Agronomic and Intercropping Performance of Newly Developed Elite Cowpea Lines for the West African Savannas

Abstract

Cowpea production in Nigeria, the world’s largest producer, is insufficient to meet domestic demand due to significant yield gaps caused by various production constraints. Several high-yielding improved cowpea varieties have been developed and disseminated among smallholder farmers to improve productivity, but their adoption is low because breeding efforts have not adequately incorporated farmers’ and consumers’ preferred traits. To address this, a study was conducted to evaluate the performance of newly developed cowpea lines and identify those with traits preferred by farmers and consumers. Twenty-four cowpea lines were evaluated in multiple environments under sole and intercropped systems in Nigeria. The study revealed significant (p < 0.001) genotypic and genotype-by-environment interaction effects for grain yield, fodder yield, and other key agronomic traits. Three genotypes consistently outperformed the standard check, with UAM15-2157-4 exhibiting a 57.6% higher grain yield and superior seed quality. UAM15-2157-4 produced the highest grain yield (1289 kg ha⁻¹) under the intercropping system. GGE biplot analysis identified UAM15-2157-4 as the most stable genotype across all tested environments. This genotype, along with other promising lines, possesses desirable traits such as Striga resistance, large seed size, and preferred seed coat color, making them suitable for release and adoption to improve cowpea productivity in the region.

1. Introduction

Cowpea [Vigna unguiculata (L.) Walp.] is a nutritionally significant grain legume widely cultivated throughout sub-Saharan Africa (SSA) for both human consumption and livestock feed. This multipurpose protein-rich (23–32%) legume crop serves as a vital source of dietary protein, energy, and essential micronutrients for over 200 million people in the region [1,2]. The average Nigerian consumes approximately 23 kg of cowpea annually, highlighting its importance as an affordable protein source often termed “poor man’s meat” [3]. Beyond human nutrition, cowpea haulms provide valuable livestock feed, particularly during the dry season, whilst simultaneously contributing to sustainable agricultural systems through biological nitrogen fixation, soil conservation, and income generation for farmers and food vendors [4].

Africa dominates global cowpea production, accounting for about 96.7% of the total production, with West Africa contributing over 95% of Africa’s production [5]. Nigeria leads cowpea production globally, accounting for 61% of Africa’s production and 48% worldwide. In 2023, Nigeria produced approximately 4.3 million tons of cowpeas from a global production of 9.8 million tons, representing 43.8% of total supply [6]. Despite being the world’s largest producer, Nigeria’s cowpea supply fails to meet domestic demand due to several production constraints. Cowpea production in Nigeria and across West Africa faces significant yield-limiting challenges resulting from both biotic and abiotic causes. Some of the major biotic stresses include insect pests such as legume pod borer (Maruca vitrata), aphids (Aphis craccivora), and flower bud thrips (Megalurothrips sjostedti); diseases caused by fungi (Colletotrichum spp.), bacteria (Xanthomonas axonopodis pv. vignicola), and viruses (Cowpea aphid-borne mosaic virus); and parasitic weeds, primarily Striga gesnerioides [7]. Of particular concern is S. gesnerioides, which can cause yield losses ranging from 30 to 100% in susceptible varieties by attaching to the host root system and extracting essential nutrients and water. Striga infestation is mostly common in the drier savannas, and it affects millions of hectares across West Africa [8].

Most smallholder farmers in Nigeria practice intercropping systems, particularly cereal–legume combinations, as a risk management strategy and to maximize land use efficiency. This practice makes most farmers prefer varieties that are suitable for intercropping combined with market demand seed quality traits. Maize–cowpea intercropping is especially common, offering multiple benefits, including enhanced soil fertility through nitrogen fixation, improved soil moisture conservation, efficient resource utilization, and increased overall system productivity [8]. Cereal–legume intercropping has gained increasing global attention as a sustainable agricultural practice, with meta-analyses showing consistent benefits in terms of enhanced nitrogen use efficiency and improved resource utilization [9]. Recent comprehensive reviews have identified yield and nitrogen fixation as the most commonly studied traits in cereal–legume intercropping systems worldwide [10]. Studies have shown that intercropping cowpea with cereals can increase land equivalent ratios (LER) by 20–60% compared to monocultures, while simultaneously improving soil physical properties, organic matter content, and nitrogen availability [11,12]. However, productivity in these systems depends significantly on the cowpea variety’s ability to tolerate shade and compete effectively for resources while maintaining biological nitrogen fixation capabilities. The reason for this is that intercropping creates competition for light, nutrients, and water [13], which can negatively impact yield if the cowpea variety is not well-suited for the system.

Another critical factor affecting cowpea adoption is varietal preference. Recent studies by Karikari et al. [14] highlight that farmers and consumers in West Africa have specific preferences regarding production and utilization traits. These preferences include adaptability to intercropping, high yield potential, drought tolerance, early maturity, Striga resistance, superior taste, and shorter cooking time. More importantly, seed quality traits such as seed coat color, eye color, and grain size directly influence consumer acceptance and market value. For example, in northern Nigeria, white, milk, brown, and red-colored cowpea varieties are preferred, while brown cowpeas (commonly called honey beans) are most popular in southern regions [3].

International and national agricultural research institutions have developed improved cowpea varieties with resistance/tolerance to various biotic and abiotic stresses. However, despite these breeding efforts, adoption rates remain low due to the mismatch between developed varieties and farmers’/consumers’ preferred traits. Therefore, developing high-yielding cowpea varieties that combine resistance/tolerance to biotic stresses with preferred seed quality characteristics and adaptation to intercropping systems is crucial for enhancing adoption of improved varieties and productivity. In this context, “elite” lines refer to breeding lines that have undergone rigorous selection and have demonstrated high performance for multiple desirable traits across different environments, making them candidates for release as new varieties [15]. The objectives of this study were to (i) assess the agronomic performance and stability of cowpea elite breeding lines across diverse environments and (ii) identify promising lines with market-preferred traits and adaptability to intercropping systems that could be recommended for release to Nigerian farmers for commercialization.

