Eighteen soybean genotypes which consist of 14 advanced breeding lines and four check cultivars were selected for this study. The genotypes were derived from several hybridizations obtained from the Indonesian soybean breeding program. The check cultivars were Detap 1 (large seed size and early maturity), Dega 1 (large seed size and early maturity), Anjasmoro (susceptible to the pod-sucking pest), and G100H (resistant to the pod-sucking pest). The experiment was conducted at the Indonesian Legumes and Tuber Crops Research Institute, which is located at Malang, East Java, Indonesia, during the dry season I and II 2022 (May to December 2022). While the climatic conditions in the experimental station were somewhat similar; we can include the data from one growing season.

The insect of N. viridula imago was collected from a soybean field at Jambegede Research Station (Malang, Indonesia) during the dry season I (May to August 2022), which is located at 8°10'30” South Latitude and 112°33'32.4” East Longitude. The type of soil was Inceptisol, elevation of 335 m above sea level, and C3 of Oldeman climate type. Adults of N. viridula were collected from mature soybean plants using a 15-inch-diameter sweep net. After collection, the insects were reared in the laboratory. Insects were reared in a mesh cage (length= 120 cm, width= 120 cm, and height= 100 cm) with a food source of fresh green beans Phaseolus vulgaris L. The first generation (F1) from the rearing of N. viridula imago was used in this study.

Soybean genotypes were grown in the screen house. Each genotype was planted in a plastic pot (Φ =18 cm) containing a 4:1 mixture of soil and manure, with two plants per pot. Planting is done following the days to maturity of each genotype to ensure that flowering occurs simultaneously. The pests and diseases were optimally controlled until 30 days after planting.

The resistance test of the soybean genotype against the N. viridula was conducted under the Free-Choice test (FC) dan No-Choice test (NC). The experiment of FC, as well as NC, was arranged as a randomized completely block design with three replicates. In the FC, 18 genotypes were enclosed in a large nylon mesh cage (length= 120 cm, width= 120 cm, and height= 100 cm). The infestation of N. viridula was done when plants reached the R5 stage (n=2 per plant or 36 adults N. viridula per cage). In the NC, each genotype was placed in a nylon mesh cage (50 cm in height and 26 cm in diameter) and arranged randomly according to replication. Each cage was then infested with a pair of newly emerged adult N. viridula when plants reached the R5 stage.

The data were collected on the agronomic traits in the NC as well as FC, consisting of plant height, number of branches/plant, number of nodes/plant, number of pods/plant, 100 seed weight, and seed yield/plant. The fully matured soybean pods (R8 stage) were used for the pod observation. The number of attacked pods was recorded. The pod damage intensity by N.viridula is calculated using the formula:

The level of resistance of the soybean genotypes to the pest N.viridula was classified using the Standard Deviation (SD) according to a method by Chiang and Talekar [24], as follows:

HR (Highly Resistant): ;

R (Resistant): ;

MR (Moderately Resistant): ;

S (Susceptible): ;

HS (Highly Susceptible):

Where:

I = Pod damage intensity of each genotype

= Mean of pod damage intensity

SD = Standard deviation

The damage intensity and agronomic data were subjected to the analysis of variance, and the mean differences were compared by LSD test (α = 5%). The relationship between pod damage intensity with the agronomic traits in the FC as well as NC was investigated using Pearson’s correlation and then visualized using the Corrplot and Hmisc packages of the RStudio program version 1.3.959 [25].

The resistance evaluation for the 18 soybean genotypes against the Southern Green Stink Bug (SGSB), which was performed under the Free-Choice test (FC) and No-Choice test (NC) showed a significant difference in the pod damage intensity and agronomic traits among genotypes (Table 1). The pod damage intensity of SGSB on FC and NC was significantly different, and each soybean genotype showed a range of damage intensity. The agronomic performance of each genotype in the FC and NC conditions was also significantly different, except for the number of branches. Agronomic performance among soybean genotypes differed significantly as well. This variation emerged as a result of the genetic backgrounds of the studied types differing, including the targeted agroecological variations [18].

