Genetics for seed traits and Mungbean yellow mosaic India virus reaction in urdbean (Vigna mungo L. Hepper)

Urdbean or blackgram is one of the important multi‐season pulse crops grown in India. Improved seed quality and resistance to Mungbean yellow mosaic India virus (MYMIV) are the important objectives of the urdbean breeding program. For this, knowledge of inheritance for seed quality traits and MYMIV resistance is essentially required. Therefore, an attempt has been made to understand the inheritance of seed traits like seed luster, seed coat color, mosaic on seed surface, and MYMIV reaction using intra‐ and inter‐specific bi‐parental populations. The inheritance results revealed that in DPU88‐31 × LBG‐685 population, the shining seed registered dominance over the dull trait, while the green seed coat color exhibited dominance over the brown seed coat color. However, the inter‐specific F2 population derived from IPU11‐02 × Pant M 6 deviated significantly from Mendelian ratios and supported severe segregation distortion for seed traits, while the mosaic on seed coat or mottling character studied in interspecific population demonstrated a dominant digenic inheritance pattern. In five inter‐ and intra‐specific F2 populations, MYMIV resistance demonstrated monogenic dominant inheritance, which was also validated in F3 generations. Furthermore, linkage analysis exhibited an association between seed luster and seed coat color, while no association was noticed between seed traits and MYMIV resistance. The normality tests revealed that seed width and 100 seeds weight were controlled by a few major genes, while other quantitative traits were governed by many genes with small additive effects. The skewness suggested that complementary gene interactions were present among genes controlling 100 seed weight in the intra‐specific cross, while duplicate gene interactions were involved in the inter‐specific cross. The identified monogenic traits and linked morphological marker after further enrichment of the linkage group could be used as an important tool in the regular breeding program.


| INTRODUCTION
Urdbean or blackgram (Vigna mungo L. Hepper) is one of the important multi-season pulse crops. It is being cultivated in several countries including India, Bangladesh, Nepal, Pakistan, Bhutan, Afghanistan, Thailand, and Myanmar. India is the largest producer and consumer of urdbean due to its high source of protein and micronutrient content, which are instrumental in providing nutritional security to increasing vegetarian population (Sen Gupta et al., 2020;Sen Gupta, Dutta, et al., 2022;. India produces about 2.45 million tonnes of urdbean annually from about 4.11 million hectares of area with average productivity of 696 kg per hectare (Project Coordinator's Report, 2020). In total pulses granary of India, urdbean contributes about 14% (24.42 million tonnes in 2020-2021). It is mainly cultivated as Kharif or rainy season crop in Northern, Central, and Southern parts of India. However, in Southern parts, it is being cultivated as a winter or rabi season crop. Most importantly, in Eastern and Southern parts of India, it is grown as a Rice fallow or Utera crop in substantial areas. In northern parts of country, it is also grown in the spring season.
Seed quality is an important aspect that plays a decisive role in consumers' preference for urdbean. Therefore, breeders have one aim to develop high-yielding cultivars with improved seed quality traits.
For this, a proper understanding of inheritance patterns of qualitative (e.g., seed coat color, seed luster, and mosaic on seed surface/mottling) and quantitative (e.g., 100 seed weight, seed length, seed width, seed thickness, geometric mean diameter, sphericity, and surface area) seed traits are basically required for breeding. In the case of seed quality traits' inheritance, very limited efforts have been made by researchers. However, Sen and Jana (1964) reported that brown seed coat color is recessive to green seed coat color and conditioned by a single gene. On the contrary, Arshad et al. (2005) found that brown seed coat color is dominant over green seed coat color. Shiny seed surface was dominant over dull seed surface (Sen & Jana, 1964).
Yellow Mosaic Disease (YMD) is a serious disease of urdbean that under congenial conditions causes up to 100% yield losses (Parihar et al., 2017). This disease particularly in Indian subcontinents is mainly caused by two different virus species, that is, Mungbean yellow mosaic virus (MYMV) and Mungbean yellow mosaic India Virus (MYMIV) (Kumar et al., 2014a). Mungbean yellow mosaic India virus is considered to be more predominant in northern, central, and eastern India and other MYMV in the peninsular region of India (Kumar et al., 2014b). However, sincere efforts have been made and many high-yielding varieties of urdbean with moderate to resistant reactions toward MYMV/MYMIV have evolved in India. To breed any variety, proper knowledge of the inheritance mechanism of targeted traits is required. So far, very limited efforts have been made to understand the inheritance of MYMIV. However, there were conflicting reports about the genetics of resistance to a yellow mosaic disease wherein claiming both resistance and susceptibility to be dominant. The monogenic dominant nature of resistance has been reported by several workers (Dahiya et al., 1977;Gupta et al., 2013;Kaushal & Singh, 1988). In some studies, the resistance was found to be mono or digenic recessive (Dwivedi & Singh, 1985;Pal et al., 1991;Reddy & Singh, 1995;Singh, 1981;Verma & Singh, 1986).
Seed traits including 100 seed weight, seed length, seed width, seed thickness, geometric mean diameter, sphericity, and surface area are important to be observed in segregating populations in urdbean particularly in inter-specific crosses, which are very common in urdbean breeding programs. Wide crosses many a times have impacts on the seed traits (Cobos et al., 2009); hence, study of the segregation pattern of these traits will be a novel one in urdbean breeding.
Hence, the present study was undertaken with the objectives to find out the inheritance of seed traits like seed luster, seed coat color, mosaic on seed surface, mungbean yellow mosaic India virus resistance (MYMIV) in biparental F 2 and F 2:3 populations as well as to record the pattern of segregation of quantitative seed traits in intraas well as inter-specific segregating urdbean populations.

