of soybean crops [35].The fact that replicates in this study significantly affected the ash content (data not shown) was in agreement with the differences in soil chemical properties, particularly the amount of Fe which is higher in replicate 1 than in other replicates (Table 2).
The total genistein and daidzein content obtained in GH 63 was considerably greater than those of Devon 1 and Devon 2, which were previously released as varieties containing considerable amounts of isoflavones [4].Detam 4, which was a black-seeded genotype, contained relatively higher total daidzein and genistein content relative to Detam 1. Genetic factor seems to be predominantly dictating the isoflavone content as both genotypes were grown at the same location, particularly the proportion of secondary metabolites in each variety.A similar phenomenon was also observed in Detap 1 as a new breed of Anjasmoro variety, which exhibited a total value of 1.8-fold greater than its parent (Table 5).The major role of genetic factors in isoflavone accumulation in soybean seed across environments is also highlighted by Pei et al. [36].
The mean of daidzein and genistein content found in this study was lower relative to those of seven soybean varieties originating from Vietnam c.a. 316 and 91 µg/g, respectively [37].However, the present study's results were higher than the values noted in four Brazil soybean genotypes c.a. 35.5-61.4µg/g and 26.2-44.4µg/g for daidzein and genistein, respectively [38] as well as in 20 promising lines adapted to acid soil that ranged from 0-10.9 µg/g for daidzein and 3.2-20.4µg/g for genistein [16] and in 218 Chinese soybean genotypes, ranging from 4.2 to 24.7 µg/g for daidzein and 4.2-31.5 µg/g for genistein [39].In particular, Daud et al. [40] reported a lower genistein content in Dena 1 variety (52.4 µg/g), while higher daidzein and genistein content was found in Devon 1 variety c.a. 191.1 and 170.4 µg/g, respectively [41].In addition to genetic variability, these differences may be due to growing conditions.High temperature and dry environment may cause a decrease in isoflavone content [42].
Both Detam 4 and Detam 1 exhibited slightly lower TFC relative to six Indonesian black-seeded genotypes (3.47 ± 1.50 mg CE/g) as reported by Yusnawan [13].The differences in seed coat colors may cause differences in TFCs among soybean seeds.The main naturally occurring flavonoid compounds in soybean seeds are anthocyanins, proanthocyanidins, and isoflavones.Anthocyanins and proanthocyanidins are predominantly condensed in black-seeded soybeans [43].Anthocyanins are not found in soybeans with yellow and green seed coat colors [43]; thus, isoflavones account for the largest proportion of TFC in these types of soybeans [43].The highest to lowest isoflavone contents in soybeans differing in seed coat colors are green, black, yellow, and yellowish green [44].No significant variation in TFC in yellow-seeded genotypes in this study may be due to compounds other than isoflavones that may also contribute to TFC.
Significant differences in TPC between black and yellow-seeded soybean genotypes as noted in the present study were also reported by some authors [18,45], and the highest value was seen in the black-seeded genotypes.TPC as a source of antioxidants is valuable for human consumption due to its beneficial health effects.The high TPC in black-seeded soybeans compared to yellow-seeded indicate that black-seeded soybeans contain high concentrations of bioactive compounds belonging to the phenolic groups.This reflects that the seed coat of soybean contributes more to TPC than the embryo and cotyledons [46].
Variations of TPC in black-seeded soybeans were also found in 20 genotypes grown in Korea with TPC ranging from 1.99 to 5.54 mg GAE/g [47] with the lowest TPC (1.99 mg GAE/g) much lower than Detam 1 (6.24 mg GAE/g).The range of TPC in the yellow-seeded genotypes (3.14-7.18mg GAE/g) was within the average value investigated in light yellow, yellow, and greenish yellow-seeded soybeans from Korea that was 4.69, 4.80, and 5.04 mg GAE/g, respectively [48].Methods of extraction and measurement of bioactive compounds may affect the different TPC in a particular sample in addition to the differences in soybean genotypes [20,48].
The antioxidant values of DPPH obtained in Detam 4 and Detam 1 (7.83 and 5.88 µmol TE/g) were lower than the antioxidant values of DPPH reported in black soybean extracts, which were between 10.99 and 20.38 µmol TE/g [18].However, a higher antioxidant activity in blackseeded genotypes compared to yellow-seeded genotypes was in line with the previous study, which was analyzed using the DPPH,2,2′-azino-bis (3-ethylbenzothiazoline-6sulfonic acid) (ABTS), and ferric reducing antioxidant power methods [49].Antioxidant activities in blackand yellow-seeded soybeans were positively correlated with anthocyanin and isoflavone contents as reported by Bursać et al. [45] and Dhungana et al. [50].Anthocyanins are predominantly present in black-seeded soybeans, while isoflavones exist predominantly in yellow-seeded soybeans [47,51].
Table 8 shows that all characteristics of chemical composition and bioactive components had a narrow genetic coefficient of variation (GCV), suggesting that 12 soybean genotypes exhibited a closed genetic relationship.This is interesting as they are primarily released for different trait purposes (derived from different parental genotypes) as listed in Table 1.A similar GCV confirmed in the present study may be due to most of the genotypes used being improved varieties.However, if we look at each characteristic, the variability occurred, such as high amounts of TFC and TPC in Detam 4 and high daidzein and genistein contents in GH 63.Thus, both genotypes can be used further as gene sources for increasing TFC and TPC (Detam 4) as well as daidzein and genistein contents (GH 63).
A previous study reported that the protein content of soybeans had a negative correlation with fat content [16].However, in present study, such a correlation was not obtained (Table 9).A narrow range of protein content as shown in Table 5 may be attributed to this finding.Both genistein and daidzein significantly affected the total isoflavone content.However, total daidzein and genistein content had a negative correlation with TFC.In fact, isoflavones belong to flavonoids [52], suggesting that the greater the value of total isoflavones, the lower the content of other flavonoid compounds.TFC showed a positive correlation with TPC, and both significantly influenced the antioxidant activity as seen in Table 9.The phenolic compounds are found to be primarily responsible for the antioxidant activity of soybeans [12].
A positive correlation obtained in this study between TPC and seed yield (Table 10) may suggest a possibility of such bioactive component contribution to the yield of soybean crops.Phenolic compounds are known to have an important role as a defensive mechanism in a plant against insects, microorganisms, and competing plants [53].
The seed size is essential regarding the end-use of soybeans.Yellow and large-seeded soybeans are favorable for the ingredient of tempe due to the bigger volume of tempe produced relative to those of small and medium-seeded soybeans [5].Based on the seed size criteria, eight genotypes belonged to yellow and large-seeded (including two promising lines of GH 63 and GH 73) were desired for tempe preparation, while 10 genotypes with yellow and large to medium-seeded were suitable for ingredients of tofu and soymilk and two black-seeded genotypes with medium size were tailored for soy sauce.This reflects that the 100-seed weight or seed size characteristic is highly of concern in releasing improved soybean varieties in Indonesia considering the preferences of soybean food processors and consumers.

