Screening of soybean genotypes for the source of soybean mosaic resistance

: Soybean mosaic virus (SMV) is the major constraints for soybean cultivation in many parts of the country. Experiments were carried out to study the seed transmission of SMV and to identify the resistant sources through screening of 30 promising soybean genotypes obtained from different sources under natural infections conditions in between 2000 and 2001. Indirect-ELISA was performed against TRSV, TMV, CPMV, CMV, BBSV, BBTMV and SMV in leaf materials, both from healthy and diseased plant materials and the results showed that the seeds and the plants in the field were free from those 6 viruses. The highest seed transmission was found in Gaurab (15.07%) followed by G-2120 and the lowest (1.5%) seed transmission was found in TG-893 followed by BS-32, CM and AGS-129. Seed mottling was found related to seed transmission. No disease incidence was found in AGS-129 but there were seed mottling (1.25%). On the other hand, genotype AGS- 160 was free from mottled seeds but produced infected seedlings. Mosaic symptom became apparent 15 days after sowing (DAS), highest at 60 DAS and disappear after 90 DAS. Seeds from seed lot containing up to 20% mottled symptom could not hamper production. Screening of 30 soybean genotypes based on the natural infections revealed that AGS-129 was identified as resistant and grading 1, 10 were identified as moderately resistant (MR) considered as grading 3 and the remaining genotypes were susceptible (S) and moderately susceptible (MS) considered as grading 7 and 5, respectively. The 10 genotypes viz . CM, BS-32, ACAGS-154, G-2261, AGS-129, Durga, Williams, AGS-160, EC-1178 and PR-164 could be used for further study to locate resistant genes against soybean mosaic virus and varietal improvement in breeding purpose. Tobacco (TRSV), Tobacco mosaic virus (TMV), Cowpea mosaic virus (CPMV), Cucumber mosaic virus (CMV), Broad bean strain virus (BBSV), Broad bean true mosaic virus (BBTMV). sample plant per plot were counted and averaged. Average pod length was collected from the harvested pods of ten randomly selected plants. The pod length was expressed in cm. The average number of seed per pod was determined by counting seeds derived from randomly taken fifty pods per sample handled above. One hundred seeds were taken randomly from the seed lot of each plot and weighted in gram using an electric balance. Recorded from date of sowing to date when most of the plants of a plot were ready to harvest. Weight of the total grains of the sample plants in a plot were taken and averaged to determine the yield per plant data. Five healthy looking and five symptom bearing plants were randomly selected from each of the unit plot. The yield of each of the plant was recorded and finally the yields per healthy looking and symptom bearing plant were calculated. Weight of the total harvested grains per plot was calculated in grams.


Laboratory Experiment 2.1.1. Seed health test
Detection of viruses on leaves and seed samples were performed in the Virology Laboratory, Professor Golam Ali Fakir Seed Pathology Laboratory, Department of Plant Pathology, Bangladesh Agricultural University (BAU) through Indirect-ELISA. Ten samples of symptom bearing seed and ten symptom bearing leaf samples of different genotypes were used to detect viruses. Indirect-ELISA was performed against the following viral antisera: Tobacco ring spot virus (TRSV), Tobacco mosaic virus (TMV), Cowpea mosaic virus (CPMV), Cucumber mosaic virus (CMV), Broad bean strain virus (BBSV), Broad bean true mosaic virus (BBTMV).

