Phenotypic and Genetic Diversity of Chickpea (Cicer arietinum L.) Accessions from Kazakhstan

Abstract

Chickpea (Cicer arietinum L.) is a key legume crop of global economic and nutritional importance, yet its cultivation in Kazakhstan is constrained by a narrow genetic base and exposure to stress-prone environments. To characterize the diversity available for breeding and conservation, 27 accessions (22 kabuli and 5 desi) were evaluated for phenotypic and molecular diversity to assess its potential for use in breeding programs. Seven agronomic traits were assessed, including plant height, the first pod’s height, the number of main stems per plant, and seed yield components. The collection showed considerable variability across traits, with the plant height ranging from 37 to 75 cm and hundred-seed weight ranging from 21 to 42 g. Strong positive correlations between the number of fertile nodes, number of seeds per plant, and yield per plant (r > 0.83) highlighted their utility as indirect selection criteria. Genotyping with 28 SSR markers revealed 110 alleles (mean 3.9 ± 0.4 per locus) with moderate polymorphism (PIC = 0.493 ± 0.089). Loci CaM00495 and TAI71 were highly informative (PIC > 0.804), while two accessions showed low polymorphism, indicating genetic uniformity. Population structure analysis grouped accessions into four highly admixed clusters. Overall, Kazakh chickpea germplasm exhibits substantial phenotypic and genetic diversity under optimal conditions, providing valuable preliminary data for selecting parental lines for future breeding programs, which should include targeted stress screening to evaluate resilience.

1. Introduction

Chickpea (Cicer arietinum L.) is a vital grain legume extensively cultivated in arid and semi-arid regions, notably in Australia, Mediterranean countries, and Africa. As the second most important pulse crop globally, it provides a rich source of plant-based protein (18–22%), carbohydrates, and essential micronutrients, including iron, zinc, and folate, thereby meeting the nutritional demands of millions of people worldwide [1,2]. In addition to its nutritional value, chickpea enhances soil fertility through symbiotic nitrogen fixation, reducing dependence on synthetic fertilizers and contributing to sustainable agricultural systems [3,4]. This dual role as a nutritional and ecological asset makes chickpea a valuable component of crop rotation systems, particularly in water-scarce environments.

Globally, chickpea is cultivated on over 14 million hectares of land, with an annual production exceeding 17 million tonnes [5] in 2023. India is the largest producer, contributing to approximately 65% of the global output, followed by Australia, Turkiye, and Pakistan [5]. In Kazakhstan, chickpea cultivation is a relatively recent but rapidly expanding practice, with production increasing from 7108 tonnes in 2003 to 16,517 tonnes in 2023 [5]. This growth is driven by the crop’s suitability to the southern and southeastern regions (Almaty, Zhambyl, and Turkistan), where dryland farming systems face unique challenges such as water scarcity, extreme temperature fluctuations, and soil degradation [6,7,8]. Despite this expansion, Kazakhstani chickpea germplasm remains largely uncharacterized, limiting the development of locally adapted cultivars. Systematically evaluating this germplasm is critical to establish a baseline for breeding programs, especially in the context of climate change, which is shifting agricultural zones and intensifying environmental stresses in Central Asia [7,8]. This study addresses this gap by providing the first comprehensive assessment of phenotypic and genetic diversity in a Kazakhstani chickpea collection, offering novel insights for regional breeding efforts aimed at enhancing productivity and sustainability.

Chickpea exists as two main market types, namely desi and kabuli. Desi chickpeas have small, dark-colored seeds and are predominant in South Asia and parts of Africa, whereas kabuli chickpeas have larger, lighter seeds and are common in the Mediterranean and Central Asia, including Kazakhstan [1,9]. The genetic diversity of cultivated chickpea is limited due to a single domestication event in the Fertile Crescent approximately 10,000 years ago, involving its wild progenitor Cicer reticulatum Ladiz [10]. According to Vavilov, the primary centers of origin for chickpea are Southwest Asia (notably Afghanistan) and the Mediterranean, with Ethiopia as a secondary center [11], while the oldest carbonized seeds and distribution of the progenitor with the closest species were found in northern Syria and southern Turkiye, respectively. This indicates that the origin center of chickpea is Asia Minor [12]. This domestication bottleneck has constrained allelic variation, limiting breeding progress for complex traits such as yield, disease resistance, and tolerance to abiotic stresses [13,14,15].

Early diversity studies in chickpea relied on agro-morphological characterization and pedigree analysis. While these were useful for basic differentiation, they revealed the crop’s narrow genetic base and provided limited resolution for detecting fine-scale variation. This genetic constraint has exacerbated susceptibility to various biotic and abiotic stresses, reducing production potential. In response, advanced genomic approaches—such as genome sequencing, QTL mapping, GWASs, marker-assisted selection (MAS), and genomic selection (GS)—have been increasingly used to dissect the genetic architecture of complex traits and accelerate breeding progress [16,17,18]. The sequencing of desi [19] and kabuli [20] genomes has been a milestone, enabling the development of thousands of molecular markers, including single nucleotide polymorphisms (SNPs) and simple sequence repeats (SSRs) [21,22].

