Abstract Cowpea (Vigna unguiculata [L.] Walp.) is an important food legume whose cultivated germplasm is generally characterized by limited genetic diversity resulting from domestication and predominant self-pollination. This study evaluated the agro-morphological and nutritional diversity of 21 farmer-maintained cowpea accessions collected from three geo-ecological zones of northern Thailand; western, eastern, and southwestern. A representative subset comprising nine landrace accessions and one commercial check was further analysed using start codon targeted (SCoT) markers to provide a preliminary assessment of molecular diversity. Considerable variation was observed among accessions for qualitative and quantitative traits. The cumulative Shannon diversity index for qualitative traits was highest in the southwestern collection (H'=3.80), followed by the eastern (H'=3.34) and western (H'=2.42) collections. Seed protein content was highest in accession SW5 (36.00 g/100 g), seed iron content was highest in accession W4 (78.19 mg/kg), and crude fibre content was highest in accession W6 (5.97 g/100 g), with different accessions exhibiting superior values for individual nutritional traits. Molecular analysis of the selected subset showed that 68% of the SCoT marker variation occurred within accessions, whereas 19% and 13% occurred among accessions within zones and among zones, respectively. The selected accessions exhibited moderate genetic differentiation (G_ST=0.41), and principal coordinate analysis indicated partial grouping according to their collection zones. Overall, this study provides baseline information on the agro-morphological, nutritional, and preliminary molecular diversity of northern Thai cowpea germplasm and identifies accessions that warrant further evaluation for germplasm conservation and future cowpea improvement.

Keywords: Germplasm characterization; Population differentiation; Phenotypic evaluation; Protein content; SCoT marker

Funding: The current work was supported by Chiang Mai University through the CMU Presidential Scholarship (2021).

Citation:  Mahir, M.Z.A., Rerkasem, B., Jamjod, S., and Pusadee, T. 2027. Agro-morphological, nutritional, and preliminary molecular diversity of northern Thai cowpea (Vigna unguiculata) landraces from three geo-ecological zones. Natural and Life Sciences Communications. 26(1): e2027013.

Graphical Abstract:

INTRODUCTION

Cowpea (Vigna unguiculata [L.] Walp.) is a drought-resilient legume of the Fabaceae that is cultivated worldwide for dry grain, tender pods, leafy vegetables, and fodder (Coulibaly et al., 2002; Xiong et al., 2016; Kebede and Bekeko, 2020). Global production is approximately 5.8 million tonnes annually (Da Silva et al., 2019), with the largest production concentrated in Africa, particularly Nigeria and Niger, although cowpea is also widely cultivated in South and Southeast Asia. Beyond its agronomic importance, cowpea is recognized as a nutritionally valuable crop because its seeds typically contain 23–35 g/100 g protein with high lysine content and good digestibility, making it an important source of dietary protein in developing regions (Boukar et al., 2019; Padhi et al., 2022). Cowpea grains are also rich in iron, crude fibre, and essential amino acids, providing nutritional benefits that complement cereal-based diets (Boukar et al., 2019; Abebe and Alemayehu, 2022; Padhi et al., 2022; Shevkani et al., 2025).

In Thailand, the yard-long bean subspecies (Vigna unguiculata ssp. sesquipedalis) occupies approximately 18,560–20,160 hectares annually (Sarutayophat et al., 2007; Benchasri et al., 2012), and Thailand is considered a secondary centre of Vigna diversity (Kaewwongwal et al., 2013). While yard-long bean is an important vegetable crop throughout Southeast Asia (Kongjaimun et al., 2012), the traditional prostrate cowpea ("Tua Lod") is commonly cultivated in upland rice-based farming systems in northern Thailand, where it contributes to soil conservation and biological nitrogen fixation (Punyalue, Jamjod, et al., 2018; Punyalue, Jongjaidee, et al., 2018). Despite its local importance, the genetic diversity of Thai cowpea landraces has received relatively little attention. Since domestication has resulted in a comparatively narrow genetic base, low within-population diversity and high inbreeding have been reported in cowpea collections from several countries, including Korea, Algeria, and China (Li et al., 2001; Ghalmi et al., 2010; Chen et al., 2017; Seo et al., 2020). Characterizing local germplasm is therefore important for its effective conservation and utilization (Sarutayophat et al., 2007; Xiong et al., 2016; Seo et al., 2020).

Previous studies have shown that cowpea exhibits low within-accession diversity but considerable variation among geographically distinct collections (Li et al., 2001; Huynh et al., 2013; Wamalwa et al., 2016; Xiong et al., 2016; Chen et al., 2017). Such diversity provides valuable information for germplasm management, conservation, and the identification of useful genetic resources for crop improvement (Fang et al., 2007; Tan et al., 2012). However, because morphological characteristics may be influenced by environmental conditions, integrating phenotypic evaluation with molecular marker analysis provides a more comprehensive assessment of diversity. Conventional PCR-based markers, including RFLP, RAPD, AFLP, SSR, and ISSR, have been widely used to investigate cowpea genetic diversity (Coulibaly et al., 2002; Patel et al., 2022). Among these marker systems, start codon targeted (SCoT) markers have demonstrated their usefulness for detecting genetic polymorphism in cowpea and other crop species (Igwe et al., 2017; Vanmathi et al., 2021).

Characterizing northern Thai cowpea landraces will improve our understanding of their agro-morphological and nutritional diversity while providing baseline information to support germplasm conservation and breeding efforts. Therefore, the objectives of this study were to (i) characterize the agro-morphological traits and nutritional composition of northern Thai cowpea landraces, (ii) quantify phenotypic diversity across geo-ecological zones, and (iii) conduct a preliminary assessment of genetic diversity and differentiation among representative accessions from the three geo-ecological zones using SCoT markers. The findings provide baseline information on the diversity of northern Thai cowpea germplasm and identify potential genetic resources for conservation and future evaluation and breeding.

