Genetic variability, heritability, genetic advance and divergence in Ethiopian cowpea [ Vigna unguiculata (L) Walp] landraces

Information on genetic variability in cowpea germplasm is important for crop improvement and for e ﬃ cient utilization of the existing genetic resources. Hence, the objectives of the present investigation were to estimate genetic variability, heritability, genetic advance and to identify divergent parents from distantly related clusters among Ethiopian cowpea accessions. The ﬁ eld experiment was carried out using 42 accessions at Abergelle Agricultural Research Center on station during the 2019 cropping season. Data were collected for 8 agronomic traits and analysis of variance revealed signi ﬁ cant differences (p<0.01) among the accessions for the traits studied. Seed yield had higher Genotypic Coe ﬃ cient of Variation (GCV) and Phenotypic Coe ﬃ cient of Variation (PCV) coupled with the highest genetic advance as percent of mean (100%). All the traits had moderate (68.01) to very high (99.98%) broad sense heritability. Further, high heritability coupled with high genetic advance as percent of mean was attained for days to ﬂ owering, grain ﬁ lling period, plant height, pod length, seed yield and thousand seed weight re ﬂ ecting the presence of additive gene action for the expression of these traits and improvement of these traits could be done through selection. The cluster analysis based on agronomic traits revealed four distinct groups at 90% similarity level. The highest inter cluster D 2 was recorded between cluster III and cluster IV (D 2 =133.69 units). The range of inter cluster distance was 15.25 to 133.69 units, respectively. In conclusion, the high genetic distance revealed among clusters has to be exploited via crossing and selection of the most divergent parents for future cowpea improvement program.


Introduction
Cowpea [Vigna unguiculata (L.) Walp., Fabaceae (2n = 2x = 22)] is an important dual purpose (food and forage) legume crop widely grown under low input production systems and in arid and semi-arid agro-ecologies of the world [1,2]. It is predominantly a self-pollinated crop, with natural crosspollination up to one percent. Cowpea could play signifi cant role in mitigating malnutrition such as micronutrient defi ciencies for poor farmers of Sub-Saharan countries [3]. It has distinct features such as its earliness in maturity, tolerance to drought, heat, acidity and low fertility, and seed types with high protein content and low cooking time [4]. It is well adapted to the drier regions of the tropics, where other food legumes do not perform well. Cowpea is eaten in the form of dry seeds, green pods, green seeds and tender green leaves. As a pulse crop, cowpea provides more than half the plant protein for human diets in some areas of semi-arid tropics is being referred to as "poor-man's meat" [5].
Cowpea grain typically contains 230-250 g/kg Crude Protein (CP) and 500-670 g/kg starch on a Dry Matter (DM) basis and cowpea forage, i.e. the crop residue after harvesting grain, 210g CP and 600g digestible dry matter per kg DM. These excellent nutritional qualities of cowpea make as one of the potential crop as a component of the cropping system and livelihood for the smallholder farmers living in drier regions of Ethiopia [6]. Thus, this crop can contribute greatly towards meeting the food requirement of people in areas where food security and malnutrition are major challenges. Regardless of the various merits of cowpea in Ethiopia, the national production and Citation: Belay  It has an approximate geographical coordinates of 13 0 14'06" N latitude and 38 0 58'50" E longitudes having an altitude 1560 meter above sea level ( Figure 1). The area is characterized by an erratic rainfall pattern with "kola" agro climatic zone. The rainy season is mono modal pattern concentrated in one season during the summer (July to August) and receives from 350-600 mm, annual precipitation. The mean minimum and maximum temperature of the area ranges from 18-42 0 C, respectively [10]. The soil texture of the specifi c site of the study area is sandy clay textural class with high available P (13.82 mg kg -1 ), very low in total N (0.08%) and low organic matter (0.72%) with a neutral pH of 7.18.

Description of the planting materials
The experimental plant materials comprised a total of 42 cowpea accessions/local landraces along with one released variety Bekur as a check was used in this study. The landraces were Ethiopian origin kindly provided by the Ethiopian Institute of Biodiversity (EIB) collected from different agroecological regions of the country, varying in altitude, rainfall, temperature, and soil type. The accession numbers and source of the genotypes are presented in Table 1.

Experimental design and crop management
The experiment was laid out using (6, 7) lattice design with three replications. The plot size was 7.2 m 2 (4m x 1.8m) with three rows of inter-row (60 cm) and intra-row (20 cm) productivity is far below the potential due to several abiotic and biotic constraints among which drought, insect pests, parasitic weeds and virus facing cowpea production to have resulted in a very low yield [7].
Though, Ethiopia is one of the centers of origin and/ or diversity of cowpea [8] and more than 66.5% of arable land is very suitable for cowpea production [9], the country has not been in a position to be benefi ted from international and continental cowpea improvement program or from the national pulse crops research. This is because low attention in research for cowpea is given as compared to other pulse crops. To harness the potential of cowpea landraces grown in Ethiopia baseline information regarding cowpea production status in the country /baseline information has been generated. Greater the variability in a population, there is the greater chance for effective selection for anticipated varieties. However, very few studies have been conducted on cowpea genetic variability using quantitative traits employed in the country. Hence, the objectives of the present study were to assess variability, heritability, genetic advance and to identify divergent parents from distantly related clusters for the future cowpea improvement program.   spacing's. The distance between plots, intra-blocks, and replications was 0.5m, 1m and 1.5m, respectively. Blended

Description of the experimental site
NPSZnB fertilizer was applied at the rate of 100 kg ha -1 during planting. Weeds were controlled periodically by hand weeding and other fi eld management and crop protection activities were done as required.

