Mining legume germplasm for genetic gains: An Indian perspective

Legumes play a significant role in food and nutritional security and contribute to environmental sustainability. Although legumes are highly beneficial crops, it has not yet been possible to enhance their yield and production to a satisfactory level. Amid a rising population and low yield levels, per capita average legume consumption in India has fallen by 71% over the last 50 years, and this has led to protein-related malnutrition in a large segment of the Indian population, especially women and children. Several factors have hindered attempts to achieve yield enhancement in grain legumes, including biotic and abiotic pressures, a lack of good ideotypes, less amenability to mechanization, poorer responsiveness to fertilizer input, and a poor genetic base. Therefore, there is a need to mine the approximately 0.4 million ex situ collections of legumes that are being conserved in gene banks globally for identification of ideal donors for various traits. The Indian National Gene Bank conserves over 63,000 accessions of legumes belonging to 61 species. Recent initiatives have been undertaken in consortia mode with the aim of unlocking the genetic potential of ex situ collections and conducting large-scale germplasm characterization and evaluation analyses. We assume that large-scale phenotyping integrated with omics-based science will aid the identification of target traits and their use to enhance genetic gains. Additionally, in cases where the genetic base of major legumes is narrow, wild relatives have been evaluated, and these are being exploited through pre-breeding. Thus far, >200 accessions of various legumes have been registered as unique donors for various traits of interest.


Introduction
Legumes of the family Fabaceae are among the most important plant groups on planet Earth. While legumes are an important source of food and nutrition, they also play an important role in improving soil health and ecosystem sustainability. Legume grains are often considered to be "the poor man's meat," as the vegetarian human population is highly dependent on legume grains for its protein needs . The "green revolution" has helped several countries to attain self-sufficiency in food, which can primarily be attributed to a manyfold increase in the production of cereals, particularly rice, wheat, and maize. However, similar advances in grain legume production have not been achieved (Figure 1), probably because legumes are less amenable to the adoption of green revolution technologies. Over 200 species of legumes are cultivated worldwide. Of these, we list the major grain legume crops, with their production and yield status and taxonomic information, in Table 1.
India is the largest producer and consumer of grain legumes globally. India's contribution constitutes around 28.12% of global grain legume production (ca. 23.37 million tonnes), and this is the output of ca. 29 million ha of cultivated land (Department of Economics and Statistics, Department of Agriculture and Farmers Welfare, Ministry of Agriculture and Farmers Welfare, GoI). Of this total, around 34% of the cultivated land (9.89 mha) is covered by cultivation of chickpea alone; this is followed by black gram (4.81 mha), pigeon pea (4.72 mha), green gram (4.61 mha), lentil (1.43 mha), and field pea (0.75 mha). Other minor legumes cultivated in India are green gram, cowpea, moth bean, grass pea, and horse gram. In 2020, global grain legume production was approximately 83.1 million tonnes; this figure is the sum of the production values of grain legumes grown globally, specifically beans (dry), chickpeas, peas (dry), lentils, cowpeas, pigeon peas, Bambara beans, and other minor pulses (www.fao.org/faostat/en). The primary contributors to this total (at 42.12 million tonnes) were the five major grain legume producing countries of India (28. 12%), Canada (9.24%), Myanmar (4.84%), Nigeria (4.47%), and the Russian Federation (4.01%). The major producing countries for each crop are given in Table 1.
Globally, over 3,800 improved cultivars of grain legumes have been released, with improvements to traits such as yield, crop duration, and nutritional qualities (Pratap et al., 2022). However, between 1961 and 2020, only a 1.5-fold increase in grain legume productivity was achieved, from 637 kg/ha to 964 kg/ha (Pratap et al., 2022; Figure 1). This can be primarily attributed to various factors, such as a narrow genetic base in cultivated gene pools, poor plant ideotype, high susceptibility to insect pests and diseases, a lack of robust seed systems, and frequent stresses from drought, heat, and flooding.
Over 850 high-yielding varieties of food legumes have been developed in India, and these are now playing a vital role in food legume production (Chauhan et al., 2016). However, the foundation of any crop breeding program is based on only a small number of parental lines, which has led to a narrow genetic base in these cultivated varieties. In a pedigree analysis, it was found that 41% of chickpea varieties had PB 7 as one of its ancestors; in pigeon pea, T 1 and T 190 appeared in 34% of varieties; and T 9 and T 1 appeared in 64% and 35% of varieties of black gram and green gram, respectively (Kumar et al., 2004). Furthermore, in the process of rigorous selection in the development of a variety, alleles conferring defense mechanisms are also lost. This is one of the reasons that the actual yield of most food legume crops is half their potential yield. Recently, drastic climatic change, to which abrupt temperature rises, erratic and heavy rainfall, frequent droughts, episodes of flooding, and rapid pest and pathogen evolution can be attributed, has exceeded the adaptation capability of modern varieties (Guo, 2022). As a result, the breakdown of resistance to biotic stress has become rather common in modern cultivars (Sharma et al., 1999;Burdon et al., 2014;Rex Consortium, 2016;Mbinda and Masaki, 2021;Hu et al., 2022;Van de Wouw et al., 2022). Therefore, ex situ collections are now being utilized to increase genetic variability in modern cultivars in order to improve their climate resilience, including via genetic gains in breeding programs. Advances in genomics, phenomics, and breeding methods are playing an important role and exerting a significant impact on legume improvement by accelerating genetic gains via enhancements to selection efficiency and the advancement of desired genotypes with high precision.
It is well understood that, in terms of enhancing the variability of a crop gene pool, landraces are the primary resource; the desired traits need to be sought out among these, as they are easy to cross and their use significantly reduces the chances of linkage drag as compared to the use of wild species. Additionally, landraces are well adapted to microclimatic niches and have several superior traits in terms of nutritional value. In this study, we have focused on the identification of desired genes and traits and their utilization in the improvement of legume crops. We also propose a comprehensive strategy for the enhancement of genetic gains (Figure 2). environments of crop cultivation, including rainfed, dryland, and coastal areas, flood-prone areas, and areas at high altitude, as well as disease hotspots and human preferences in terms of nutritional qualities, aesthetics, and cultural values, have played important roles in the development and deployment of diverse germplasm. Although diversity has been continually developing and has been sustained through traditional practices over the last several thousand years, crop diversity has recently come under threat due to increasing pressure arising from demographic, sociocultural, and technological changes.
India is rich center of diversity for several cultivated crops, including important legumes such as chickpea (Cicer arietinum), moth bean (Vigna aconitifolia), rice bean (Vigna umbellata), cowpea (Vigna unguiculata), yard-long bean (Vigna unguiculata subsp. sesquipedalis), green gram (V. radiata), black gram (V. mungo), horse gram (Macrotyloma uniflorum), and dolichos bean (Lablab purpureus) (Zeven and Zhukovsky, 1975;Hawkes, 1983). The development of extensive and organized germplasm collections and conservation activity in India began only after the establishment of the National Bureau of Plant Genetic Resources (NBPGR) in 1976 (Rana et al., 2016). Since then, around 63,000 accessions of legumes have been collected and conserved in ex situ conditions ( Table 2). The organization is continuously enriching its collections based on gap analysis with respect to earlier collections established within India and also introducing accessions from abroad. Globally, over 0.7 million legume germplasms, including their crop wild relatives (CWRs), are conserved in 276 gene banks distributed worldwide (WIEWS, 2022).

Utilization of grain legume germplasm for crop improvement
Crop evolution in early times was based entirely on appearance and performance in terms of agro-morphological traits, and these are still the primary focus of plant breeders and researchers. During the domestication process and subsequent structured breeding programs, genotypes with greater biotic and abiotic stress tolerance are often unintentionally selected, but agronomic traits have always been the prime target for selection. Landraces, which are locally adapted cultivars with a high level of genetic variability developed by farmers over the years, are the primary source of such traits in modern breeding programs. However, in terms of the utilization of germplasm from gene banks, it has become difficult to identify a manageable number of accessions with the desired levels of variability and traits. Recognizing this challenge, Frankel (1984) proposed the concept of a core collection, a minimum number of representative accessions representing maximum variability across the entire collection. Since then, several crop-specific diverse core sets have been developed (Table 3), and this has accelerated the utilization of gene bank collections. A number of significant  (Schreinemachers et al., 2019;Khine et al., 2021). studies conducted to date in the area of trait identification and utilization are discussed below, presented in crop-wise fashion, and promising trait-specific accessions are summarized in Table 4 (biotic stress resistance) and Table 5 (abiotic stress tolerance). We also find that in the process of breeding modern varieties, the focus on yield per se has eventually led to a gradual decrease in the nutritional qualities of new varieties. Comparative studies on the nutritional composition of landraces and traditional cultivars in various crops, such as vegetables and fruits (Davis et al., 2004), wheat (Fan et al., 2008, the potato (White et al., 2009), the common bean (Celmeli et al., 2018), and green gram (Ebert et al., 2017), have indicated that the improved varieties are poorer than the older varieties in terms of nutritional value. Therefore, recognizing the significance of nutritional value and of the availability of nutritional variability in germplasm, we also discuss the important nutritional characteristics of each legume crop in the following sections.

