Identification of diverse agronomic traits in chickpea (Cicer arietinum L.) germplasm lines to use in crop improvement

A chickpea (Cicer arietinum L.) collection consisting of 445 Kabuli‐type collected from different locations of Iran was evaluated based on six qualitative and 14 quantitative morphological descriptors. Chickpea accessions were widely varied in plant height, the number of pods per plant, plant dry weight, 100‐grain weight, grain yield, and harvest index. Grain yield was positively correlated with the number of pods per plant (r = 0.39**), pod fertility percentage (r = 0.42**), dry weight per plant (r = 0.88**), and harvest index (r = 0.30**). The highest value for the Shannon–Weaver diversity index was observed in growth habit (0.98) and leaf color (0.88). The first and second components of the principal component analysis (PCA) explained 17.97% and 16.20% of the total variations, respectively. The cluster analysis results revealed that the accessions with higher pod, leaflet, and peduncle length were grouped in cluster I, whereas cluster II indicated the dominant contribution for the number of pods per plant, plant dry weight, and grain yield. The highest phenotypic coefficient of variation (PCV) and genotypic coefficient of variation (GCV) were observed for grain yield per plant (44.3% and 48.3%), the number of pods per plant (44.3% and 45.1%), and dry weight per plant (44.4% and 44.8%). The heritability estimates were also more significant for the number of pods per plant, plant dry weight, and harvest index. These accessions might be used in the chickpea breeding programs to expand high‐yielding Kabuli‐type cultivars with a broad genetic base.


| INTRODUCTION
Chickpea (Cicer arietinum L.) is an important grain legume crop with an annual production of $15 M tonnes, which is globally cultivated in $15 M ha with average productivity of 1000 kg ha À1 (Bapurao et al., 2018;FAO, 2022). India (65%), Pakistan (7%), Turkey (7%), and Iran (3%) are considered the main chickpea-producing countries in Asia (FAO, 2014). Chickpea with the ability to biological nitrogen fixation at a rate up to 140 kg N ha À1 year À1 , plays an essential role in preserving soil fertility and health (Flowers et al., 2010). Two types of chickpeas, including desi and Kabuli, are recognized; the Kabuli-type has owl-shaped, large cream-colored grains, whereas the desi-type has angular-shaped (Upadhyaya et al., 2006). The desi-type is mainly consumed in Asia and accounts for nearly 80% of global chickpea production and equally plays a role in the total chickpea trade (Archak et al., 2016).
Models for estimating genetic diversity, including molecular markers or morphological traits, can be diverse (Rao et al., 2007).
Such quantitative characteristics as cluster analyses and principal component analysis (PCA) provide estimations of genetic diversity (Ghafoor et al., 2001); these have been utilized successfully in classifying and measuring the genetic pattern variations in plant germplasms (Seid et al., 2021). The PCA and cluster analysis provide methodical and efficient ways of estimating the genetic diversity of agromorphological traits in plants, including chickpea (Gupta et al., 2011;Kayan & Adak, 2012;Parameshwarappa et al., 2011). Upadhyaya et al. (2006) developed a chickpea core collection consisting of 1956 accessions to increase the utilization of chickpea genetic resources in breeding programs, representing 84% and 100% of the variation range of the entire collection in plant height, 100-grain weight, numbers of pods per plant, days to maturity, and grain yield. Working on 13 chickpea accessions, Khan et al. (2006) reported that although the genotypic coefficient of variation was relatively low for days to flowering, days to maturity, plant height, and the number of pods per plant, it was high for 100-grain weight.
Besides the lack of adapted varieties, several biotic and abiotic stresses contribute to the fluctuations in chickpea yield (Rasool et al., 2015). Ascochyta blight (Ascochyta rabiei), phytophthora root rot (Phytophthora medicaginis), Fusarium wilt (Fusarium oxysporum f. sp. Ciceri), and Botrytis gray mold (Botrytis cinerea Pers. ex Fr.) (Ahmad et al., 2005;Knights et al., 2008;Nabati et al., 2021;Singh et al., 2008), and drought, extreme temperatures, and salinity are among biotic and abiotic stresses, respectively, that significantly limit the chickpea productivity (Jha et al., 2014;Nabati et al., 2021). Furthermore, low genetic variation for yield, yield components, and the resistance to major diseases are the main limitation to reaching high yield potential (Malik et al., 2014). Therefore, the wide use of few and closely related germplasm in crop improvement may lead to the vulnerability of newly developed cultivars to pests and diseases . As a result, it is necessary to identify the different germplasm characteristics of a plant because it provides information about the available variety and helps to choose one or more specific traits (Kathiresan, 2000). in 2020 (36 15 0 N and 59 38 0 E, and an altitude of 985 m). Seeds of each accession were sown on four rows (with and within row space of 50 and 5 cm, respectively) in plots of 4 m long and 1.5 m apart. The plots were surface irrigated twice (once after planting and once at the flowering stage). Hand weeding was performed twice in mid-April and mid-May. To control the legume pod borer (Heliothis viriplaca), indoxacarb (Sc 15%; 200 ml ha À1 ) and carbaryl (Wp 85%, 3 Kg ha À1 ) were foliar applied twice at the flowering stage at one-week intervals, respectively. Climate data during the experiment are presented in were considered for characterization per a list of descriptors presented by Mahajan et al. (2000). The qualitative traits were growth habit (GH), seed color (SC), seed shape (SS), seed texture