Data collection
Some phenological (days to fl owering, days to maturity, grain fi lling period), morphological (plant height, pod length) and yield and related traits (seed yield, thousand seed weight, number of seeds per pod) of each genotype was collected following the descriptor for cowpea developed by the International Board for Plant Genetic Resources [12]. The data collected on plot basis were days to fl owering, days to maturity, grain fi lling period (days), thousand seed weight (g) and seed yield (g). In addition, the data collected on individual plant

Data analysis
Data for agronomic traits were subjected to analysis of variances (ANOVA) for lattice design procedures of SAS Version 9.2 [13] to test the presence of signifi cant differences among genotypes. Variability among accessions was estimated using genotypic variances and coeffi cients of variations as suggested by Burton and De vane [14] and these components of variance ( 2 p,  2 e,  2 g) were used for the estimation of coeffi cients of variation (PCV, GCV) as described by Singh and Chaudhary [15]. 6. Genetic Advance (GA) for selection intensity (K) at 5%

Genotypic Variance GV=
was computed according to Allard [16] as given here: GA = K 2 pH 2 , where, GA = expected genetic advance, K = the standardized selection differential at 5% selection intensity (K=2.063),  2 p = is phenotypic standard deviation on mean basis and H 2 = heritability in broad sense.

Genetic advance as percentage of population means
(GAM) was also estimated with the methods described by Johnson, et al. [17] to compare the extent of the predicted advance of different traits under selection using the following formula: GAM=  [20]. The contributions of each of the traits to divergence were estimated as described as Sharma [21] with the formula [CTIC=
Furthermore, high heritability coupled with genetic advance were recorded for days to fl owering, grain fi lling period, plant height, pod length, seed yield and thousand seed weight, indicates additive gene action control the expression inheritance of these traits in cowpea [32]. A similar result was reported by Thorat and Gadewar [27], Sharma, et al. [33] and Das, et al. [34]for seed yield and Khan, et al. [28] for the number of pods per plant and Reshma, et al. [23] for plant height, seed yield per plant, number of pods per plant, pod length and hundred seed weight in cowpea. Overall, the estimates of heritability (H 2 ), genetic advance as percent of mean (GAM); genotypic coeffi cients of variation (GCV) and phenotypic coeffi cient of variation (PCV) were high for seed yield and thousand seed weight which are critical to identify potential for development of superior cowpea genotypes and/ or improvement of population through selection.

Clustering of accessions
The tested accessions were grouped in four different clusters with the number of accessions per cluster varying from 6 to 16 ( Table 3). The covariance matrix gave hierarchical clustering ( Figure 2) using average linkage method and the appropriate number of clusters was determined from the values of Pseudo F and Pseudo T 2 statistics among 42 cowpea accessions (Table 4). Cluster I was the largest cluster comprising 16 accessions, followed by clusters II and IV that contained 10 and 10 accessions, respectively, where cluster III contained the smallest accession (6) number. Of the 16 accessions grouped in cluster I, 43.75%, 25%, 12.5%, 12.5% and 6.25% of accessions originated from Oromia, Tigray, Amhara, SNNP and Gambella regions, respectively.

Expected genetic advance for selection
Genetic advance is a measure of predetermined progress under artifi cial selection program. According to Jonhson, et al. [17] the value of GAM is categorized as low (< 10%), moderate (10-20%) and high (> 20%). In this study, the highest GAM was recorded for seed yield (100%) followed by 100-seed weight (68.16%), grain fi lling period (39.65%), pod length (31.25%), days to fl owering (30.05%) and plant height (28.42%), indicating that these traits are governed by additive genes and selection will be rewarding for improvement of cowpea for these traits. In agreement with the current study, high GAM for   (Tables   3,4).
The main reasons for the grouping of accessions of the same origin into different clusters could be the exchanges of germplasm by farmers among neighboring regions, natural and artifi cial selection, genetic enrichment, genetic drift and environmental variation. Furthermore, Tesfaye, et al. [35] reported no relationship between genetic origin and /or diversity and geographic distribution.

Mean performance of clusters
The mean value of 8 agronomic traits per cluster is presented on On the basis of overall mean performance, cluster IV showed the best performance for most important traits including seed yield. Therefore, cluster IV would be preferable for selection

Genetic divergence analysis
The standardized Mahalanobis D 2 (square distances) statistics showed that there is high genetic distance and highly signifi cant variation at P≤0.01 and P≤0.05 among the four clusters (Table 6)

Conclusion
The result of the study revealed the existence of signifi cant (p<0.01) genetic variability among Ethiopian cowpea landraces.
Seed yield had higher Genotypic Coeffi cient of Variation (GCV) and phenotypic coeffi cient of variation (PCV) coupled with the highest genetic advance as percent of mean (100%). Further, high heritability coupled with high genetic advance as percent of mean was recorded for days to fl owering, grain fi lling period, plant height, pod length, seed yield and thousand seed weight refl ecting the presence of additive gene action for the expression of these traits and improvement of these traits could be done through selection. The cluster analysis based on agronomic traits revealed four distinct groups at 90% similarity level. The highest inter cluster D 2 was recorded between cluster III and cluster IV (D 2 =133.69 units). The range of inter cluster distance was 15.25 to 133.69 units, respectively. Therefore, based on the present fi ndings, it can be conclude that the high genetic distance revealed among clusters has to be exploited via crossing and selection of the most divergent parents for future cowpea improvement program.