Among these, SSR markers are particularly valuable in chickpea research due to their co-dominant inheritance, high polymorphism, reproducibility, and wide genome coverage. They have been widely applied in genetic diversity assessment, linkage mapping, QTL identification, and MAS [23,24,25]. SSR-based diversity studies reveal allelic richness, identify genetically distinct groups, and facilitate selecting parental lines for breeding programs [26]. Such analyses have uncovered significant variations between desi and kabuli chickpea and identified genotypes with superior agronomic traits [26,27,28,29,30], offering pathways to broaden the genetic base and improve stress tolerance and yield stability [31].

In Kazakhstan, evaluating local germplasm is essential to understanding its phenotypic and genetic diversity, which can inform the development of improved cultivars. The country’s agroecological conditions, marked by extreme temperatures, limited rainfall, and, in some areas, saline soils, necessitate further stress-specific evaluations. SSR markers have been successfully applied in genetically characterizing local legumes, including soybean [32] and chickpea [33], highlighting their utility in resource-limited settings. At the same time, phenotypic diversity—encompassing traits such as plant height, seed size, flowering time, and stress tolerance—is equally crucial for breeding, as it directly influences yield and adaptability [34]. In Kazakhstan, where climatic stresses are severe, phenotypic variation is indispensable in identifying resilient genotypes [35,36]. Integrating phenotypic and molecular data enables breeders to target specific traits, such as drought tolerance and high seed yield, to develop cultivars suited to local needs.

Despite advances in next-generation sequencing (NGS) technologies, such as SNP arrays and genotyping-by-sequencing (GBS), SSRs remain a cost-effective and practical choice for genetic diversity studies, especially in resource-constrained environments [26]. Their minimal infrastructure requirements make them suitable for large-scale germplasm screening and developing core collections. In chickpea, SSRs have been instrumental in mapping 100-seed weight QTLs [37], QTLs associated with seed traits [38], drought tolerance [39], and ascochyta blight resistance [40], supporting MAS pipelines. However, despite the global progress in chickpea diversity studies, systematic assessments integrating both phenotypic and SSR-based genetic data remain scarce for Kazakhstan’s germplasm. This lack of integrated analysis limits the ability to design targeted breeding strategies for local environments.

The aim of this study was to evaluate both genetic and phenotypic diversity in a previously uncharacterized collection of chickpea accessions from Kazakhstan using polymorphic SSR markers and detailed phenotypic assessments under optimal conditions. To our knowledge, this is the first integrated analysis of Kazakh germplasm, providing a critical baseline for breeders in Central Asia, where climate change and shifting agricultural zones necessitate locally adapted cultivars. The specific objectives were to (1) characterize variation in key agronomic traits; (2) quantify allelic richness and polymorphism across accessions; (3) detect genetically distinct clusters and assess population structure; and (4) identify representative accessions for future breeding and conservation programs.

2. Materials and Methods

2.1. The Collection and Field Experiments

A panel of 27 chickpea accessions from Kazakhstan, including cultivars and breeding lines, was used to analyze phenotypic and genetic diversity (

Table 1. List of chickpea accessions used for phenotyping and genotyping in the study.

Accession ID	Accession Name	Type
CPKZ-03	13-B	kabuli
CPKZ-18	1-liniya	kabuli
CPKZ-09	28-B	desi
CPKZ-19	2-liniya	desi
CPKZ-10	32-B	kabuli
CPKZ-20	3-liniya	desi
CPKZ-21	7-liniya	kabuli
CPKZ-22	8-liniya	desi
CPKZ-04	F00-21	kabuli
CPKZ-01	F92-52	kabuli
CPKZ-05	F97-121	kabuli
CPKZ-32	F97-25/1	kabuli
CPKZ-12	F97-60	kabuli
CPKZ-06	F98-108c	kabuli
CPKZ-16	Icarda	kabuli
CPKZ-15	Kamila (check cultivar)	kabuli
CPKZ-26	Karabalykskaya-1	desi
CPKZ-17	Luch	kabuli
CPKZ-07	Miras07	kabuli
CPKZ-13	Nurly80	kabuli
CPKZ-27	Rozana	kabuli
CPKZ-14	Satti	kabuli
CPKZ-08	Sen-sen	kabuli
CPKZ-02	TH45-1-01	kabuli
CPKZ-25	Vektor	kabuli
CPKZ-24	Volzhanin	kabuli
CPKZ-23	Zavolzhski	kabuli

). The collection comprised accessions of two seed types, namely desi (n = 5) and kabuli (n = 22). The kabuli type predominated, reflecting its greater prevalence in Central Asia, including Kazakhstan, where local breeding programs are primarily oriented toward this type.

The experimental fields of the Kazakh Research Institute of Agronomy and Plant Growing (KRIAPI, Almalybak, Almaty region) were used for the experiments in 2024 (Figure S1). The local cultivar Kamila (CPKZ-15) was used as a control cultivar. Chickpea accessions were sown using a nearest-neighbor randomized complete block design (nn-RCBD) with randomly assigned accessions. Each accession was grown in individual 1 m² plots (25 cm spaces between neighboring plots) in two independent replications. Seven important agronomic traits, including morphology parameters (plant height—PH, cm; the first pod’s height—FPH, cm; and the number of main stems per plant—NMSP, count) and seed yield components (the number of fertile nodes—NFN, count; the number of seeds per plant—NSP, count; the hundred-seed weight—HSW, g; and the seed yield per plant—YP, g) of chickpea, were assessed. The field trials and trait assessment were performed according to standard protocols [41].

