Genetic Variability, Heritability, Correlation and Path Coefficient Analysis of Growth and Yield Traits of Cowpea [Vigna unguiculata (L.) Walp] Parental Genotypes and their F1 Crosses

Cowpea [Vigna unguiculata (L.) Walpers], also known as black-eyed pea or southern pea, is a self-pollinating crop belonging to the family Fabaceae. It is a popular legume grown for multiple utility purposes. Three cowpea genotypes were grown in accordance with a completely randomized design, replicated four times. The F1 populations were generated by crossing B138 with speckled grey Lecheng (SGL) and or ER7, which were crossed in all possible combinations to produce reciprocal crosses. Seventeen successful pods were generated from the 124 crosses, giving 13.7% success rate. There were two SGL X B138, two B138 X SGL, four B138 X ER7 and nine ER7 X B138 progenies generated from the crosses. These were then planted with their parents for evaluation of genetic variability, heritability, correlation, and path coefficient analysis of growth, yield, and yield-related attributes. High heritability, along high genetic advance, was reported for pod length, days to flowering, 100-seed weight, and seed yield per plot, indicating that these traits were governed by additive gene effects and that selection for these attributes would be beneficial. The phenotypic coefficient of variation (PCV) was consistently greater than the equivalent genotypic coefficient of variation (GCV), indicating that environmental influences have a role in the characteristic under study. The highest GCVs were recorded for number of leaves (23.82), seed yield per plant (32.51), leaf area (38.28), plant height (54.53) and seed yield per plot (80.17). The highest PCV was found in for number of pods per plant (23.96), seed yield per plant (32.72), leaf area (55.88), number of leaves (62.41), plant height (79.88), and seed yield per plot (80.17). The findings indicate that these traits could be utilized favorably to successfully identify and select preferred genotypes.

Introduction

Cowpea [Vigna unguiculata (L.) Walpers] is an annual crop mostly produced in the dry agroecosystems of the tropics in Latin America, Africa, and South Asia. Cowpea is farmed and consumed in several ways, depending on local preferences. Cowpea plants are multipurpose and are the most widely adapted, versatile, and nutritious legume crop [1]. It is commonly identified as ‘vegetable meat’ due to its high protein content (20%–25%), 50%–67% starch content, and multiple utility, drought tolerance, and nitrogen-fixing crop-nature, which make it a prominent grain legume [1]. The variability of the breeding material determines the effectiveness of a breeding program. Selection is also successful when there is a considerable level of genetic variability between individuals in the same group [2]. Knowledge of the nature and magnitude of genetic variability existing in any crop species performs a critical role in designing a suitable breeding method [3]. Previous reports [4] stated that genetic divergence among genotypes is more pronounced due to two factors: incorporating genetically diverse parents into the hybridization program enhances the probability of generating a strong heterotic effect. The segregating generation of crosses involving distantly related parents may exhibit a broad range of variability. In cowpea, quantifying the magnitude of variability in characters and classifying the varieties into groups will help researchers identify potential distinct genotypes with contrasting characters that can be used as parents for effective yield improvement in breeding.

Cowpea breeding is mostly dependent on the size and nature of interactions between genotypic and environmental variables. It is critical to partition the observed variability into heritable and nonheritable components in order to understand measures such as the genetic coefficient of variation, heritability, and genetic advance [3]. The genotypic coefficient of variation (GCV) and phenotypic coefficient of variation (PCV) measure the variability existing within the germplasms. High values of GCV and PCV suggest the possibility of large variations. Understanding the meaning of GCV and PCV values is essential for proficient utilization in agricultural development programs [3]. The choice of selection schemes to predict selection benefits and assess the relative significance of genetic effects is influenced by the degree of heritability of traits [5].

Another key element in selection programs is how much correlation exists between highly heritable characters and important characters, such as yield. The correlation could determine the direction and progress of the selection process. However, if the relationship between two attributes is the result of the indirect effects of other traits, then correlation coefficients, though beneficial in measuring the size and direction of trait associations, may be deceptive [6]. Therefore, the path coefficient is a method for studying the direct and indirect effects of correlated components of a complex trait because it measures the direct influence of one variable on another. Determining the interrelatedness of grain yield components can lead to a better understanding of the direct and indirect effects of particular components [6], [7]. The purpose of this study was to determine the genetic variability, correlations between different growth and yield traits, and their impact on cowpea grain production, both directly and indirectly, as well as to determine desirable traits for future breeding to improve yield.

Materials and Methods

The research experiment was conducted at the Botswana University of Agriculture and Natural Resources (BUAN) Content Farm, Sebele (12 km north of the Gaborone city center: altitude of 992 m, latitude of 24.33.40 S and longitude of 25.56.37 E) Botswana. The experiment was conducted under the greenhouse environment during August–September 2022 and January–April 2023. The F₁ populations were generated by crossing B138 with speckled grey Lecheng (SGL) and or ER7, which were crossed in all possible combinations as well as to produce reciprocal crosses. B138 is a late-maturing (91-day) landrace with reddish brown seeds. ER7 and SGL have high-intensity earliness characteristics, and their days to maturity range between 61 and 68 days. SGL is speckled grey in colour, while ER7 is creamy white in colour. The parental genotype seeds were planted in perforated plastic pots (5 L) in accordance with a completely randomized design replicated four times. Irrigation was performed at two-day intervals as per the plant requirement until maturity. Weeding was performed when needed.

At flowering, flower buds of the female plant that were ready to open were picked, and a pair of forceps was used to mechanically emasculate them. To prevent cross-contamination, the forceps were sanitized by immersing them in 70% ethanol after each emasculation step. For recognition, the emasculated flowers were labeled and tagged. The emasculated flower was bagged to discourage unintended pollination by various modes of pollination and pollinators [8]. On the following morning, between 06:30 am and 08:30 am, pollen was collected from the freshly opened flowers of the recurrent parents to pollinate the emasculated flower. To make sure the pollen stuck to the stigma of the emasculated flower, the pollen grains were gently rubbed over it. As soon as pollination was complete, the stigma was put back on the keel and sealed with the wing and petals to prevent desiccation [9]. The pollinated flowers were also bagged to avoid any cross-pollination [8]. After successful fertilization occurred and the pods developed to physiological maturity, the F₁ seeds were harvested, threshed, and stored properly for the next season. The F₁ seeds and the parental seeds were subsequently grown from the January to April 2023 season for evaluation.

Data Collection and Analysis

Data were collected on growth, yield, and yield-related traits using cowpea descriptors [10] for plant height (PH), number of branches (NOB), number of leaves (NOL), leaf area (LA), days to first flowering (DF), days to 50% flowering (DFF), days to first pod maturity (DFPM), days to 90% pod maturity (DNPM), days to maturity (DM), number of pods per plant (NOPP), pod length (PL), pod width (PW), number of seeds per plant (NOSP), 100 seed weight (HSW), seed yield per plant (SYP) and seed yield per plot (SYPP). The data were collected from nine plants of each parental line and F₁ lines. The data were analysed using the variability package in R programming language software.