The pod damage intensity caused by N. viridula in 18 soybean genotypes under the Free-Choice test (FC) and No-Choice test (NC) were presented in Figs. (1 and 2), respectively. The intensity of pod damage caused by SGSB ranged from 20.09 to 46.40% under FC, with an average of 31.96%, and from 25.63 to 67.63% under NC. This shows that the NC condition generates 38.42% more selection pressure than the FC condition. A previous study also obtained a higher damage intensity in the NC than in the FC [2]. This was reasonable because individual insect pests have the opportunity to choose their host plant in the FC, meanwhile, the NC forced the insect pest to feed the host plant, without other choices of hosts.

Fig. (1). The pod damage intensity caused by N. viridula in 18 soybean genotypes under Free-Choice test (FC). The color of the bar represents the resistance level: blue = resistant (R), green = moderately resistant (MR), yellow = susceptible (S), red = highly susceptible (HS).

Fig. (2). The pod damage intensity caused by N. viridula in 18 soybean genotypes under No-Choice test (NC). The color of the bar represents the resistance level: blue = resistant (R), green = moderately resistant (MR), yellow = susceptible (S), red = highly susceptible (HS).

Based on the pod damage intensity, the resistance level of 18 soybean genotypes to SGSB under FC (Fig. 1) was classified into four categories: three resistant genotypes, seven moderately resistant genotypes, four susceptible genotypes, and four highly susceptible genotypes. In the NC (Fig. 2), three soybean genotypes were classified as resistant, six genotypes were moderately resistant, six genotypes were susceptible, and three genotypes were highly susceptible. The check cultivar of G100H was susceptible to SGSB under NC but moderately resistant in the FC. A study reported that G100H was resistant to the pod-sucking bug R.linearis [7], and has moderate resistance to the pod pest of armyworm [26]. The G100H is the progeny from the IAC 100 and Himeshirazu cross [26]. Meanwhile, the IAC-100 has previously been found to be resistant to a variety of insects, including the neotropical stink bug complex (E. heros, N. viridula, and P. guildinii) [22].

In this study, different resistance reactions were exhibited by each soybean genotype in the NC as well as FC. Other studies also showed differences in resistance response to the stink bug among soybean cultivars [18, 27]. These differences in resistance levels between genotypes could be due to oxidative stress response and isoflavonoid production following an attack by stink bugs [3, 28], the digestive activity and organic compounds in watery saliva [12], different content of soluble leaf phenolics [29], cultivar differences [30], and the developmental stage of the plant [31] as well as the stinkbug [32]. According to a study, the pentatomid feeding habit, mouthpart morphology, and saliva all contribute to the high level of seed damage [12].

The evaluation for SGSB resistance in 18 soybean genotypes under the NC (Fig. 1) revealed a higher intensity of pod damage than the FC (Fig. 2). Three soybean genotypes (G2, G12, G16) were resistant to SGSB under the FC, meanwhile three other genotypes (G5, G10, G11) were resistant to SGSB under the NC. A study found that the resistance genotype with higher pod penetration resistance had a lower peroxide content after stink bug attack, and higher guaiacol peroxidase (GPOX), catalase (CAT), and superoxide dismutase (SOD) activities in seeds [3, 17]. Those three soybean genotypes that were classified as resistant (G2, G12, G16) in the FC were changed to highly susceptible, moderately resistant, and moderately resistant, respectively. Meanwhile, three soybean genotypes were identified as resistant in the NC, two of which changed to moderately resistant and one genotype to highly susceptible in the FC. Accordingly, the NC provides a higher resistance selection pressure than the FC, thus the soybean genotypes G10 (Anj/Rjbs-305) and G11 (Anj/Rjbs-306) were suggested as resistant genotypes to SGSB. Another study found genotypes IAC-100, V00-0742, V00-0842, and V99-1685 are less preferred by the southern green stink bug, and genotypes IAC-100, PI 558040, and V00-0870 were suggested as a source of genetic traits for resistance by reducing seed weight loss caused by southern green stink bug feeding [15]. The use of insect resistance genotypes could be an effective method to reduce the application of chemical pesticides and promote eco-friendly pest control and environmental protection.