| Plant materials
Seven urdbean genotypes (DPU88-31, LBG-685, Uttara, PLU-272, IPU11-02, Pant M 6, Lalitpur local) were used to generate five segregating populations (DPU88-31 Â LBG-685, IPU11-02 Â Pant M 6, Uttara Â PLU-272, IPU11-02 Â Lalitpur local, Uttara Â Lalitpur local). DPU88-31, an advanced breeding line, had dull seed luster, and brown seed coat with mosaic on the seed surface, while another parent LBG685 had shiny seed luster and green seed coat with mosaic on the seed surface. In another parent, IPU11-02 had dull seed luster and brown seed surface with mosaic, while its contrasting parent Pant M 6 (mungbean) had shiny seed luster and green seed surface without mosaic. F 1 seeds were grown along with their parents and advanced to F 2 and F 3 generations. Phenotypic data were recorded, and the inheritance pattern of seed quality traits was studied in each generation. Details of the donors are provided in Table 1.

| Development of F 2 and F 3 populations
Populations derived from crosses between trait specific donors (DPU88-31 Â LBG-685, IPU11-02 Â Pant M 6, Uttara Â PLU-272, IPU11-02 Â Lalitpur local, Uttara Â Lalitpur local) generated fresh F 1 s. F 1 s were raised during rainy season of 2018 and 2019 to develop the F 2 and F 3 populations, respectively, in the Main farm of ICAR-IIPR, Kanpur (26.28 N and 80.21 E). The field trials were conducted by following recommended package of practices; no insecticide was sprayed in these plant populations as that might have affected the whitefly population, which is responsible for the spread of the MYMIV disease. Row to row spacing was maintained at 30 cm and plant to plant spacing was approximately 10 cm. No irrigation was applied to these field trials from crop sowing to harvesting. Susceptible border rows (LBG-685) were grown along the populations to invite more disease incidence.

| Data recording in F 2 and F 3 populations
The populations (DPU88-31 Â LBG-685, IPU11-02 Â Pant M 6) were phenotyped for seed luster, color, and mottling. These two populations were also used for recording quantitative seed-related traits like 100 seed weight, seed length, seed width, geometric mean of diameter, sphericity, and surface area of seed.