MATERIALS AND METHODS

Plant materials
Twenty-one farmer-maintained cowpea accessions were obtained from the Highland Research and Development Institute (HRDI), Chiang Mai, Thailand. The accessions originated from three geo-ecological zones of northern Thailand and comprised nine accessions from the western zone (W), five from the eastern zone (E), and seven from the southwestern zone (SW) (Figure 1 and Table 1). After collection, the seeds were manually cleaned, labelled using their accession codes and available passport data, and grouped according to their zones of origin. Several accessions were collected from the same village or collection site but represented different farmer-maintained landraces and were therefore maintained as separate accessions. The elevation of the original collection sites varied among accessions, as shown in Table 1. However, elevation was included only as passport data describing the original collection sites and was not treated as an experimental factor in the present study because all accessions were evaluated under the same field conditions.

Figure 1. Sampling locations of 21 cowpea accessions in northern Thailand grouped into three geo-ecological zones: Western (W), Eastern (E), and Southwestern (SW). Several accessions were collected from the same geographic location and therefore share identical map coordinates.

Table 1. Collection sites, geographic coordinates, elevation, and geo-ecological zones of the 21 farmer-maintained cowpea (Vigna unguiculata [L.] Walp.) accessions collected in northern Thailand.

Note: Elevation represents the altitude of the original collection site obtained from passport information. All accessions were evaluated under the same field conditions at Chiang Mai University.

Field experiment

A field experiment was conducted at the Agricultural Innovation Research, Integration, Demonstration, and Training Center, Chiang Mai University, Thailand (18° 45' 33.145" N, 98° 55' 41.507" E) (Supplementary Figure S3). Twenty-one local cowpea accessions and one commercial cultivar used as a check were evaluated in a randomized complete block design with three blocks. The experimental area measured 11 × 6 m and was divided into plots of 1.0 × 1.2 m. Each plot contained eight planting hills, with spacing of 30 cm between hills within rows and 50 cm between rows. Three seeds were sown per hill and thinned to one plant per hill one week after emergence. Pre-emergence weed control was conducted using alachlor, followed by manual weeding at three and five weeks after planting. No fertilizer was applied during the growing period because the field evaluation was conducted under uniform low-input management conditions. The same field preparation and crop management practices were applied to all accessions. Therefore, the observed performance represents accession responses under the specific field and management conditions of this study.

Agro-morphological characterization

Agro-morphological characterization was conducted according to the cowpea descriptors of the International Board for Plant Genetic Resources (International Board for Plant Genetic Resources, 1983). A total of 22 traits were recorded, including ten qualitative traits, such as seed coat characteristics and flower colour, and twelve quantitative traits, such as pod length, seed dimensions, and hundred-seed weight. Quantitative plant traits were measured from five randomly selected competitive plants in each plot of each block. The mean of the five plants within a plot was used as the experimental value for the corresponding accession in each block. Consequently, the three field blocks represented three independent biological replicates in the statistical analysis.

Seed nutritional analysis

Mature seeds were harvested separately from each experimental block, cleaned, and oven-dried at 60°C until a constant weight was obtained. The dried seeds from each block were milled separately into a fine powder and analysed at the Central Laboratory, Faculty of Agriculture, Chiang Mai University, Chiang Mai, Thailand. Thus, the seed samples from the three field blocks represented three biological replicates for each accession. Accession E2 was excluded from nutritional analysis because the harvested seed quantity was insufficient to complete all laboratory determinations. Seed protein content was determined using the Kjeldahl method (AOAC, 2005) and calculated by multiplying total nitrogen concentration by a conversion factor of 6.25. For iron determination, ground seed samples were dry-ashed and the resulting ash was dissolved in acid before measurement using an atomic absorption spectrophotometer (AA240FS, Agilent Technologies, Santa Clara, CA, USA) according to standard procedure (AOAC, 2005). Crude fibre content was determined using the Weende method (AOAC, 2005). Any repeated measurements performed on the same block sample were regarded as analytical replicates and were not treated as independent biological replicates.

Selection of accessions for preliminary molecular analysis

Because molecular characterization was intended as a preliminary assessment rather than a comprehensive survey of all 21 landraces, a stratified subset was selected from the three geo-ecological zones. Three accessions were selected from each zone: W3, W6, and W9 from the western zone; E3, E4, and E5 from the eastern zone; and SW1, SW3, and SW6 from the southwestern zone. One commercial cultivar was included as a reference, resulting in a total of ten entries for molecular analysis. The local accessions were selected after the agro-morphological evaluation to include contrasting phenotypic profiles within each zone, particularly differences in plant growth habit, pod characteristics, and seed morphology. The same number of accessions was selected from each zone to provide balanced geographical representation. The molecular subset was therefore not randomly selected and was not intended to represent the complete genetic diversity of all 21 accessions.

DNA extraction and SCoT amplification

Young leaf tissue was collected separately from five individual plants of each selected accession. Genomic DNA was extracted using a cetyltrimethylammonium bromide-based method originally described by Doyle and Doyle (1987). The extracted DNA from each plant was analysed individually, and the five plants represented within-accession samples. Polymerase chain reaction amplification was performed using start codon targeted markers in a total reaction volume of 10 µL following the procedures described by Collard and Mackill (2009) and Igwe et al. (2017). The amplification programme consisted of an initial denaturation at 94°C for 5 min, followed by 30 cycles of denaturation at 94°C for 30 s, primer annealing for 1 min at the primer-specific temperatures shown in Table 2, and extension at 72°C for 1 min. A final extension was performed at 72°C for 10 min. The PCR products were mixed with 1 μL of bromophenol blue loading dye and separated by electrophoresis on 1.5% (w/v) agarose gels stained with MaestroSafe™ Nucleic Acid Stain (Maestrogen, Hsinchu City, Taiwan). Amplified fragments were visualized under ultraviolet illumination using a BioDoc-It™ 2 Imaging System (Analytik Jena, Jena, Germany). Only clearly resolved and reproducible bands were included in the subsequent marker scoring.