Variability Calculations

Genotypic and Phenotypic Coefficients of Variation

Based on estimations of genotypic and phenotypic variance, the genotypic and phenotypic coefficients of variation were calculated using [11]: where $\overline{X}$ is the general mean, σg is the genotypic standard deviation, and σp is the phenotypic standard deviation. GCV and PCV were categorized as low (<10%), moderate (10%–20%) or high (>20%), as suggested by Burton and DeVane [11].

$$\begin{array}{rl}& Genotypiccoefficientofvariation(GCV)\\ & =GCV\left(\mathrm{\%}\right)=\frac{\sigma g}{\overline{X}}\times 100\end{array}$$

$$\begin{array}{rl}& Phenotypiccoefficientofvariation(PCV)=\\ & PCV\left(\mathrm{\%}\right)=\frac{\sigma p}{\overline{X}}\times 100\end{array}$$

Heritability

For each character, the broad-sense heritability was determined using the following formula, which was proposed by Burton and DeVane [11]: where σ² g is the genotypic component of variance, σ² p is the phenotypic component of variance.

$$\begin{array}{r}{h}^{2}b\left(\mathrm{\%}\right)=\frac{{\sigma }^{2}g}{{\sigma }^{2}p}\times 100\end{array}$$

The range of heritability in the broad sense was classified as low (<30%), moderate (30%–60%) or high (60% and above), as suggested by Johnson et al. [12].

Genetic Advance (GA)

Genetic advance is the anticipated genetic gain of a superior individual under a specific level of pressure for selection. The genetic basis of each character was calculated using the formula provided by Johnson et al. [12]: where GA is Genetic advance, h² (b) is heritability in a broad sense, k is Selection differential, which is equivalent to 2.06 at 5% intensity of selection and σp is Phenotypic standard deviation. Furthermore, the genetic advance as a percentage of the mean was processed by using the following formula:

$$\begin{array}{r}GA=K\times \sigma p\times h^2\left(b\right)\end{array}$$

$$\begin{array}{r}GAaspercentofmean=\frac{GA}{Grandmean}\times 100\end{array}$$

The genetic advance percentages were categorized as low (0%–10%), moderate (10.1%–20%) or high (>20.1%), as suggested by Johnson et al. [12].

Correlation Coefficient Analysis

Character association, or a character’s change brought about by another character’s change, is shown via correlation coefficient analysis. Using the covariance technique, correlation coefficients between genotype and phenotype were determined for different variables [13]. Calculating correlation coefficients allowed us to assess the strength of the association between the characters and yield, as well as among the yield characters. The phenotypic and genotypic relationships between yield and other characters were computed using the following formula:

$$\begin{array}{r}{r}_g\left(xy\right)=\frac{Co{v}_g\left(xy\right)}{\sqrt{{\sigma }_g^2}\left(x\right).{\sigma }_g^2\left(y\right)}\end{array}$$

$$\begin{array}{r}{r}_p\left(xy\right)=\frac{Co{v}_p\left(xy\right)}{\sqrt{{\sigma }_p^2}\left(x\right).{\sigma }_p^2(y)}\end{array}$$

The genotypic and phenotypic correlation coefficients are denoted by r_g (x y) and r_p (x y), respectively. Cov_g and Cov_p are the genotypic and phenotypic covariance of xy, respectively. σ² g and σ² p are the genotypic and phenotypic variances of x and y, respectively. The significance of the correlation coefficients was verified by comparing the phenotypic correlation coefficients to the table values [14] at (n-2) degrees of freedom at the 5% and 1% levels, where ‘n’ denotes the total number of pairs of observations used in the calculation.

Path Coefficient Analysis

Path coefficient analysis was utilized to calculate the direct and indirect contributions of various attributes to yield, following the recommendation of [15] and described by [16]. For the purpose of assessing different direct and indirect effects, the following simultaneous equations were created and described. The following simultaneous equations were solved in order to find the path coefficients:

where r1y is simple correlation coefficient between ×1 and y and dependent character Y, P1y is direct effect of ×1 on y and the dependent character Y, r12P2y is Indirect effect of ×1 on y through ×2, r12 is correlation coefficient between ×1 and ×2, r1k Pky is indirect effect of ×1 only through the kth variable.

$$\begin{array}{r}r1y=P1y+r12P2y+r13P3y+\dots \dots \dots +r1kPky\end{array}$$

Likewise, equations for r2y, r3y, r4y, and even longer were obtained.

Solving simultaneous equations allowed for the calculation of the direct and indirect effects. Apart from the direct and indirect effects, the residual effect was determined by the following formula:

where R² is P_1yr_1y + P_2yr_2y + P_3yr_3y + …………. P_iyr_iy

$$\begin{array}{r}Residualeffect\left(P_{ry}\right)=1-R^2\end{array}$$

P_ry – Residual effect

P_1y – Direct effect of ×1 on y

r_1y – Correlation coefficient of ×1 and y

P_2y – Direct effect of ×2 on y

r_2y – Correlation coefficient of ×2 and y.

P_3y – Direct effect of ×3 on y

r_3y – Correlation coefficient of ×3 and y

P_iy – Direct effect of ×i on y

r_iy – Correlation coefficient of ×i and y

where P_ry is residual effect, P_1y is direct effect of ×1 only, r_1y is correlation coefficient of ×1 only

$$\begin{array}{r}{P}_{ry}=\sqrt{1-P_{1y}}{r}_{1y}+P_{2y}{r}_{2y}+\dots P_{ky}{r}_{ky}\end{array}$$

The characters were considered to be negligible (0.00 to 0.09), low (0.10 to 0.19), moderate (0.20 to 0.29), high (0.30 to 0.99) or very high (>1.00).

Results

Genetic Variability Traits for Growth, Yield and Yield Related Traits of Cowpea Parental Lines and their F1 Genotypes

Analysis of variance (ANOVA) of growth traits exhibited significant variation between the 3 parental lines and 6 F₁ and 11 reciprocal F₁ generations for the studied characters (Fig. 1). For the data pertaining plant height, the F₁ and reciprocal crosses of B138 X SGL recorded a significantly higher plant height compared to both their parental lines (Fig. 1A). Similarly, all the F₁ crosses between B138 X ER7 had significantly higher plant height compared to both their parents, except for BE-1F₁ (64.03 cm), which only performed better than B138. Furthermore, two reciprocal crosses between ER7 and B138 (EB-3F₁ and EB-5F₁) showed significant difference in plant height (111.07 cm and 93.26 cm, respectively) compared to both parents, EB-1F₁ (62.67 cm), EB-2F₁ (63.42 cm), EB-4F₁ (55.88 cm), EB-6F₁ (54.52 cm), and EB-8F₁ (63.19 cm) plant heights were only higher than that of B138 parent only. When compared outside their parental lines, BS-2F1 line had the highest plant height (113.3 cm), which was higher than all the parents: B138 (50.18 cm), ER7 (87.33 cm), and SGL (51.02 cm). The minimal plant height was recorded at 42.69 cm (EB-9F₁), below all the parental lines.