Agronomic traits
Development of the first core collection in the domain of legumes was reported by Hannan et al. (1994) with the objective of making use of chickpea collections. In this study, a diverse set of 505 chickpea accessions was designated as a core set; this was derived from 7,613 accessions conserved in the Western Regional Plant Introduction Station (WRPIS), USDA. Later, the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) developed a core set of 1,956 accessions based on information on the geographic origins and on 13 morphological traits for 16,991 accessions (Upadhyay and Ortiz, 2001). Following this exercise, Upadhyay et al. (2007a) identified 28 early-maturing chickpea germplasm lines having wide geographical distribution. Based on multi-location trials of the core set, ICC 16641, ICC 16644, ICC 11040, ICC 11180, and ICC 12424 were further identified as

FIGURE 2
Strategy to enhance genetic gains through utilization of advanced phenotyping tools, efficient operational tools, and advanced selection methods and technologies. A strategy to achieve higher genetic gains by broadening the genetic base through the infusion of increasing levels of variability from diverse sources into the target breeding populations is illustrated. The integration of improved crossing program strategies and advanced tools for phenotyping, operations, and desired genotype selection will further enhance the genetic gains made. This strategy will help with the attainment of greater genetic gains along with enhanced crop adaptability to changing climatic conditions.
(2), Atylosia (18), Rhynchosia aurea (1) Frontiers in Genetics frontiersin.org 05 terms of agronomic traits were identified (Archak et al., 2016). The characterization of a large number of accessions also provides the opportunity to identify rare and unique morphotypes, which sometimes turns out to be very useful. For example, in the study carried out by Archak et al. (2016), accession IC486088 was found to have upright podding behavior, which makes it a potential donor that could be used in altering chickpea plant type  ( Figure 3). To enhance the utilization of such unique germplasm of economic or scientific value in crop improvement programs, these are registered with a national germplasm registration facility, i.e., the Germplasm Registration and Information System (GRIS; http://www.nbpgr.ernet.in:8080/registration/AboutUs.aspx). As of December 2022, a total of 28 unique accessions of chickpea have been registered by this facility. These unique traits help in the development of plant types and/or high-yielding cultivars. For example, a unique determinate phenotype was identified in BGD 9971; this is considered an important trait in the alteration of chickpea plant type (Hegde, 2011;Ambika et al., 2021). In the F2 line of an inter-specific cross, ICC 5783 (C. arietinum) × ICCW 9 (C. reticulatum), 3 to 9 flowers per flowering node were observed; this is an important trait for improving chickpea plant type and yield (Gaur and Gour, 2002). Finally, to reduce harvesting and threshing time and cost, chickpea genotypes with the erect plant type were identified and are being used to develop chickpea cultivars suitable for mechanical harvesting (Vishnu et al., 2020). GBM 2 and NBeG 47 (Dheera) are the first two such chickpea varieties that have been released (http://dpd.gov.in/Varieties/Chickpea%20varieties. pdf) (Figure 3).

Biotic stress
In chickpea, the major diseases are fusarium wilt (Fusarium oxysporum f. sp. ciceris), ascochyta blight (Ascochyta rabiei (Pass.) Lab.), collar rot (Sclerotium rolfsii Sacc.), dry root rot (Rhizoctonia bataticola (Taub.) Butler), and botrytis gray mold (Botrytis cinerea Pers. Ex. Fr.), and the major pest is pod borer (Helicoverpa armigera Hubner). Chickpea germplasm screening programs have identified rather plentiful instances of donor germplasm resistant to fusarium wilt. However, robust resistant donor sources for dry root rot, botrytis gray mold, collar rot, and pod borer are lacking; thus, germplasm use could result in the identification of moderately resistant donors for these diseases (Pandey et al., 2004;Reddy et al., 2016). The ICAR-National Bureau of Plant Genetic Resources (ICAR-NBPGR) has evaluated over 2,500 accessions for resistance to botrytis gray mold, collar rot, and dry root rot under artificial inoculation and field conditions, but only a few moderately resistant accessions have been identified, such as IC244185, IC251727, ICC6881, and IC350842 for BGM; IC270930, IC95064, IC350829, IC95100, IC209375, IC83805, IC487359, and IC83991 for collar rot; and IC413984, IC397375, IC487359, IC506915, and ICC4295 for dry root rot (unpublished data). This is a sign of the narrow genetic base of the cultivated germplasm. Similarly, no resistance sources have yet been identified for pod borer. Gayacharan U. et al. (2020) have identified several robust resistance sources (viz., IC275447, IC117744, EC267301, IC248147, and EC220109) for ascochyta blight disease using the sick plot method following artificial inoculation in multiple environments and seasons; these are now being utilized in national chickpea breeding programs. Pande et al. (2006) have also identified several other promising chickpea accessions (viz., ICC 17211, IG 69986, IG 70030, IG 70037, and IG 70038), which have shown combined tolerance against ascochyta blight and botrytis gray mold diseases. Finally, Singh (1997) has listed several of the important sources identified at the International Center for Agricultural Research in the Dry Areas (ICARDA) and at the ICRISAT. Various important sources of biotic stress resistance are listed by Singh et al. (2022) and are also presented in Table 4 and Figure 4.

Abiotic stress
In chickpea, the major abiotic stresses are terminal drought, terminal heat stress, and low temperatures during the late vegetative stage. The Northern Plains of India, which was once the most favorable zone for chickpea production, has faced a drastic decline in production of this crop due to a sharp rise in the minimum night temperature (Basu et al., 2009). Terminal heat stress is also a major challenge in the expansion of chickpea cultivation to rice-fallow land, of which there are around 11.7 million ha in the country (Singh N et al.,201a). Therefore, in order to tackle this problem, new sources of tolerance are being sought to enable the development of shortduration and heat-tolerant varieties. Basu and coworkers (2009) screened chickpea germplasm and identified several highly heattolerant chickpea lines, viz., ICCV 92944 (JG14), ICCV 37, ICC67, JKG 1, GCP 101, and PG 12. A reference collection for heat stress tolerance has also been developed and screened at the reproductive stage (Krishnamurthy et al., 2011). The researchers observed broad genetic variation in heat-responsive traits, and later identified 10 heat stress tolerance lines under field conditions (Devasirvatham et al., 2015). A mini-core collection has also been screened for resistance to drought stress; five highly tolerant accessions (namely, ICC867, ICC 1923, ICC9586, ICC12947, and ICC14778) were identified (Krishnamurthy, et al., 2010). A germplasm line (ICC4958) developed by Saxena et al. (1993) has been extensively used for breeding drought-tolerant varieties. In-depth molecular analysis of the ICC4958 line has identified a QTL-hotspot region that harbors several traits related to drought tolerance . Certain other drought stress tolerant lines have been identified under field conditions (Ganjeali et al., 2011). A mini-core set was also screened for salinity stress resistance under pot conditions and artificial application of saline water (100 mM), which led to the identification of 10 highly tolerant accessions (Serraj et al., 2004). Genotypes ICCV 00104, ICCV 06101, CSG8962, and JG62 have also been identified as promising in terms of salinity tolerance . Additionally, a total of 3,276 germplasm lines of chickpea were evaluated against cold stress at the ICARDA, Tel Hadya, Syria, between 1981 and 1987; 21 lines were found to be tolerant of cold stress (Singh et al., 1989). Choudhary and coworkers (2018) list several popular donors that represent the major sources for improvement of chickpea tolerance to abiotic stress. An extensive list of chickpea germplasm lines that have been identified as promising in relation to various abiotic stresses is also given in Table 5.
Large-scale nutritional profiling has not yet been carried out for legumes, primarily due to a lack of high-throughput nutritional profiling platforms. However, nutrient-specific donors with high mineral content have been identified, such as for Zn (MG-13, MG-17), Ca (PI518255, PI358934), and P (PI339154), and these can be used for biofortification of modern chickpea cultivars (Constantini et al., 2021). In an analysis of 79 accessions, one (LEGCA728) was identified as having high lutein content (28.32 μg g −1 ), and distinct morphotypes were identified as superior in terms of high concentration of specific nutrients (Serrano et al., 2017). In this study, it was observed that nutritional variation is associated with seed morphology. Black and brown seeded varieties were found to have higher dietary fiber content, ranging from 18.0 to 22.1 g 100 g −1 , and higher polyunsaturated fatty acid (PUFA) content (67.0 g 100 g −1 of total fatty acids; Summo et al., 2019). Accessions with brown coloring also have high water absorption capacity (1.9 g water g −1 of flour), which makes these varieties suitable for mixing with cereal flours to produce nutritionally rich cereal-based food products. The vitamins, minerals, and fibers present in chickpea grains promote their utilization for many health benefits. Finally, carotenoid concentration (with the exception of lycopene) has been found to be higher in wild germplasm as compared to cultivated types (Jukanti et al., 2012).

Agronomic traits
Lentil is one of the eight founder crops of agriculture (Ambika et al., 2022) and the most nutritious cool season legume cultivated in many farming systems worldwide. Lentil is divided into two categories Frontiers in Genetics frontiersin.org 09 #FIGS: focused identification of ermplasm strategy.