2. Materials and Methods

2.1. Plant Materials and Genotype Selection

The initial plant materials consisted of one hundred and forty (140) advanced cowpea breeding lines developed by the International Institute of Tropical Agriculture (IITA), the Institute for Agricultural Research (IAR) at Ahmadu Bello University (ABU) (Samaru, Zaria, Nigeria), and the Joseph Sarwuan Tarka University Makurdi (JOSTUM) (Makurdi, Benue State, Nigeria). In 2021, these lines were first evaluated in a Striga-endemic field in Jigawa State, Nigeria (12.01° N, 9.58° E), to screen for resistance to the parasitic weed Striga gesnerioides. To ensure uniform infestation, each planting hole was artificially inoculated with approximately 500 germinable S. gesnerioides seeds [16]. Striga emergence and severity were recorded weekly. Based on this screening, twenty-four (24) elite lines exhibiting high levels of resistance to Striga and desirable agronomic characteristics were selected for further multi-location evaluation (Table 1).

Table 1. Description of cowpea genotypes used in the multi-location study.

No	Genotype Name	Origin	Seed Colour	Description
1	IAR-09-1042-4	IAR/ABU	White	Breeding line
2	IAR-17-1015	IAR/ABU	White	Breeding line
3	IARD-17-10177	IAR/ABU	White	Breeding line
4	IT08K-126-19	IITA	White	Breeding line
5	IT08K-150-12	IITA	White	Variety
6	IT14K-2188-1	IITA	White	Breeding line
7	IT16K-1685-2	IITA	White	Breeding line
8	IT16K-1968-3	IITA	White	Breeding line
9	IT16K-2287-4	IITA	White	Breeding line
10	IT16K-2556-2	IITA	White	Breeding line
11	IT17K-1095-2-1	IITA	White	Breeding line
12	IT17K-1266-2-1	IITA	White	Breeding line
13	IT17K-1403-1-1	IITA	White	Breeding line
14	IT17K-1707-2-2	IITA	White	Breeding line
15	IT17K-1809-4	IITA	White	Breeding line
16	IT17K-2024-4	IITA	White	Breeding line
17	IT17K-2357-1	IITA	White	Breeding line
18	IT17K-3217-1	IITA	White	Breeding line
19	IT17K1802-1	IITA	White	Breeding line
20	UAM14-126-L2	UAM	White	Breeding line
21	UAM14-126-L28	UAM	White	Breeding line
22	UAM14-126-L29	UAM	White	Breeding line
23	UAM14-126-L31-1	UAM	White	Breeding line
24	UAM15-2157-4	UAM	Light Brown	Breeding line

IAR/ABU: Institute of Agricultural Research, Ahmadu Bello University, Zaria; IITA: International Institute of Tropical Agriculture, Ibadan; UAM: University of Agriculture, Makurdi.

2.2. Multi-Location Evaluation for Agronomic Performance

The 24 selected elite cowpea lines, along with a standard check variety (IT08K-150-12), were evaluated in five environments. These environments comprised three locations across Nigeria’s major cowpea-growing zones over two years: Makurdi (7°41′ N, 8°37′ E; Southern Guinea savanna) in 2022 and 2023, Biu (10°36′ N, 12°11′ E; Northern Guinea savanna) in 2022 and 2023, and Zaria (11°04′ N, 7°43′ E; Northern Guinea savanna) in 2023. The environmental conditions for each location and year are summarized in Table S1.

In all environments, the trials were laid out in a randomized complete block design with three replications at each location. Each experimental unit consisted of four rows, each 4 m long, with an inter-row spacing of 0.75 m and an intra-row spacing of 0.25 m. Three seeds were planted per hill and later thinned to 2 plants per stand approximately two weeks after sowing (WAS), resulting in a final plant population density of 106,666 plants ha⁻¹.

A compound fertilizer (NPK 15:15:15) mixed with 2 bags (100 kg) of Single Superphosphate was applied to supply the optimum nutrient requirements of 15 kg/ha N, 30 kg/ha P, and 15 kg/ha K at 2 WAS. Weed control was achieved through a combination of pre-emergence herbicide application (Pendimethalin at 1.5 kg a.i. ha⁻¹) followed by manual weeding at 3 and 6 WAS to maintain weed-free plots throughout the growing period. Insect pest management involved three applications of Cyper-DiForce at the rate of 30 g L⁻¹ + dimethoate 250 g L⁻¹ (EC) during the flower bud formation, full flowering, and podding stages. The insecticide was applied using a knapsack sprayer.

2.3. Evaluation for Adaptation to Intercropping

Following the multi-location trials, twelve of the most promising lines were selected based on their agronomic performance (high grain and fodder yield), stability, and market-preferred seed traits. These twelve lines, along with three commercial varieties as checks (IT89KD-288, FUAMPEA 3, and FUAMPEA 4), were evaluated for their suitability in a maize–cowpea intercropping system. The experiment was conducted at the Zaria and Makurdi research stations during the 2024 growing season.

The experiment was laid out using a split plot design, with the cropping system (sole cowpea vs. maize–cowpea intercrop) as the main plot and the fifteen cowpea genotypes as the sub-plots. The medium-maturing maize variety, Comp 3-DT, was used as the cereal component. In the intercrop plots, maize was planted at a spacing of 0.75 m × 0.50 m with two plants per stand (53,333 plants ha⁻¹). The cowpea lines were relayed into the standing maize at 6 weeks after the maize was planted. Cowpea was planted in two stands between the maize rows at a spacing of 0.25 m, resulting in a population of 106,666 plants ha⁻¹. In the sole crop plots, cowpea was planted at the same time and density as in the intercrop plots, using a spacing of 0.75 m × 0.25 m.