Fig. 1. Parental genotypes and their F₁ crosses: (A) plant height, (B) number of branches, (C) number of leaves, and (D) leaf area. Error bars indicate standard deviations. Different lowercase letters indicate p ≤ 0.05, as determined by analysis of variance. BS-1F₁ and BS-2F₁ are the F₁ crosses of B138 X SGL. SB-1F₁ and SB-2F₁ are the reciprocal F₁ crosses of SGL X B138. BE-1F₁, BE-2F₁, BE-3F₁, and BE-4F₁ are the F₁ crosses of B138 X ER7 while EB-1F₁, EB-2F₁, EB-3F₁, EB-4F₁, EB-5F₁, EB-6F₁, EB-7F₁, EB-8F₁, and EB-9F₁ are the reciprocal F₁ crosses of ER7 X B138.

In terms of number of branches (Fig. 1B), all the F₁ lines (BS-1F₁ and BS-2F₁) and reciprocal lines (SB-1F₁ and SB-2F₁) had a greater number of branches than B138 (5.24). However, SGL parental line displayed a significantly higher number of branches compared to all its F₁ lines. The F₁ and reciprocals of B138 X ER7 also recorded a greater number of branches compared to B138 parental line. Notably, EB-9F₁ was the sole genotype with a higher number of branches (7.67) compared to both the parental lines in contrast to EB-2F₁ (4.58) and EB-4F₁ (5.17) that recorded a smaller number of branches than both parental lines. Across the parental lines, the least and maximum number of branches per plant was observed from the progenies of ER7 X B138, which ranged from 4.58 (EB-2F₁) to 7.67 (EB-9F₁).

The F₁ and reciprocal lines resulting from crosses between B138 X SGL as well as SGL X B138, exhibited a higher number of leaves per plant (Fig. 1C) compared to one parental line B138 (23.62). SGL parental line had a greater number of leaves compared to all the progenies at 33.42. Two of the F₁ crosses of B138 X ER7 (BE-1F₁ and BE-2F₁) displayed a greater number of leaves compared to both their parents (30.5 and 28.25, respectively). The BE-4F₁ (25.67) performed better than B138 parental line, while BE-3F₁ (22.33) had fewer leaves than both B138 and ER7 parental lines. The reciprocal crosses of ER7 X B138 parental lines, namely EB-8F₁ (31.17) and EB-9F₁ (46.00) exhibited more number of leaves than the parental lines while EB-1F₁ (23.92) and EB-3F₁ (25.50) recorded more number of leaves than B138 parental line only. The other reciprocal lines had a smaller number of leaves compared to both their parental lines. Across all the studied lines, EB-9F₁ (46.00) obtained the highest number of leaves per plant, while the least was obtained by EB-2F₁ (21.67).

Significant differences among the genotypes were recorded for leaf area (Fig. 1D) for all the parents, F₁ lines, and reciprocal crosses. All the F₁ and reciprocal crosses exhibited greater leaf area than all their parental lines. The leaf area across the studied genotypes ranged from 60.20 cm² (EB-9F₁) to 126.96 cm² (BS-1F₁). B138 (57.06 cm²) as the parental lines across all the lines being the one with the least leaf area.

Data for days to flowering (Table I) significantly varied among the different genotypes studied. The F₁ and the reciprocal crosses of SGL XB138 flowered 34 days after sowing earlier than both parental lines except for BS-2F_1, which was the earliest to flower at 33 days after sowing. For the F₁ crosses of B138 X ER7, they flowered at 34 days after sowing, except for BE-2F1, which flowered at 37 days. For the reciprocal crosses of ER7 X B138, EB-3F₁ was the earliest to flower at 33 days after sowing, while EB-1F1, EB-2F1, and EB-4F₁ flowered at 38 days after sowing. EB-7F₁, EB-8F₁, and EB-9F₁ flowered at 39 days. EB-6F₁ was the latest to flower at 40 days after sowing, later than ER7 parent only. Across the studied lines, the shortest days to flowering were recorded for BS-2F₁ and EB-3F₁. There was no significant difference for SB-1F₁, SB-2F₁, BS-1F₁, BE-1F₁, BE-3F₁, and BE-4F₁, as they flowered at 34 days after sowing. The late-flowering genotype was the parental line B138 at 46.44 days. The genotypes studied also recorded the same trend with number of days to 50% flowering. The number of days to 50% flowering varied from 35 days (recorded by BS-2F₁) to 50 days (recorded by B138).

Table I. Mean Physiological Traits of Cowpea Parental Genotypes and their F₁ Crosses

Genotypes	Pedigree	DF	DFF	DFPM	DNPM	DM
B138	Landrace	46.44 a	50 a	61 b	69.22 a	72.44 a
ER7	Variety	38.11 d	40.89 de	57.44 c	61.56 c	63.89 c
SGL	Landrace	40.44 b	44 b	62.89 a	65.11 b	67.78 b
BS-1F1	B138 X SGL	34 g	37 g	49 f	54 ef	56 f
BS-2F1	B138 X SGL	33 g	35 h	50 ef	54 ef	56 f
SB-1F1	SGL X B138	34 g	36 gh	50 ef	53 f	56 f
SB-2F1	SGL X B138	34 g	37 g	51 ef	54 ef	58 e
BE-1F1	B138 X ER7	34 g	37 g	50 ef	54 ef	56 f
BE-2F1	B138 X ER7	37 ef	39 f	52 e	55 e	57 ef
BE-3F1	B138 X ER7	34 g	37 g	51 ef	55 e	57 ef
BE-4F1	B138 X ER7	34 g	37 g	50 ef	54 ef	56 f
EB-1F1	ER7 X B138	38 de	40 ef	55 d	59 d	62 d
EB-2F1	ER7 X B138	38 de	40 ef	61 b	66 b	68 b
EB-3F1	ER7 X B138	33 g	37 g	49 f	53 f	56 f
EB-4F1	ER7 X B138	38 de	40 ef	62 ab	66 b	68 b
EB-5F1	ER7 X B138	36 f	37 g	50 ef	54 ef	57 ef
EB-6F1	ER7 X B138	40 bc	42 cd	61 b	66 b	68 b
EB-7F1	ER7 X B138	39 cd	43 bc	61 b	66 b	68 b
EB-8F1	ER7 X B138	39 cd	42 cd	61 b	66 b	68 b
EB-9F1	ER7 X B138	39 cd	43 bc	55 d	59 d	62 d
R2		0.99	0.99	0.99	0.99	1.00
MSe		0.44	0.85	1.63	0.89	0.48
GM		38.04	40.91	56.23	60.60	63.13
CV		1.75	2.26	2.27	1.56	1.10
LSD		0.39	0.54	0.74	0.55	0.40