FIGURE 3
Highlights of various important agro-morphological variations. Genotype (ICC16358) with a large number of branches per plant (A); genotype (IC486088) having upright peduncle and pods (B); genotype (ICC15559) with two to three flowers/peduncle (C); genotype (EC398937) with greater pod length (>15 cm) and a higher number of seeds/pod (D); leaflet size variation (E); leaf size variation (F); genotype with short internode length and compact phenotype (G); genotype (IC24417) with erect and tall growth habit (H); an erect genotype (NBeG 47) in chickpea (I); early-maturing (IC347181) (J) and erect type (VLG 39) (K) genotypes of horse gram; sona mung with bright yellow seeds having superior visual appeal (L) in green gram; a common bean germplasm (IC341862) having pea-shaped, bright white-colored seed with superior visual appeal (M); and a pigeon pea genotype with determinate growth habit (N).
Frontiers in Genetics frontiersin.org 11 based on seed size, i.e., microsperma (seed diameter 2-6 mm) and macrosperma (seed diameter 6-9 mm), with 100 seed weight ranging from 1.5 to 8.0 g. In order to identify new germplasm for various agromorphological traits, extensive germplasm exploration, characterization, and evaluation programs have been undertaken globally. As a result, several trait-specific donors have been identified and used to develop improved varieties. High biomass, good plant standing, higher seed weight, and number of pods/ peduncle are considered to be crucial traits for yield enhancement in lentil. With the availability of such genotypes having tall (>30 cm), erect growth habits and good standing ability with good ground clearance (>15 cm), high-yielding varieties that are suitable for mechanical harvesting have been developed, such as ILL590, ILL1005, ILL6037, ILL6212, ILL6994, ILL7155, and ILL7947 Kumar et al., 2013). Germplasm lines for early flowering (IC560333, IC559639, IC560111, and IC560148), high biomass (IC559744, IC559608, IC559767, and IC560040), a large number of primary branches (IC559870, IC318881, IC398688, and IC560182), and high yield (IC398094, IC560212, IC560332, and IC560206) have also been identified through gene bank germplasm characterization (Gautam et al., 2013). Mishra et al. (2022b) identified PMF-1, PMF-2, PMF-3, and PMF-4 as producing multiple flowers per peduncle, which is an important trait in lentil breeding. The GRIS portal indicates the registration of accessions for a range of important unique traits, such as extended funiculus, which helps with rapid water uptake (IC317520); multiple flowers and pods per peduncle (IC241473); early flowering and maturity (IC241532); and extra bold seeds (EC499760). A core set of 287 accessions was developed for lentil using diversity documentation on 3,068 accessions conserved at the WRPIS, USDA, by Simon and Hannan (1995). Promising germplasm lines have been identified for various agronomic traits, such as seedling vigor, earliness, number of pods/peduncle, number of pods/plant, and seed weight (Singh., 1995). In another large-scale characterization conducted at the ICAR-NBPGR, accessions were characterized on 26 agro-morphological traits, and a core set of 170 accessions was developed (Tripathi et al., 2021a). Kumar et al. (2013) have also highlighted important lines for traits such as winter hardiness, short duration type, mechanical harvesting, and higher seed weight.

Biotic stress
The major diseases in the lentil crop are fusarium wilt (Fusarium oxysporum f. sp. lentis (Fol)), root rot complex, rust (Uromyces viciae-fabae), stemphylium blight (Stemphylium botryosum), powdery mildew (Erysiphe spp.), and ascochyta blight (Ascochyta lentis); the major pests are pod borer (Etiella zinkenella), aphids (Aphis craccivora), and seed weevil (Bruchus spp.). Several studies have been conducted to identify resistant donors, some of which are listed in Table 4. For example, 12 accessions were identified as resistant to fusarium wilt out of 196 landraces screened under both field and controlled conditions (Pouralibaba et a., 2015). In another study, 93 accessions were screened under three different screening conditions (specifically, a hotspot location, field sick plot, and artificial greenhouse Several promising newly identified resistance donors: IC486215 (A) for resistance against dry root rot in chickpea; IC275447 (B) for resistance against ascochyta blight in chickpea; IC118998 (C) for yellow mosaic disease (YMD) resistance in green gram; and IC278261 (D) for powdery mildew resistance in field pea.

Abiotic stress
Drought, heat, cold, frost, salinity, and waterlogging are the major abiotic stresses affecting lentil cultivation around the world. Several studies have been conducted to identify germplasm tolerant to these stresses (Table 5). In one such study, 166 lentil accessions were screened under field conditions, and six lines (ILL 7835,ILL 6075,ILL 6362,ILL 7814,ILL 7835,and ILL 7804) were identified for combined heat and drought stress tolerance . The Focused Identification of Germplasm Strategy (FIGS) was used to select 162 accessions for screening against heat and combined heat-drought stresses under field conditions at two locations (El Haddad et al., 2020); one germplasm line (IC621470) has been registered for drought tolerance on the GRIS portal. Based on a salt stress tolerance index, several promising accessions (ILL 5845, ILL 6451, ILL6788, ILL 6793, and ILL 6796) were identified in a screening of 133 accessions under artificial conditions (Ashraf and Waheed, 1990). Accessions ILL52,  ILL465, ILL 1878, ILL 1918, ILL7115, ILL7155, ILL468, ILL590,  ILL662, ILL669, ILL780, ILL857, ILL975, WA8649041, and WA8649090 have been selected for winter hardiness (Erskine et al., 1981;Hamdi et al., 1996). Additionally, Stoddard et al. (2006) have identified ILL5865 and ILL1878 as lines with good levels of tolerance to freezing. In Australia, in order to expand lentil cultivation, 310 accessions were screened in soils with a high boron concentration; accessions originating from Afghanistan and Ethiopia were found to perform comparatively well under these conditions. Boron-tolerant accessions ILL213A and ILL2024 were also recorded as having higher biomass than boron-intolerant accessions (Hobson et al., 2006).

Nutritional quality
Lentil has a high protein content (20%-27%; Zaccardelli et al., 2012) and contains 2%-3% fat, 50%-65% starch, and 8%-9% soluble sugars (Jood et al., 1998). Lentil protein is considered to be among the most beneficial, as it has good Leu/Ile and Leu/Lys ratios (1.24-1.98 and 1.08-2.03, respectively), high digestibility (~83%), and strong potential for use in food products (Jarpa-Parra, 2018). Among pulses, lentil is also one of the richest sources of Zn and Fe. A screening of over 2000 cultivated and wild germplasm has revealed a wide range of variation in Fe (42-168 ppm) and Zn (22-101 ppm; Mehra et al., 2018), with accessions EC78933 and EC 78414 found to have particularly high Fe and Zn content, respectively. In one of the experiments conducted by , 41 genotypes were examined for stability of Zn and Fe content over three locations; L 4704 (136.91 mg/kg grain) and VL 141 (81.542 mg/kg grain) were found to be promising in relation to Fe and Zn, respectively. A germplasm line (IC317520) with an extended funicle has also been identified; this is expected to be associated with shorter cooking time (Tripathi et al., 2021b). Several genotypes have been registered with GRIS: IC208329 and IC208326 for high protein content (27.4%-28.5%), and IC0616579 for high iron 136.91 (mg/kg grain) and zinc (71.69 mg/kg grain) content.

Common bean 3.3.1 Agronomic traits
Common bean is an economically important legume and is cultivated worldwide. In order to assess phenotypic variability in the ex-situ collections of the Indian National Gene Bank, 4,274 accessions were characterized on 22 traits, and a good range of variation was observed in leaf length, leaf width, pod length, number of pods per plant, number of seeds per pod, and seed weight . Promising accessions were identified for early flowering (IC370764), pod length (IC328871, EC271552), pods/plant (EC500299), early maturity (EC0944456), a large number of seeds/ pod (IC383008), etc. In another study, 203 accessions of a core collection were examined for seed quality traits and to identify promising germplasm lines (Saba et al., 2016). The Andean Diversity Panel (ADP), a regional core collection comprising 396 accessions, with the majority originating from the Andean region, was established in order to enhance germplasm utilization in the region's common bean improvement program (Cichy et al., 2015a). The ADP consists primarily of popular cultivars, breeding lines, and landraces. The CIAT gene bank conserves over 40,000 common bean accessions, making it the largest collection in the world. In an evaluation of a core set of 1,414 accessions, Amirul et al. (2006) observed wide variability in their morphological, biochemical, and nutritional traits. Through 12 multi-environment trials, a recent study has also identified four specific germplasm from 481 breeding lines with notable agronomic traits; the authors also developed a model to predict genotypic performance under different environmental conditions (Keller et al., 2020).

Biotic stress
The common bean is affected by many bacterial, fungal, and viral diseases, as well as insect pests. Several studies have been conducted to identify resistant sources in common bean germplasm (Table 4). A recent study has identified 14 accessions resistant to angular leaf spot Frontiers in Genetics frontiersin.org 13 (Phaeoisariopsis griseola) under field conditions (Rezene and Mekonin, 2019). Peña et al. (2013) screened 274 germplasm lines under artificial and pot conditions and identified two lines (PI 310668 and PI 533249) showing resistance against damping-off disease (Rhizoctonia solani). A set of 248 accessions of wild bean (Phaseolus spp.) were screened under greenhouse, pot, and artificial conditions against fusarium root rot (Fusarium cuneirostrum) and fusarium wilt (Fusarium oxysporum), resulting in the identification of 21 and 16 lines resistant to fusarium root rot and fusarium wilt, respectively (Haus et al., 2021). Urrea and Harveson (2014) carried out screening of 467 germplasm lines against bacterial wilt (Curtobacterium flaccumfaciens pv. flaccumfaciens) under pot and artificial conditions, and identified PI 325691 as a resistant line to the disease. The GRIS portal also indicates that several accessions have been registered as resistant to important diseases, such as anthracnose (IC0341862, IC635031, and IC635032), white mold disease (EC271515 and IC278744), and bean common mosaic virus (IC340947 and IC0360831).