2.4. Data Collection and Analysis

Data were collected on days to 50% flowering, days to 95% maturity, pods per plant, 100-seed weight, grain yield (kg ha⁻¹), and fodder yield (kg ha⁻¹). Disease incidence was visually assessed on a scale of 0–5 based on the severity of the leaf area affected (0 = no symptoms, 5 = severe symptoms with >75% leaf area affected). Striga resistance was assessed based on emergence counts. For the intercropping trial, the Land Equivalent Ratio (LER) was calculated. All the data collected were subjected to Analysis of Variance (ANOVA) using JMP statistical software (Version 17). A combined analysis was performed with genotype considered as a fixed effect, while replications nested within environment and their interactions were treated as random effects. Mean separation was conducted using the Student–Newman–Keuls (SNK) test at a 5% probability level.

Hierarchical cluster analysis was performed using JMP software to classify the cowpea genotypes based on their response to S. gesnerioides infestation and grain yield performance. Correlation matrix between key agronomic traits in cowpea based on Spearman Rank correlations was performed using R Software (4.4.3) [17].

The multi-location yield data were further analyzed using GGE (Genotype and Genotype × Environment Interaction) biplot analysis with Plant Breeding Tools (PBTools) version 1.4 to identify high-yielding and stable genotypes across environments. The GGE-biplot stability was calculated using the statistical model of Yan [18], represented as follows:

$$Y_{ij}- {\beta }_j={\lambda }_1{\xi }_{i1}{\eta }_{j1}+ {\lambda }_2{\xi }_{i2}{\eta }_{j2}+ {\epsilon }_{ij}$$

where $Y_{ij}$ is the mean yield of the ith genotype in the jth environment, ${\beta }_j$ is the mean yield across all genotypes in the jth environment, ${\lambda }_1$ and ${\lambda }_2$ are the singular values for PC1 and PC2, respectively; ${\xi }_{i1}$ and ${\xi }_{i2}$ are the PC1 and PC2 eigenvectors, respectively, for the ith genotype; ${\eta }_{j1}$ and ${\eta }_{j2}$ are the PC1 and PC2 eigenvectors, respectively, for the jth environment; and ${\epsilon }_{ij}$ is the residual of the model associated with the ith genotype in the jth environment.

For the intercropping experiment, Partial Land Equivalent Ratio (LER) was calculated for cowpea and maize using the following formula:

LER = (Yic/Ysc) + (Yim/Ysm)

where Yic and Ysc are the yields of cowpea in intercrop and sole crop, respectively, and Yim and Ysm are the yields of maize in intercrop and sole crop, respectively. LER values greater than 1.0 indicate yield advantage of intercropping over sole cropping.

The relative advantage of intercropping of cowpea compared to sole cropping was calculated using the method described by Deb and Dutta [19]. The partial land equivalent ration (PLER) was calculated using the equation as follows:

Partial LER (PLER_c) for cowpea

PLER_c = Y_cm in intercropping/Y_c in sole cropping

Here, Y_cm is the yield of “cowpea under intercropping with maize” and Y_c is the yield of “cowpea” grown under sole cropping.

3. Results and Discussion

3.1. Screening of the Breeding Lines for Striga Resistance

The Striga evaluation trial conducted under artificially Striga infested conditions showed variable responses of the 140 breeding lines to Striga infestation. Cluster analysis grouped the cowpea genotypes into seven distinct clusters based on Striga emergence and grain yield (Figure 1). While clusters 4, 5, 6, and 7 showed varying levels of Striga infestation, cluster 4 recorded the highest count of 58.6 Striga plants per plot and the lowest grain yield of 160.2 kg/ha. Clusters 1, 2, and 3 had zero Striga emergence, and they exhibited a good level of resistance to S. gesnerioides and produced higher grain yields. This result corroborates other studies on cowpea that report severe yield reduction due to Striga infestation [16,20]. Among the three resistant clusters, cluster 1 had only one genotype, UAM15-2157-4, which produced the highest grain yield (1780.0 kg/ha) and demonstrated complete resistance to Striga in addition to the superior yield performance. Cluster 2 had 12 genotypes with moderate yields, while cluster 3 had 46 genotypes, but with low grain yields.

Figure 1. Tree diagram of 140 genotypes for 2 studied variables (yield and Striga) using hierarchical cluster analysis.

The complete absence of Striga emergence on UAM15-2157-4 throughout the growing season coupled with highest grain yield indicates a robust and durable resistance mechanism. This complete resistance to S. gesnerioides is particularly valuable in the context of smallholder farming systems in sub-Saharan Africa because it can significantly increase cowpea yields in smallholder farming systems, protecting the crop from the parasitic weed’s devastating effect and improving food security [16,21]. Unlike cultural or chemical control methods that offer partial and temporary solutions, host plant resistance provides a more economical and efficient control method compared to other options, as the parasitic weed is a major threat to the production of cowpea in the dry savannas, a vital food and income source for the region. The Striga resistance demonstrated in UAM15-2157-4 may involve a hypersensitive response that prevents Striga attachment and development. A similar mechanism had been reported in other resistant cowpea genotypes [16]. Based on genotypes’ response to Striga and grain yield potential, twenty-four cowpea genotypes were selected for multi-location trials to assess their performance under various environmental conditions. The trials aimed to identify genotypes that are high-yielding, are stable across locations, and produce good yields under intercropping systems.