A significant difference was obtained for a number of days to first pod maturity amongst the studied genotypes (Table I). BS-1F₁ (49 days) and BS-2F₁ (50 days) recorded number of days to first pod maturity earlier than both their parents. The F₁ reciprocal crosses, SB-1F1 (50 days) and SB-2F1 (51 days), also reached first pod maturity earlier than both parents. For the F₁ crosses of B138 X ER7; BE-1F₁ and BE-4F₁ reached first pod maturity at 50 days while BE-2F₁ at 52 days and BE-3F₁ at 51 days. There was no significant difference between the parental line B138 and the reciprocal crosses, EB-2F1, EB-6F1, EB-7F1, and EB-8F1, as they both reached first pod maturity at 61 days after sowing, but a significant difference was observed for ER7 parent that reached first pod maturity at 57.44 days after sowing. Across the studied genotypes, days to first pod maturity ranged from 49 days (recorded by EB-3F₁) to 62.89 days (recorded by SGL).

Data for days to 90% pod maturity (Table I) showed no significant difference for BS-1F₁, BS-2F_1, and SB-2F₁ as they reached 90% pod maturity at 54 days, except for SB-1F₁ that reached 90% pod maturity at 53 days after sowing. The F₁ progenies of B138 X ER7 reached 90% pod maturity at 54 days (BE-1F₁ and BE-4F₁) and 55 days (BE-2F₁ and BE-3F₁). The reciprocal crosses of ER7 X B138 showed a significant variation among themselves. EB-3F₁ was the first to reach 90% pod maturity at 53 days, followed by EB-5F₁ at 54 days after sowing. EB-1F₁ and EB-9F₁ recorded 59 days to 90% pod maturity. The above progenies reached days to 90% pod maturity earlier than both their parental lines. There was no significant difference amongst EB-2F₁, EB-4F₁, EB-6F₁, EB-7F₁, and EB-8F₁ as they all recorded 66 days to 90% pod maturity. These progenies reached days to 90% pod maturity later than ER7 (that recorded 61.56 days) but earlier than B138 (that recorded 69.22 days) parental lines. Across the studied genotypes, days to 90% pod maturity ranged from 53 to 69.22 days.

Data recorded for days to maturity (Table I) showed no significant difference for BS-1F1, BS-2F1, and SB-1F₁ as they reached maturity at 56 days after sowing. SB-2F₁ reached maturity at 58 days after sowing. For the F₁ crosses of B138 X ER7, slight variation was recorded. BE-1F₁ and BE-4F₁ recorded 56 days to maturity, while BE-2F₁ and BE-3F₁ recorded 57 days to maturity. These progenies reached maturity earlier than both their parental lines. For the reciprocal crosses, EB-3F₁ and EB-5F₁ were the earliest to reach maturity at 56 and 57 days, respectively. No significant difference was observed for EB-1F₁ and EB-9F₁ as they reached maturity at 62 days. No variation was recorded for EB-2F₁, EB-4F₁, EB-6F₁, EB-7F₁, and EB-8F₁ as they all reached maturity at 68 days later than ER7 (63.89 days) but earlier than B138 (72.44 days) parental lines. Across the studied genotypes, days to maturity ranged from 56 to 72.44 days after sowing.

Data recorded for a number of pods per plant (Table II) showed significant variation for the parental lines and their offspring. BS-1F₁ (11.67) recorded more number of pods compared to SGL (9.11) parental line only. BS-2F₁ (14) and the two reciprocal F₁s; SB-1F₁ (14) and SB-2F₁ (14.33) had more pods compared to both parental lines. The F₁ crosses of B138 X ER7 were significantly different from both parents except for BE-4F1, which recorded the same number of pods with the ER7 (13.33) parental line. BE-1F₁ (15.33), BE-2F₁ (14.67), and BE-3F₁ (15) were not significantly different from each other. The reciprocal crosses recorded a wide range of significant differences among each other. EB-1F₁ (13), EB-2F₁ (11.33), EB-4F₁ (11.33), EB-5F₁ (12.67), EB-6F₁ (10.67) recorded less number of pods per plant compared to both their parental lines. EB-3F₁ (15), EB-7F₁ (14.67), EB-8F₁ (15), and EB-9F₁ (21.33) recorded a significantly higher number of pods per plant than both their parental lines. Across the studied genotypes, the number of pods per plant ranged from 9.11 to 21.33.

Table II. Mean Yield-Related Traits of Cowpea Parental Genotypes and their F₁ Crosses

Genotypes	Pedigree	NOPP	PL (cm)	PW (cm)	NOSP	HSW (g)	SYP (g)	SYPP (kg)
B138	Landrace	13.78 bcd	17.1 b	2.69 a	15.89 ab	12.66 g	10.9 e	0.196 a
ER7	Variety	13.33 cd	18.46 a	2.6 ab	15.33 abc	10.03 q	8.97 h	0.182 b
SGL	Landrace	9.11 g	18.76 a	2.68 a	14.44 c	8.78 r	4.53 l	0.154 c
BS-1F1	B138 X SGL	11.67 def	17.5 b	2.73 a	14.33 cd	12.13 i	7.57 j	0.013 r
BS-2F1	B138 X SGL	14 bcd	17.2 b	2.63 ab	15.33 abc	12.14 h	8.43 i	0.031 o
SB-1F1	SGL X B138	14 bcd	17.2 b	2.67 ab	13 d	11.77 n	6.8 k	0.012 s
SB-2F1	SGL X B138	14.33 bc	17.03 b	2.63 ab	15 abc	12.88 e	13.15 d	0.025 p
BE-1F1	B138 X ER7	15.33 b	16.77 b	2.57 abc	15.33 abc	11.82 l	12.59 d	0.046 l
BE-2F1	B138 X ER7	14.67 bc	16.97 b	2.67 ab	16 ab	12.86 f	18.42 a	0.034 n
BE-3F1	B138 X ER7	15 bc	17.07 b	2.6 abc	16.33 a	11.78 m	6.52 k	0.023 q
BE-4F1	B138 X ER7	13.33 cde	16.9 b	2.57 abc	15 abc	12.14 h	14.25 c	0.055 i
EB-1F1	ER7 X B138	13 cdef	17.23 b	2.67 ab	16.33 a	11.85 k	7.02 jk	0.041 m
EB-2F1	ER7 X B138	11.33 ef	17.13 b	2.67 ab	16.33 a	13.3 c	12.77 d	0.066 f
EB-3F1	ER7 X B138	15 bc	17.2 b	2.67 ab	14.33 cd	12.13 i	9.6 g	0.053 j
EB-4F1	ER7 X B138	11.33 ef	16.9 b	2.53 bc	16.33 a	13.09 d	16.32 b	0.09 d
EB-5F1	ER7 X B138	12.67 cdef	16.87 b	2.57 abc	15.33 abc	11.46 o	11.18 e	0.06 g
EB-6F1	ER7 X B138	10.67 fg	16.67 b	2.43 c	15.67 abc	12 j	10.59 ef	0.057 h
EB-7F1	ER7 X B138	14.67 bc	16.77 b	2.53 bc	14.67 bc	16.47 a	12.66 d	0.067 e
EB-8F1	ER7 X B138	15 bc	16.63 b	2.43 c	15.33 abc	11.09 p	9.43 gh	0.051 k
EB-9F1	ER7 X B138	21.33 a	16.87 b	2.63 ab	16.33 a	13.39 b	10.36 f	0.057 h
R2		0.96	0.90	0.77	0.88	1.00	1.00	1.00
MSe		1.93	0.26	0.01	0.78	0.00	0.13	0.00
GM		13.31	17.38	2.62	15.31	11.80	10.03	0.09
CV		10.43	2.94	3.89	5.76	0.00	3.59	0.00
LSD		0.81	0.30	0.06	0.51	0.00	0.21	0.00