| Statistical analysis
The experiment was carried out in a randomized complete block design replicated three times. The descriptive statistics-range, mean, and standard deviation-of the accessions were computed. A chisquare test for qualitative traits was performed to evaluate the similarity of the distribution frequencies in chickpea collections. The Shannon-Weaver diversity index (H 0 ) was measured using the phenotypic frequencies of qualitative characters (Shannon & Weaver, 1949).
Correlation coefficients were estimated to determine the level of the interrelationship between the traits.
The PCA and cluster analyses were also conducted to analyze morphogenetic traits (Sneath & Sokal, 1973). The cluster analysis was performed using Ward's minimum variance dendrogram clustering (Ward, 1963). The PCA was performed to determine the characters accounted for the total variation. The data were standardized before computing principal calculating analysis. The estimation of genetic parameters, which included phenotypic coefficient of variation (PCV), genotypic coefficient of variation (GCV), and heritability (H 2 ), as a whole, gives an idea of the mode of gene action in the expression of a character (Burton & Devane, 1952;Jandong et al., 2020;Johnson et al., 1955;Majidi et al., 2009).
Environmental Variance (Ve) and error mean square (EMS) Genotypic, phenotypic, and environmental coefficients of variation were calculated as follows: where PCV% = phenotypic coefficient of variation, Vg = genotypic variance, Vp = phenotypic variance, GCV% = genotypic coefficient of variation, and X = average trait.
Heritability (H2) on entry mean basis was calculated as follows: Expected genetic advance (GA) was calculated as: where K = 2.06 at 5% selection intensity, H2 = heritability, and The statistical analyses were performed using Excel, SAS 9.1, and JMP 4.1.

| Agromorphological variation in chickpea germplasm
Means and ranges for 14 quantitative characteristics of the 445 chickpea accessions are presented in Table 1. The days to 50% emergence in 98% of chickpea accessions were between 50 and 60 days. A wide range was also observed for the days from 50% emergence to 50% flowering: <50 days in 33% of chickpea accessions (145 samples), between 50 and 60 days in 65% of the accessions (288 samples), and >60 days in 12 accessions (Table 1) Table 2). The H 0 of the growth habit (0.98) and leaf color (0.88) showed the highest value, indicating the high diversity of these traits (Table 2).

| Yield and yield components
Chickpea germplasm showed a wide range of yield and yield compo-  Figure 3a).
There was considerable variation in plant dry weight and grain yield in chickpea accessions (Table 1). The harvest index of chickpea accessions also widely ranged from 4 to 95 (Table 1). The results showed that 51% of the accessions had 11-20 g, 38% had >10 g, and 11% (50 accessions) had >21 g grain yield per plant ( Figure 5).

| Cluster analysis
Based on a grain yield of >21 g plant À1 , 50 accessions of chickpeas were selected, and cluster analysis was performed. Clusters analysis showed the distribution of accessions according to the adjusted mean of 14 traits ( Table 3). The accessions were grouped into six clusters ( Figure 4). The highest mean values for 100-grain weight, plant dry weight, and grain yield were recorded from cluster II (MCC2100, MCC1460). Cluster V was classified by days to 50% emerging, pod per plant, and filled pod percentage. The maximum mean values of days from 50% emerging to 50% flowering, plant height, and harvest index were recorded from accessions in clusters III, VI, and VI, respectively (Table 3).

| PCA
The    (Table 5). A positive correlation was also observed between the number of days to emergence with the days to flowering, plant height, 100-grain weight, and plant dry weight (Table 5).