Conclusion
This study reflects that soybean agronomic characteristics, seed chemical composition, and bioactive components were greatly dictated by soybean genotype.Regarding the seed coat color, size, and protein content criteria, eight, ten, and two studied genotypes were respectively tailored for ingredients of tempe, tofu, soy milk, and soy sauce.Significant phenotypic correlations of TPC with most agronomic characteristics suggest that both traits can be used as criteria for soybean breeding selection.Two high-yielding promising lines (GH 63 and GH 73) with large seed size and early maturity superiority are potential as soybean variety candidates adapted/ tolerant to those regions with water shortage conditions.Nutritional and health benefits of soybeans can be improved by using two genotypes, namely Detam 4 (black-seeded) and GH 63 (yellow-seeded) as the gene resources.Promising lines derived from both genotypes either the black-or yellow-seeded would enrich the farmers' and processors' choices of superior soybeans.
and Development for research facility support as well as Purwono, Sutrisno, Fajar, Lina Kusumawati, Ria Gustina, Yulius Eko Laksamana Samba, Amri Amanah, and Intarti for their technical and administrative assistance.

Figure 2 :
Figure 2: The weather conditions during the period of soybean cultivation (2020).

Table 1 :
Breeding status, physical, and specific characteristics of 12 soybean genotypes

Table 2 :
Chemical properties of soil collected from Jambegede Experimental Station, Malang, Indonesia (July 2020)

Table 5 :
Chemical composition and bioactive components of 12 soybean genotypes

Table 6 :
Genetic variability and heritability of agronomic characteristics

Table 8 :
Genetic variability and heritability of seed characteristics