Pod length (Table II) was not significantly different amongst the F1 and reciprocal crosses of all the progenies studied and the B138 parental line, which recorded 17.1 cm. ER7 (18.46 cm) and SGL (18.76 cm) parental lines were significantly different from their progenies. Across the studied genotypes, pod length ranged from 16.63 cm (EB-8F₁) to 18.76 cm (SGL). With regard to number of seeds per pod (Table II) BS-2F₁ (15.33) and SB-2F₁ (15) were significantly higher than SGL (14.44) parental lines but not significant from B138 (15.89). BS-1F₁ (14.33) and BS-1F₁ (13) recorded less number of seeds per pod compared to both their parental lines. No significant difference amongst the parental lines and the F₁ crosses of B138 X ER7 was recorded with regard to number of seeds per pod. BE-1F₁, BE-2F₁, BE-3F₁, and BE-4F₁ recorded 15.33, 16, 16.33, and 15 number of seed per pod, respectively. For the reciprocal crosses, EB-1F₁, EB-2F₁, EB-4F₁ and EB-9F₁ were not significant amongst themselves as they recorded 16.33 number of seeds per pods higher than both parental lines. EB-3F₁ (14.33), EB-5F₁ (15.33), EB-6F₁ (15.67), EB-7F₁ (14.67), and EB-8F₁ (15.33) recorded slight variation of number of seeds per pod than both the parental lines. Across the studied genotypes, the number of seeds per pod varied from 13 to 16.33.

Data recorded for 100 seed weight (Table II) for the F₁ generations of B138 X SGL (BS-1F1 and BS-2F1 at 12.13 g and 12.14 g, respectively) were higher than both the parental lines. For the reciprocals, SB-1F1 (11.77 g) recorded a higher seed weight than SGL (8.78 g) but less than B138 (12.66 g) parental line, whereas SB-2F1 (12.88 g) recorded a significantly higher seed weight compared to both parental lines. For the F₁ crosses of B138 X ER7; BE-2F₁ (12.86 g) recorded more seed weight than both the parental lines. BE-1F₁ (11.82 g), BE-3F₁ (11.78 g) and BE-4F₁ (12.14 g) recorded less number of seed weight than B138 (12.66 g) but more seed weight compared to ER7 (10.03 g). F₁ reciprocal crosses of ER7 X B138 recorded significant seed weight amongst themselves and their parental lines. EB-2F₁ (13.3 g), EB-4F₁ (13.09 g), EB-7F₁ (16.49 g), and EB-9F₁ (13.39 g) recorded more seed weight than both the parental lines. EB-1F₁ (11.85 g), EB-3F₁ (12.13 g), EB-5F₁ (11.46 g), EB-6F₁ (12 g), and EB-8F₁ (11.09 g) recorded less seed weight than B138 (12.66 g) but more than ER7 (10.03 g). Across the studied genotypes, 100-seed weight ranged from 8.78 g to 16.47 g.

Results pertaining to seed yield per plant (Table II) showed significant differences across all the studied parental lines and their progenies. BS-1F₁ (7.57 g), BS-2F₁ (8.43 g), and SB-1F₁ (6.8 g) recorded more seed yield per plant than SGL (4.53 g) but less than B138 (10.9 g) parental line. SB-2F₁ (13.15 g) recorded more seed yield per plant compared to both the parental lines. For the F₁ crosses of B138 X ER7; BE-1F₁ (12.59 g), BE-2F₁ (18.42 g) and BE-4F₁ (14.25 g) recorded more seed yield per plant than both the parental lines except for BE-3F₁ (6.52 g) that recorded less seed yield per plant compared to both parental lines. The reciprocal F1 crosses of ER7 X B138 recorded significantly higher seed yield per plant compared to all the parental lines except for EB-1F₁ (7.02 g), which recorded less seed yield per plant than both parental lines. Across all the studied genotypes, seed yield per plant ranged from 4.53 g to 18.42 g.

Data for seed yield per plot (Table II) displayed highly significant differences among the studied parental lines and their progenies. BS-1F₁ (0.013 g), BS-2F₁ (0.031 g), SB-1F₁ (0.012 g), and SB-2F₁ (0.025 g) recorded less seed yield per plot compared to B138 (0.196 g) and SGL (0.154 g) parental lines. The F₁ genotypes of B138 X ER7 and the reciprocal F₁ progenies recorded significantly lower seed yield per plot compared to both parental lines. Across the studied genotypes, pod yield per plot varied from 0.012 g to 0.196 g.

The phenotypic and genotypic variances were portioned into phenotypic coefficient of variation (PCV) and genotypic coefficient of variation (GCV), as shown in Fig. 2. High PCV coupled with high GCV was observed for seed yield per plant (32.72 and 32.51, respectively), leaf area (55.88 and 38.28, respectively), number of leaves (62.41 and 23.82, respectively), plant height (79.88 and 54.53, respectively) and seed yield per plot (80.17 and 80.17, respectively). Moderate PCV and GCV were observed for the 100-seed weight (12.14 and 12.14, respectively). In this study, low PCV and GCV were observed for days to 90% pod maturity (9.86 and 9.85, respectively), days to 50% pod maturity (9.67 and 9.65, respectively), days to maturity (9.35 and 9.34, respectively), days to 50% flowering (9.08 and 9.06 respectively), days to flowering (9.07 and 9.07, respectively), number of seeds per pod (7.90 and 4.29 respectively), pod width (3.70 and 2.71, respectively) and pod length (3.43 and 3.00, respectively).