Abiotic stress
The common bean is severely affected by abiotic stresses, such as cold, drought, heat, and salinity, and not a great deal of work has been carried out to identify trait-specific donors, with a few exceptions. Urrea and Porch (2009) screened 277 accessions of P. vulgaris and P. acutifolius under terminal drought stress conditions at Mitchell. The G35346 line has been identified for aluminum (Al) tolerance and used to transfer Al tolerance to common bean varieties (Butare et al., 2012). Tepary bean (P. acutifolius Gray), a relative of common bean, is known to have comparatively better tolerance for drought and subzero temperatures; on this basis and through preliminary screening of tepary bean accessions, one accession (W6 15578) has been identified as a potential donor for tolerance of both these stresses (Souter et al., 2017). Additionally, Dasgan and Koc (2009) screened 64 lines at 125 mM NaCl to identify salt-tolerant donor lines; a good level of variation was observed, and five highly tolerant genotypes were identified: Yalova 5, TR68587, Kibris Amerikan, Magnum, and Yerhammadisi.

Nutritional quality
Common bean is an excellent source of protein, dietary fiber, vitamins, and minerals. Its grains are a rich source of water-soluble vitamins, particularly thiamin, riboflavin, niacin, and folic acid. Analysis of a Chilean bean core collection of 246 accessions revealed protein content ranging from 183.5 to 259.7 g kg −1 , Fe content from 68.9 to 152.4 mg kg −1 , and Zn content from 27.9 to 40.7 mg kg −1 (Paredes et al., 2009). Kaur et al. (2009) also studied the physicochemical, hydration, textural, and cooking properties of common bean, observing a wide range of variation in terms of seed density (0.51-2.15 g/ml), hydration capacity (0.03-0.62 g/seed), hydration index (0.16-0.97), swelling capacity (1.24-1.93 ml/seed), cooking time (50-120 min), and amylose content (0.09%-5.02%). Another study revealed the ranges of variation in common bean for antioxidant activity (5.5%-44.9%), starch content (17.4%-40.7%), size of starch granules (1.64-176 μm), rapidly digestible starch (11.1%-19.5%), slowly digestible starch (8.5%-17.3%), and resistant starch (63.9%-76.1%; . Common bean is well known for rich diversity in seed coat color, and this color plays a major role in the selection, taste, and palatability of particular genotypes. Therefore, to investigate the relationship between color and protein and mineral content, a study was conducted in 100 genotypes having carioca, black, and other grain color patterns (Silva et al., 2012). The results indicated that black-colored beans are richer in protein, iron, and zinc; carioca grains are richer in manganese and magnesium; and grains of other colors are rich in calcium. Ciat-A-257, Bolinha, Iapar 81, Linea 29, and Roxo PV were found to be rich in protein (28.95%-30.40%). Additionally, 206 accessions from the Andean Diversity Panel were evaluated on cooking time, and five accessions (ADP0367, ADP0521, ADP0469, ADP0518, and ADP0452) were identified as promising in terms of shorter cooking time (Cichy et al., 2015b). Germplasm was also compared on nutritional composition and cooking characteristics with its closely related cultivated species, the tepary bean (Phaseolus acutifolius), in order to identify superior donors, as the latter species is highly tolerant to abiotic stresses . The results of this study indicated that there were no species-level differences on most nutritional parameters, with the exception of shorter cooking times for tepary bean accessions .

Agronomic traits
Pigeon pea is a legume of Indian origin (Ambika et al., 2022), and India remains its largest producer and consumer (Bohra et al., 2012). Pigeon pea is a resource-rich crop in terms of genetic and genomic resources, whole genome sequencing information, availability of trait-specific germplasm, genetic stocks, etc. The largest collection of pigeon pea germplasm is currently conserved at the ICRISAT gene bank (13,632 acc.), followed by the Indian National Gene Bank, ICAR-NBPGR (11,210 acc.); these collections are the major resources for trait identification and crop improvement. To enhance germplasm utilization, a set of 1,290 pigeon pea accessions (Reddy et al., 2005) has been developed, followed by a mini-core set of 146 accessions (Upadhyaya et al., 2006) and a composite core set of 1,000 accessions, plus a reference set of the most diverse 300 accessions (Upadhyaya et al., 2011a). An exceptionally good level of phenotypic variation has been observed for traits such as pods/plant, number of racemes, plant height, seed yield/plant, and days to maturity (Reddy et al., 2005). Promising accessions included in the pigeon pea composite core set have been listed for important economic traits, such as early flowering, a large number of pods/plant, seed weight, and yield/ plant (Upadhyaya et al., 2011b). A vast amount of variability in flowering period has been observed, and a number of genotypes have been reported to show exceptionally short and long flowering durations. ICPL 90011 is reported to be an extra-short duration genotype with the lowest photoperiod sensitivity (Silim et al., 2007). Diverse trait-specific germplasms have been identified for use as potential sources for improvement programs (Upadhyaya et al., 2007c;Mir et al., 2014;Yohane et al., 2020). As of December 2022, 55 pigeon pea germplasms have been registered with the GRIS portal for a range of unique traits, including genetic male sterility (IC296750), cytoplasmic genetic male sterility (IC471860,  IC471861, IC296590, IC296592, IC555904

Biotic stress
Pigeon pea production is adversely affected by many insects and diseases, such as wilt (Fusarium udum Butler), sterility mosaic virus (PPSMV) disease, phytophthora blight (Phytophthora drechsleri f. sp. cajani), Gram pod borer (Helicoverpa armigera), pod fly [Melanagromyza obtusa (Malloch)], and spotted pod borer [Maruca vitrata (Geyer)]. In one experiment, Saxena et al. (2002) evaluated 271 accessions under natural field conditions, and found that disease severity scores in pigeon pea germplasm ranged from 3 to 9. Screening against fusarium wilt and sterility mosaic disease (SMD) was carried out under artificial conditions for multiple seasons, resulting in the identification of several resistant accessions, viz., ICP 6739, ICP 8860, ICP 11015, ICP 13304, and ICP 14819 (Sharma et al., 2012). Accession ICPW 94 of the wild species C. scarabaeoides has been identified as resistant to all isolates of SMD, and is used in crossing programs (Hema et al., 2014). Earlier similar studies were also conducted using petiole grafting and artificial conditions in search of donors for SMD resistance in pigeon pea (Kulkarni et al., 2003;. Several sources of resistance for various biotic stresses in pigeon pea are listed in a review by Sultana et al. (2021) and in Table 4.

Abiotic stress
Pigeon pea is considered to be a drought-tolerant crop due to its deep root system and wide range in maturity period, which allows it to fit into a wide range of environments and cropping systems (Choudhary et al., 2011). Major abiotic stresses limiting pigeon pea productivity are waterlogging, drought, low temperatures (<10°C), and photoperiod sensitivity. Through several germplasm evaluation programs in pigeon pea, a number of popular donors have been identified; these are major sources for abiotic stress tolerance (Choudhary et al., 2018). In one study, 96 pigeon pea accessions were identified for early flowering; these are considered potential sources for the breeding of early-maturing pigeon pea varieties in order to avoid terminal drought and heat stress (Upadhyaya et al., 2011b). Sultana et al. (2013) screened 272 pigeon pea lines for waterlogging stress tolerance under laboratory and field conditions, and identified 12 lines tolerant to waterlogging. Similarly, in another study conducted under pot conditions for multiple seasons, 24 pigeon pea accessions were identified as waterlogging-tolerant (Krishnamurthy et al., 2012). Various other sources for resistance to important abiotic stresses are also listed in a review by Sultana et al. (2021) and in Table 5.

Field pea 3.5.1 Agronomic traits
Field pea is cultivated in over 100 countries for fresh and dry grains and for fodder. Over 31,000 germplasm accessions of Pisum are conserved ex situ in various gene banks, including the Australian Grains Genebank, Australia; the Western Regional Plant Introduction Station, USDA, United States of America; the Leibniz Institute of Plant Genetics and Crop Plant Research, Germany; and the ICAR-NBPGR, New Delhi, India. Although a limited number of large-scale studies have been conducted on the agro-morphological characterization of field pea and for trait identification, several studies nevertheless indicate a substantial amount of phenotypic variability on qualitative as well as quantitative traits, such as days to 50% flowering, seed weight, plant height, and number of pods/plant (Azmat et al., 2011;Bhuvaneswari et al., 2017). The accessions IPF-5-19, EC 8495, HUDP-15, and DDR-30 have been found to show promise in terms of seed yield (Bhuvaneswari et al., 2017). Singh et al. (2010) evaluated 71 accessions on agronomic performance and seed and flower characteristics, identifying promising accessions in terms of early flowering (IC279013), early maturity (IC394017), a large number of pods/cluster (IC279195), longer pods (IC279013), pods/plant (IC219027), seed yield (IC279082), and seeds/pod (IC394028). Genotypes with five flowers per peduncle (VRPM-901-5) and three flowers per peduncle at multiple flowering nodes have been reported in garden pea, which could be highly useful in field pea improvement (Devi et al., 2018). Several unique trait-specific pea accessions have been registered in the GRIS portal, such as IC296677 (leafletless, dual purpose, and high-yielding), IC296678 (dwarf, leafletless), IC296737 (male sterile line governed by a single gene), IC279125 (bold seed with 50.14 g 100 seed weight), IC0610501 and IC630592 (≥ three pods/peduncle), IC636671 and IC640781 (extra-early flowering), and EC414478 (extended funicle).