3.2. Agronomic Performance of Elite Cowpea Lines

The combined analysis of variance for the 24 cowpea lines across the five environments revealed highly significant differences (p < 0.0001) among the genotypes for grain yield, fodder yield, 100-seed weight, days to first flowering, and days to 50% flowering (Table 2). Studies on cowpea diversity have revealed extensive variation among genotypes [22]. The genotype × environment (G × E) interaction was also significant for these traits, indicating that the genotypes responded differently to the varying environmental conditions. This is particularly critical in the diverse agro-ecological zones of Nigeria, where environmental conditions can vary significantly [23]. Several studies have reported significant GEIs for growth and agronomic traits of cowpea [24]. Similar G × E interactions have been reported in other legume crops across different continents, emphasizing the universal importance of multi-environment testing in breeding programs [25]. Environmental effects were significant for all traits, highlighting the influence of location-specific factors. Popoola et al. [24] also reported significant differences in grain yield, 100-seed weight, and flowering time in cowpea genotypes evaluated across diverse environments. The significant genotype effect for traits indicates substantial genetic variability within the tested cowpea lines, providing opportunities for selection of superior performers.

Table 2. Mean squares of agronomic trait of cowpea across five locations.

SOV	df	Days to First Flower	Days to 50% Flowering	Days to 95% Pod Maturity	Pods/Plants	Seed/Pod	100-Seed Weight	Grain Yield kg/ha	Fodder Yield kg/ha
Rep(Environment)	10	232.10 **	17.04	24.26 *	4.58	177.99	5.37	59,4954.80 **	685,830.50
Genotype (G)	23	57.08 **	86.67 **	8.50	3.71	111.32	116.17 **	533,449.96 **	1,668,775.96 **
Environment (E)	4	4763.93 **	3061.46 **	3616.42 **	271.30 **	58,890.12 **	129.17 **	26,891,006.75 **	3,6259,677.50 **
G × E	92	18.03 **	24.89 **	9.19	3.75 *	69.53	10.45 **	20,750.55 **	803,406.98 **
Error	230	9.34	15.59	9.06	3.54	135.82	3.84	81,187.00	492,023.00

* and ** = significant at 5% and 1% level of probabilities.

The agronomic performance of the 24 elite cowpea lines evaluated across the five environments is presented in Table 3. The days to 50% flowering ranged from 43.0 to 53.3 days, 100-seed weight from 17 to 31 g per 100 seeds, and grain yield from 869 for IT14K-2188-1 to 1674 kg ha⁻¹ for UAM15-2157-4. UAM15-2157-4 produced a 57.6% higher grain yield than the standard check, IT08K-150-12 (1109.38 kg ha⁻¹). Similarly, UAM15-2157-4 recorded the highest 100-seed weight (30.8 g), which was significantly higher than the check (17.5 g). Cowpea varieties are often categorized based on their 100-seed weight, varying between small, medium, large, and very large seeds. This characteristic is valued by farmers and consumers alike, affecting both market preferences and agronomic performance. Though farmers’ and market preferences for seed size vary across the different regions of Nigeria, consumers in southwestern Nigeria often have preferences for larger seed size. Thus, they pay premiums in the market for larger seed size, particularly with brown seed coat color, making it a desirable seed quality trait. This desirable attribute coupled with its seed coat texture (light brown) is expected to make it appealing to both farmers and consumers. While the light-brown seed of UAM15-2157-4 is highly valued in southern Nigeria, the white-seeded varieties like IAR-17-1015 and IT17K-1266-2-1 also perform well and cater to the preferences of northern Nigerian markets. The commercialization strategy should therefore be region-specific, promoting different lines in different markets to maximize adoption and impact. This approach addresses the shortcomings of previous breeding efforts that did not adequately consider regional consumer preferences. Although UAM15-2157-4 was the top performer, other lines such as UAM14-126-L28 and IAR-09-1042-4 also showed promising yields, providing a valuable pool of genetic material for future breeding efforts. This diversity is crucial for developing a portfolio of varieties adapted to different regional preferences and farming systems rather than relying on a single genotype.

Table 3. Grain yield and other agronomic traits of cowpea genotypes evaluated across years at three locations in Nigeria, 2022–2023.

Genotype	Days to First Flower	Days to 50% Flowering	Days to 95% POD Maturity	Seeds/Pod	Pods/Plant	100-Seed Weight (g)	Grain Yield kg/ha	Fodder kg/ha
IAR-09-1042-4	48.27	57.40	78.83	9.43	29.23	16.79	1087.06	1778.34
IAR-17-1015	43.40	52.80	77.83	8.72	26.82	18.25	901.80	1476.10
IARD-17-10177	42.13	49.67	79.17	9.27	23.00	19.45	951.28	1206.12
IT08K-126-19	44.67	52.33	80.25	9.29	25.05	18.92	1141.23	1435.55
IT08K-150-12	46.47	53.73	79.25	9.77	28.24	17.93	1109.38	1667.21
IT14K-2188-1	46.27	53.87	78.50	9.25	25.12	17.73	869.79	1365.00
IT16K-1685-2	47.13	54.93	77.92	9.39	27.66	17.66	1207.85	1525.55
IT16K-1968-3	48.73	57.60	78.75	9.25	22.96	20.91	1051.98	1718.88
IT16K-2287-4	44.40	53.47	77.17	9.86	27.78	18.87	1015.49	1446.11
IT16K-2556-2	47.20	55.53	79.83	9.90	25.16	21.79	1424.75	1687.22
IT17K-1095-2-1	45.33	56.60	79.17	9.83	31.39	20.04	1060.76	1831.66
IT17K-1266-2-1	48.73	57.80	78.83	9.63	26.08	18.59	1291.83	1555.55
IT17K-1403-1-1	46.53	54.07	77.83	10.36	30.03	18.53	1125.72	1458.89
IT17K-1707-2-2	45.67	56.33	79.00	9.75	31.48	20.17	1034.78	1230.00
IT17K-1802-1	45.93	54.53	79.00	10.37	23.70	20.71	1136.57	1493.33
IT17K-1809-4	47.13	56.60	79.25	9.05	23.70	21.60	1333.37	1615.89
IT17K-2024-4	47.87	57.73	80.00	10.37	25.97	18.57	1297.75	1608.33
IT17K-2357-1	50.13	57.80	79.00	10.42	26.79	20.87	1361.54	1571.66
IT17K-3217-1	47.13	58.47	78.33	9.07	29.95	19.41	1107.88	1428.33
UAM14-126-L2	45.60	54.73	78.08	10.98	23.29	21.97	1406.59	1967.22
UAM14-126-L28	48.93	59.20	79.08	10.31	30.96	19.43	1021.59	2322.77
UAM14-126-L29	46.87	57.07	79.67	9.91	30.88	23.34	992.53	2243.88
UAM14-126-L33-1	44.36	52.57	79.45	9.45	30.25	19.79	1161.94	1527.97
UAM15-2157-4	49.93	58.87	80.00	10.39	21.11	30.82	1674.11	2566.66
Mean	46.62	55.57	78.93	9.75	26.94	20.09	1156.98	1655.34
Heritability	0.684	0.713	0.02	0.016	0.064	0.910	0.611	0.519
SE ±	2.20	1.88	2.33	0.78	9.15	0.67	177.62	263.05