Fig. 2. Phenotypic and genotypic coefficient of variation for growth and yield traits in cowpea parental genotypes and their F1 crosses.

The highest heritability in the broad sense (Fig. 3) was recorded for pod length (76.39%), seed yield per plant (98.7%), days to first pod maturity (99.49%), days to 50% flowering (99.63%), days to maturity (99.86%), days to 90% pod maturity (99.88%), days to flowering (99.9%), 100 seed weight (100%) and 100% seed yield per plot. High heritability coupled with high genetic advance as a percentage of the mean was reported for 100-seed weight (100% and 25.02%, respectively), seed yield per plot (100% and 165.15%, respectively), days to 90% pod maturity (99.88 and 20.29%, respectively), and seed yield per plant (98.70% and 66.53%, respectively). Moderate heritability coupled with moderate genetic advance as a percentage of the mean was recorded for a number of pods per plant (35.21% and 17.38%, respectively). Low heritability coupled with low genetic advance as a percentage of the mean was obtained for number of seeds per pod (29.49% and 4.80%, respectively).

Fig. 3. Heritability and genetic advance as percentage of the mean for growth and yield traits in cowpea parental genotypes and their F1 crosses.

Trait Associations

Significant and positive correlations at the genotypic level (Table III) were observed between days to flowering and days to 50% flowering (0.97^**), days to first pod maturity (0.82^**), days to 90% pod maturity (0.87^**), days to maturity (0.89^**), number of seeds per pod (0.48^*) and seed yield per plot (0.74^**). Days to 50% flowering had a positive significant correlation with days to first pod maturity (0.80^**), days to 90% pod maturity (0.85^**), days to maturity (0.87^**), and seed yield per plot (0.76^**). The number of days to first pod maturity was positively correlated with the number of days to 90% pod maturity (0.98^**), days to maturity (0.98^**), and the seed yield per plot (0.62^**).

Table III. Genotypic Correlation Coefficient for Yield Characters of Cowpea Parents and their Offspring

Traits	DF	DFF	DFPM	DNPM	DM	NOSP	HSW	SYP	SYPP
DF	1	0.97**	0.82**	0.87**	0.89**	0.48*	0.03 ns	0.03 ns	0.74**
DFF		1	0.8**	0.85**	0.87**	0.39 ns	0.07 ns	−0.01 ns	0.76**
DFPM			1	0.98**	0.98**	0.43 ns	0.05 ns	0.04 ns	0.62**
DNPM				1	1**	0.46*	0.11 ns	0.05 ns	0.65**
DM					1	0.45*	0.1 ns	0.03 ns	0.67**
NOSP						1	0.23 ns	0.46*	0.21 ns
HSW							1	0.54*	−0.27 ns
SYP								1	−0.04 ns
SYPP									1

The number of days to 90% pod maturity was significantly positively correlated with days to maturity (1.00^**), the number of seeds per pod (0.46^*), and the seed yield per plot (0.65^**). A significant positive correlation was also detected between the number of days to maturity and the number of seeds per pod (0.45^*) and seed yield per plot (0.67^**). The number of seeds per pod was significantly positively correlated with seed yield per plant (0.46^*). A significant positive correlation was observed between 100-seed weight and seed yield per plant (0.54^*).

A significant and positive association at the phenotypic level (Table IV) was observed between days to first flowering and days to 50% flowering (0.97^**), days to first pod maturity (0.82^**), days to 90% pod maturity (0.87^**), days to maturity (0.89^**), number of seeds per pod (0.48^*) and seed yield per plot (0.74^**). Days to 50% flowering had a significant positive correlation with days to first pod maturity (0.80^**), days to 90% pod maturity (0.85^**), days to maturity (0.87^**), and seed yield per plot (0.76^**). A positive and significant correlation was recorded between days to first pod maturity and days to 90% pod maturity (0.98^**), days to maturity (0.98^**), and seed yield per plot (0.62^**). Days to 90% pod maturity recorded a positive and significant correlation with days to maturity (1.00^**), number of seeds per pod (0.46^*), and seed yield per plot (0.65^**). Days to maturity recorded positive and significant correlation with number of seeds per pod (0.45^*) and seed yield per plot (0.67^**). Number of seeds per pod with seed yield per plant (0.46^*), 100-seed weight with seed yield per plant (0.54^*).

Table IV. Phenotypic Correlation Coefficients for Yield Characters of Cowpea Parents and their Offspring

Traits	DF	DFF	DFPM	DNPM	DM	NOSP	HSW	SYP	SYPP
DF	1	0.97**	0.82**	0.87**	0.89**	0.48*	0.03 ns	0.03 ns	0.74**
DFF		1	0.8**	0.85**	0.87**	0.39 ns	0.07 ns	−0.01 ns	0.76**
DFPM			1	0.98**	0.98**	0.43 ns	0.05 ns	0.04 ns	0.62**
DNPM				1	1**	0.46*	0.11 ns	0.05 ns	0.65**
DM					1	0.45*	0.1 ns	0.03 ns	0.67**
NOSP						1	0.23 ns	0.46*	0.21 ns
HSW							1	0.54*	−0.27 ns
SYP								1	−0.04 ns
SYPP									1

Direct and Indirect Effects of Different Traits on the Pod Yield Per Plant

The number of days to flowering, 50% flowering, first pod maturity, 90% pod maturity, and days to maturity had strong positive direct effects (0.599, 0.583, 0.492, 0.520, and 0.533, respectively) on the pod yield per plant (Table V). Its indirect effects through other characters were high for days to 90% pod maturity (0.241) and days to maturity (2.279). The correlation coefficients between these factors and seed yield per plant were positive. The other indirect effects were low and negligible. The number of days to maturity had a positive direct effect (0.533), with a positive correlation of 0.013 with the pod yield per plant. A strong positive indirect effect was shown through days to 90% pod maturity (0.276). A strong negative indirect effect was observed in the days to the first pod maturity (−2.975). The other indirect effects were low and negligible.