Table 2. SCoT primer sequences with GC content (%) and annealing temperature (Tm °C)

Data analysis

Qualitative traits and Shannon diversity index

Shannon's diversity index (H) was calculated separately for each qualitative trait following Shannon and Weaver (1949):

where p_i is the frequency of the ith phenotypic class and k is the number of observed phenotypic classes for that trait.

To facilitate comparisons among traits with different numbers of phenotypic classes, the Shannon diversity index was standardized by its theoretical maximum, H_max=ln(k), to obtain the relative Shannon–Weaver diversity index:

following Hennink and Zeven (1990). The overall diversity value for each geo-ecological zone was calculated as the sum of the standardized Shannon diversity indices across the ten qualitative traits. Consequently, the cumulative diversity values are not restricted to a maximum of 1 and should be interpreted as an overall diversity score rather than as a single normalized Shannon diversity index.

Figure 2. Total Shannon diversity indices (H′) for qualitative morphological traits across four zones: commercial cultivar, western, eastern, and southwestern zones. Higher values indicate greater within zones phenotypic diversity

Quantitative traits and seed nutritional composition

The experimental unit for agro-morphological and nutritional traits was the individual plot within each block. Data from the three blocks were treated as three biological replicates for each accession. Statistical analyses were conducted using Statistix 8 (Analytical Software, 2003) and R version 4.1.0 (R Core Team, 2021) with the agricolae package (de Mendiburu, 2021). Descriptive statistics were initially calculated for all quantitative traits. Differences among accessions were subsequently examined using analysis of variance based on the randomized complete block design, with accession treated as the main effect and block included to account for field variation. When the overall accession effect was significant at P < 0.05, mean comparisons were performed using Fisher’s protected least significant difference test at α =0.05. The protected procedure was applied only after a significant overall analysis of variance to reduce the likelihood of conducting unnecessary pairwise comparisons. Because the test is less conservative than procedures such as Tukey’s test, the resulting mean groupings were interpreted as exploratory comparisons among accessions. Principal component analysis was performed using XLSTAT 2021 (Addinsoft, 2021) to summarize multivariate variation among accessions based on the measured quantitative traits. Variables were standardized before analysis because they were expressed in different units. The PCA was used to describe major patterns of trait variation and accession dispersion and was not treated as independent evidence of geographic population structure.

SCoT marker scoring and genetic diversity analysis

Clearly amplified SCoT fragments were scored for each individual plant as present (1) or absent (0), producing a binary data matrix. Because SCoT markers are dominant, the presence and absence of bands were treated as dominant marker states rather than directly observed codominant alleles. Population genetic statistics, including the observed number of alleles, effective number of alleles, expected heterozygosity, genetic differentiation (G_ST), and the indirect estimate of gene flow (N_m) were calculated using PopGene version 1.32 (Yeh et al., 1997). The N_m value was reported as a conventional indirect estimate derived from (G_ST). Because this estimate relies on simplifying assumptions that may not be fully applicable to predominantly self-pollinating species or to a non-randomly selected subset, it was interpreted cautiously and was not considered a direct measurement of contemporary seed or pollen movement. Additional diversity statistics, Shannon’s information index, analysis of molecular variance, and principal coordinate analysis were calculated using GenAlEx version 6.5 (Peakall and Smouse, 2012). Analysis of molecular variance was used to partition marker variation among geo-ecological zones, among accessions within zones, and within accessions.

Genetic relationships among the ten entries were examined using Nei’s genetic distance in NTSYSpc version 2.10e (Rohlf, 2000). A dendrogram was constructed using the unweighted pair-group method with arithmetic mean through the sequential agglomerative hierarchical and nested clustering procedure. The UPGMA and principal coordinate analyses were used to describe genetic relationships within the selected molecular subset. They were not used to infer the genetic structure of the full set of 21 northern Thai cowpea accessions.

RESULTS

Agro-morphological variation

Qualitative traits

The 21 farmer-maintained cowpea accessions and one commercial check exhibited variation in the ten qualitative traits evaluated (Table S1). The magnitude of variation differed among traits and accessions, as indicated by the standardized Shannon diversity index (H').

Kidney-shaped seeds were predominant among the evaluated materials. However, W6, E1, and E3 also contained kidney-to-rhomboid or globose-to-rhomboid seed forms and showed relatively high within-accession diversity for seed shape, with (H') values of 0.31, 0.31, and 0.20, respectively. Seed coat colour included brown, reddish-brown, black, bicoloured, and white classes, with (H') values ranging from 0.14 to 0.39 among accessions exhibiting variation. Darker seed coat colours were frequently observed among accessions from the eastern and southwestern zones, whereas W5 and W8 each showed an (H') value of 0.29 for this trait.

Most accessions had smooth seed coats. Semi-rough seed coats occurred in E5 and SW1, with H' values of 0.20 and 0.31, respectively, while semi-wrinkled forms were observed in some accessions. SW7 showed the highest within-accession diversity for seed coat texture (H'=0.39).

Prostrate growth habit was the most common form. W8 was semi-erect, W4 was climbing, and SW5 showed an intermediate growth habit. Terminal leaflet shape varied from sub-hastate to sub-globose, whereas leaflet colour ranged from pale to dark green and leaflet texture ranged from soft to intermediate. Most accessions had violet flowers, while E2 and SW7 had white flowers. Fresh pod colour was predominantly green, although purple or variegated pods were observed in SW2 and SW5. Dry pod colour ranged from tan to purple (Supplementary Figure S1).

The commercial check had an erect growth habit and smooth brown seeds and showed no within-entry variation for the qualitative traits evaluated. In contrast, several farmer-maintained accessions contained more than one phenotypic class, particularly for seed shape, seed coat colour, and seed coat texture.

When diversity was summarized by geo-ecological zone, the cumulative standardized Shannon diversity score across the ten qualitative traits was highest in the southwestern zone (H'=3.80), followed by the eastern (H'=3.34) and western (H'=2.42) zones (Table 3). The commercial check had a cumulative score of 0.00. These cumulative values represent the sum of the standardized indices across traits and therefore are not restricted to a maximum value of 1.

Table 3. Standardized Shannon diversity indices (H') of ten qualitative morphological traits for the commercial cultivar and the three geo-ecological zones.