High heritability was observed for all traits except seeds per pod and pod per plant (Table 3). The heritability of grain yield was 0.6 while 100-seed weight was 0.9. The higher broad-sense heritability observed may indicate that a larger portion of the observed variations may be heritable. Popoola et al. [24] also reported the highest broad-sense heritability for 100-seed weight and fodder yield. The grain yield performance of the genotypes at each individual location is presented in Table 4. The consistent high yield of UAM15-2157-4 in different locations suggests broad adaptation to diverse growing conditions, which is an essential trait for varieties intended for widespread adoption. The performances of the genotypes for other agronomic traits in each location are presented in Tables S2–S6.

Table 4. Grain yield of cowpea genotypes evaluated at three locations in Nigeria, 2022 and 2023.

	Locations
	BIU		MAKURDI		SAMARU	Grain Yield Advantage of UAM15-2157-4 (%)
Genotype	2022	2023	2022	2023	2023
IAR-09-1042-4	419.43	2566.67	1000.67	631.83	816.72	35.1
IAR-17-1015	641.64	1796.39	1093.1	697.53	280.33	46.1
IARD-1710177	797.19	1976.11	1115.27	777.47	90.39	43.2
IT08K-126-19	1455.5	1886.11	1056.33	655.97	652.22	31.8
IT08K-150-12	1119.95	1716.67	1080.03	684.53	945.72	57.6
IT14K-2188-1	652.75	1952.78	785.77	523.7	433.94	48.0
IT16K-1685-2	1597.16	2128.89	830.23	593.63	889.33	27.9
IT16K-1968-3	1177.73	2016.67	727.1	606.57	731.83	37.2
IT16K-2287-4	652.75	2230.56	1256.37	602.73	335.06	39.3
IT16K-2556-2	1561.05	2297.22	1706.13	639.57	919.78	14.9
IT17K-1095-21	1119.4	1800	935.07	597.83	851.5	36.6
IT17K-1266-2-1	1513.83	2011.11	1427.93	558.4	947.89	22.8
IT17K-1403-1-1	1394.39	1858.33	985.83	623.9	766.17	32.8
IT17K-1707-2-2	805.53	2277.78	962.5	677.17	450.94	38.2
IT17K-1802-1	1366.61	2225	856.33	493.23	741.67	32.1
IT17K-1809-4	1322.17	2594.44	1528.67	588.87	632.72	20.4
IT17K-2024-4	1394.39	2252.78	960	737.9	1143.67	22.5
IT17K-2357-1	738.86	2916.67	1649.77	754.07	748.33	18.7
IT17K-3217-1	733.3	2278.89	1167.17	836.47	523.56	33.8
UAM14-126-L2	1579.38	2572.22	1026.17	896.07	959.11	16.0
UAM14-126-L28	1038.85	2358.33	246.57	479.6	984.61	39.0
UAM14-126-L29	931.63	1957.22	633.5	691.5	748.78	40.7
UAM14-126-L33-1	1158.29	1950	1151.5	798.93	544.23	33.1
UAM15-2157-4	1663.27	2945	1595.07	967.23	1200	0.0
Mean	1118.13	2190.24	1074.05	671.45	722.44

The reaction of the lines to major pests and diseases of cowpea is presented in Table 5. UAM15-2157-4 recorded low severity scores for Cercospora leaf spot (0.5) and bacterial blight (0.5) and zero counts for S. gesnerioides and Alectra. Furthermore, no significant symptoms of Septoria leaf spot, pod scab, or virus infection were observed for the line. The complete absence of Striga and Alectra emergence in all test environments confirmed the robust and stable nature of UAM15-2157-4 resistance to these parasitic weeds, which pose a great threat to cowpea production in the savannas.

Table 5. Reaction of genotypes to disease incidence and parasitic weeds.