Table V. Path Coefficient Analysis for Yield Characters of Cowpea Parents and their Offspring at the Genotypic Level

Traits	DF	DFF	DFPM	DNPM	DM	NOPP	PL	PW	NOSP	HSW	SYP
DF	0.599	−0.191	−2.506	0.241	2.279	−0.023	0.180	0.103	0.043	0.009	0.011
DFF	0.583	−0.196	−2.425	0.235	2.227	−0.004	0.219	0.069	0.036	0.019	−0.006
DFPM	0.492	−0.156	−3.048	0.272	2.499	−0.058	0.247	0.305	0.039	0.013	0.017
DNPM	0.520	−0.166	−2.991	0.278	2.548	−0.052	0.129	0.293	0.042	0.029	0.021
DM	0.533	−0.171	−2.975	0.276	2.561	−0.049	0.153	0.263	0.041	0.027	0.013
NOPP	−0.097	0.005	1.275	−0.104	−0.903	0.139	−0.797	−0.027	0.024	0.119	0.066
PL	0.068	−0.027	−0.477	0.023	0.248	−0.070	1.578	−0.366	−0.032	−0.176	−0.225
PW	−0.082	0.018	1.237	−0.108	−0.894	0.005	0.768	−0.752	−0.027	−0.026	−0.110
NOSP	0.285	−0.077	−1.321	0.128	1.143	0.036	−0.547	0.221	0.091	0.059	0.190
HSW	0.020	−0.014	−0.148	0.031	0.265	0.064	−1.065	0.074	0.020	0.261	0.223
SYP	0.016	0.003	−0.122	0.014	0.081	0.022	−0.857	0.200	0.042	0.141	0.415
Residual effect: −0.0227

Table VI. Path Coefficient Analysis for Yield Characters of Cowpea Parents and their Offspring at the Phenotypic Level

Traits	DF	DFF	DFPM	DNPM	DM	NOPP	PL	PW	NOSP	HSW	SYP
DF	−0.053	0.370	−1.500	0.285	1.541	−0.006	0.071	0.025	0.003	−0.008	0.010
DFF	−0.052	0.381	−1.454	0.277	1.502	−0.001	0.098	0.017	0.002	−0.016	−0.005
DFPM	−0.044	0.303	−1.830	0.321	1.688	−0.015	0.113	0.076	0.002	−0.011	0.015
DNPM	−0.046	0.322	−1.792	0.328	1.722	−0.013	0.054	0.074	0.003	−0.025	0.019
DM	−0.047	0.330	−1.783	0.326	1.733	−0.012	0.061	0.066	0.002	−0.023	0.012
NOPP	0.005	−0.007	0.461	−0.073	−0.363	0.058	−0.231	0.007	0.001	−0.060	0.030
PL	−0.005	0.051	−0.281	0.024	0.144	−0.018	0.738	−0.120	−0.002	0.131	−0.180
PW	0.005	−0.025	0.528	−0.092	−0.436	−0.001	0.338	−0.253	−0.001	0.016	−0.078
NOSP	−0.014	0.078	−0.397	0.082	0.401	0.005	−0.153	0.025	0.010	−0.027	0.099
HSW	−0.002	0.027	−0.088	0.036	0.179	0.016	−0.435	0.018	0.001	−0.222	0.201
SYP	−0.001	−0.005	−0.073	0.016	0.055	0.005	−0.354	0.054	0.003	−0.119	0.376
Residual effect: 0.1313

The number of seeds per pod had a moderate positive direct effect (0.285) on the pod yield per plant, and the correlation coefficient was positive (0.190). A strong positive indirect effect of days to maturity (1.143) was observed on the pod yield per plant. A negligible positive direct effect was shown for the number of seeds per pod (0.005) and pod width (0.005) on the seed yield per plant, with positive and negative correlation coefficients (0.030) and (−0.078), respectively. All the other traits had negligible negative direct effects on the pod yield per plant, with a low positive correlation coefficient (0.010). Days to flowering showed a high positive indirect effect via days to maturity (1.541), with a high negative through days to first pod maturity (−1.500) on pod yield per plant.

Discussion

Genetic Variability

Genetic variation in cowpea is essential for cowpea improvement programs. A narrow genetic base can lead to vulnerability in improved varieties and limit genetic gains [17]. Studies have shown that genetic variation in cowpea is vital for various traits, such as seed size, plant height, leaf area, which is linked to adaptation in plants [18]. Furthermore, the genetic variability among cowpea landraces is crucial for preserving local varieties and forming the basis for developing improved varieties [19]. From this study, significant variability among the genotypes was recorded for plant height, number of branches, number of leaves per plant, and leaf area between parental material, F₁, and reciprocal progenies.

On this study, higher plant height was observed in BS-2F₁, SB-2F₁, BE-3F₁, and EB-3F₁ progenies, with BS-2F₁ being the highest at 113.3 cm. These progenies performed better than their parental genotypes under study in this trait. Plant height influenced the competitive ability of cowpea genotypes with weeds. Studies have shown that taller plants often exhibit better resilience to water scarcity, making plant height a key factor in determining the drought tolerance of cowpea varieties [20]. From the previous results, Devi and Jayamani [2] recorded a wide range of variations in plant height (30.8 cm–76.7 cm). In cowpea breeding, plant height is often considered alongside other important traits, such as yield components. A positive correlation between plant height and yield has been observed in this study, suggesting the potential of considering plant height as a trait to improve cowpea yield. This was found in SB-2F₁ that recorded significant amount of height and yield per plant of 13.15 g.

The number of branches in cowpea is an important trait that significantly influences yield potential and agronomic traits. As reported in this study almost all the studied progenies recorded more number of branches than the recurrent parent (B138). A range of a number of branches was recorded between 4.58 and 7.67. Studies have demonstrated that the number of branches per plant is positively correlated with pod yield in cowpea, indicating that an increased number of branches can lead to higher yields [21]. As indicated by this study, the progeny (EB-9F₁) that showed the highest number of branches also recorded more number of pods per plant (21.33) compared to other progenies and parental lines. Additionally, the number of productive branches is among the agronomic traits selected to enhance cowpea yield potential [22]. In cowpea breeding programs, increasing grain yield is a primary objective, and the number of branches is a trait that can significantly contribute to achieving this goal [23].

Leaf area is an important parameter in various agronomic and physiological studies of cowpea plants [24]. It was reported by Takai et al. [25] that thicker leaves with a larger surface area are essential for enhancing leaf photosynthesis as they can efficiently capture light energy due to having more chlorophyll per unit leaf area. From this study, all progenies had more leaf area as compared to their parental lines with most of them recording more seed yield per plant compared to their parental lines. This indicated that further improvement of these lines could also improve their yield potential.

For days to flowering, the earliest progenies to flower were BS-2F₁ and EB-3F_1, while the late flowering was the parental line B138. Early flowering in cowpea is a crucial trait that significantly impacts plant development and yield potential. From this study, progenies (BS-1F1, SB-1F1, BE-1F1, and BE-4F1) that flowered early at 34 days after sowing were also the earliest to reach pod maturity at 56 days. This emphasized the importance of early flowering for timely pod development [26]. Previous studies have also shown that crossing cowpea lines with early flowering traits can lead to progenies combining desirable attributes, such as early flowering and determinate maturing characteristics [27], [28]. Similar studies also recorded a wide range in variation in flowering time, Devi and Jayamani [2] observe days to 50% flowering in cowpea to be 35.5 to 56.5 days. These results were consistent with conclusions of Lazaridi et al. [29], where they reported different days to flowering of the landraces BGE038478 (85.15 days) and BGE038479 (82.80 days), as well as the breeding line IT97K-499-35 (74.94 days) in comparison to the other cowpea accessions they studied. Similarly, [30] stated that days to 50% flowering in cowpea genotypes studied ranged from 46.00–71.00 days after sowing.