Abiotic stress
Cold, frost, salinity, and heat stresses are the major sources of abiotic stress in field pea crop production. Many studies have taken up the aim of developing lines tolerant to abiotic stresses. In one such study, five field pea germplasm (ATC 104, ATC 377, ATC 968, ATC 3992, and ATC 4204) were identified as frost-tolerant at the reproductive stage through screening of 84 accessions under controlled environmental conditions (Shafiq et al., 2012). Screening of 3,672 pea germplasm lines under field conditions led to the Frontiers in Genetics frontiersin.org 15 identification of 214 cold-tolerant lines (Zhang et al., 2016). Additionally, 780 accessions were screened for salinity stress tolerance under artificial conditions (Leonforte et al., 2013). Finally, in a recent study, IPFD 11-5, Pant P-72, P-1544-1, and HUDP 11 were identified as heat-tolerant lines based on evaluation under timely-and late-sown field conditions (Lamichaney et al., 2021).

Nutritional quality
Field peas in general have lower protein content (~25%), very low fat content (~0.1%), and very high carbohydrate content (~70%). Major yield-attributing traits in field pea are pods/plant, number of grains/pod, and seed weight. In one study, 94 pea genotypes were examined for pea carotenoid content; higher carotenoid content (10-27 μg/g) was observed in accessions with green cotyledons, and comparatively low carotenoid content (5-17 μg/g) in accessions with yellow cotyledons (Ashokkumar et al., 2015). Pea grains have comparatively higher antioxidant activity than chickpeas. Promising field pea accessions have also been identified in terms of mineral content, such as Zn (IG52442, IG134828), Cu (IG116297, IG52442) and Ca (IG51520, IG52442), by Costantini et al. (2021). Additionally; Singh et al. (2010) have identified lines with shorter cooking time (IC260344) and observed that the genotypes that absorb more water and swell more during soaking require less cooking time. The authors have also identified IC320964 as superior in terms of ash content (3.73%), and several other accessions as promising in terms of their physicochemical properties. In a nutritional analysis of 96 accessions from diverse collections at the USDA National Germplasm Center, Pullman, WA, a wide range of variation was observed in mineral micronutrient content (Hacisalihoglu et al., 2021). An atypical morphotype having extended funicle (EC0414478) was identified in pea germplasm, and this accession was found to be associated with faster water uptake in comparison to the checks included (Tripathi et al., 2021b); this is likely to help with the development of pea cultivars with shorter cooking times.

Agronomic traits
Cowpea is a multi-purpose grain legume (yielding grains, green pods, and leaves) and is widely cultivated in Asia, Africa, and America. It is considered to be one of the best suited crops for hotter, semi-arid agroclimatic conditions, as it requires less water and also grows well in sandy soils. The germplasm conserved in various gene banks has exhibited a good amount of genetic variability, which enables it to grow in various agroclimatic regions and in various soil types. To enhance the utilization of cowpea germplasm, over 12,000 accessions of cowpea were characterized on 28 agro-botanical descriptors at the International Institute of Tropical Agriculture (IITA), Ibadan, and a core set of 2,062 accessions was developed (Mahalakshmi et al., 2007). In another study, 4,000 accessions were characterized in multi-location trials by the ICAR-National Bureau of Plant Genetic Resources (unpublished records). A great deal of variability was observed in plant and seed morphology. Gerrano et al. (2015) identified germplasm lines having desirable grain yield characteristics, such as Fahari, IT93K129-4, Glenda, and vegetable cowpea dakama cream; Nkhoma et al. (2020) identified lines Bubebe, CP411, CP421, CP645, Chimponogo, and MS1-8-1-4 as highyielding and genetically divergent among 90 genotypes studied, making them ideal parental lines. Cowpea genotypes IT96D-604, 93K-619-1, IT97K-569-9, and IT99K-1060 have also been identified as highyielding (Goa et al., 2022).

Biotic stress
The major diseases affecting cowpea are cowpea mosaic virus (CpMV) disease, Cercospora leaf spot (CLS), brown blotch (Colletotrichum capsici), and bacterial blight (Xanthomonas axonopodis pv. vignicola), while the major pests are pod borer, aphids, thrips and bruchids. The severity of these biotic factors varies with agro-climatic zone and growing conditions. Although cowpea is one of the more prominent legume crops and the largest of the Vigna group, it has not received commensurate research attention. As a result, cowpea improvement has suffered from a lack of reliable donors for resistance to many of these biotic factors. Nonetheless, efforts have recently been undertaken in this direction, and several important and promising donors for resistance to a small number of these biotic stresses have been identified; these are listed in Table 4. In a study that aimed to identify resistant donors for aphid (Aphis craccivora), cultivated germplasm (105 accessions) and wild germplasm (92 accessions) were screened under greenhouse conditions; only a single accession (TVNu 1,158) was identified as a resistant line (Souleymane et al., 2013). The findings of this study also indicated that both the cultivated and the wild relatives of this crop have poor genetic bases. Boukar et al. (2019) identified 14 lines having resistance to bacterial blight (Xanthomonas axonopolis pv. vignicola) under artificial inoculation. In another study, 225 germplasm lines were screened against CpMV, CLS, and cowpea rust (Uromyces vignae), resulting in the identification of promising accessions for resistance to these pathogens (Deshpand et al., 2010; Table 4). Finally, Tripathi et al. (2020) identified EC528425 and EC528387 as tolerant to bruchid (Callosobruchus maculatus) through the screening of 103 cowpea lines using a 'no-choice' test method.

Abiotic stress
The major abiotic stresses are drought, heat stresses, and poor soil fertility, especially in sub-Saharan Africa (SSA), where cowpea is grown as a major crop, as well as soil salinity in almost all irrigated areas worldwide (Horn and Shimelis, 2020). Several studies have been conducted to identify resistant donors (see Table 5). Five lines with superior drought stress tolerance (viz., TVu1436, TVu9693, TVu12115, TVu14632, and TVu15055) have been identified using the water withdrawal method under field conditions (Fatokun et al., 2012), while Dagupan Pangasinan, UCR 369, and Negro have been identified as tolerant to waterlogging at the seedling stage (Olorunwa et al., 2022). Accessions EC472250, EC472267, EC472285, EC472286, EC472289, and Pusa Komal have been identified as tolerant to heat stress through screening in multiple seasons under field conditions (Mishra et al., 2005). Accessions PI582422, 09-529, PI293584, and PI582570 have been identified as tolerant to salinity stress under artificial screening conditions through imposition of salinity stress (150 mM NaCl) at the seed germination stage (Ravelombola et al., 2017). Other similar studies have also identified lines tolerant to salt stress using different NaCl concentrations (150 mM and 120 mM) at the germination and seedling stages (Ravelombola et al., 2018;Dong et al., 2019). Based on screening of 155 cowpea lines in 200 mM NaCl, several promising lines were identified as salt tolerant, i.e., PI354686, PI353270, Frontiers in Genetics frontiersin.org 16 PI354666, PI354842 PI548785, PI582466, PI339599, and 09-697 .

Nutritional quality
Cowpea is a major source of nutrition in sub-Saharan Africa, Asia, and Latin America. Based on nutritional profiling of 100 breeding lines on a dry weight basis, a significant range of variation has been observed in terms of protein content (22.9%-32.5%), ash content (2.9%-3.9%), fat content (1.4%-2.7%), and carbohydrate content (59.7%-71.6%; Nielsen et al., 1993). Genotypes also vary in 50% cooking time, which ranges from 21.1 to 61.9 min, and promising donors have been identified, such as IT83S-872 for protein content, IT84S-2085 and IT86D-466 for ash content, and IT85F-2805 for shortest cooking time (Nielsen et al., 1993). In a study aiming to investigate nutritional variability in immature cowpea pods, 22 genotypes were analyzed on various nutritional composition parameters; genotypes such as ITOOK-1060, TVU-14196, and 98K-5301 were found to be superior on such parameters as Mg, Na, Mn, Boron, Al, Zn, Cu, K, P, and protein (Gerrano et al., 2017). The fresh young leaves of cowpea are also consumed in several countries; therefore, analyses have been conducted of the nutritional composition of 15 varieties and the sensory attributes of 10 varieties (Ahenkora et al., 1998). In this study, nutrient concentration in cowpea leaves on a dry weight basis was found to range from 303.8 to 468.9 mg/100 g for phosphorus, from 33.5 to 148.0 mg/100 g for ascorbic acid, and from 27.1% to 34.7% for protein.

Agronomic traits
Black gram is a grain legume of Indian origin, primarily cultivated in South Asian regions. Although black gram is an important legume, its productivity level is very poor compared to that of other legumes, which can mainly be attributed to a lack of good plant ideotypes and resistance sources for major diseases in its cultivated gene pool Shanthi et al., 2019;Subramaniyan et al., 2022). Therefore, to identify donors for desired agro-morphological traits, 484 accessions have been characterized on qualitative and quantitative traits; a good deal of variation was observed in flowering and maturity period and in yield (Ghafoor et al., 2001). Recently, 840 accessions of black gram were also characterized, resulting in the identification of promising germplasm lines in terms of early flowering (IC343936, IC436615), synchronous flowering (IC73523, IC396032, IC485444), pod length (IC438379), number of seeds/pod (IC472051_2, IC565238), and seed weight (IC485605_2, IC485588) (Gayacharan et al., 2022). For novel trait generation in black gram, gamma-irradiated mutants were generated using black gram cultivars ADT 3, Co 6, and TU 17-9, which have exhibited high plant yield (Dhasarathan et al., 2021). Additionally, RBU1012 and Pant U-19 have been found to be the most stable genotypes in terms of yield when evaluated under field conditions . The GRIS portal indicates that unique germplasms of black gram such as IC296878 (dwarf with ground pod bearing habit), IC553269 (brown pods with yellow seeds), IC594172 (male sterile flowers with protruded stigma and crumpled petals), IC594173 (sympodial pod-bearing habit), IC426765 (photosensitive), and IC636672 (extra-early maturing) have been registered for important traits.