2.2. DNA Extraction and SSR Genotyping

The DNA was extracted from 5-day-old seedlings of chickpea accessions in three replicates using the DNeasy Plant Pro Kit (QIAGEN, Hilden, Germany). Genotyping of the chickpea panel was conducted using 28 SSR markers associated with important agronomic traits (

Table 2. The list of SSR markers used for genotyping.

SSR Marker	Associated Trait(s)	Reference
CaGM00495	Number of main stems per plant, number of seeds per plant	[42]
CaM0803	Plant height, seed yield per plant, 100-seed weight	[42]
CaSTMS2	Plant height, seed yield per plant	[43]
CaSTMS21	Seed yield per plant	[43]
ICCM0019b	Seed yield per plant	[44]
ICCM0034	First pod’s height, number of main stems per plant	[44]
ICCM0043	Plant height, height of lower pod	[44]
ICCM0089a	100-seed weight	[44]
ICCM0105	Plant height	[42]
ICCM0120b	Number of seeds per plant, 100-seed weight	[44]
ICCM0127	First pod’s height, number of main stems per plant	[44]
ICCM0190a	Seed yield per plant	[44]
ICCM0191	Seed yield per plant	[44]
ICCM0192a	First pod’s height, number of main stems per plant	[44]
ICCM0202b	100-seed weight	[44]
ICCM0243c	First pod’s height, number of main stems per plant	[44]
ICCM0249	First pod’s height, number of main stems per plant	[42]
NCPGR19	Number of seeds per plant, seed yield per plant	[45]
NCPGR223	100-seed weight	[42]
NCPGR7	Seed yield per plant	[45]
STMS11	Seed yield per plant	[42]
TA130	100-seed weight	[46]
TA200	First pod’s height, number of main stems per plant	[46]
TA22	Plant height, number of seeds per plant, seed yield per plant, 100-seed weight	[46]
TA46	Seed yield per plant	[46]
TA71	First pod’s height, number of main stems per plant	[46]
TA72	Seed yield per plant	[46]
TAA170	Seed yield per plant	[42]

and Table S1).

The PCR conditions were optimized in order to provide high amplification efficiency and accuracy. The PCR was performed in a total volume of 20 µL, comprising 20 ng of genomic DNA, 1 U of Taq polymerase, 0.2 mM of each deoxyribonucleotide triphosphate (dNTP), 10 pM of each primer, 1.5 mM of magnesium chloride (MgCl₂), and a standardized 1× Taq buffer solution. The reactions were performed using a SimpliAmp Thermal Cycler (Thermo Fisher Scientific, Singapore) with an initial denaturation step at 94 °C for 3 min, followed by 40 cycles of 94 °C for 30 s, annealing temperature (Ta °C) for 50 s, and 72 °C for 1 min 40 s. The final extension step was at 72 °C for 5 min. The PCR products were separated on a QIAxcel Connect System for capillary electrophoresis (QIAGEN, Hilden, Germany) using a QIAxcel DNA High Resolution Kit and QX Alignment Marker 15 bp/3 kb, as well as a QX Size Marker (50 bp/1 kb). Samples were processed using the standard OH500 method with an injection time of 20 s.

2.3. Statistics, Genetic Diversity, and Population Analysis

To statistically analyze the phenotypic data, R version 4.5.0 was used to calculate the minimum, maximum, mean, standard deviation (SD), and coefficient of variation (CV) values. The R package ggrepel version 0.9.5 was applied for UPGMA dendrogram construction, while ggplot2 version 3.5.1 and GGally version 2.2.1 were used for principal component analysis (PCA) and Pearson’s correlation analysis, respectively.

The polymorphism information content (PIC) index was calculated according to Botstein et al. [47]. Based on genetic variability data, markers with PIC > 0.50 were classified as highly informative, those with 0.25 < PIC ≤ 0.50 as informative, and PIC ≤ 0.25 as non-informative. To assess the discriminatory power of each marker, the number of alleles per locus (Na), the number of effective alleles (Ne), Shannon’s information index (I), expected heterozygosity or gene diversity (h), and unbiased expected heterozygosity (uh) were calculated using GenAlEx version 6.5 [48]. The same diversity indices, along with the percentage of polymorphic loci per accession (%P), were calculated for each accession using the same software.

The population structure of the studied chickpea collection was evaluated using the three following complementary approaches: Bayesian clustering, PCA, and pairwise genetic distance analysis. Bayesian clustering was performed in STRUCTURE version 2.3.4 using a Bayesian Markov chain Monte Carlo (MCMC) approach with admixture and correlated allele frequency models [49]. The number of hypothetical clusters (K) ranging from 1 to 10 was tested using 100,000 burn-in iterations followed by 100,000 recorded iterations. The optimal K value was determined by analyzing ΔK values with CLUMPAK [50]. PCA plots and the heatmap of the pairwise genetic distance matrix were generated using the R packages ggplot2 and pheatmap, respectively.

3. Results

3.1. Phenotypic Diversity of Chickpea Collection

The phenotypic evaluation of a collection comprising 27 chickpea cultivars and breeding lines from Kazakhstan, including 5 desi and 22 kabuli types (Figure 1), revealed significant variation across seven agronomic traits.

The data are summarized for desi, kabuli, and the entire collection, with the minimum, maximum, mean, standard deviation (SD), and coefficient of variation (CV, %) provided for each trait (

Table 3. Descriptive statistics of agro-morphological traits.