Agronomic character performance differed between genotypes and resistance evaluation methods (FC and NC) (Fig. 3). In the FC, the average plant height was 33.80 cm, the number of branches/plant was 2.94, the number of nodes/plant was 14.56, the number of pods/plant was 33.80, 100 seed weight was 15.76 g, and seed yield/plant was 17.75 g. On the other hand, in the NC, the average plant height was 34.08 cm, the number of branches/plant was 3.20, the number of nodes/plant was 14.81, the number of pods/plant was 34.08, 100 seed weight was 14.28 g, and seed yield/plant was 9.95 g.

Fig. (3). The agronomic performance of 18 soybean genotypes under Free-Choice test (FC) and No-Choice test (NC): (a) Number of pods/plant, (b) Plant height, (c) Number of branches/plant, (d) Number of nodes/plant, (e) 100 seed weight, and (f) Seed yield/plant.

Yield is considered one of the most important traits of soybeans. A stink bug attack can lead to a major reduction in yield quantity and quality [12]. In this study, the susceptible genotypes had a low yield. Other studies reported that stink bugs result in fewer pods, fewer seeds per pod, smaller seeds, altered fatty acid compositions, and poorer quality soybean seeds [33, 34]. The soybean crop may also exhibit irregular growth, delayed maturity, and foliar retention as a result of a severe stink bug infestation [34]. Using their piercing-sucking stylets, stink bugs feed on soybeans by puncturing the seed pods and releasing poisonous digestive enzymes that induce tissue damage or even seed abortion [12].

A lower ratio of pod damage and seed yield are two major indicators of a plant's pest resistance [35]. The soybean genotypes G10 (Anj/Rjbs-305) and G11 (Anj/Rjbs-306) have low pod damage intensity, and a good agronomic trait, particularly for 100 seed weight and seed yield. Both of these genotypes produced large seeds, with only a minor difference in yield between FC and NC. The novelty of this research lies in the fact that in addition to being resistant to SBSB, the soybean genotypes G10 (Anj/Rjbs-305) and G11 (Anj/Rjbs-306) have the potential for high seed yield and are supported by another important characteristic, namely the relatively large seed size. Since the NC gave a higher selection pressure than the FC, the genotypes that have been shown to be resistant to NC may also develop high resistance in the field. The superior and resistant genotypes in this study will be continued to the next step of breeding. However, the evaluation of these improved breeding lines for insect resistance in field conditions for two years needs to be considered in further research.

The quantity and quality of soybean seeds determine the economic value of the soybean. The stink bug attack results in a reduction in seed quality and quantity. In this study, the pod damage intensity had a significant negative correlation with seed yield in the FC (r = -0.502*) as well as in the NC (r = -0.557*) (Figs. 4). Accordingly, using SGSB-resistant cultivars will prevent soybean seed yield losses.

In the FC, seed yield was positively correlated with PHG and SSZ. The PHG significantly supported the NOP, and NOP was supported by NOB and NON. In this case, it is demonstrated that the PHG plays an indirect role in seed yield. In the NC, seed yield was only significantly determined by the NON. However, the NON was positively correlated with NOP, and the NOP has a significantly strong correlation with the PHG. A study by Kuswantoro et al. [18] found that resistance to insect pest N. viridula was significantly correlated with all agronomic characteristics, except the number of branches per plant. Resistance to N. viridula was found to have a negative correlation with days to maturity, duration of the reproductive phase, number of unfilled pods, and weight of 100 seeds. Further study in pod-sucking bug R. linearis by Adie et al. [2] found a significant negative correlation between pod damage intensity and the number of pods/plant in the NC, but there was no significant correlation in the FC. In pecky rice, there was no significant difference in the sucking frequency of N. viridula on mature husks of CRR-99-95 W and the check genotypes. These findings indicate that CRR-99-95 W has no antixenosis effect against rice stink bugs [36].

Fig. (4). Heatmap of Pearson correlation coefficient for observed traits of 18 soybean genotypes under (a) Free-Choice test (FC) and (b) No-Choice test (NC). The different colors represent Pearson correlation coefficients, and the diverse circle sizes represent the P value. Blue color indicates positive correlations and red indicates negative correlations. The darker the color, the higher the correlation between the two traits, with a significant level being < ± 0.05. DI = pod damage intensity, PHG = plant height, NOB = number of branches/plant, NON = number of nodes/plant, NOP = number of pods/plant, SSZ = 100 seed weight (seed size), SYP = seed yield/plant.