Genotype	Cercospora Leaf Spot (0–5)	Bacteria Blight (0–5)	Septoria (0–5)	Pod Scab	Virus Severity (0–5)	Aphid (0–5)	Maruca (0–5)	Striga Count (m2) @ 63 DAP	Alectra Count (m2) @ 63 DAP
IAR-09-1042-4	1.0	1.0	0.7	0.5	0.5	0.6	0.5	0.0	1.5
IAR-17-1015	1.5	0.5	0.8	0.5	0.8	0.6	0.5	2.0	11.2
IARD-17-10177	0.8	1.5	0.5	0.5	0.7	0.6	0.5	0.0	0.8
IT08K-126-19	1.2	2.3	0.8	0.5	0.5	0.6	0.5	0.0	0.0
IT08K-150-12	0.8	2.3	0.8	0.5	0.7	0.6	0.5	0.0	0.0
IT14K-2188-1	0.8	0.5	0.7	0.7	0.5	0.6	0.5	0.0	0.0
IT16K-1685-2	1.2	0.7	0.5	0.5	0.5	0.7	0.7	0.0	0.8
IT16K-1968-3	0.7	0.5	0.5	0.5	0.5	0.6	0.5	0.0	0.0
IT16K-2287-4	1.3	0.8	0.7	0.5	0.8	0.6	0.5	0.0	0.0
IT16K-2556-2	1.2	1.2	0.8	0.5	0.5	0.6	0.5	0.0	0.0
IT17K-1095-2-1	1.0	1.2	0.7	0.5	0.5	0.6	0.5	0.0	0.0
IT17K-1266-2-1	1.2	1.5	0.7	0.5	0.5	0.6	0.5	0.0	0.0
IT17K-1403-1-1	1.2	0.8	0.8	0.5	0.5	0.6	0.8	0.0	0.0
IT17K-1707-2-2	1.0	0.7	0.7	0.7	0.5	0.6	0.5	0.0	0.0
IT17K-1809-4	1.0	1.3	0.7	0.5	0.5	0.6	0.5	0.0	0.0
IT17K-2024-4	1.2	1.8	0.8	0.5	0.5	0.6	0.5	0.0	0.0
IT17K-2357-1	1.0	0.7	0.7	0.5	0.7	0.6	0.5	0.0	0.0
IT17K-3217-1	1.2	2.0	0.8	0.5	0.5	0.6	0.5	0.0	0.0
IT17K1802-1	1.0	1.5	0.8	0.5	0.5	0.6	0.5	0.0	0.0
UAM14-126-L2	1.0	0.7	0.7	0.7	0.5	0.6	0.7	0.0	0.0
UAM14-126-L28	0.8	0.5	0.8	0.5	0.5	0.6	0.5	0.0	0.0
UAM14-126-L29	0.7	0.5	0.5	0.5	0.5	0.6	0.5	0.0	0.0
UAM14-126-L31-1	1.3	0.5	0.7	0.5	0.5	0.6	0.5	0.0	0.0
UAM15-2157-4	0.5	0.5	0.7	0.7	1.3	0.6	0.5	0.0	0.0
Mean	1.0	1.1	0.7	0.5	0.6	0.6	0.5	0.1	0.6
SE ±	0.228	0.378	0.230	0.097	0.141	0.048	0.084	0.441	1.028

3.3. Correlation Among Traits of the Cowpea Genotypes

The correlation analysis revealed significant relationships between grain yield and other agronomic traits (Figure 2). Grain yield showed a significant positive correlation with pods per plant (r = 0.58 *), seeds per pod (r = 0.70 *), and days to 50% flowering (r = 0.51 *). These positive correlations suggest that these are important yield components that can be targeted in selection programs. This aligns with the findings of Popoola et al. [24], who also reported similar relationships. A weak, non-significant positive correlation was observed between grain yield and 100-seed weight (r = 0.05) and fodder yield (r = 0.03). The weak correlation between grain yield and 100-seed weight in our study suggests that while large seeds are a critical market trait, selecting for seed size alone may not necessarily lead to higher yields. This highlights the complexity of yield as a quantitative trait and the need for a balanced selection approach that considers both yield components and market-preferred traits. Similar findings have been reported by other authors working on cowpea [26]. Days to maturity also showed a strong positive correlation with days to flowering.

Figure 2. Phenotypic correlation coefficients among yield components. The correlation moves from −1 (red) to +1 (blue). Dfflw: days to first flowering; D50flw: days to 50% flowering; D95mat: days to 95% maturity; pod_plt: pods per plant; Sd-pod: seeds per pod; sd100: 100-seed weight; Gyield: grain yield (kg ha⁻¹); Fyield: fodder yield (kg ha⁻¹).

3.4. Stability of the Cowpea Genotypes

The GGE biplot analysis proved to be a powerful tool for visualizing the G × E interaction and identifying superior genotypes. This methodology has been widely adopted in international breeding programs and has proven effective across various crops including sorghum, wheat, and other legumes [27,28]. The GGE biplot approach is particularly valuable for plant breeders as it provides both graphical and numerical outputs that facilitate decision-making in multi-environment trials [29]. The GGE biplot analysis, which explained 78.5% of the total variation (PC1 = 45.1%, PC2 = 30.7%), was used to visualize the performance and stability of the genotypes across the test environments (Figure 3). UAM15-2157-4 (G24) and IT16K-2556-2 (G10) were positioned farthest to the right on the biplot, indicating their superior mean yield performance across environments. Conversely, genotypes like IT16K-1685-2 were identified as low-yielding and unstable. UAM15-2157-4 exhibited a shorter vertical projection from the average environment axis, demonstrating greater yield stability compared to IT16K-2556-2. According to Yan et al. [30], a long projection of a genotype to the AEA line indicates a greater contribution to GEI while genotypes with a short projection line into the AEA are considered stable. This combination of high grain yield and stability makes UAM15-2157-4 particularly valuable for cultivation across diverse growing conditions.

Figure 3. A biplot showing the mean performance and stability of 24 cowpea genotypes evaluated at three locations in Nigeria in 2022 and 2023. The red dots indicate the environments. G1: IAR-09-1042-4; G2: IAR-17-1015; G3:IARD-17-10177; G4:IT08K-126-19; G5:IT08K-150-12; G6:IT14K-2188-1; G7:IT16K-1685-2; G8:IT16K-1968-3; G9:IT16K-2287-4; G10:IT16K-2556-2; G11:IT17K-1095-2-1; G12:IT17K-1266-2-1; G13:IT17K-1403-1-1; G14:IT17K-1707-2-2; G15:IT17K-1802-1; G16:IT17K-1809-4; G17:IT17K-2024-4; G18:IT17K-2357-1; G19:IT17K-3217-1; G20:UAM14-126-L2; G21:UAM14-126-L28; G22:UAM14-126-L29; G23:UAM14-126-L33-1; G24:UAM15-2157-4.