| Statistical analysis
All the recorded data about seed traits and MYMIV resistance were subjected to Chi-square analysis using Microsoft excel. Data pertaining to quantitative seed traits (100 seed weight, seed length, seed width, the geometric mean of diameter, sphericity, and surface area) were subjected to descriptive statistics analysis as well as test for normality (Shapiro-Wilk p at 5% level of significance) (The Jamovi Project, 2019). The Q-Q (Quantile-quantile) plots and histograms were also prepared based on the quantitative seed traits in two F 2 populations (The Jamovi Project, 2019).

| Seed luster, color, and mottling
In the present study, inheritance of seed luster, color, and mottling traits were studied in two F 2 populations. One population derived from a cross between DPU88-31 and LBG-685 (Urdbean Â Urdbean) comprised of 346 individual plants of which 258 were Shiny (AA, Aa) types and 88 were Dull (aa) types seeds. Chi-square test (p ≤ 0.05) found non-significant for 3:1 segregation ratio, which further indicates monogenic dominant gene action for seed luster in urdbean. Similarly, in another F 2 population of 125 individuals derived from an interspecific cross, that is, IPU11-02 Â Pant M 6 (Urdbean Â Mungbean) in which 101 individuals had shiny seeds and 24 individuals had dull seeds and this segregation pattern was non-significant at p ≤ 0.05 through chi-square test further suggesting monogenic inheritance for this trait.
Inheritance of seed coat color was examined in an intra-specific cross DPU88-31 Â LBG-685 (Urdbean Â Urdbean). Segregation pattern in F 2 population of 346 individuals exhibited 260 green mosaic (AA, Aa) types and 86 brown mosaic (aa) seed types. Chi-square test (p ≤ 0.05) was found non-significant for 3:1 segregation ratio, which indicated monogenic dominant inheritance for seed coat color in urdbean. However, similar results were obtained in the F 2 population of an interspecific cross IPU11-02 Â Pant M 6 (Urdbean Â Mungbean).
In this population, 81 brown mosaic and 44 green mosaic seeded types were obtained, which were significantly different at p ≤ 0.05 probability from expected value fitted into 3:1 ratio. The genetics of mosaic on seed coat or mottling character was studied only in an interspecific cross IPU11-02 (mottling seed coat) Â Pant M 6 (smooth seed coat). Seeds of F 1 plants derived from this cross had mottling seed coat. In F 2 population comprising of 125 individuals, 114 individual plants had mottling seed coat, while 11 individual plants had smooth seed coat. The chi-square test confirmed goodness fit of this segregation pattern into 15:1 ratio as X 2 value (0.513) was nonsignificant at p ≤ 0.05. This indicated dominant digenic inheritance for this trait in urdbean Â mungbean crosses.

| MYMIV resistance
Inheritance of MYMIV resistance was studied in five different F 2 populations derived from inter-and intra-specific crosses, which were developed from crossing between highly resistant genotypes and highly susceptible urdbean genotypes (Table 1). Each F 2 population had a size of 143 to 346 individuals and segregated for MYMIV resistance and susceptibility. Segregation pattern of resistant and susceptible individuals in each population was in 3:1 ratio as X 2 value was found to non-significant at ≤0.05 probabilities ( Further inheritance of MYMIV resistance was confirmed in the F 3 generation. For this, individual plant progeny of each F 2 population was screened for MYMIV disease under field conditions. In F 3 , resistant, susceptible, and segregating progenies were observed, which had a good fit in to 1:1:2 ratios as the chi-square test showed non-T A B L E 1 Characteristics of urdbean genotypes used as parents in the present study