Note: H' values for individual traits were standardized using H'=H/ln(k), where k is the number of observed phenotypic classes for each trait. Total H' represents the sum of the standardized diversity indices across the ten qualitative traits. Minor differences between the summed values and Total H' are due to rounding.

Trait-specific patterns differed among zones. The eastern accessions showed the highest diversity for seed shape (H'=0.60), whereas the southwestern accessions showed relatively high diversity for dry pod colour (H'=0.69) and seed coat texture (H'=0.58). The western accessions showed no variation in seed coat texture or flower colour within the evaluated set (H'=0.00). Terminal leaflet colour was variable in all three zones, with H' values of 0.35, 0.59, and 0.45 in the southwestern, eastern, and western zones, respectively. Growth habit and fresh pod colour were uniform among the eastern accessions evaluated.

Quantitative trait variation

Analysis of variance showed significant differences among accessions for the measured quantitative traits (Table S2). Fisher’s protected least significant difference test was applied only to traits for which the overall accession effect was significant at (P < 0.05).

Principal component analysis was used to summarize multivariate variation in the twelve quantitative traits (Figure 3). The first two principal components accounted for 44.64% of the total variation, with PC1 and PC2 explaining 27.71% and 16.93%, respectively.

Figure 3. Principal component biplot of 21 northern Thai cowpea accessions and one commercial cultivar based on twelve quantitative agronomic traits. Vectors show trait loadings on PC1 (27.7%) and PC2 (16.9%). Colour coding: western (blue), eastern (yellow), southwestern (pink), and commercial (white)

PC1 was positively associated with terminal leaflet length, terminal leaflet width, stem girth, pod length, pod width, and hundred-seed weight. PC2 was associated mainly with days to 50% flowering, branch number, raceme number, seed width, and seed thickness. These loadings indicate that vegetative size and pod and seed characteristics contributed substantially to the separation of accessions along the first two components.

The accession scores were broadly distributed across the PCA biplot. Several western accessions were positioned on the positive side of PC1 and were associated with relatively large vegetative structures, pods, or seed weight. Eastern accessions were more dispersed along PC2, while southwestern accessions occupied both intermediate and dispersed positions. The commercial check was located near the centre of the biplot.

Pod length and leaflet dimensions were oriented in a similar direction in the biplot, indicating a positive association within the multivariate dataset. Stem girth and hundred-seed weight also had similar vector directions. In contrast, seed thickness and hundred-seed weight were oriented in different directions. These vector orientations indicate associations within the PCA ordination.

Although some accessions originating from the same zone occurred near one another, the first two principal components explained less than half of the total quantitative variation. Therefore, the PCA was interpreted as a description of trait variation among accessions rather than evidence of distinct geographic grouping.

Seed protein, iron, and crude fibre contents

Analysis of variance revealed significant differences among the 20 analysed cowpea accessions and one commercial cultivar for seed protein, iron, and crude fibre contents (P < 0.05; Table 4). Accession E2 was excluded from the nutritional analysis because the harvested seed quantity was insufficient to complete all laboratory determinations.

Table 4.  Seed protein, iron, and crude fibre contents of 20 farmer-maintained cowpea accessions and one commercial cultivar. Values are presented as mean ± SD.

Note: Values are presented as mean ± SD (n = 3 biological replicates). Means within each column followed by different letters are significantly different according to Fisher's protected least significant difference (LSD) test at P < 0.05. Accession E2 was excluded from the nutritional analysis because the harvested seed quantity was insufficient to complete all laboratory determinations.

Seed protein content ranged from 23.50 to 36.00 g/100 g. Accession SW5 had the highest protein content (36.00 ± 0.14 g/100 g), which was significantly higher than all other analysed accessions and the commercial cultivar according to Fisher's protected LSD test. The lowest protein content was observed in SW7 (23.50 ± 0.21 g/100 g).

Seed iron content ranged from 28.79 to 78.19 mg/kg. W4 had the highest iron concentration (78.19 ± 1.41 mg/kg), followed by SW3 (71.82 ± 0.52 mg/kg) and the commercial cultivar (62.34 ± 1.75 mg/kg). The lowest iron concentration was recorded in W8 (28.79 ± 2.13 mg/kg).

Crude fibre content ranged from 0.82 to 5.97 g/100 g. W6 had the highest crude fibre content (5.97 ± 0.10 g/100 g), followed by SW3 (4.98 ± 0.35 g/100 g) and W8 (4.95 ± 0.02 g/100 g), whereas W4 had the lowest value (0.82 ± 0.06 g/100 g).

Overall, different accessions exhibited superior values for different nutritional traits. SW5 showed the highest protein content, W4 had the highest iron content, and W6 had the highest crude fibre content under the field conditions of this study. These accessions represent potentially useful genetic resources for further evaluation. However, because the experiment was conducted at a single location during one growing season, their nutritional performance should be confirmed under multiple environments before being considered for breeding or cultivar development.

Genetic diversity within selected entries

The SCoT analysis included nine farmer-maintained accessions selected from the three geo-ecological zones and one commercial check. Because the farmer-maintained accessions were selected on the basis of contrasting agro-morphological profiles, the molecular results represent a preliminary assessment of the selected subset and not a comprehensive characterization of all 21 accessions. Variation in SCoT marker diversity was observed among the ten entries (Table 5). The mean observed number of alleles (Na) was 1.08, and the mean effective number of alleles (Ne) was 1.22. E4 and SW6 had the highest values of Na=1.24 and Ne=1.29. E3 had the lowest values, with Na=0.84 and Ne=1.12. Shannon’s information index averaged 0.19 and ranged from 0.11 in E3 to 0.25 in E4 and SW6. Expected heterozygosity ranged from 0.07 in E3 to 0.17 in E4 and SW6, with an overall mean of 0.13. The number of polymorphic bands ranged from 33 in E3 to 65 in E4 and SW6, while the mean percentage of polymorphic loci was 37.88% (Supplementary Figure S2). Thus, within the selected molecular subset, E4 and SW6 showed the highest marker diversity, whereas E3 showed the lowest. These diversity estimates are specific to the plants, accessions, and SCoT loci included in the analysis and should not be extrapolated to all cowpea germplasm from the corresponding zones.