Data recorded in relation to number of days to maturity indicated that there was an improvement in the F1 progenies as compared to their parental lines. Almost all the progenies reached maturity earlier than their parental lines; this displays that the progenies will be harvested before the drought season approaches, allowing breeders to optimize the varieties they release to protect their yield potential. Similar studies like Devi and Jayamani [2] recorded days to maturity range at 63.5 to 92 days after sowing, while [30] days to maturity ranged from 75.00 to 105 days in cowpea genotypes studied. A great variability was recorded amongst the genotypes for number of pods per plant, pod length, pod width, number of seeds per pod, 100 seed weight, seed yield per plant and seed yield per plot. These results are in line with the results of [31], who revealed significant differences between cowpea parents and their hybrids for all the quantitative traits studied. In other cowpea studies, [30] observed that number of pods per plant range from 8.00–34.00, pod length from 11.02 cm to 22.03 cm, number of seeds per pod ranged from 7.75 to 16.25 while yield per plant ranged from 6.80 g to 42.35 g from the 55 cowpea genotypes studied.

From this study, high PCV and GCV were recorded for seed yield per plant, leaf area, number of leaves, plant height, and seed yield per plot. This suggests that the genotypes for these variables under study exhibit a higher degree of variation. In other studies, [30] reported high PCV coupled with high GCV for the number of pods per plant (30.56 and 24.86, respectively) and grain yield per plant (34.34 g and 30.67 g, respectively). Moderate PCV and GCV were observed for the seed yield per plant, days to 50% flowering, number of pods per plant, and days to maturity. This moderate variability indicates moderate amount of variability for the traits in their studied genotypes. The results obtained are in line with what was found by Kavyashree et al. [30], who observed moderate PCV and GCV for 100 seed weight (14.85 and 11.94, respectively) in cowpea genotypes. In contrast, Manggoel et al. [6] recorded the highest PVC and GCV for 100-seed weight (41.46 and 38.47, respectively), followed by number of pods per plant (29.88 and 28.23, respectively). Low PCV and GCV were observed for days to 90% pod maturity, days to 50% pod maturity, days to maturity, days to 50% flowering, days to flowering, number of seeds per pod, pod width, and pod length. This indicates a low amount of variability for the traits in their studied genotypes. Similarly, Manggoel et al. [6] recorded low PVCs, and GVCs were recorded for a number of seeds per pod (24.11 and 20.66, respectively) and pod length (19.81 and 15.79, respectively).

The traits (pod length, days to flowering, 100 seed weight, and seed yield per plot) recorded high heritability coupled with high genetic advance as a percent of the mean; these showed that these traits are slightly influenced by the environmental effects and controlled by additive gene action [32]. Therefore, superior results can be obtained via direct selection centered on these attribute’s phenotypic performance. High heritability and high GAM were also reported by [2], [30], [32]. The number of pods per plant showed moderate heritability and genetic advance as a per cent mean, this may suggest the influence of non-additive gene action and the significant influence of the environment on the expression of the genes. These characters can be exploited through manifestation of dominance and epistatic components through heterosis breeding [5]. Thus, breeders should employ an effective breeding method that takes advantage of both additive and non-additive gene effects simultaneously.

Trait Associations

From this study, significant correlations were recorded at genotypic and phenotypic levels for the characters studied. In relation to these results, a high positive correlation between the number of pods per plant, the number of seeds per pod, and seed weight was reported by Ajayi et al. [33], implying that selecting these attributes will result in enhanced yield. In a study by Meena et al. [34], 72 cowpea germplasms were evaluated, and the results indicated that seed yield per plant was significantly positively correlated with days to 50% flowering, plant height, number of pods per plant, pod length, number of seeds per pod and 100-seed weight at the phenotypic level. If the selection is made from any of the component traits, simultaneous gain will be achieved for all the traits.

For all the traits studied, each had two path actions, viz., a direct effect on pod yield per plant and an indirect effect through other characters. Crop improvement programs can be effectively developed by using path coefficient analysis, an efficient method of dividing genotypic correlation coefficients into the direct and indirect effects of individual traits on yield [3]–[20]. From the results obtained, the studied characters showed strong direct effect with pod yield per plant. This result indicates an actual correlation and the effectiveness of direct selection based on these traits.

When a strong positive indirect effect was observed over the days to 90% pod maturity, the positive correlation appeared to be caused by the indirect effects. The selection process simultaneously takes the indirect causal aspect into account. The strong direct effects of the traits in this study suggest that these characteristics are good yield-enhancing indicators. However, the higher indirect values may be neutralized in most instances by negative indirect effects via other characters, which can lead to low and non-significant genotypic correlations with yield. Hence, selection for such characters may not enhance yield improvement. For traits which the correlation coefficient is negative, but the effect is positive and high, a restricted simultaneous selection model is used. In order to utilize the direct effect, the restrictions must be put in place to alleviate the unwanted indirect effects. When both the direct effect and the correlation coefficient are negative, selections based on those characteristics are dropped. These results were in agreement with those of [2], [21], [22].

Conclusion

There is a significant degree of genetic variability and heritability in the characters under-studied in relation to the genotype materials used in this study, which supports selection in the genotypes studied for improvement. To breed cowpea varieties for high production, it would be helpful to know the level of genetic variability computed for various traits. In conclusion, high heritability alongside high genetic advance as a percent mean observed for days to flowering, pod length, 100-seed weight, and seed yield per plot suggested that these traits were primarily controlled by additive gene action and least affected by environmental factors. In order to further enhance the development of high-yielding cowpea varieties, direct selection based on these traits would be beneficial. Thus, the genotypes that show improved performance for these traits can be used in future breeding programs.

Characters that could be used in cowpea breeding programs include days to flowering, days to 50% flowering, days to first pod maturity, days to 90% pod maturity, days to maturity, and number of seeds per pod because they significantly increase yield. These traits were identified when high genetic variability components and heritability estimates were used in combination with a strong direct effect on seed yield and significant positive correlations.

F₁ genotype EB-3F₁, SB-2F₁ and BS-2F₁ could be selected for improvement of plant height, while EB-9F₁ could be best used for improvement of the number of branches in cowpea improvement programs. SB-2F₁ and BS-1F₁ may be selected and used when studying how leaf area correlates with yield. The SB-1F₁, BS-1F₁, BS-2F₁, BE-1F₁, BE-4F₁, and EB-3F₁ generations might be selected to further improve the earliness of cowpea plants, as they were able to reach maturity earlier during this study.