Desi (n = 5)
Trait	Min	Max	Mean	SD	CV (%)
Plant height (PH, cm)	50	75	62	10.5	17.0
First pod’s height (FPH, cm)	30	36	33	2.6	8.0
Number of main stems per plant (NMSP, count)	1.5	3.5	2.7	0.9	32.1
Number of fertile nodes (NFN, count)	10	57	33	18.9	58.0
Number of seeds per plant (NSP, count)	12	57	33	18.4	55.9
Seed yield per plant (YP, g)	5.8	18.7	10.9	5.9	54.0
100-seed weight (HSW, g)	24	38	31	5.6	1.8
Kabuli (n = 22)
Trait	Min	Max	Mean	SD	CV (%)
Plant height (PH, cm)	37	70	52	7.3	14.2
First pod’s height (FPH, cm)	11	36	25	5.4	21.9
Number of main stems per plant (NMSP, count)	1.3	3.9	2.5	0.7	28.0
Number of fertile nodes (NFN, count)	16	81	37	14.6	39.1
Number of seeds per plant (NSP, count)	17	81	37	14.6	39.0
Seed yield per plant (YP, g)	5.4	22.8	11.2	4.2	37.6
100-seed weight (HSW, g)	21	42	29	5.5	1.9
Whole collection (n = 27)
Trait	Min	Max	Mean	SD	CV (%)
Plant height (PH, cm)	37	75	53	8.7	16.3
First pod’s height (FPH, cm)	11	36	26	5.9	22.3
Number of main stems per plant (NMSP, count)	1.3	3.9	2.6	0.7	28.4
Number of fertile nodes (NFN, count)	10	81	37	15.2	41.6
Number of seeds per plant (NSP, count)	12	81	37	15.1	41.2
Seed yield per plant (YP, g)	5.4	22.8	11.2	4.4	39.8
100-seed weight (HSW, g)	21	42	29	5.5	1.9

Min—minimum; Max—maximum; SD—standard deviation; CV—coefficient of variation.

).

The chickpea collection exhibited substantial phenotypic variation across seven agronomic traits, with PH ranging from 37 to 75 cm (mean = 53 cm, SD = 8.7, CV = 16.3%), FPH from 11 to 36 cm (mean = 26 cm, CV = 22.3%), and HSW from 21 to 42 g (mean = 29 g, CV = 1.9%). Observational trends suggest that desi types (n = 5) generally displayed taller plants (mean = 62 cm) and larger seeds (mean = 31 g) compared to kabuli types (n = 22, mean PH = 52 cm, mean HSW = 29 g), while kabuli types tended to have higher NFN (mean = 37) and NSP (mean = 37). These trends should be interpreted cautiously due to the limited sample size of desi types, which may influence observed variability (e.g., higher CV for NFN = 58.0% and NSP = 55.9% in desi vs. 39.1% and 39.0% in kabuli). The entire collection displayed considerable diversity across all traits, highlighting its potential for breeding programs.

A correlation analysis was performed in order to assess the relationships among agronomic and agro-morphological traits, providing insight into potential indirect selection criteria for breeding (Figure 2).

The correlation analysis revealed distinct relationships among the measured traits, with several exhibiting strong and statistically significant associations. A strong positive correlation was observed between NFN and NSP (r = 1.000, p < 0.001). Strong positive correlations were also found between NFN and YP (r = 0.837, p < 0.001) and between NSP and YP (r = 0.836, p < 0.001). A moderately strong positive correlation was noted between PH and FPH (r = 0.740, p < 0.001). Weaker but statistically significant positive correlations were identified between PH and both NFN (r = 0.475, p < 0.05) and NSP (r = 0.475, p < 0.05). Other relationships showed weaker or non-significant correlations.

A PCA biplot and an unweighted pair group method with arithmetic mean (UPGMA) dendrogram were constructed to visualize the relationships between the seven measured variables and the two chickpea types (Figure 3). The two main axes accounted for 46.1% and 23.1% of the total variance, respectively, for a cumulative variance of 69.2% (Figure 3A).

PC1 primarily separates groups based on YP and HSW, which exhibit the longest vectors and contribute most significantly to the variation, with NFN and NSP also showing strong correlations along this axis. Kabuli types generally show higher values for these traits, clustering on the positive side of PC1, while desi types are more dispersed, with some on the negative side, indicating lower values. PC2 is mainly associated with PH and FPH, which are correlated and drive variation along this axis. Desi types exhibit greater variation in PC2, with some individuals showing higher PH and FPH. HSW, pointing negatively along PC1, is negatively correlated with NFN, NSP, and YP. The small angle between NFN, NSP, and YP vectors indicates strong positive correlations among these traits.

The dendrogram shows two major clusters at a Euclidean distance of approximately 100 (Figure 3B). The first major cluster is a diverse group containing both kabuli and desi accessions, suggesting a high degree of phenotypic similarity regardless of type. The second major cluster, which branches off at a distance of approximately 80, is more complex and can be further subdivided. One large sub-cluster is composed primarily of kabuli accessions, which are tightly grouped, indicating high phenotypic resemblance. Another sub-cluster contains a mix of both types but is also dominated by kabuli accessions. A distinct sub-cluster is formed by accessions CPKZ_12, CPKZ_22, CPKZ_07, CPKZ_05, and CPKZ_02, where CPKZ_22 is a desi type that groups closely with kabuli types. The dendrogram revealed significant intermixing of the two chickpea types in the collection. This suggests that phenotypic traits do not strictly align with the desi and kabuli classifications and that there is considerable phenotypic diversity within each group.