Table 5. Genetic diversity parameters estimated from SCoT marker analysis of the selected cowpea accessions.

Note: n, number of individuals analysed; Na, observed number of alleles; Ne, effective number of alleles; I, Shannon's information index; He, expected heterozygosity; uHe, unbiased expected heterozygosity; PPB, number of polymorphic bands; P%, percentage of polymorphic bands.

Analysis of molecular variance

Analysis of molecular variance partitioned the SCoT marker variation into three hierarchical components (Table 6). Among-zone differences accounted for 13% of the total molecular variation, differences among accessions within zones accounted for 19%, and variation within accessions accounted for 68%. The among-zone component was statistically significant at P < 0.001. The relatively large within-accession component indicates that variation among the five sampled plants within the selected accessions contributed substantially to the total marker variation. However, the AMOVA results do not identify the biological processes responsible for this variation. In particular, the present analysis cannot distinguish the relative contributions of seed mixture, residual heterogeneity, mutation, occasional outcrossing, or other factors. The among-zone component indicates that some molecular differentiation was associated with the zone classifications in this selected subset. Nevertheless, because only three non-randomly selected accessions per zone were examined, the magnitude of regional differentiation should be regarded as preliminary.

Table 6. AMOVA results partitioning genetic diversity among regions, among accessions within zones, and within accessions (df, SS, MS, variance components % variance, P - value).

Population-level diversity statistics

Total gene diversity across the selected entries was H_T=0.21, while the mean within-group diversity was H_S=0.12 (Table 7). The coefficient of genetic differentiation was G_ST=0.41, indicating differentiation in SCoT band frequencies among the groups used in the analysis. The indirect estimate of gene flow was N_m=0.71. This value was calculated from G_ST under the assumptions implemented in PopGene and was not obtained from direct measurements of seed or pollen movement. It is therefore presented for comparison with other marker-based studies and should be interpreted cautiously, particularly because cowpea is predominantly self-pollinating and the molecular accessions were not randomly sampled.

Table 7. Population genetic statistics: H_T (total diversity), H_S (within-population diversity), G_ST (differentiation), N_m (gene flow).

UPGMA and principal coordinate analyses

The UPGMA dendrogram showed relationships among the ten selected entries based on SCoT marker similarity (Figure 4). The farmer-maintained accessions were broadly arranged into groups containing western, eastern, and southwestern entries. The commercial check occurred on a separate branch from most farmer-maintained accessions. The three western accessions W3, W6, and W9 occurred within the same major grouping. E3, E4, and E5 were also positioned relatively close to one another, as were SW1, SW3, and SW6. Similarity coefficients within parts of the dendrogram ranged from 0.91 to 0.99. However, these groupings were based on only three selected accessions from each zone and should not be interpreted as definitive regional gene pools.

Principal coordinate analysis provided a complementary representation of relationships among the selected entries (Figure 5). The first two axes accounted for 58.99% of the total molecular variation, with PCo1 explaining 41.59% and PCo2 explaining 17.40%. Entries from the three zones occupied partly differentiated positions, although some overlap was present. The partial separation observed in the PCoA was consistent with the significant among-zone component detected by AMOVA. Nevertheless, the PCoA, UPGMA, and AMOVA results apply only to the selected molecular subset. Collectively, these analyses suggest that geographic origin may be associated with part of the molecular variation among the selected accessions, but they do not establish geographic origin as the principal determinant of genetic relationships across the full northern Thai cowpea collection.

Figure 4. UPGMA dendrogram of genetic relationships among the nine selected northern Thai cowpea accessions and the commercial control based on SCoT markers.

Figure 5. Principal coordinate analysis (PCoA) of the selected cowpea accessions from the western, eastern, and southwestern geo-ecological zones and the commercial cultivar based on SCoT markers. PCo1 and PCo2 explained 41.59% and 17.40% of the total molecular variation, respectively.

DISCUSSION

The present study demonstrated substantial agro-morphological and nutritional variation among northern Thai cowpea landraces. Considerable differences were observed in qualitative traits, quantitative agronomic characteristics, and seed nutritional composition despite the generally narrow genetic background reported for cultivated cowpea (Ghalmi et al., 2010; Chen et al., 2017; Seo et al., 2020). These findings indicate that northern Thai landraces retain useful phenotypic variation that may contribute to germplasm conservation and crop improvement.

Protein concentration varied from 23.50 to 36.00 g/100 g, which falls within the range previously reported for cowpea germplasm from other regions (Boukar et al., 2011; Padhi et al., 2022). Likewise, iron content (28.79–78.19 mg/kg) and crude fibre content (0.82–5.97 g/100 g) varied considerably among accessions. Different accessions consistently exhibited superior values for different nutritional traits, with SW5 showing the highest protein concentration, W4 the highest iron concentration, and W6 the highest crude fibre content. These results indicate that nutritionally valuable variation exists within northern Thai cowpea germplasm. However, because the experiment was conducted at a single location during one growing season, the stability of these nutritional characteristics across environments remains to be determined.

Qualitative characterization also revealed appreciable variation in seed morphology, seed coat colour, seed coat texture, growth habit, and pod characteristics. The cumulative Shannon diversity index was highest in the southwestern collection, followed by the eastern and western collections. Because Shannon diversity was calculated independently for each qualitative trait and subsequently summed across traits, these values represent cumulative phenotypic diversity rather than a normalized diversity index. Similar variation in seed and pod characteristics has been reported in previous cowpea diversity studies (Seehalak et al., 2006), demonstrating that traditional landraces continue to represent potential sources of phenotypic diversity.