| Linkage analysis among seed traits and MYMIV resistance
Linkage among seed traits and MYMIV resistance has also been studied in segregating F 2 populations. For this, four F 2 populations were also phenotyped for their reactions to MYMIV incidence.
Among these, two F 2 populations derived from crosses, DPU88-31x LBG685 and IPU11-02 Â Pant M 6 showed an association between seed luster and seed coat color. Numbers of F 2 plants observed for four classes of these two seed traits; namely, shiny green, dull green, shiny brown, and dull brown were found significantly different from the mendelian digenic segregation ratio, 9:3:3:1 as X 2 values for two crosses (371.39 and 116.26, respectively) were observed significantly different at ≤0.05 probabilities (Table 4).
However, no association was observed between seed traits and MYMIV resistance in two F 2 populations phenotyped for MYMIV reaction as X 2 values for 9:3:3:1 digenic ratio were non-significant at ≤0.05 probabilities for four different classes found between seed luster and MYMIV reaction and between seed coat color and MYMIV reaction in F 2 generation (Table 4). Therefore these results suggested that seed traits are independent of MYMIV reaction in urdbean.
3.4 | Inheritance of quantitative seed traits in segregating F 2 population (DPU88-31 Â LBG685) In the present study, inheritance of various quantitative traits like 100 seed weight, seed length, seed width, geometric mean of T A B L E 2 Segregation ratio in F 2 population for seed traits and MYMIV resistance in different crosses under field trials at hot-spot location (Kanpur)  diameter, sphericity, and surface area was also studied in two F 2 cross DPU88-31 Â LBG685 and IPU11-02 Â Pant M 6 (Tables S1 and S2 and Figures 1-4). In first population, 100 seed weight ranged from 2.43 to 4.91 g with a mean value of 3.70 g. Seed length was varied from 3.81 to 5.21 mm with a mean value of 4.53 mm. Seed width was ranged from 3 to 4.28 mm with a mean value of 3.72 mm. Seed thickness was varied from 2.62 to 3.74 with a mean value of 3.24 mm. Genometric mean of diameter was ranged from 3.21 to 4.28 mm with a mean value of 3.79 mm. Another seed trait, sphericity was ranged between 9.71 to 23.10 mm with a mean value of 16.2 mm 3 . Seed surface area ranged between 129 and 230 mm 2 with a mean value of 181 mm 2 . Skewness was negative for seed length, seed width, seed width, seed thickness, surface area, and positive for 100 seed weight and sphercity. Kurtosis was lowest in case of 100 seed weight and highest in case of seed width. Test of normality of the segregating population was performed by Shapiro-Wilk Test, histogram (Figures 1 and 3) and Q-Q plot analysis (Figures 2 and 4), which clearly illustrated that except seed width all the traits had followed normal distribution. F I G U R E 1 Histogram of seed traits (seed weight, seed length, seed width, seed thickness, geometric mean diameter, seed sphericity, and total surface area) in F 2 population of DPU88-31 Â LBG685 seed thickness, surface area, and spherecity. Kurtosis was positive for 100 seed weight, seed length, and sphericity and negative for seed length, seed width, seed width, seed thickness, and surface area. Test of normality of the segregating population was conducted and observed that only 100 seed weight was not following normal distribution.

T A B L E 3 Segregation ratio in F 3 population for MYMIV resistance in different crosses
F I G U R E 2 Q-Q plots of seed traits (seed weight, seed length, seed width, seed thickness, geometric mean diameter, seed sphericity, and total surface area) in F 2 population of DPU88-31 Â LBG685 F I G U R E 3 Histogram of seed traits (seed weight, seed length, seed width, seed thickness, geometric mean diameter, seed sphericity, and total surface area) in F 2 population of IPU11-02 Â Pant M 6 4 | DISCUSSION