3.2. Genotyping and Genetic Diversity of the Collection

In total, 28 SSR markers were used to genotype the chickpea collection from Kazakhstan. Of these, twenty-three markers were polymorphic, revealing two to eight alleles per locus, while the remaining five markers were monomorphic. A representative fragment of the electropherogram is shown in Figure 4.

Genetic diversity assessment using 28 SSR markers revealed allelic richness in the chickpea collection from Kazakhstan, with a mean number of alleles (Na) of 3.929 ± 0.466 and number of effective alleles (Ne) of 2.747 ± 0.289 (

Table 4. Allelic variation in the microsatellite markers used for genotyping chickpea accessions.

SSR	Variation in Product Size (bp)	Na	Ne	I	h	uh	PIC
CaGM00495	276–375	8	6.031	1.927	0.834	0.865	0.842
ICCM0105	300–317	2	1.990	0.691	0.497	0.516	0.499
TAA170	200–300	7	3.015	1.467	0.668	0.693	0.683
STMS11	229	1	1.000	0.000	0.000	0.000	0.000
NCPGR223	254–277	3	1.931	0.768	0.482	0.500	0.491
ICCM0249	139–155	2	1.849	0.652	0.459	0.476	0.444
TA130	185–250	6	4.000	1.552	0.750	0.778	0.743
ICCM0019b	100–130	2	1.324	0.410	0.245	0.254	0.252
ICCM0120b	158–194	7	4.667	1.694	0.786	0.815	0.787
ICCM0192a	275–327	6	3.698	1.470	0.730	0.757	0.724
TA200	246–274	2	1.960	0.683	0.490	0.508	0.483
NCPGR7	200–213	2	1.324	0.410	0.245	0.254	0.252
NCPGR19	300–311	2	1.690	0.598	0.408	0.423	0.417
CaSTMS2	245–300	5	4.404	1.542	0.773	0.802	0.779
CaSTMS21	169–188	2	1.774	0.628	0.436	0.452	0.417
TA22	188–262	7	5.026	1.759	0.801	0.831	0.790
TA46	136–177	6	3.960	1.529	0.747	0.775	0.733
TA71	194–257	8	5.158	1.816	0.806	0.836	0.804
ICCM0127	300–376	7	4.612	1.711	0.783	0.812	0.793
ICCM0191	129–147	2	1.324	0.410	0.245	0.254	0.252
ICCM0243c	217–243	4	3.350	1.295	0.702	0.728	0.691
TA72	244–287	5	3.063	1.332	0.673	0.698	0.664
ICCM0202b	200–250	6	3.564	1.420	0.719	0.746	0.713
ICCM0190a	185–209	4	2.190	0.968	0.543	0.563	0.554
CaM0803	139	1	1.000	0.000	0.000	0.000	0.000
ICCM0034	270	1	1.000	0.000	0.000	0.000	0.000
ICCM0043	298	1	1.000	0.000	0.000	0.000	0.000
ICCM0089a	205	1	1.000	0.000	0.000	0.000	0.000
Mean	-	3.929	2.747	0.955	0.494	0.512	0.493
SE	-	0.466	0.289	0.123	0.055	0.057	0.089

Notes: Na—number of alleles; Ne—number of effective alleles; I—Shannon’s information index; h—expected heterozygosity or gene diversity; uh—unbiased expected heterozygosity; PIC—polymorphism information content; SE—standard error.

).

Based on criteria established in chickpea diversity studies, Na values are classified as moderate when ranging from three to five alleles per locus compared to higher values (>5) in diverse global collections or lower values (<3) in highly selected cultivars [26,51]. The mean expected heterozygosity (h = 0.494 ± 0.055) and Shannon’s information index (I = 0.955 ± 0.123) indicate moderate genetic diversity, as values of h > 0.60 and I > 1.5 are typically reported for wild-relative introgressed panels, while h < 0.30 and I < 0.7 indicate low diversity in domesticated germplasm [24,51,52]. The mean polymorphism information content (PIC) was 0.493 ± 0.089, with values ranging from 0.000 to 0.842. PIC values are classified as highly informative (>0.50), moderately informative (0.25–0.50), or non-informative (<0.25) [47]. Markers CaM00495 (PIC = 0.842), ICCM0120b (PIC = 0.787), and TA71 (PIC = 0.804) were highly informative, while five markers (e.g., STMS11, CaM0803) were non-informative (PIC = 0.000), indicating their limited utility for diversity studies [26]. These metrics collectively suggest moderate allelic richness suitable for breeding, with highly informative markers providing robust tools for genetic differentiation.

The genetic diversity of 27 chickpea accessions was assessed using SSR genotyping, and the results are summarized in

Table 5. Genetic diversity parameters of 27 chickpea accessions based on SSR genotyping.