Dry inspection of seeds and transmission study of SMV by seedling symptom test
Sorting of mottled seeds were done from 400 randomly taken seeds from the seed lot and counted the symptom bearing seeds and healthy-looking seeds in the laboratory. The percentage of symptom bearing seeds in the seed lot for every genotype under study was calculated. In addition, 100 seed weight of symptom bearing seeds and healthy-looking seeds were taken by an electric balance and calculated out the reduction of seed weight (%) due to SMV infection. The experiment was conducted in the net house and the soil was collected from BAU, Mymensingh. After drying the soil, decomposed cow dung was mixed with collected soil (1:1) and earthen pots of 30 cm dia. were filled two third portions with the mixture. Chemical fertilizers were not used in the pot soil. The pots were arranged in RCBD with 3 replications. The soybean seeds were sown on December 15, 2000. Twenty-five seeds were sown in each pot. As check, five pots were planted with only symptom bearing seeds and three pots with apparently healthy seeds, sowing 10 seeds per pot. Irrigation was done regularly in the pots by supplying water in the trays on which the pots were placed in the net house. The data recorded for each variable was averaged to obtain mean values and analysis of variance was performed using these mean values. Duncan's Multiple Range Test (DMRT) was performed for all the variables to locate the difference between them following Steel and Torrie (1960). Multiple regression model was also used to estimate the disease incidence, yield per plant, yield of healthy-looking plant, yield of symptom bearing plant and yield per plot. For the estimation of disease incidence, the multiple regression model was: Yi = a + ∑bjXij + ei Where, i =  X=1,2,3,4,5,6,7,8,9,10,11 and 12 represents Disease incidence, Plant height, No. of effective nodules per plant, Pod/plant, Pod length, Seeds/pod, 100 seed weight, Days to maturity, Germination (%), Seed transmission (%), Seed infection (%), Reduction in seed weight (%) respectively. In addition, a = Constant; b = Regression coefficient; c = Random error distributed as N (0, 2).

Field Experiment 2.2.1. Experimental site and layout
The field experiment was conducted at the experimental field laboratory of the Department of Genetics and Plant Breeding, Bangladesh Agricultural University, Mymensingh during the period from November 2000 to April 20. Urea, Triple Super Phosphate (TSP), Muriate of Potash (MoP), Gypsum, Zinc Sulphate were applied to @ 40, 120, 60, 25 and 5 kg/ha respectively. All fertilizers except half portion of urea were applied at the time of final land preparation. The remaining half of urea was applied as top dressing at vegetative stage. Rhizobial inoculum was mixed with the seeds @ 25g/Kg seeds just before sowing. The entire field was divided into 90unit plots for this study. The plots were arranged following Randomized Complete Block Design (RCBD) with 3 replications. The plot size was 3m × 2m with 10 rows of 3 m long each. For each material there were ten rows of plants having 30 cm distance between the rows and aprox. 5 cm distance between the plants. The distance between the blocks were 50 cm. The genotypes were designated as treatment.

Intercultural Operations and Harvesting
Thinning out of seedlings was done 15 days after sowing (DAS) after taking account of the symptom bearing seedlings. Two times weeding was done in the crop field, one at 30 DAS and another at 50 DAS. No irrigation was done in the field. The plants were harvested at the physiological maturity stage. Such maturity came with yellowing of leaves with completion of leaf shedding and the pod color mostly became dark brown. The varieties were harvested at different dates as they reached maturity. Plants were harvested leaving the root part in the field.

Data collection 2.3.1. Disease incidence and Disease severity
To record the incidence of mosaic diseases in the experimental plots 6 times regular inspection of the plots was made at 15 days interval starting from 15 DAS. Three lines from each unit plot were selected randomly and number of symptoms bearing plants from the total plants were recorded each time. The data were averaged for each plot and finally the result was expressed as % plant infection indicating the incidence of disease. Soybean mosaic virus disease severity was recorded at 60 DAS (flowering stage) following a 1-9 disease rating scale used for scoring MYMV (Mughbean Yellow Mosaic Virus) by Singh et al. (1988). Ten affected plants of each unit plot were randomly selected for collecting the data on diseases severity. Mottling of leaves on 0.1-5% and 5.1-10% of the plants were found resistant and moderately resistant respectively. Meanwhile, mottling and yellow discoloration of leaves on 10.1-25% and 25.1-50% of the plants were moderately susceptible and susceptible respectively. Severe yellow mottling on more than 50% and up to 100% of the plants were found to be stunted of plants as well as failure of flowering and fruit set occurred in highly susceptible plants. To investigate the mosaic disease progress in the field average of date wise disease incidence was made. From this data a disease progress curve was prepared.