The biplot also offered practical recommendations by identifying which genotypes were best suited for specific environments, information that is invaluable for making targeted variety deployment decisions [31]. The “which-won-where” polygon view showed that the environments were separated into different sectors, with specific genotypes performing best in each (Figure 4). Genotype UAM15-2157-4 was the vertex genotype in the sector containing the BIU 2023, MKD 2022, MKD 2023, and SMR 2023 environments, indicating it was the highest-yielding line in those locations. This remarkable performance across multiple locations and years underscores the broad adaptation and yield reliability of this variety. The biplot also offered practical recommendations by identifying which genotypes were best suited for specific environments, information that is invaluable for making targeted variety deployment decisions.

Figure 4. Performance of 24 cowpea genotypes across five test environments in Nigeria during 2022 and 2023, highlighting the best and worst performers. G1:IAR-09-1042-4; G2: IAR-17-1015; G3:IARD-17-10177; G4:IT08K-126-19; G5:IT08K-150-12; G6:IT14K-2188-1; G7:IT16K-1685-2; G8:IT16K-1968-3; G9:IT16K-2287-4; G10:IT16K-2556-2; G11:IT17K-1095-2-1; G12:IT17K-1266-2-1; G13:IT17K-1403-1-1; G14:IT17K-1707-2-2; G15:IT17K-1802-1; G16:IT17K-1809-4; G17:IT17K-2024-4; G18:IT17K-2357-1; G19:IT17K-3217-1; G20:UAM14-126-L2; G21:UAM14-126-L28; G22:UAM14-126-L29; G23:UAM14-126-L33-1; G24:UAM15-2157-4. E1:BIU 2022; E2:BIU 2023;E3: MAKURDI 2022; E4:MAKURDI 2023; E5: SAMARU 2023.

The analysis of environment vectors (Figure 5) revealed that BIU 2022, BIU 2023, and MKD 2022 had the longest vectors, indicating their high discriminating ability for differentiating genotypes. BIU 2023 had the smallest angle with the average environment axis, making it the most representative test environment for generalizing genotype performance. Informative environments are used in selecting desirable genotypes as they provide useful information about the performance of a genotype [32,33]. These findings are particularly useful for identifying optimal testing locations for future variety evaluations.

Figure 5. GGE biplot showing the discriminativeness ability and representativeness of the 5 environments. G1:IAR-09-1042-4; G2: IAR-17-1015; G3:IARD-17-10177; G4:IT08K-126-19; G5:IT08K-150-12; G6:IT14K-2188-1; G7:IT16K-1685-2; G8:IT16K-1968-3; G9:IT16K-2287-4; G10:IT16K-2556-2; G11:IT17K-1095-2-1; G12:IT17K-1266-2-1; G13:IT17K-1403-1-1; G14:IT17K-1707-2-2; G15:IT17K-1802-1; G16:IT17K-1809-4; G17:IT17K-2024-4; G18:IT17K-2357-1; G19:IT17K-3217-1; G20:UAM14-126-L2; G21:UAM14-126-L28; G22:UAM14-126-L29; G23:UAM14-126-L33-1; G24:UAM15-2157-4. E1:BIU 2022; E2:BIU 2023; E3: MAKURDI 2022; E4:MAKURDI 2023; E5: SAMARU 2023.

3.5. Cowpea Performance in Intercropping Systems

Significant differences were observed among the genotypes for 100-seed weight, grain yield, and fodder yield (Table 6). This indicates that genetic variation exists among the genotypes for these traits. The cropping system also resulted in significant yield differences. The interaction effects of genotype × cropping system and genotype × cropping system × location were significant for grain and fodder yields. Research has consistently shown that the success of intercropping systems depends significantly on genotype selection, with significant genotype × cropping system interactions being a common finding across different legume–cereal combinations [34]. The performance of the selected cowpea lines under the maize intercropping system is presented in Figure 6 and Table 7. The evaluation of cowpea lines under maize–cowpea intercropping provided critical insights into their adaptability to mixed cropping systems, which are predominant among smallholder farmers in Nigeria. Under sole cropping, UAM15-2157-4 yielded 1566.9 kg/ha, significantly outperforming other tested varieties, including the popular varieties used as checks, IT89KD-288 (1201.4 kg ha⁻¹) and FUAMPEA 4 (1378.2 kg ha⁻¹) (Figure 6). These varieties are among the few improved cowpea varieties with an indeterminate growth habit used for cereal–cowpea intercropping by farmers in the savannas of Nigeria.

Table 6. Mean squares of grain yield and agronomic traits under cropping systems across two locations.

SOV	df	Number of Branches per Plant	Number of Peduncles per Plant	100-Seed Weight (g)	Grain Yield kg/ha	Fodder Yield kg/ha
Rep(Location)	2	2.96 *	0.10	14.30 *	22,084.20	1,190,274.50
Genotype (G)	14	1.26	4.09	92.64 **	408,958.34 **	1,248,104.64 *
Cropping system (CS)	1	0.18	607.95 **	79.54 **	4,297,133.60 **	167,962,814.00 **
G × CS	14	0.90	3.52	2.68	56,845.31 **	718,141.07
Location (L)	1	92.58 **	3289.67 **	23.32 *	1,710,185.50 **	11,331,577.00 **
G × L	14	0.89	1.18	6.37	236,972.11 **	2,313,856.14 **
CS × L	1	8.64 **	693.60 **	45.76 **	1,711,159.90 **	41,041,264.00 **
G × CS × L	14	1.31	4.42	4.39	48,619.59 **	13,64,608.71 **
Error	58	0.86	3.07	3.59	22,280.00 *	563,496.00 **

* and ** = significant at 5% and 1% level of probabilities.