Biotic stress
Urdbean leaf crinkle disease (ULCD) and mungbean yellow mosaic disease (MYMD) are the two major diseases of the black gram crop. Yield losses may reach or exceed 60%, depending on the susceptibility of the host plant, if the crop is affected in its early vegetative stage. Nevertheless, unlike green gram, black gram has a high level of resistance against MYMD in its cultivated gene pool, as revealed in a large-scale evaluation under field and artificial conditions conducted during 2019 and 2020 (unpublished records). Several black gram sources of MYMD resistance, identified on the basis of field screening, are highlighted in a review published by Mishra et al. (2020a). Urdbean leaf crinkle virus (ULCV) disease has spread across all the cropping systems in India, and yield losses can reach 100% if the disease outbreak occurs at the early growth stage under favorable weather and host genotype conditions (Biswas et al., 2009). Resistance sources for ULCV have been reported by several researchers (Ashfaq et al., 2007;Gautam et al., 2016); several such sources for this and other diseases are listed in Table 4. In the GRIS database, accessions IC0570267, IC0570268, IC0570269, IC11613, IC636672, IC0144901, and IC485638 are registered as MYMD resistant, and IC0585931 as bruchid resistant. Powdery mildew (Erysiphe polygoni) and Cercospora leaf spot (Cercospora canescens) are the other major diseases of black gram. The major pests affecting this crop include spotted pod borer (Maruca testulalis r), whitefly (Bemicia tabaci), bruchids (Callosobruchus chinensis. and C. maculatus), and nematodes (Meloodogyne incognita, M. javanica, and Heterodera cajani), for which reliable sources of resistance are lacking. Bruchids begin infesting the crop during the pod maturity stage, and they are the cause of up to 90% of produce losses (Soundararajan et al., 2012).

Abiotic stress
The crop is grown in a rainfed environment under tropical and sub-tropical climatic conditions. Therefore, terminal drought and heat, as well as waterlogging, are the major constraints on black gram production. Salinity is another problem, particularly in arid and semi-arid regions. Only a small number of studies have examined the potential for improvement of black gram in terms of resistance to abiotic stresses. Saline-tolerant lines, such as BARI Mash-1 (Hasan et al., 2017), VNBG 017, AUB 3, and AUB 20 (Priyadharshini et al., 2019), have been identified as promising under artificial screening conditions. Under natural waterlogging conditions during a germplasm characterization program, a small number of germplasm lines have been identified as tolerant; these were further evaluated under artificial waterlogging conditions, and accessions IC530491 and IC519330 were found to be tolerant to waterlogging (Bansal et al., 2019). In another study, 26 genotypes were analyzed under waterlogging stress; a large amount of variation was observed in various quantitative traits, and BU Acc 25, BU Acc 17, and BU Acc 24 were identified as the strongest performers in terms of yield (Rana et al., 2019). In terms of drought stress tolerance, cultivars VBN4 and K1 have been identified as promising based on protein and biochemical analyses (Sai and Chidambaranathan, 2019).

Nutritional quality
Black gram grains are a rich source of protein (22%-26%) and moderately high in calories (ca. 350 cal/100 g), carbohydrates (ca. 56.6%), and fat (1.1%-1.2%) (Panhwar, 2005;Suneja et al., 2011). They also contain vitamins, viz., Vit. B1 (0.42 mg/100 g), Vit. B2 (0.37 mg/ 100 g), Niacin (2 mg/100 g), and minerals, viz., Ca (185 mg/100 g), Fe Frontiers in Genetics frontiersin.org (8.7 mg/100 g), and P (345 mg/100 g) (Panhwar, 2005). However, a limited amount of germplasm has been nutritionally profiled for the identification of nutrient-rich lines. Black gram is reported to exhibit a substantial amount of variation in nutrient content between the whole grain and its milled fraction (Girish et al., 2012). A small number of genotypes among 26 investigated, such as Shekhar 2, have been found to have high Fe and Zn content, and genotypes Yakubpur, PU 31, IPU 99-200, and PDU 1 have been found to have high polyphenol content . There is a need for large-scale nutritional profiling to develop an understanding of nutritional variability in the germplasm and to identify superior genotypes with minimal levels of anti-nutritional factors to enhance the palatability of the crop.

Green gram 3.8.1 Agronomic traits
Green gram is a highly nutritious and palatable grain legume cultivated in Asia, primarily for its grains. Green gram cultivation faces constraints such as a narrow genetic base in the cultivated gene pool, a lack of ideal plant type, and many biotic and abiotic stresses. In order to identify new donors, 1,532 ex situ collections of green gram conserved in the Indian National Gene Bank were characterized, potential donors for certain agro-morphological traits were identified, and a core set of 152 accessions was also developed (Bisht et al., 1998). A good level of variation was observed in branch length, nodulation, number of pods bearing a peduncle, number of pods per plant, and yield per plant. The World Vegetable Center, Taiwan, holds over 6,700 accessions of green gram, which have been utilized for the development of a core set of 1,481 accessions based on geographic stratification and clustering of genotypes on eight phenotypic traits (Schafleitner et al., 2015). This core set was genotyped using 20 microsatellite markers, and a minicore set of 289 accessions was developed; this is now extensively utilized for trait identification. In another large-scale characterization and preliminary evaluation of green gram germplasm, promising germplasm lines were identified in terms of early flowering (EC398944, EC398883), synchronous flowering (EC396115, IC76414, and IC488968), greater pod length (EC398937), seed weight (EC398903, EC398884, and EC396413), etc. (Gayacharan K. et al., 2020). Recently, the entire green gram ex situ collection of the Indian National Gene Bank has been characterized in multi-location trials, and a diverse core set of 400 accessions has been developed (unpublished records). Photoperiod-insensitive genotypes (EC 318985-319057) have also been identified in green gram Pratap et al., 2019). The GRIS portal lists a number of accessions with unique traits, such as a photosensitive nature (IC546478), high seed weight (IC418452), early maturity (IC0589309, IC589310, IC39289, and IC639796), and penta-foliate leaves (IC296679).

Biotic stress
Green gram production is affected by biotic stresses such as yellow mosaic disease (YMD), pod borers, and storage pests. YMD is a comparatively new disease in green gram and is spreading rapidly into new areas, which is a cause for concern. In YMD-susceptible genotypes of green gram, yield losses up to 85% are reported (Karthikeyan et al., 2014), but it has been observed that losses may reach 100% if the crop is infected at seedling stage. Resistance sources are lacking in the entire cultivated gene pool of the crop, as revealed in a field screening of 4,100 accessions at New Delhi (a YMD hotspot location). However, variability in the severity of the disease is observed according to multiple factors, such as genotypic constitution, vector population load, weather conditions, presence of multiple virus strains, etc. (unpublished record). Similar reports have also made by other researchers based on germplasm screening (Shad et al., 2006). Several resistant sources for YMD are listed in a review (Mishra et al., 2020a). There are also several reports of YMD resistance in green gram under field conditions (Iqbal et al., 2011;Mohan et al., 2014;Nainu and Murugan, 2020). Aside from YMD, Duraimurugan et al. (2014) identified four lines (viz. LM 131, V 1123, LM 371, and STY 2633) as resistant against bruchid beetle based on a 'free choice' and 'no choice' test method. Spotted pod borer (Maruca vitrata) also causes severe damage to the crop, if not controlled at the appropriate stage of crop growth, and there are no resistant sources available for this pest. Sandhya et al. (2014) have reported that KM-9-128, KM-9-136, RMG-492, LGG-5, and LGG-538 are tolerant to Maruca vitrata following field screening of 110 genotypes.

Abiotic stress
Green gram is primarily grown under rainfed conditions; thus, abiotic stresses such as drought, waterlogging, heat, and salinity affect crop production (Singh and Singh, 2011). In general, reliable tolerant donors for these abiotic stresses are lacking in this crop. Forty-one elite lines were screened for heat stress tolerance under late-sown conditions; of these, five lines (viz., EC693357, EC693358, EC693369, Harsha, and ML1299) showed heat stress tolerance . Additionally, IC333090 and IC507340 were found to be drought tolerant, out of 100 lines screened under hydroponics conditions (Meena et al., 2021). Mung bean lines OBGG-2013-9 and OBGG-2013-14 have also been reported to exhibit cold tolerance (Kabi et al., 2017).
3.9 Horse gram 3.9.1 Agronomic traits Horse gram is one of the least utilized and least studied legumes. The crop is known for its nutritional and therapeutic value, and is Frontiers in Genetics frontiersin.org primarily cultivated in hill states and dry areas of southern India. This crop has failed to attract the attention of breeders and researchers due to a lack of ideal ideotypes and morphological variation. A small number of characterization and evaluation studies have been conducted, indicating comparatively wide variation in terms of plant height, pod length, seed test weight, and pods per plant . Additionally, genotypes CRHG-6 and CRHG-8 are of the non-shattering type, which has been developed through mutation breeding (Salini et al., 2014). In a characterization study examining seven qualitative and quantitative traits in 66 horse gram genotypes, a good amount of variability was observed for pod length and pods per plant (Gomashe et al., 2018). Priyanka et al. (2021) studied 12 quantitative traits across 252 horse gram genotypes, reporting that the highest yield was 65.61 g per plant. Another characterization and evaluation study of 51 accessions led to the identification of several promising accessions (viz., S44/L23, S56/ L29, S8/L4, S96/L49, and S29/L14) in terms of early flowering and maturity (Kaundal and Kumar, 2021). In the GRIS portal, only one accession (IC587788) is registered for high fodder yield.