Yield and yield contributing characters
Ten plants were randomly selected from each unit plot except the border lines for collection of data and the plants were tag-marked. Data for the designated yield and yield contributing characters were taken on plot and individual plant basis. The plant heights were measured from ground level to the tip of the main stem after harvest. It was expressed in cm. Number of effected nodules were counted from ten randomly selected plants of each plot and were averaged over per plant. Both fertile and empty pods from each of the sample plant per plot were counted and averaged. Average pod length was collected from the harvested pods of ten randomly selected plants. The pod length was expressed in cm. The average number of seed per pod was determined by counting seeds derived from randomly taken fifty pods per sample handled above. One hundred seeds were taken randomly from the seed lot of each plot and weighted in gram using an electric balance. Recorded from date of sowing to date when most of the plants of a plot were ready to harvest. Weight of the total grains of the sample plants in a plot were taken and averaged to determine the yield per plant data. Five healthy looking and five symptom bearing plants were randomly selected from each of the unit plot. The yield of each of the plant was recorded and finally the yields per healthy looking and symptom bearing plant were calculated. Weight of the total harvested grains per plot was calculated in grams.

Laboratory experiment 3.1.1. Seed health test
Indirect-ELISA confirmed that all genotypes were from TRSV, TMV, CPMV, CMV, BBSV and BBTMV (Table 2). Unfortunately, antiserum for SMV was not available in the Plant Virology Laboratory, Professor Golam Ali Fakir Seed Pathology Centre, Bangladesh Agricultural University, Mymensingh during the study (2000)(2001). Thus, SMV was confirmed by ELISA at DGISP at Copenhagen, Denmark. The results of ELISA carried out with dried leaf materials; both from healthy and diseased plant confirmed the presence of SMV in different genotypes.

Percentage of SMV symptoms bearing seeds in different genotypes
Among 30 genotypes highest (64%) symptoms bearing seed was found in BS-14 followed by THINUNG-154 (28.25%), G-2261 (26.75%), GAURAB (24.25%) while AGS-160 showed no symptoms bearing seed followed by AGS129 (1.25%), PR-164 (1.25%), DURGA (1.75%), BS-16 (2%) and remaining others genotype revealed lower moderate level of symptoms bearing seeds. Above mentioned results were calculated based on the data obtained from number of symptom bearing seed and number of healthy looking seed (Table 2). In this experiment, results of ELISA tests showed that the seeds and the plants in the field were free from TRSV, TMV, CPMV, CMV, BBSV and BBTMV. However, visual symptoms were confirmed by indirect ELISA and results from Andayani et al. (2011) indicated that soybean plants infected with SMV produced mosaic symptom as shown by positive reaction in indirect ELISA with SMV antiserum (Table 2).

Performance of soybean genotypes depending on seed germination, transmission, infection and reduction in seed weight
Considering germination percentage, BS-14 was recorded highest (98.33%) germination followed by Acadian (97.66%), Williams (97.66%), Sau-Luis (96.67%), G-2261 (96.67%), G-2120 (96.67%), ACAGS-154 (96.67%), EC-1178 (95%) and Gaurab (95%) exhibiting statistically similar. BS-29 showed lowest (36.67%) germination while others were in moderate amount of germination (Table 3). With a view to percent seed transmission, highest (15.07%) seed transmission was found in Gaurab which was statistically similar to G-2120 and the lowest was found in TG-893 (1.5%) which was statistically similar to AGS-129, BS-32, and CM (Table 3). According to Balgude et al. (2012), SMV was accounted for 6-8% seed transmission in soybean, whereas in our case seed transmission was found 1.3 to 16%. Despite the presence of mottled or symptom bearing seeds (1.25%) in the working sample of AGS-129 (Table 6) there was no disease incidence in the field. The reason is that when the seed coat is only infected (producing typical mottling symptom) the seed transmission to seedling does not take place as Khetarpal et al. (1992) showed, to transmit from seed to seedling, an embryo must be infected. On the other hand, AGS-129 may be a very good tolerant genotype, not really resistant. That's why it did not produce any symptom. A positive correlation was found between the percentage of symptom bearing seed and seed transmission indicating one percent increase in symptoms bearing seed leads to 0.075 percent seed transmission, when symptoms bearing seed was more than 60% then seed transmission was 19.314 ( Figure  2E). Moreover, a negative correlation was found between the percentage of symptom bearing seed and seed germination revealing one percent increase in symptoms bearing seed accounted for 0.160% reduction of germination ( Figure 1A). In terms of percent seed infection, BS-29 was responsible for highest (64%) infection while minimum seed infection was found in AGS-160 followed by AGS-129 (1.25%), PR-164 (1.25%). The highest reduction in seed weight was recorded in Gaurab (56.65%) and the lowest seed weight reduction was found in AGS-160 followed by PB-1 (0.36%), others genotypes were recorded moderate to lower reduction in seed weight. Negative correlation was observed between the percentage of reduction in seed weight and yield (g/plant) showing one percent decrease of reduction in seed weight resulted in 0.0089% decrease of yield ( Figure 2D). Earlier research revealed that identifying sources of YMV resistance is a viable strategy for addressing this viral disease. Numerous researchers have previously identified similar types of genotype analyses (Kumar et al., 2008;Talukdar et al., 2013;Baruah et al., 2014).