| Seed luster, color, and mottling
The present study demonstrated a monogenic dominant inheritance for seed luster and seed coat color in the intra-specific (urdbean Â urdbean) and inter-specific (urdbean Â mungbean) crosses. Sen and Jana (1964) reported that the shiny trait was dominant over dull seed luster in urdbean as observed in the present study. They also proposed allelic notion for this pair of contrasting characters: D = shiny smooth, d = dull rough (Sen & Jana, 1964). In the past, inheritance of seed coat color in urdbean and mungbean was studied and found to be monogenic dominant (Dwivedi & Singh, 1985;Singh, 1982). In another study, brown seed coat color was reported dominant over green seed coat color (Arshad et al., 2005). Mosaicism in urdbean is due to the presence of deep blue pigment in the epidermal layer of palisade cells, in irregular patches (Pandey et al. 1989). The mosaicism is of the constant type, monogenically dominant over its absence, as found in mungbean by Sen and Ghosh (1959). Monogenically inherited traits could be used as a marker for the identification of true hybrids and approximating the magnitude of crossing in self-pollinated crops like urdbean (Senapati & Roy, 1990). However, detection of heterozygous is not possible through morphological markers in case of complete dominance. Our results portrayed that green mosaic was dominant over brown mosaic in an intra-specific cross (urdbean Â urdbean), whereas in an inter-specific cross (urdbean Â mungbean) mottling on seed coat (mosaic) was dominant over smooth seed coat.
It was observed as a digenic dominant gene action (15:1) at 5% level of significance. Linkage analysis further demonstrated a tight linkage between shiny green and dull brown types in intra-specific (urdbean Â urdbean) and inter-specific (urdbean Â mungbean) crosses. In an earlier study, Sen and Jana (1964) also reported linkage between base-color of the seed coat and seed surface type.

| MYMIV resistance
In the present study, F 1 s derived from MYMIV resistant and susceptible parents were resistant, indicating resistance in urdbean to be dominant. On the contrary, the dominance of susceptibility in urdbean has been reported by Singh (1980) and Verma and Singh (1986). However, considering molecular markers and analysis of F 2 and F 3 generations, it was already reported in urdbean that a single dominant gene controls the resistance reaction (Gupta et al., 2013). Further, in the present study while analyzing F 2:3 populations the segregation ratio among segregating and parental types also confirmed the monogenic dominant gene control for this disease in urdbean. Linkage analysis showed that genetic control of different seed traits was in no way related to MYMIV resistance in urdbean. In earlier studies, one major recessive gene has also been reported for different viral diseases in grain legumes (Reeder et al., 1972). However, in contrast to this, Verma and Singh (1986) reported that two recessive genes are involved in imparting resistance against MYMV (Mungbean Yellow Mosaic Virus) in urdbean. It is highly desirable to transfer yellow mosaic virus resistance into mungbean from urdbean and Vigna sublobata as resistance to YMV in mungbean is found very rarely (Pal et al., 1991). Inheritance to F I G U R E 4 Q-Q plots of seed traits (seed weight, seed length, seed width, seed thickness, geometric mean diameter, seed sphericity, and total surface area) in F 2 population of IPU11-02 Â Pant M 6 YMV resistance was reported digenic recessive in the intra -(mungbean Â mungbean) and interspecific (mungbean Â urdbean and mungbean Â Vigna sublobata) crosses (Pal et al., 1991).