Accession	Na	Ne	I	h	uh	%P
1liniya	1.107	1.086	0.068	0.048	0.071	10.71%
2liniya	1.036	1.029	0.023	0.016	0.024	3.57%
28-B	1.107	1.086	0.068	0.048	0.071	10.71%
32-B	1.429	1.386	0.254	0.167	0.250	32.14%
3liniya	1.179	1.143	0.114	0.079	0.119	17.86%
7liniya	1.214	1.171	0.136	0.095	0.143	21.43%
8liniya	1.393	1.343	0.238	0.159	0.238	32.14%
F92-52	1.107	1.086	0.068	0.048	0.071	10.71%
F97-121	1.393	1.314	0.250	0.175	0.262	39.29%
F97-25/1	1.500	1.400	0.318	0.222	0.333	50.00%
F97-60	1.393	1.314	0.250	0.175	0.262	39.29%
F98-108c	1.214	1.171	0.136	0.095	0.143	21.43%
F00-21	1.214	1.171	0.136	0.095	0.143	21.43%
Sen-sen	1.464	1.414	0.277	0.183	0.274	35.71%
TH45-1-01	1.500	1.414	0.312	0.214	0.321	46.43%
Volzhanin	1.107	1.086	0.068	0.048	0.071	10.71%
Kamila	1.500	1.429	0.306	0.206	0.310	42.86%
Karabakykskaya-1	1.179	1.143	0.114	0.079	0.119	17.86%
Miras07	1.464	1.400	0.283	0.190	0.286	39.29%
Nurly80	1.429	1.343	0.273	0.190	0.286	42.86%
Rozana	1.321	1.286	0.192	0.127	0.190	25.00%
Satti	1.393	1.329	0.244	0.167	0.250	35.71%
Icarda	1.179	1.143	0.114	0.079	0.119	17.86%
Vektor	1.143	1.129	0.085	0.056	0.083	10.71%
Zavolzhski	1.071	1.057	0.045	0.032	0.048	7.14%
13-B	1.500	1.414	0.312	0.214	0.321	46.43%
Luch	1.571	1.500	0.345	0.230	0.345	46.43%
Mean	1.301	1.252	0.187	0.128	0.192	27.42%
SE	0.018	0.016	0.011	0.008	0.011	2.69%

Notes: Na—number of alleles; Ne—number of effective alleles; I—Shannon’s information index; h—expected heterozygosity or gene diversity; uh—unbiased expected heterozygosity; %P—percentage of polymorphic loci per accession; SE—standard error.

.

The Na per accession was 1.301 ± 0.018, and Ne per accession was 1.252 ± 0.016. The mean I was 0.187 ± 0.011, indicating the overall allelic diversity and richness across the accessions. The mean h was 0.128 ± 0.008, while the mean uh was 0.192 ± 0.011. The percentage of polymorphic loci (%P) per accession ranged from 3.57% to 50.00%, with a mean of 27.42 ± 2.69%. Several accessions showed higher levels of genetic diversity. For instance, F97-25/1 and Luch had the highest %P of 50.00% and 46.43%, respectively, and also showed high values for Na and Ne. Conversely, accessions like 2linya and Zavolzhski displayed diversity below 10%, with %P values of 3.57% and 7.14%, respectively.

3.3. Population Structure of Collection

Genotypic data from 28 SSR markers for 27 chickpea accessions were used to assess the population structure using Bayesian clustering in STRUCTURE (Figure 5), PCA, and pairwise genetic distance analysis (Figure 6).

The distribution of Delta K values across K values from 2 to 9 determined the highest peak at K = 4 (Delta K = 25.256), indicating the most likely number of genetic clusters within the 27 chickpea accessions of the two types (Figure 5A). Secondary peaks were noted at K = 2 and K = 5. The STRUCTURE-based population structure of the 27 chickpea accessions is illustrated in Figure 5B using a bar plot at K = 4, where each accession is represented by a vertical bar segmented into four colors corresponding to the inferred genetic clusters. The proportion of membership in each cluster varies across accessions, with notable admixture observed. For instance, accessions CPKZ_07, CPKZ_12, and CPKZ_19 exhibit high proportions (>0.7) in a single cluster, while others like CPKZ_08, CPKZ_17, and CPKZ_20 show more balanced contributions from multiple clusters, indicating significant genetic admixture within the collection. The plot reveals that the kabuli and desi types do not form distinct separate clusters but instead share a complex genetic heritage.

A PCA plot of the first two principal components (PC1 and PC2) accounts for 22.71% and 18.66% of the total variance, respectively (Figure 6A). These percentages suggest that the genetic variation is distributed across multiple components rather than being dominated by the first two.

The plot visually confirms the high level of genetic similarity and admixture identified in the population structure analysis. The desi and kabuli accessions are not separated into distinct groups. Instead, they are highly intermingled across the biplot, suggesting a lack of clear genetic differentiation between the two types.

A heatmap and dendrogram (Figure 6B) illustrate the pairwise genetic distances among the 27 accessions. The heatmap shows several small, tightly knit clusters of genetically similar accessions. However, the overall pattern shows a high degree of intermixing, with no clear segregation of the desi and kabuli types into large distinct clusters. This is consistent with the results from the population structure and PCA analyses.

Thus, the combined results from the STRUCTURE, PCA, and pairwise genetic distance analyses indicate the presence of four main genetic clusters among the 27 chickpea accessions, with substantial admixture and no clear genetic separation between the desi and kabuli types. This suggests a shared and complex genetic background within the collection.