Principal component analysis summarized the quantitative variation among accessions, with the first two principal components accounting for 44.64% of the total variation. Traits related to vegetative growth, pod characteristics, and seed size contributed most strongly to the first principal component, whereas flowering time and several reproductive traits contributed mainly to the second component. Although some accessions from the same geo-ecological zone occurred near one another in the PCA, the first two components explained less than half of the total variation. Therefore, the PCA should be interpreted primarily as a summary of quantitative trait variation rather than evidence of distinct geographic grouping.

The SCoT analysis provided a preliminary assessment of molecular diversity within a representative subset of the evaluated germplasm. Expected heterozygosity averaged 0.13, which is consistent with previous reports describing relatively low genetic diversity in cultivated cowpea based on different molecular marker systems (Chen et al., 2017; Seo et al., 2020). Among the selected accessions, E4 and SW6 exhibited the highest levels of marker diversity, whereas E3 showed comparatively lower diversity. Because only three accessions were selected from each geo-ecological zone based on contrasting agro-morphological characteristics, these molecular results should be regarded as representative of the analysed subset rather than the entire northern Thai cowpea collection.

The AMOVA indicated that 68% of the molecular variation occurred within accessions, whereas 32% occurred among accessions. This relatively high within-accession component is noteworthy because cowpea is predominantly self-pollinating and would generally be expected to show limited within-accession variation. Several factors may contribute to this pattern. Farmer-maintained accessions may retain residual heterogeneity through historical seed management, seed mixing, or occasional outcrossing. However, the result should also be considered in relation to the marker system used in this study. SCoT markers are dominant and cannot distinguish heterozygous from homozygous dominant genotypes, which limits the resolution of within-accession genetic variation. Therefore, the observed within-accession variation should not be attributed to a single biological process and should be interpreted cautiously. Further studies using a larger set of accessions and higher-resolution codominant or sequence-based markers would help clarify this pattern.

The indirect estimate of gene flow (N_m = 0.71) was calculated from G_ST using the assumptions implemented in PopGene. Because this estimate was derived from dominant molecular markers and a non-random subset of accessions, it should be interpreted cautiously and not regarded as a direct estimate of contemporary pollen or seed movement. Likewise, the UPGMA and principal coordinate analyses suggested partial grouping of the selected accessions according to their collection zones, consistent with the AMOVA results. However, these analyses describe relationships only among the selected molecular subset and should not be interpreted as evidence of genetic structure across all northern Thai cowpea landraces.

Collection sites also differed considerably in elevation. However, elevation was recorded only as passport data describing the original collection sites and was not included as an experimental factor. Because all accessions were evaluated under the same field conditions at a single location, the present study cannot determine whether elevation at the original collection sites contributed to the observed agro-morphological, nutritional, or molecular patterns. Further multi-environment evaluation would be required to examine the possible influence of elevation and other environmental variables.

The present study has several limitations. First, the molecular analysis included only a representative subset of the evaluated accessions and therefore provides only a preliminary assessment of genetic diversity. Second, all field evaluations were conducted at a single location during one growing season, preventing assessment of the stability of accession performance across environments. Third, only protein, iron, and crude fibre were determined; other nutritional components, including essential amino acids and anti-nutritional compounds, were not evaluated. Finally, the study did not investigate consumer preference or sensory characteristics that may influence utilization of these landraces.

Despite these limitations, the present study provides baseline information on the agro-morphological, nutritional, and molecular diversity of northern Thai cowpea germplasm. The identified variation among accessions offers useful information for germplasm conservation and provides candidate materials for multi-environment evaluation and breeding studies.

CONCLUSION

The agro-morphological and nutritional evaluation of 21 farmer-maintained cowpea accessions from northern Thailand revealed substantial variation in qualitative traits, quantitative characteristics, and seed nutritional composition. Several accessions showed high values for specific nutritional traits under the conditions of this study, including SW5 for protein content, W4 for iron content, and W6 for crude fibre content. These accessions represent potentially useful materials for further evaluation and germplasm utilization. The preliminary SCoT marker analysis of nine selected farmer-maintained accessions and one commercial check detected molecular variation within and among the selected entries. Most of the molecular variation occurred within accessions, while smaller proportions were detected among accessions within zones and among zones. UPGMA and principal coordinate analyses suggested partial grouping associated with geographic origin; however, these patterns apply only to the non-randomly selected molecular subset and do not confirm geographic population structure across the complete northern Thai cowpea collection. Overall, this study provides baseline information on the agro-morphological, nutritional, and preliminary molecular diversity of northern Thai cowpea germplasm. The observed variation supports the value of conserving these farmer-maintained accessions and provides a basis for selecting materials for multi-environment evaluation, comprehensive molecular characterization, and cowpea improvement.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the technical support provided by members of the CMUPN Laboratory, Faculty of Agriculture, Chiang Mai University. The current research was supported by the Graduate School, Chiang Mai University through the CMU Presidential Scholarship (2021). Special thanks are extended to Dr. Adirek Punyalue of the Highland Research and Development Institute (HRDI) for generously providing cowpea accessions and associated collection data. We also appreciate the valuable contributions of farmers who maintain these important genetic resources in northern Thailand.

AUTHOR CONTRIBUTIONS

Muhammad Zahrulakmal Ahmad Mahir: Methodology (Lead), Formal Analysis (Lead), Investigation (Equal), Data Curation (Equal), Writing – Original Draft (Lead), Writing – Review & Editing (Equal), Visualization (Equal); Benjavan Rerkasem: Writing – Review & Editing (Equal); Sansanee Jamjod: Conceptualization (Equal), Methodology (Supporting), Validation (Equal), Investigation (Equal), Data Curation (Supporting), Writing – Review & Editing (Supporting), Visualization (Supporting), Supervision (Supporting); Tonapha Pusadee: Conceptualization (Lead), Methodology (Lead), Validation (Lead), Formal Analysis (Supporting), Investigation (Equal), Data Curation (Equal), Writing – Original Draft (Equal), Writing – Review & Editing (Equal), Visualization (Lead), Supervision (Lead), Project Administration (Lead), Funding Acquisition (Lead). All authors have read and agreed to the published version of the manuscript.