| Inheritance of quantitative seed traits in F 2 population
Nature of gene action (Fisher et al., 1932) and number of genes (Robson, 1956) controlling quantitative traits were studied in the intraspecific (DPU88-31 Â LBG685) and inter-specific (IPU11-02 Â Pant M 6) crosses through skewness and kurtosis analysis, respectively. Kurtosis is negative or close to zero in the absence of gene interaction and is positive in the presence of gene interactions (Choo & Reinbergs, 1982;Kotch et al., 1992;Pooni et al., 1977). A normal distribution with kurtosis 0 is known as mesokurtic, kurtosis with less than 0 is called platykurtic and kurtosis with greater than 3 is known as leptokurtic. The traits having leptokurtic kurtosis are under the control of a large number of genes, while traits having platykurtic distribution are controlled by fewer genes. In the present study, Kurtosis was positive for 100-seed weight in intra-and inter-specific crosses but it was highest in the inter-specific cross compared to intra-specific cross. These results indicated presence of gene interaction for this trait, which was higher in the case of the inter-specific cross (Choo & Reinbergs, 1982).
Moreover, a large number of genes control this trait. However, selection intensity and maximum genetic gain can be higher under complementary gene interaction than duplicate (additive Â additive) gene interactions (Choo & Reinbergs, 1982). For this, skewness was studied in the present study and found negative in an intra-specific cross and positive in an inter-specific cross for seed length, seed width, seed width, seed thickness, and surface area. These results showed that urdbean genome has presence of duplicate gene interactions for controlling these traits, while genome of mungbean used in inter-specific cross had different genes with complementary gene interactions (Choo & Reinbergs, 1982). However, for 100 seed weight positive and negative skewness was observed in intra-specific cross and inter-specific cross, respectively, in the present study. This indicates that average complementary interactions is present among genes conferring 100 seed weight in intra-specific cross, while duplicate gene interactions are involved in inter-specific cross for this trait (Choo & Reinbergs, 1982).
Therefore, a maximum genetic gain can be obtained in respect of these traits with positively skewed distribution through intense selection from the existing variability in intra-specific cross (Roy, 2000).
During the test of normality of the segregating population observed that seed width in intra-specific cross and 100 seed weight in inter-specific cross do not follow normal distribution, while remaining traits followed. This indicates that two traits are under the control of some major genes while other quantitative traits following normal distributions are governed by many genes with small additive effects.
For QTL mapping of targeted traits in a population various factors are considered including deviation from normality and the degree of skewness (RebaÏ, 1997). Nonparametric or logistic regression or standard regression interval mapping method are used if the trait is not normally distributed. However, the interval mapping was found to be useful with regard to loss of power compared to other two approaches in many cases (RebaÏ, 1997). Non-parametric and logistic regression may be useful in highly skewed population traits (RebaÏ, 1997).

| CONCLUSION
The overall results revealed that the seed luster and seed coat color were governed by monogenic dominant gene action in urdbean, while the mosaic on seed coat or mottling character which was examined through interspecific (urdbean Â mungbean) hybridization witnessed dominant digenic inheritance pattern. Furthermore, five F 2 population synthesized through inter-and intra-specific crosses demonstrated monogenic dominant inheritance pattern of MYMIV resistance in urdbean which was also confirmed in F 3 generations. Since morphological markers are inadequate in urdbean, these traits can be used as a morphological marker for recognition of true hybrids and assessment of the magnitude of crossing in urdbean. Further, linkage analysis showed an association between seed luster and seed coat color, while no association was observed between seed traits and MYMIV resistance. The identified association in the present investigation is suggested to be utilized for preliminary mapping of the genome as there is a dearth of this type of information in urdbean. The utility of the mapped marker could be grasped when loci affecting QTLs including diseases and other economically important genes would be added to the linkage group. The closely linked markers could be used in the regular breeding program by providing a mean of selection in the lack of nurseries and screening processes that can be more expensive and time-consuming. The normality tests indicated that only two traits, that is, seed width and 100 seeds weight were under the control of some major genes, while other quantitative traits were governed by many genes with small additive effects.
In the case of 100 seed weight, complementary gene interactions were present among genes in the intra-specific cross, while duplicate gene interactions were involved in the inter-specific cross. Thus, the maximum genetic gain can be achieved for these traits through intense selection from the existing variability in the intra-specific crosses.

CONFLICT OF INTEREST
The authors declare that no conflict of interest exists among authors.

AUTHOR CONTRIBUTIONS
DSG conceived the idea, designed, and supervised the experiments; JK performed the statistical analysis; DSG and JK made the initial draft of the manuscript; AK helped in recording data of field experiments; AKP edited the primary draft; AL and PKK reviewed and edited the final draft of the manuscript; and SPD and SG reviewed and edited the final draft.

DATA AVAILABILITY STATEMENT
Data are available on request to the corresponding author.