4. Discussion

4.1. Phenotypic Diversity of Germplasm from Kazakhstan

The phenotypic evaluation of the 27 chickpea accessions from Kazakhstan revealed a moderate level of variation across seven key agronomic traits under optimal conditions, establishing a critical baseline for targeted breeding in a previously uncharacterized regional germplasm collection. This variation is particularly valuable in Central Asia, where climate change is intensifying challenges such as limited rainfall, extreme temperature fluctuations, and soil degradation, necessitating locally adapted cultivars [7,8]. The data provide novel insights into the phenotypic diversity of Kazakhstani chickpea, supporting the identification of accessions with favorable traits for future breeding programs, with further evaluation under stress conditions needed to confirm their suitability. Observational trends suggest that desi types (n = 5) generally exhibited taller plants (mean PH = 62 cm), higher first pod heights (mean FPH = 33 cm), and larger seeds (mean HSW = 31 g) compared to kabuli types (n = 22, mean PH = 52 cm, mean FPH = 25 cm, mean HSW = 29 g). Conversely, kabuli types tended to show higher numbers of fertile nodes (mean NFN = 37) and seeds per plant (mean NSP = 37). These trends are limited by the small sample size of desi types, which may contribute to higher observed variability in yield-related traits such as NFN (CV = 58.0%), NSP (CV = 55.9%), and YP (CV = 54.0%) compared to kabuli types (CV = 39.1%, 39.0%, and 37.6%, respectively). The substantial diversity across the entire collection highlights its value for breeding programs targeting stress resilience and productivity. This division of trait emphasis underscores the value of maintaining both seed types in breeding programs, as they contribute complementary traits that can be exploited for hybrid vigor and stress adaptation.

The high coefficients of variation (CV) for the number of fertile nodes (41.6%) and number of seeds per plant (41.2%) in our collection (Table 3) highlight the influence of genotype. These high CVs indicate that yield components in chickpea are highly responsive to environmental variability, making them useful indicators of stability under stress-prone conditions. This observation is consistent with findings from diverse chickpea collections, where yield components such as total seeds per plant exhibited CVs nearing 35% under variable growth conditions [53,54]. Such results emphasize that phenotypic plasticity is an important breeding target in semi-arid regions, where fluctuations in rainfall and temperature are common.

Correlation analysis further identified strong positive associations between NFN, NSP, and YP (r > 0.83, p < 0.001) (Figure 2), suggesting these traits as reliable indirect selection criteria for yield improvement. This finding is particularly valuable for breeders, as direct yield selection can be unstable under variable environments, whereas correlated traits such as node and seed number provide more reliable indicators of genetic potential. Such correlations have been corroborated in other Kazakh germplasm collections of chickpeas, where seed weight per plant was positively correlated with the number of productive nodes, number of branches, and number of pods per plant, as well as the number of productive nodes being strongly positively correlated with the number of pods per plant [33]. These consistent correlations across different germplasm panels reinforce the potential of using trait-based selection in marker-assisted breeding strategies for enhanced productivity.

In our dataset, phenotype-based PCA (69.2% variance explained) and UPGMA indicated substantial intermixing of the desi and kabuli accessions (Figure 3). The lack of strict clustering between market classes suggests that agro-morphological traits alone are insufficient to differentiate between seed types in modern breeding lines. A similar lack of strict separation between the two types has been widely reported at the genomic level, reflecting shared pedigrees and desi × kabuli introgression in modern breeding and supporting the plausibility of the phenotypic overlap we observed [55]. This intermixing underscores the influence of breeding practices that deliberately blur distinctions between market classes in order to combine desirable traits [56].

The phenotypic evaluation of the 27 chickpea accessions from Kazakhstan revealed substantial variation across seven key agronomic traits, highlighting the considerable potential for targeted breeding in local germplasm. Such variation is of particular importance in regions like Central Asia, where extreme temperature fluctuations, limited rainfall, and soil degradation exert strong selective pressures. Overall, the phenotypic diversity observed in this study supports the selection of high-performing accessions like Kamila (CPKZ-15) for YP and F98-108c (CPKZ-06) for HSW, offering immediate value for Kazakhstan’s dryland farming systems, where water scarcity amplifies trait variability. Moreover, the diversity found in traits such as PH and NSP provides a useful resource for improving both stress tolerance and productivity in environments facing increasing climatic uncertainty.

4.2. Genetic Diversity and Population Structure

Genetic diversity assessment using 28 SSR markers revealed allelic richness in the chickpea collection from Kazakhstan, with a mean number of alleles (Na = 3.929 ± 0.466) and number of effective alleles (Ne = 2.747 ± 0.289) (Table 4). These values indicate moderate allelic richness, as Na values of 3–5 alleles per locus are typical for domesticated chickpea germplasm compared to >5 for diverse global or wild-introgressed collections and <3 for highly selected cultivars [26]. This moderate diversity reflects the genetic bottleneck associated with chickpea’s single domestication event but suggests sufficient variability for breeding [51]. Highly informative markers such as CaGM00495 (Na = 8, PIC = 0.842) and TA17 (Na = 8, PIC = 0.804) are classified as highly informative (PIC > 0.50), contrasting with monomorphic markers like STMS11 and CaM0803 (PIC = 0.000), which are non-informative (PIC < 0.25) [47]. These highly informative markers are valuable for differentiating accessions and identifying unique alleles, as reported in similar studies [26]. These metrics align with Indian chickpea germplasm studies, where PIC ranged from 0.27 to 0.74 (mean 0.66) across 30 SSRs, emphasizing the continued efficacy of SSR markers in capturing genetic variation despite domestication constraints [51].