CONFLICT OF INTEREST

The authors declare no known competing financial interests or personal relationships that could have influenced the work reported in the paper. The funding sources were not involved in study design, data collection, analysis or interpretation, manuscript preparation or the decision to submit the article for publication.

REFERENCES

Abebe, B.K., and Alemayehu, M.T. 2022. A review of the nutritional use of cowpea (Vigna unguiculata L. Walp) for human and animal diets. Journal of Agriculture and Food Research. 10: 100383. https://doi.org/10.1016/j.jafr.2022.100383

Addinsoft. 2021. XLSTAT statistical and data analysis solution (Version 2021) [Computer software]. https://www.xlstat.com

Analytical Software. 2003. Statistix 8 user's manual.

AOAC International. 2005. Official methods of analysis. (18^th ed.). Association of Official Analytical Chemists.

Benchasri, S., Bairaman, C., and Nualsri, C. 2012. Evaluation of yardlong bean and cowpea for resistance to Aphis craccivora koch in southern part of Thailand. Journal of Animal and Plant Sciences. 22(4): 1024–1029.

Boukar, O., Massawe, F., Muranaka, S., Franco, J., Maziya-Dixon, B., Singh, B., and Fatokun, C. 2011. Evaluation of cowpea germplasm lines for protein and mineral concentrations in grains. Plant Genetic Resources. 9(4): 515–522. https://doi.org/10.1017/S1479262111000815

Boukar, O., Belko, N., Chamarthi, S., Togola, A., Batieno, J., Owusu, E., Haruna, M., Diallo, S., Umar, M.L., Olufajo, O., et al. 2019. Cowpea (Vigna unguiculata): Genetics, genomics and breeding. Plant Breeding. 138(4): 415–424. https://doi.org/10.1111/pbr.12589

Chen, H., Chen, H., Hu, L., Wang, L., Wang, S., Wang, M.L., and Cheng, X. 2017. Genetic diversity and a population structure analysis of accessions in the Chinese cowpea [ Vigna unguiculata (L.) Walp.] germplasm collection. The Crop Journal. 5(5): 363–372. https://doi.org/10.1016/j.cj.2017.04.002

Collard, B.C.Y. and Mackill, D.J. 2009. Start codon targeted (SCoT) polymorphism: A simple, novel DNA marker technique for generating gene-targeted markers in plants. Plant Molecular Biology Reporter. 27(1): 86-93. https://doi.org/10.1007/s11105-008-0060-5

Coulibaly, S., Pasquet, R.S., Papa, R., and Gepts, P. 2002. AFLP analysis of the phenetic organization and genetic diversity of Vigna unguiculata L. Walp. reveals extensive gene flow between wild and domesticated types. Theoretical and Applied Genetics. 104(2): 358–366. https://doi.org/10.1007/s001220100740

da Silva, A.C., Santos, D. da C., Teixeira Junior, D.L., da Silva, P.B., dos Santos, R.C., and Siviero, A. 2019. Cowpea: A strategic legume species for food security and health. In: Jimenez-Lopez, J.C., Clemente, A., editors. Legume seed nutraceutical research. London (UK): IntechOpen. p. 47–65. https://doi.org/10.5772/intechopen.79006

de Mendiburu, F. 2021. Agricolae: Statistical procedures for agricultural research (R package version 1.3-5). https://CRAN.R-project.org/package=agricolae

Doyle, J.J. and Doyle, J.L. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin. 19(1): 11-15

Fang, J., Chao, C.-C.T., Roberts, P.A., and Ehlers, J.D. 2007. Genetic diversity of cowpea [Vigna unguiculata (L.) Walp.] in four West African and USA breeding programs as determined by AFLP analysis. Genetic Resources and Crop Evolution. 54(6): 1197–1209. https://doi.org/10.1007/s10722-006-9101-9

Ghalmi, N., Malice, M., Jacquemin, J.-M., Ounane, S-M., Mekliche, L., and Baudoin, J.-P. 2010. Morphological and molecular diversity within Algerian cowpea (Vigna unguiculata (L.) Walp.) landraces. Genetic Resources and Crop Evolution. 57(3): 371–386. https://doi.org/10.1007/s10722-009-9476-5

Hennink, S. and Zeven, A.C. 1990. The interpretation of Nei and Shannon–Weaver within population variation indices. Euphytica. 51(3): 235–240. https://doi.org/10.1007/BF00039724

Huynh, B-L., Close, T.J., Roberts, P.A., Hu, Z., Wanamaker, S., Lucas, M.R., Chiulele, R., Cissé, N., David, A., Hearne, S., et al. 2013. Gene pools and the genetic architecture of domesticated cowpea. The Plant Genome. 6(2): 1-8. https://doi.org/10.3835/plantgenome2013.03.0005

Igwe, D.O., Afiukwa, C.A., Ubi, B.E., Ogbu, K.I., Ojuederie, O.B., and Ude, G.N. 2017. Assessment of genetic diversity in Vigna unguiculata L. (Walp) accessions using inter-simple sequence repeat (ISSR) and start codon targeted (SCoT) polymorphic markers. BMC Genetics. 18(1): 98. https://doi.org/10.1186/s12863-017-0567-6

International Board for Plant Genetic Resources. 1983. Descriptors for cowpea. https://hdl.handle.net/10568/72871

Kaewwongwal, A., Jetsadu, A., Somta, P., Chankaew, S., and Srinives, P. 2013. Genetic diversity and population structure of Vigna exilis and Vigna grandiflora (Phaseoleae, Fabaceae) from Thailand based on microsatellite variation. Botany. 91(10): 653–661. https://doi.org/10.1139/cjb-2013-0029

Kebede, E. and Bekeko, Z. 2020. Expounding the production and importance of cowpea (Vigna unguiculata (L.) Walp.) in Ethiopia. Cogent Food & Agriculture. 6(1): 1769805. https://doi.org/10.1080/23311932.2020.1769805

Kongjaimun, A., Kaga, A., Tomooka, N., Somta, P., Vaughan, D.A., and Srinives, P. 2012. The genetics of domestication of yardlong bean, Vigna unguiculata (L.) Walp. ssp. unguiculata cv.-gr. sesquipedalis. Annals of Botany. 109(6): 1185–1200. https://doi.org/10.1093/aob/mcs048