Field experiment 3.2.1. Disease incidence and disease severity
The highest (15.3%) disease incidence was calculated in G-2120 followed by Bs-29 (9%) and BS-23 (8.3%) which were statistically similar. No disease incidence was recorded in AGS-129 and rest of the genotypes showed lower to moderate level of incidence (Table 6). Disease incidence differences among the genotypes were found highly significant which is in an agreement with the findings of Bachkar et al. (2019) as well. On the other hand, AGS-129 may be a very good tolerant genotype, not really resistant. That's why it did not produce any symptom. It was also interesting to note that genotype AGS-160 did not have mottled seeds in the working sample but disease incidence (1.5%) due to seed transmission was observed (Table 6).This reveals the fact that, unmottled seeds of seeds producing without seed coat symptom are not necessarily healthy. Similar findings were obtained by Parakh et al. (1994). Leaves showing different levels of severity as compared with the healthy leave [ Figure 1 (A-E)]. Disease severity was recorded according to Singh et al. (1988) which conceded that out of 30 genotypes only one genotype (AGS-129) showed resistance with a grading scale one while ten genotypes (PR-164, AGS-160, G-2261, Williams, Durga, Ec-1178, ACAGS-154, BS-32, BS-10 and CM) were under the group of grading three exhibiting moderately resistance (MR). Ten genotypes (Colombus, BS-11, THINUNG-154, G-2120, BS-23, BS-15, BS-16, BS-60, Sau-Luis and TG-893) were recorded as moderately susceptible having grading scale five while nine (Gaurab, Acadian, PB-1, BS-13, BS-14, Cobb, BS-29, CH-1 and BS-17) genotypes were recorded as susceptible having grading scale seven (Table 5). Bachkar et al. (2019) screened 36 varieties against the disease, among them two were resistant, seven moderately resistant; twenty moderately susceptible, five susceptible while two varieties recorded highly susceptible reaction to SMV. Akhtar et al. (1992) conducted an SMV resistance screening on twelve cultivars. Four cultivars (Crow ford, Cico, Zane, and 80-B4007) were shown to be viral resistant. Zheng et al. (2000) evaluated 348 soybean accessions for resistance to soybean mosaic virus (SMV) using SMV3, a highly pathogenic strain from north east China. 113 accessions were found to be very resistant, 113 to be moderately resistant, and 122 to be vulnerable. Shrirao et al. (2009) tested 16 genotypes and discovered that 14 were completely resistant to soybean mosaic virus (SMV) and two were extremely resistant. Positive correlation was observed between the percentage of disease incidence and seed transmission which showed one percent increase of seed transmission resulted in 0.547 percent increases of disease incidence ( Figure 2B). SMV strain, soybean cultivar, soybean development stage, and the incidence of infection all play a role in the magnitude of yield loss caused by SMV (Ross, 1983).