Biotic stress
Horse gram is susceptible to various biotic stresses and still lacks resistant donors for use in crop improvement programs. Only a small number of studies have been carried out to identify resistant donors for selected diseases and pests, such as YMD, wilt, anthracnose, and storage pests. Parimala et al. (2011) identified horse gram accessions AK-38, HG-14, HG-52, HG-59, HG-63, HG-75 as having resistance against horse gram YMD, and AK-38 and HG-46 as resistant to powdery mildew disease. In another study, accessions HG 63, HG 58, HG 50, and Palem 2 were identified as resistant to wilt disease (Durg, 2012). Accession IC470275 has also been identified as resistant to anthracnose disease (Colletotrichum dematium) (Sankar et al., 2015). Finally, horse gram lines Palem-1, Palem-2, AK-21, and NSB-27 have been identified as resistant against Callosobruchus chinensis, a storage pest (Divya et al., 2012;Divya et al., 2013). Accession IC587786 is registered on the GRIS portal as resistant to anthracnose disease.

Abiotic stress
Horse gram germplasm have been screened against abiotic stresses, such as drought, salinity, moisture, and heavy metal stress. Several germplasm lines, such as M-249 and HPK-4, have shown resistance against drought stress (Bhardwaj and Yadav, 2021). Yasin et al. (2014) identified a line tolerant to moisture stress, namely D13. The variety Paiyur-2 was found to have high proline and glycine betaine content and lower lipid peroxidation under salinity stress (Kanagaraj and Sathish, 2017). This genotype was further tested for antioxidant activity, and was found to exhibit enhanced antioxidant activity under salinity stress (Desingh and Kanagaraj, 2019). Separately, the same Paiyur-2 variety was found to be promising for salinity tolerance, and heavy metal tolerance was observed in Madhu (for chromium) and in HGR-4 (for nickel; Dhali et al., 2021;Edulamudi et al., 2021). Based on a screening of 88 germplasm lines for biochemical parameters, accessions TCR491, IC110286, IC56145, and IC53641 were identified as suitable for environments imposing drought stress (Sharma and Chahota, 2022).

Nutritional quality
Horse gram is used as a food and fodder crop and is known for its medicinal and therapeutic uses. It provides protein (17.9%-25.3%), carbohydrates (51.9%-60.9%), lipids (0.58%-2.06%), and vitamins, such as riboflavin, niacin, and vitamin C (Jha et al., 2022). The protein content of horse gram is relatively high compared to that of green gram, black gram, dry peas, the kidney bean, chickpea, or pigeon pea (Longvah et al., 2017). As consumers have become more health conscious, consumption of sprouted horse gram seeds has increased. In the seeds, albumin-globulin is a major contributor (~75.27-78.76%) to the total protein content. The seeds are low in fatty acid content but rich in dietary fiber, required for proper functioning of the lower intestine (Kawale et al., 2005). Horse gram seeds also exhibit antioxidant activity and radical scavenging activity. In a recent study, metabolic profiling was conducted for 96 accessions of horse gram, which were derived from 700 accessions spread across the entirety of India (Gautam and Chahota, 2022). Tremendous variability in protein content was observed, with the lowest protein content (13%) being found in IC120837 and TCR-1439, while a related wild species (Macrotyloma sar-gharwalensis) had the highest protein content (40%). Accessions IC280031 and IC139356 were found to be most nutritive, as the largest number of metabolites (44) was observed for these among the 96 lines selected in an analysis using 1 H NMR spectroscopy (Gautam and Chahota, 2022). An earlier study had also identified the species Macrotyloma sar-gharwalensis (IC212722) as containing 34.88% protein (Yadav et al., 2004).

Moth bean 3.10.1 Agronomic traits
Moth bean is considered to be a hardy crop suitable for hot, arid regions. In addition to this, it also helps to reduce soil erosion, particularly in sandy deserts, due to its extensive root system and profuse foliage cover. Most local cultivars continue to have wild traits, such as pod-shattering, a trailing and spreading growth habit, asynchronous maturity, and a photo-sensitive nature. Few studies have been conducted to explore the existing variability in the gene pool. However, a good amount of phenotypic variability in moth bean was reported by Chaudhary et al. (2021) in an evaluation of 40 genotypes on 10 morphological traits. In another study, accessions IC 36607, IC 39675, IC 251908, IC 36563, and IC 36245 exhibited higher seed yield as compared to checks (Vir and Singh, 2015). Similarly, 50 genotypes of moth bean were phenotyped on 12 quantitative traits and exhibited high levels of variability (Sahoo et al., 2019). Additionally, Singh S et al. (2017) have developed a variety (RMO 257) with superior agronomic traits, such as plant height, dry matter accumulation, and seed yield. A small number of moth bean genotypes are registered in the GRIS portal for early maturity (IC432859 and IC120963) and for single stem formation (IC432859).

Biotic stress
The major biotic stresses are YMD, leaf crinkle disease (LCD), bacterial leaf spot (Xanthomonas phaseoli), Cercospora leaf spot (Cercospora dolichi), charcoal rot (Macrophomina phaseolina Tassi), pod borer, and bruchids. Accessions IC36522 and IC36217 have been identified as YMD resistant as a result of screening under field conditions in multiple seasons . Meghwal et al. (2015) screened 204 germplasm lines and identified 14 accessions resistant to YMD. Resistance to leaf crinkle virus and Cercospora leaf spot has also been reported in the crop Frontiers in Genetics frontiersin.org . Vir and Singh (2015) identified LCD resistance in moth bean under field conditions in multiple seasons.