Figure 6. Grain yield of cowpea lines under intercropping and sole cropping combined across 2 locations in 2024. Significant difference was observed among the genotypes at p < 0.05.

Table 7. Grain yield and other agronomic traits under sole cropping and intercropping combined across locations in 2024.

	Number of Branches per Plant		Number of Peduncles per Plant		Fodder Yield (kg/ha)		100-Seed Weight (g)
VARIETY	Sole	Inter	Sole	Inter	Sole	Inter	Sole	Inter
UAM15-2157-4	3.80	3.65	12.15	7.13	4291.68	2037.50	29.95	31.08
FUAMPEA 3	3.60	3.20	9.70	6.40	4252.78	1736.79	24.98	27.13
FUAMPEA 4	3.65	3.05	12.55	6.35	2752.08	1557.63	26.98	26.75
IT89KD-288	3.90	3.40	11.00	6.90	4609.74	2134.03	16.78	20.18
UAM 2-1	3.70	2.40	12.80	4.98	3784.03	1240.96	28.33	32.80
UAM 3-1	3.65	3.30	13.15	8.30	3811.82	1535.42	23.78	25.30
UAM 4-1	3.00	2.35	12.55	8.00	3909.74	1728.48	24.25	25.15
UAM 5-1	2.63	3.10	11.75	7.40	5078.48	2215.29	22.85	24.70
UAM 20-1	3.25	3.20	11.50	6.90	4828.47	2033.34	21.03	23.33
UAM 22-1	3.75	4.40	11.15	7.15	4635.40	1535.40	22.33	23.65
UAM 24-1	2.75	4.25	11.60	8.15	3995.13	2569.44	21.00	22.80
UAM 25-1	4.10	3.95	11.30	8.65	3395.46	1962.03	21.88	22.28
UAM 26-1	2.98	3.40	12.25	7.90	4345.16	1884.72	21.95	23.30
UAM 28-1	3.30	3.65	11.85	9.05	4254.88	1390.98	19.10	20.33
UAM 29-1	3.20	2.80	13.10	7.63	4221.53	1111.79	23.18	24.00
Mean	3.42	3.34	11.89	7.39	4144.42	1778.25	23.22	24.85
SE ±	0.2933	0.5590	1.0403	0.5820	392.7887	352.0880	1.0403	0.8190

More importantly, when intercropped with maize, UAM15-2157-4 had high productivity, yielding 1289.7 kg ha⁻¹, which represented 82.3% of its yield in sole cropping. In contrast, other varieties experienced substantially lower yield penalties under intercropping, which ranged from 65 to 75% compared to sole cropping. The ability of UAM15-2157-4 to produce high yields even when relayed into a standing maize crop indicates its tolerance to shading and effective competition for resources. This is a critical trait for intercropping success, as competition for light is often a major limiting factor for legume productivity in cereal–legume mixtures [35]. This result is consistent with the earlier findings by Lithourgiis et al. [36], who reported that cowpea varieties suitable for intercropping must have efficient use of resources and competitive advantage. Cereal–legume intercropping can contribute significantly to sustainable intensification, with benefits extending beyond yield to include improved soil health and enhanced biodiversity [37].

The variety’s 100-seed weight remained high under intercropping (31.1 g) compared to sole cropping (30.0 g), suggesting minimal impact of intercropping on seed development and filling. This stable seed size under different cropping systems is particularly valuable for farmers who practice intercropping but still aim to produce large-seeded cowpeas that command premium market prices and demand pull. This is similar to the report of Omoigui et al. [35] and Gondwe et al. [38].

Land Equivalent Ratio (LER) calculations for the maize–cowpea intercropping systems revealed values ranging from 1.24 to 1.53, indicating substantial land use efficiency advantages of intercropping over sole cropping. The highest LER (1.53) was calculated for UAM15-2157-4, demonstrating that this variety not only performs well in intercropping but also contributes to maximizing overall system productivity. This finding aligns with research by Namatsheve et al. [8], who reported that appropriate cereal–cowpea intercropping systems can increase land use efficiency by 20–60% compared to sole cropping. The high LER values confirm the overall productivity advantage of the intercropping system, reinforcing the importance of breeding for varieties adapted to this practice.

The superior performance of UAM15-2157-4 under intercropping has significant implications for smallholder farming systems in Nigeria, where intercropping serves multiple purposes, including risk mitigation, food security enhancement, and adaptation to climate variability. By maintaining high productivity under intercropping, this variety offers farmers the opportunity to maximize returns from limited land resources while benefiting from the ecological advantages of cereal–legume intercropping, such as enhanced soil fertility, improved moisture conservation, and reduced pest pressure.

4. Conclusions

This study successfully identified several elite cowpea lines with high yield potential, stability across diverse environments, and traits preferred by farmers and consumers in West Africa. The genotype UAM15-2157-4 was identified as the most outstanding line, combining high grain yield, broad stability, resistance to Striga gesnerioides, and desirable seed characteristics. Its superior performance in both sole and intercropping systems makes it a prime candidate for release to farmers. In addition to UAM15-2157-4, other genotypes like UAM14-126-L28 and IAR-09-1042-4 also demonstrated significant potential, providing a valuable set of genetic resources for future breeding programs. The findings underscore the importance of a multi-trait selection approach that integrates agronomic performance, stability, stress tolerance, and market-preferred traits to accelerate the adoption of improved cowpea varieties and enhance food security in the region.