Yield of healthy-looking plant and symptom bearing plant
Highest yield of healthy-looking plant was found in CM (9.977%) followed by TG-893 (9.943%), ACAGS-154 (9.897%), BS-10 (9.723%), BS-32 (9.573) and G-2261 (9.387%) exhibiting statistically identical while the lowest was found in BS-29 (3.330%). Highest yield of symptom bearing plant was found in BS-10 (8.74%) and lowest yield was found in BS-29 (0.903%) followed by BS-16 (1.720%) ( Table 6). A negative correlation was observed between seed transmission and yield conceded that one percent increase in seed transmission reduces 0.245 g yield/ plant ( Figure 2B). Understanding the relationship between grain yield and other characteristics aids in selecting the most suitable plant type (Adiya et al., 2011). There is considerable genetic variation across soybean genotypes in terms of leaf area (cm), days to flowering initiation, days to blooming, and days to maturity, Plant height (cm) Pods/plant, Branches/plant, 100-seed weight (g), Seed yield/ plant (g) Oil content (%) were observed by various researchers in different countries (Sihag et al., 2004, Chettri et al., 2005, Muhammad et al., 2003, Malik et al., 2006.

Plant height and Number of effective nodules per plant, Pod/plant, Pod length and Seeds per pod
Considering plant height, the top most height (69.60 cm) was recorded in Gaurab followed by THINUNG-154 (64.13cm), BS-11 (59.32%), EC-1178 (56.82%) and ACAGS-154 (56.82%) representing statistically identical data while the minimal plant height (12.63 cm) was found in PB-1 which were statistically similar to BS-60, BS-23, BS-16, BS-29, Acadian, BS-10 and BS-13. Moreover, others genotype exhibited moderate plant height (Table 7). PR-164 showed best (32.23) number of effective nodules per plant while the lowest was found in AGS-129 (3.95), remaining genotypes showed lower to moderate number of effective nodules per plant. Highest number of pods per plant was found in ACAGS-154 which was statistically similar to EC-1178, Williams, Acadian, G-2261, Gaurab, Thinung-154, Sau-Luis, G-2120 and CM. The lowest number of pods per plant was found in PB-1 which was statistically similar to BS-16 and BS-29 (Table 7). The highest (4.747) pod length was found in BS-32 followed by BS-15 (4.653), BS-11 (4.643), AGS-160 (4.593), Durga (4.520), AGS-129 (4.467), BS-29 (4.373), Cobb (4.330), PR-164 (4.300), BS-60 (4.277) and TG-893 (4.23) while the lowest was found in PB-1 (2.897), rest of the genotypes resulted moderate pod length( Table 7). The highest number of seeds per pod was found in Williams (3.023) which was statistically similar to G-2120, G-2261, Acadian, ACAGS-154, Thinung-154, Gaurab and CM whereas the lowest was found in BS-17 (1.167). Genotypes were accounted for insignificant results considering plant height and number of effective nodules per plant, pod/plant, pod length and seeds per pod, as these genotypes showed moderate susceptible (MS) to susceptible (S) reaction which is in accordance with the results of Naveesh et al. (2020), they also observed susceptible genotypes showing pronounced yellow mottling discolouration of leaves, reduction in leaf size and stunting of plants and reduction in pod size. According to Baruah et al. (2014), found Pod weight was substantially and positively linked with the number of seeds and pods/plant. Malik et al. (2011) found a similar result in terms of days to maturity and days to flowering. Baruah et al. (2014) also revealed that increased seed yield/plant was associated with increase in 100-seed weight which in turn showed negative correlation with number of pods/plant and seeds/plant.   Table 3. Performance of 30 selected soybean genotypes: percentage of seed germination, transmission, infection and reduction in seed weight.

Conclusions
SMV's growing impact on soybeans underscores the importance of introducing SMV resistance into Bangladeshi soybeans.11 genotypes viz. AGS-129, AGS-160, G-2261, Williams, CM, Durga, EC-1178, ACAGS-154, BS-32, BS-10 and PR-164 cultivars can be selected to locate resistant genes against SMV while the remaining cultivars experienced an increase in disease severity and incidence. However, more research is needed to identify the resistance sources of these soybean genotypes and to further validate them before they can be applied in breeding programs.