Abiotic stress
Moth bean is one of the best-suited crops for arid and semi-arid environments, and is highly tolerant to drought and heat stress.
Although only a small number of moth bean accessions have been investigated, a good amount of variation in the germplasm has been revealed in terms of resistance to drought and heat stress; this could be exploited for crop improvement in order to sustain its productivity amid climate change (Vir and Singh, 2015;Pal et al., 2020). Accessions such as IC129177, IC103016, IC415139, IC 415155, IC36157, Maru moth, and Jadia, which have been identified as tolerant to drought stress, can serve as donors for crop improvement (Malambane and Bhatt, 2014). Additionally, in a separate study, lines IC103016, IC36011, and IC36157 have been identified as promising for drought tolerance (Sachdeva et al., 2016).
4 Role of genomics in enhancing grain legume germplasm utilization and attaining higher genetic gains The first step toward enhancing the utilization of grain legume crop germplasm accessions for trait discovery and subsequent genetic improvement requires thorough and extensive genotypic and phenotypic characterization of such accessions using large-scale data (Rasheed et al., 2017). The numerous germplasm resources (including landraces, wild accessions, cultivated varieties, and breeding lines) available for diverse grain legume crop species, representing diverse agro-climatic regions of the world, have been stored efficiently at various national and international gene banks and repository centers. Considering the difficulties involved in genotypic and phenotypic characterization of this vast set of available germplasm resources, efforts are currently being made to develop core and minicore collections in the case of several legumes by identifying the greatest amount of genetic diversity that can be represented with a minimal number of accessions (Table 3). This is where the crucial role of genomics comes into play, especially as a means of producing realistic estimates of the level of molecular diversity existing among germplasm accessions, which enables efficient screening of unduplicated authentic accessions in the process of constructing core collections of grain legumes.
Tremendous technological advances made over the last decade in sequencing and other high-throughput sequence-and array-based genotyping assays have supplied much-needed momentum to germplasm characterization. Draft and reference whole genomes, resequencing information, and global transcriptome information have now been decoded for many important grain legume crop plants using first-generation Sanger sequencing and nextgeneration sequencing (NGS)-based second-generation short read and third-generation long read sequencing assays; the results of these are now publicly accessible (Table 6; Michael and Jackson, 2013; http://www.embl.de; http://www.ebi.ac.uk; http://www.ddbj. nig.ac.jp; http://www.phytozome.org). Grain legumes were once considered to be resource-poor crops, but recent efforts by national and international organizations has altered this trend. As a result, a vast amount of genome sequence information, including whole genome sequences, is available in public databases (Table 6). This sequence information has since been used to understand the genomic features and evolutionary characteristics of the crops in question, and also to develop a vast range of genomic resources, including molecular markers (Garg et al., 2022). Among several sequence-based molecular markers that have been made available, simple sequence repeats (SSRs) and single-nucleotide polymorphisms (SNPs) have occupied a central position, finding extensive use in allelic diversity screening and other genomics-assisted crop improvement programs due to their genome-wide distribution, multi/bi-allelic nature, and amenability to high-throughput detection and genotyping assays (Lateef, 2015). The availability of several high-throughput genotyping platforms and the rapid evolution of the chemical techniques that they employ has enhanced the precision and pace of large-scale mining and genotyping of SSR and SNP markers (Kujur et al., 2015). Highthroughput genotyping of SSR and SNP markers in a larger set of germplasm accessions and core or mini-core collections of grain legumes has been expedited via the use of various array-based and NGS assays, especially automated fragment analyzer (ABI3730xl automated DNA sequencer), Illumina GoldenGate, and Infinium assays; the Fluidigm dynamic array; KASP (KBioScience Allele-Specific Polymorphism) profiling; MALDI-TOF; the Affymetrix GeneTitan SNP Chip array; and Genotyping-By-Sequencing (GBS) assay (Varshney et al., 2013a;Jain et al., 2013). Among these, the MALDI-TOF, Illumina GoldenGate, and Infinium assays, SNP Chip Array, and KASP profiling have come to be considered highly advantageous and are utilized widely for high-throughput genotyping of previously mined SNP markers in many crop plants Hiremath et al., 2012). In particular, GBS assay has been extensively utilized for simultaneous genome-wide discovery and genotyping of SNPs in diverse plant species (Sonah et al., 2013;Spindel et al., 2013). Its development has thus expedited the mining of novel functional allelic variants and their large-scale validation and genotyping at the whole genome level for efficient trait association mapping of diverse small-and large-genome grain legume crop plants.
Using the aforementioned high-throughput marker-based genotyping strategies, along with large-scale multi-environment phenotyping information, sets of 211, 146, 184, and 289 germplasm accessions belonging to core or mini-core collections of chickpea, pigeon pea, groundnut, and green gram, respectively, have been developed. These have been collated based on the 16,991, 13,632, 15,490, and 6700 accessions available for these respective crop species as a result of the efforts of international institutes such as the IRRI, ICRISAT, USDA, and the World Vegetable Center Zhang et al., 2011;Schafleitner et al., 2015). These core or mini-core germplasm resources, readily available for many grain legume crop plants, are the primary sources of trait discovery once these collections have been thoroughly characterized genotypically and phenotypically for diverse traits of agronomic importance, including yield, (a)biotic stress tolerance, and nutritional quality traits. Under this perspective, candidate gene-based association mapping and genomewide association studies (GWAS) relying on the large-scale genotyping of informative SNP markers and robust field phenotyping information on these core or mini-core germplasm lines (i.e., association panels) are now considered to be an effective approach for the identification of major and minor genes/QTLs and alleles regulating both simple qualitative and complex quantitative traits in grain legume crop plants (Varshney et al., 2011). Candidate gene-based association mapping, which is carried out using genotyping information from SNPs in various coding and regulatory sequence components of Frontiers in Genetics frontiersin.org 20 genes in a trait-specific association panel, plays a significant role in the identification of genes/QTLs controlling yield, nutritional quality, and stress tolerance traits in grain legume crops. With the availability of high-throughput genome-wide SSR and SNP marker-based genotyping information on germplasm lines belonging to an association panel, the GWAS has now become a routine approach for high-resolution scanning of the whole genome to identify target genomic regions, including genes/QTLs (both major and minor) associated with traits of agricultural importance in many grain legume crops . The trait-influencing molecular signatures once identified using trait association mapping are significant for their potential utilization for genomics-(marker-)assisted crop improvement programs.
The delineated molecular signatures regulating traits of agronomic importance have been utilized for introgression, combining, and pyramiding into selected grain legume crop genotypes of interest through traditional and advanced genomics-assisted breeding approaches in order to develop superior crop varieties in terms of yield and stress tolerance (Varshney, 2016). Recently, a chickpea cultivar 'Pusa JG16' has been released in India as a drought-hardy cultivar; this was developed through genomics-assisted breeding utilizing a QTL-hotspot region from ICC4958 (Bhardwaj et al., 2021). Introgressions of functional natural genetic variations and of favorable genes, QTLs, alleles, and chromosomal segments identified from a larger set of grain legume germplasm accessions (including landraces and wild species), particularly for yield and stress component traits, have been transferred into the cultivated genetic background for improvement of the relevant crop through the use of such approaches as introgression lines (ILs), advanced-backcross QTL (AB-QTL) analysis, association genetics, and multi-parent advanced generation intercross (MAGIC) populations (Roorkiwal et al., 2020;Bohra et al., 2021). For example, the 'Geletu' chickpea variety was developed through marker-assisted back-crossing (MABC) and released in Ethiopia; it provided a yield advantage of 15% over the check variety 'Teketay' and 78% over a local check (https://www. icrisat.org/). The MABC technique has been used in the development of introgression of QTLs into elite cultivars in order to develop introgressed lines (Bharadwaj and Yadav, 2012;Varshney et al., 2013b;Barmukh et al., 2022). The 'Pusa Chickpea 20211' variety is another example in which resistance genes for multiple races (foc 1,2,3,4, and 5) of fusarium wilt have been stacked through MABC in the mega desi chickpea variety 'Pusa 391' (Bharadwaj et al., 2022). Molecular tags associated with major effects on qualitative and quantitative trait regulation have now been transferred into diverse grain legume crop genotypes for their genetic enhancement through marker-assisted selection (MAS), including MABC and markerassisted foreground and background selection. The identification of a QTL-hotspot region in linkage group 4 (CaLG04) in chickpea that harbors major QTLs for multiple drought adaptive traits, followed by its introgression into elite chickpea cultivars, is an excellent example of genomics-assisted breeding (Barmukh et al., 2022). This region accounts for 58.2% of explained phenotypic variation and a 16% yield enhancement under drought conditions in introgressed lines, which is primarily attributed to improvements in root traits, such root length, density, surface area, and volume (Varshney et al., 2013b;Roorkiwal et al., 2020;Bharadwaj et al., 2021).
Complications in the domain of genetic background effects/ epistasis and linkage drag of QTLs, as well as minor effects of both minor and major QTLs/genes on complex trait regulation, have impeded the use of the traditional MAS (QTL-MAS) approach in the genetic enhancement of grain legume crops on complex quantitative traits. To overcome these complexities, many novel advanced genomics-assisted breeding approaches are currently emerging, such as marker-assisted recurrent selection (MARS), MAGIC, and genomic/genome-wide (haplotype) selection. These involve the transferal and pyramiding of the favorable alleles of minor-effect genes/QTLs controlling complex quantitative traits for the genetic enhancement of grain legume crop plants in terms of yield, nutritional quality, and stress tolerance.  have identified superior haplotypes through whole genome sequencing (WGS) of 3,336 accessions, both cultivated (3,171) and wild (195), for important traits relating to yield enhancement; these can then be introgressed into elite chickpea cultivars. The study also identified target genomic regions for the purging of deleterious alleles, which can be achieved through genomics-assisted breeding and/or gene editing . Similarly, superior haplotypes have been identified in pigeon pea based on the 292 pigeon pea genotypes of a reference set that included breeding lines, landraces, and wild species (Sinha et al., 2020). In this study, haplotype-phenotype association analysis for drought-responsive traits resulted in the identification of promising haplotypes (C. cajan_23080-H2, C. cajan_30211-H6, C. cajan_26230-H11, and C. cajan_26230-H5) for three genes regulating five drought component traits (Sinha et al., 2020). Genomic selection (GS) and integrated genomic-enviromic prediction (iGEP) are other promising strategies that can be used to improve genetic gain in legume crops (Heffner et al., 2009;Xu et al., 2022). GS uses a smaller training population that is well genotyped and phenotyped, while iGEP uses additional data on genotype-environment interactions to build a prediction model. These models are then used to predict the true breeding values of selecting particular candidates based on multiomics data, big data technology, and artificial intelligence (Xu et al., 2022). Thus, genomics plays an integral role in improving genetic gain in modern agricultural practices and has immense potential to expedite future grain legume crop improvement programs. The traditional and novel genomics-assisted breeding approaches that are now available provide clues to the quantitative dissection of complex trait regulation, and thus have potential to expedite studies of the genetic enhancement of complex traits in diverse grain legume crops.

Future outlook
Grain legumes are a major source of food and nutrition globally. However, it has not been possible to enhance gain legume production to the required level, primarily due to the narrow genetic base of most of the legume crops, coupled with changing climatic conditions. Most grain legumes are lacking in desired plant ideotypes and resistance sources for various biotic and abiotic stresses. The genetic diversity conserved in gene banks globally is a major resource for crop breeding programs, but it is utilized only marginally. Therefore, in order to broaden the genetic base of grain legume crops and enhance the genetic gains made in improvement programs, conventional approaches and modern scientific tools should be integrated in a phased and carefully judged manner. The first phase should focus on the search for desired traits and the infusion of diversity into the cultivated gene pool through use of landraces and CWRs. The second phase should focus on the utilization of advanced selection tools, such as genomics, high-throughput precision phenotyping, and artificial intelligence, to exploit the hidden potential of the available genetic diversity; and the third phase should involve technologies such as mutational breeding, genome editing, and transgenic technologies to improve, modify, or introgress any novel or alien traits that are not available in the entire crop gene pool. We presume that enrichment of the genetic diversity of cultivated grain legume gene pools, along with simultaneous improvements in their yield and plant type with the aid of advanced scientific tools, will enhance grain legume crop yields to the required level.

Author contributions
JR conceptualized the topic addressed in the manuscript. Gayacharan, SP, HV, and RY collated the literature and wrote the sections on their respective areas of expertise. NM extended her assistance in grammatical correction, reference editing, and the creation of figures. Gayacharan compiled the contributions received from all authors, and finally the manuscript was edited by JR and Gayacharan. All authors reviewed the final draft of the manuscript and gave their approval.

Funding
The authors thank the GEF project "Mainstreaming agricultural biodiversity conservation and utilization in agricultural sector to ensure ecosystem services and reduce vulnerability," implemented in India by the UN Environment (Project Code: A 1265), for providing funding support.