The mean expected heterozygosity (h = 0.494 ± 0.055) and Shannon’s information index (I = 0.955 ± 0.123) further support moderate diversity, as h values of 0.30–0.60 and I values of 0.7–1.5 are typical for cultivated chickpea compared to h > 0.60 and I > 1.5 for wild-relative panels and h < 0.30 and I < 0.7 for low-diversity collections [24,51,52]. These metrics align with Indian chickpea studies, where the PIC ranged from 0.27 to 0.74 (mean 0.66) across 30 SSRs, confirming the efficacy of SSR markers in capturing variation despite domestication constraints [51]. These findings suggest that while Kazakh germplasm has adequate diversity to support selection, its expansion with wild relatives or exotic introductions could further enhance its utility. Per-accession diversity varied, with F97-25/1 and Luch showing high polymorphic loci percentages (50.00% and 46.43%, respectively) and Na values (1.500–1.571) (Table 5), indicating greater diversity, likely due to admixture. Conversely, accessions like 2-liniya and Zavolzhski exhibited low diversity (%P < 10%, Na < 1.1), consistent with intense selection or genetic drift, as reported in studies of selected cultivars [51]. This pattern mirrors SSR-based findings in Iranian germplasm, where landraces demonstrated higher allele numbers (mean = 4.2) and marker informativeness compared to cultivars, highlighting the progressive loss of diversity through modern breeding [57].

Population structure analysis revealed a complex genetic landscape, with STRUCTURE identifying four clusters (K = 4, ΔK = 25.256) (Figure 5A) and showing significant admixture with no clear separation between the desi and kabuli types. This observation was reinforced by PCA (41.37% variance explained) (Figure 6A) and pairwise distance analyses (Figure 6B), all of which showed high overlap between the two seed types. Such patterns are consistent with findings in Ethiopian chickpea germplasm: for instance, Getahun et al. (2021) [58] identified six genetic subpopulations with high admixture across market types, and Admas et al. (2021) [59] reported that clustering did not correspond to desi–kabuli classifications due to seed exchange and shared breeding pedigrees. Similarly, the PCA biplot showed intermingled distribution, consistent with global panels where PC1 and PC2 captured 20–30% variance without clear desi–kabuli bifurcation, reflecting environmental adaptation over market type [60]. The dendrogram’s mixed clusters corroborate this, resembling South Asian collections where SSR-based phylogenies revealed intermixing due to introgression and shared ancestry [26].

These findings collectively indicate that Kazakh chickpea germplasm harbors considerable genetic diversity under optimal conditions, supporting the selection of parental lines for breeding programs aimed at improving yield and agronomic performance. Future evaluations under stress conditions, such as drought or salinity, are needed to assess the potential of these accessions for challenging environments. While the phenotypic evaluation showed a clear separation between the desi and kabuli types based on seed traits, the molecular analysis revealed a more complex population structure that was not strictly aligned with market type. This suggests that while agro-morphological traits are distinct, there may be shared genetic backgrounds or historical gene flow between the types within this specific germplasm collection. This integration enhances breeding precision, as demonstrated in global studies combining SSR genotyping with trait evaluation to select core collections for marker-assisted selection.

The current results provide foundational insights for conserving and utilizing Kazakh chickpea diversity, identifying accessions such as 32-B and Kamila as promising candidates for yield enhancement due to their favorable agronomic traits under optimal conditions. These accessions form a valuable basis for breeding programs, but further evaluation under stress conditions is essential to assess their suitability for challenging environments.

Beyond their agronomic and genetic significance, the diversity observed in the Kazakh chickpea germplasm also carries important economic implications. Chickpea is increasingly valued as a low-input, high-value crop for dryland systems, requiring less fertilizer compared to cereals while contributing to soil fertility through biological nitrogen fixation. Developing improved varieties with enhanced yield stability and stress tolerance could reduce production risks for farmers in Kazakhstan’s semi-arid regions, lowering the reliance on costly inputs and stabilizing household incomes. Moreover, expanding the chickpea sector may diversify national agricultural exports, create new opportunities in value-added processing, and support regional food security under changing climate and market conditions.

5. Conclusions

This study revealed a moderate level of phenotypic and genetic diversity within the Kazakh chickpea germplasm collection, providing a foundation for breeding programs aimed at improving agronomic traits under optimal conditions. Observational trends suggest that desi types (n = 5) generally displayed robust agro-morphological traits, such as taller plants and larger seeds, while kabuli types (n = 22) tended to exhibit higher reproductive traits, such as greater numbers of fertile nodes and seeds per plant. These observations are limited by the smaller sample size of desi types, which may influence trait variability. SSR-based genetic analysis revealed moderate allelic richness and a complex population structure, comprising four admixed clusters. These findings address the research objectives by identifying breeding potential in accessions such as Kamila (CPKZ-15) and 32-B (CPKZ-10), which exhibited favorable agronomic traits, such as high YP = 22.8 g for Kamila and NSP = 55.7 for 32-B, under optimal conditions. The integration of phenotypic and molecular data highlights the effectiveness of a combined approach for characterizing chickpea diversity and selecting parental lines for breeding. The results provide a foundation for conserving both seed types and developing improved cultivars, but further evaluation under arid and semi-arid conditions, including water scarcity and soil degradation, is essential to confirm their performance in challenging environments. Beyond their biological value, these findings highlight the economic potential of chickpea improvement for Kazakhstan. Developing stress-resilient cultivars can strengthen food self-sufficiency, reduce losses in drought-prone areas, and lower input costs. As a profitable rotation crop that enriches soils, chickpea breeding from this germplasm can support more sustainable and competitive farming, reinforcing its role as a strategic crop for the national economy.