Li, C-D., Fatokun, C.A., Ubi, B., Singh, B.B., and Scoles, G.J. 2001. Determining genetic similarities and relationships among cowpea breeding lines and cultivars by microsatellite markers. Crop Science. 41(1): 189–197. https://doi.org/10.2135/cropsci2001.411189x

Padhi, S.R., Bartwal, A., John, R., Tripathi, K., Gupta, K., Wankhede, D.P., Mishra, G. P., Kumar, S., Archak, S., and Bhardwaj, R. 2022. Corrigendum: Evaluation and multivariate analysis of cowpea [Vigna unguiculata (L.) walp] germplasm for selected nutrients—mining for nutri-dense accessions. Frontiers in Sustainable Food Systems. 6: 986734. https://doi.org/10.3389/fsufs.2022.986734

Patel, C.G., Patel, P.T., Dharajiya, D.T., Patel, A.B., Zala, H.N., Khatri, A.B., and Tiwari, K.K. 2022. Morphological and molecular characterization of cowpea (Vigna unguiculata L. Walp.) genotypes. Genetika. 54(2): 633–648. https://doi.org/10.2298/GENSR2202633P

Peakall, R. and Smouse, P.E. 2012. GenAlEx 6.5: Genetic analysis in Excel. Population genetic software for teaching and research-an update. Bioinformatics. 28(19): 2537-2539. https://doi.org/10.1093/bioinformatics/bts460 

Punyalue, A., Jamjod, S., and Rerkasem, B. 2018. Intercropping maize with legumes for sustainable highland maize production. Mountain Research and Development. 38(1): 35–44. https://doi.org/10.1659/mrd-journal-d-17-00048.1

Punyalue, A., Jongjaidee, J., Jamjod, S., and Rerkasem, B. 2018. Legume intercropping to reduce erosion, increase soil fertility and grain yield, and stop burning in highland maize production in northern Thailand. Chiang Mai University Journal of Natural Sciences. 17(4): 265-274. https://doi.org/10.12982/CMUJNS.2018.0019

R Core Team. 2021. R: A language and environment for statistical computing [Computer software]. R Foundation for Statistical Computing. Vienna. https://www.R-project.org

Rohlf, F.J. 2000. NTSYSpc: Numerical taxonomy and multivariate analysis system, version 2.1. Setauket, New York: Exeter Software.

Sarutayophat, T., Nualsri, C., Santipracha, Q., and Saereeprasert, V. 2007. Characterization and genetic relatedness among 37 yardlong bean and cowpea accessions based on morphological characters and RAPD analysis. Songklanakarin Journal of Science and Technology. 29(3): 591-600. 

Seehalak, W., Tomooka, N., Waranyuwat, A., Thipyapong, P., Laosuwan, P., Kaga, A., and Vaughan, D.A. 2006. Genetic diversity of the Vigna germplasm from Thailand and neighboring regions revealed by AFLP analysis. Genetic Resources and Crop Evolution. 53(5): 1043–1059. https://doi.org/10.1007/s10722-004-7939-2

Seo, E., Kim, K., Jun, T.-H., Choi, J., Kim, S.-H., Muñoz-Amatriaín, M., Sun, H., and Ha, B.-K. 2020. Population structure and genetic diversity in Korean cowpea germplasm based on SNP markers. Plants. 9(9): 1190. https://doi.org/10.3390/plants9091190

Shannon, C.E. and Weaver, W. 1949. The mathematical theory of communication. The University of Illinois Press.

Shevkani, K., Shivani, B., Dhaka, S.S., and Patil, C. 2025. Cowpeas for sustainable agriculture and nutrition security: An overview of their nutritional quality and agroeconomic advantages. Discover Food. 5(1): 109. https://doi.org/10.1007/s44187-025-00382-x

Tan, H., Tie, M., Luo, Q., Zhu, Y., Lai, J., and Li, H. 2012. A review of molecular markers applied in cowpea (Vigna unguiculata L. Walp.) breeding. Journal of Life Sciences. 6: 1190–1199.

Vanmathi, S., Arulbalachandran, D., and Soundarya, V. 2021. Effects of gamma radiation on quantitative traits and genetic variation of three successive generations of cowpea (Vigna unguiculata (L.) Walp.). Plant Science Today. 8(3). https://doi.org/10.14719/pst.2021.8.3.1054

Wamalwa, E.N., Muoma, J., and Wekesa, C. 2016. Genetic diversity of cowpea (Vigna unguiculata (L.) Walp.) accession in Kenya gene bank based on simple sequence repeat markers. International Journal of Genomics. 2016: 8956412. https://doi.org/10.1155/2016/8956412

Xiong, H., Shi, A., Mou, B., Qin, J., Motes, D., Lu, W., Ma, J., Weng, Y., Yang, W., and Wu, D. 2016. Genetic diversity and population structure of cowpea (Vigna unguiculata L. Walp). PLoS One. 11(8): e0160941. https://doi.org/10.1371/journal.pone.0160941

Yeh, F.C., Yang, R.C., Boyle, T.B.J., Ye, Z., and Mao, J.X. 1997. POPGENE, the user-friendly shareware for population genetic analysis. Molecular Biology and Biotechnology Centre, University of Alberta, Canada.   

OPEN access freely available online

Natural and Life Sciences Communications

Chiang Mai University, Thailand. https://cmuj.cmu.ac.th

Supplementary material

Figure S1. Representative agro-morphological variation observed among northern Thai cowpea landraces. Flower colour, fresh pod colour, dry pod colour, seed colour variation.

Figure S2. Representative SCoT marker amplification profiles generated using primer SCoT28. The DNA size marker is shown in the leftmost lane, while accession codes are indicated above each sample lane.

Figure S3. The field planting site for the assessment.

Table S1. Morphological variation among 21 accessions and commercial control showing Shannon diversity indices (H’) for ten qualitative traits.