| DISCUSSION
For effective and efficient germplasm collections in breeding programs, the description of agronomically important traits is an essential prerequisite (Byrne et al., 2018). Thus, the evaluation of morphological traits, individually or in combinations, is necessary to identify populations with desired characteristics (Archak et al., 2016).
Descriptor development is one of the first systematic attempts to document plant species diversity. A genetic statistical method for evaluating intraspecies diversity through morphological genes was proposed by Smiryaev and Bocharnikova (2002). Similar studies were carried out to develop descriptors for medicinal and aromatic herbs (Singh et al., 2003), fruit trees (subtropical and tropical) (Mahajan et al., 2002), cereals (Mahajan et al., 2000), and vegetables (Srivastava et al., 2001). Higher values of variances in plant height, 100-grain weight, and the number of pods per plant were also reported by previous studies in chickpea accessions. This indicates the importance of these traits in increasing plant productivity (Archak et al., 2016;Khan et al., 2006;Malik et al., 2014).
Low variability observed in some traits, namely, days to 50% emergence and days from 50% emergence to 50% flowering, indicates the limitation of selection based on these traits in the accessions. Analyzing yield components and their relative contribution to yield provides a better opportunity to select high-yielding accessions (Malik et al., 2014). More than 1000 morphological markers have been identified in barley (Hordeum vulgare) (Dhanapal & Govindaraj, 2015). A set of descriptors for Jatropha curcas and guayule (Parthenium argentatum Gray) were also developed based on the germplasm collected around India by Sunil et al. (2013) and Coffelt and Johnson (2011), respectively.
The frequency distribution of the six qualitative descriptors (plant growth habit, seed color, seed shape, testa texture, flower color, and leaf color) revealed a different degree of variation for different traits.
The Shannon-Weaver diversity index (H 0 ) is used in genetic studies as a convenient measurement of both allelic richness and evenness (Upadhyaya et al., 2006). The H 0 index was estimated to compare phenotypic variety. The value of H 0 for growth habit and leaf color showed the highest values, indicating the high diversity of those traits.

T A B L E 5 Correlation coefficients of quantitative descriptors in Kabuli collections
The values of genotypic and phenotypic variation coefficients for the traits given in Table 5 indicated significant variations among the accessions. The degree of variation in phenotypic traits associated with genetic variation is measured by heritability (Barreto et al., 2021). It suggests the likelihood of a selection benefit and/or a correlation response based on a range of environmental variables related to crop production (Malosetti et al., 2013;Nayak et al., 2014).
The heritability values of the number of pods per plant, dry weight per plant, and harvest index in the present study were above 0.96, 0.92, and 0.90, respectively, suggesting strong genetic components for the studied traits. High heritability accompanied by a low level of relationship between PCV and GCV for all traits indicated less environmental influence on their expression; hence, it can be used in establishing the distinctness between the chickpea accessions. These results were in agreement with the findings of Noor et al. (2003) and Farshadfar et al. (2013). Mushtaq et al. (2013) also reported the maximum heritability estimates for days to flowering, days to maturity, pods per plant, the total weight of plants, secondary branches per plant, plant height, 100-grains weight, and grain yield. Also, low heritability was also evidenced by low values of PCV and GCV for days to 50% emergence, the number of branches per plant, the leaf length, and the peduncle length. When the heritability estimates of qualitative and quantitative traits are high (i.e., >60%), the phenotypic appearance would provide a close assay of the genotypic value, and selection on the basis of phenotypic performance alone may be effective. If the heritability is low (i.e., <30%), the environmental influence is high in the expression of those traits (Joshi et al., 2018).
The genetic gain, which can be expected by the selection of a character, is estimated by the GA (Joshi et al., 2018). High heritability coupled with the GA observed for harvest index and percentage of fertile pods indicated the presence of additive gene action for this trait (Patil & Phadnis, 1977). Whereas high heritability coupled with low GA as observed for days from 50% emerging to 50% flowering, the number of leaflets per leaf, the number of leaflets per leaf, and the number of pods per plant revealed the presence of non-additive gene action. Bicer and Şakar (2007) have also reported that plant height showed moderate to low heritability, and environmental conditions played a major role in this trait. The present findings were also in agreement with the previous research (Ali et al., 2008;Patil et al., 2010).
The grain yield of a plant is determined by the interaction of various traits and is influenced by the genetic makeup and environment. In

CONFLICT OF INTEREST
The authors declare that there is no conflict of interest.

ETHICS APPROVAL
Not applicable.

DATA AVAILABILITY STATEMENT
None.