Evaluating cowpea mutant genotypes for grain yield and nutritional value in South Africa By Moshieng Ntswane Dissertation submitted in fulfilment of the requirements in respect of the degree Master of Science in Agriculture in the Department of Plant Sciences (Plant Breeding) Faculty of Natural and Agricultural Sciences University of the Free State Bloemfontein November 2022 Supervisor: Dr. NW Mbuma Co-supervisors: Prof. M Labuschagne Dr. SF Shandu i DECLARATION “I, Moshieng Ntswane, declare that the Masters research dissertation that I herewith submit for the Master of Science in Agriculture degree qualification at the University of the Free State (UFS), is my independent work, and that I have not previously submitted it for a qualification at another institution of higher education.” 24 November 2022 Signature Date ii ACKNOWLEDGEMENTS I would like to thank Almighty God for strength and guidance throughout the journey of my life as a whole. I would also like to thank the following individuals and organisations:  My supervisor Dr. Ntombokulunga Wedy Mbuma and co-supervisors Prof. Maryke Labuschagne and Dr. Siphiwokuhle Funani Shandu for their mentorship and guidance throughout the research journey.  My late co-supervisor Dr. Maletsema Alina Mofokeng who laid out the foundation of the research; May her soul continue to rest in peace.  Mr. Paul Rantso, research technician from Agricultural Research Council Grain Crops (ARC-GC) for his persistent support and assistance with field layout, planting, trial management and data collection.  The University of the Free State (UFS) Plant Breeding students for assisting me with field planting and data collection.  The UFS for tuition fees bursary, approving the research project and the ethical clearance.  The ARC-GC for providing the plant materials and funding the research project.  North West Department of Agriculture and Rural Development (DARD) for providing personnel to assist with field work.  Agricultural Sector Education Training Authority (AgriSETA) for providing research funding.  National Research Foundation (NRF) for funding the research project.  Corteva Agriscience for providing funding.  My grandmother Makgalita Ntswane and my family as a whole for love and support throughout my MSc studies. iii TABLE OF CONTENTS DECLARATION....................................................................................................................... i ACKNOWLEDGEMENTS .................................................................................................... ii TABLE OF CONTENTS ...................................................................................................... iii LIST OF TABLES ............................................................................................................... viii LIST OF FIGURES ................................................................................................................. x LIST OF ABBREVIATIONS AND SI UNITS .................................................................... xii ABSTRACT .............................................................................................................................. 1 CHAPTER 1 ............................................................................................................................. 3 INTRODUCTION.................................................................................................................... 3 1.1 Aim of the study.............................................................................................................. 5 1.2 Objectives ........................................................................................................................ 6 1.3 References ....................................................................................................................... 6 CHAPTER 2 ........................................................................................................................... 11 LITERATURE REVIEW ..................................................................................................... 11 2.1 Origin, history and production of cowpea crop ........................................................ 11 2.2 Origin, history and production of cowpea in South Africa ...................................... 13 2.3 Factors affecting productivity of cowpea ................................................................... 15 2.4 Cowpea taxonomy, botany and genetics .................................................................... 15 2.5 Nutritional value/importance and functional compounds ........................................ 17 2.5.1 Proteins and amino acids ........................................................................................ 17 2.5.2 Mineral concentration ............................................................................................. 18 2.5.3 Starch, fibre and oligosaccharides .......................................................................... 19 2.5.4 Bioactive polyphenols and antioxidants .................................................................. 20 2.6 Anti-nutritional compounds ........................................................................................ 21 2.7 History of cowpea breeding in the world ................................................................... 22 2.8 History of cowpea breeding in South Africa .............................................................. 23 2.9 Selection methods in cowpea breeding ....................................................................... 23 2.9.1 Pure-line breeding ................................................................................................... 24 2.9.2 Pedigree selection .................................................................................................... 24 2.9.3 Backcross selection .................................................................................................. 25 2.9.4 Mutation breeding ................................................................................................... 25 2.10 Genetic variance, heritability and predicted selection gains .................................. 25 iv 2.11 Correlations and principal component analysis ...................................................... 27 2.12 Genotype by environment interactions and stability analysis ............................... 28 2.13 References ................................................................................................................... 29 CHAPTER 3 ........................................................................................................................... 49 PHENOTYPIC DIVERSITY AMONG COWPEA MUTANTS AND NORMAL GENOTYPES FOR GRAIN YIELD AND YIELD COMPONENTS .............................. 49 3.1 Abstract ......................................................................................................................... 49 3.2 Introduction .................................................................................................................. 50 3.3 Materials and methods ................................................................................................ 51 3.3.1 Plant material and experimental sites ..................................................................... 51 3.3.2 Trial design and management ................................................................................. 53 3.3.3 Data collection ........................................................................................................ 53 3.3.4 Data analysis ........................................................................................................... 53 3.4 Results ........................................................................................................................... 55 3.4.1 ANOVA for single trials of cowpea mutants and normal genotypes ....................... 55 3.4.2 Combined ANOVA for cowpea mutants and normal genotypes .............................. 55 3.4.3 Combined ANOVA for groupings according to origin ............................................ 56 3.4.4 Mean values of cowpea mutants and normal genotypes across environments ....... 61 3.4.5 Mean values of cowpea groupings across environments ........................................ 62 3.4.6 Phenotypic correlation ............................................................................................ 68 3.4.7 Principal component analysis (PCA) ...................................................................... 70 3.4.8 Clustered heat map of measured traits in cowpea mutant and normal genotypes .. 72 3.5 Discussion ...................................................................................................................... 74 3.5.1 Analysis of variance for cowpea mutants and normal genotypes ........................... 74 3.5.2 Combined ANOVA for groupings according to origin ............................................ 75 3.5.3 Mean values of cowpea mutants and normal genotypes across environments ....... 75 3.5.4 Mean values of cowpea groupings across environments ........................................ 76 3.5.5 Phenotypic correlation ............................................................................................ 77 3.5.6 Principal component analysis (PCA) ...................................................................... 77 3.5.7 Clustered heat map of measured traits in cowpea mutant and normal genotypes .. 78 3.6 Conclusions ................................................................................................................... 79 v 3.7 References ..................................................................................................................... 79 CHAPTER 4 ........................................................................................................................... 84 VARIATION IN SEED PROTEIN, SELECTED MINERALS, PHYTIC ACID AND POTENTIAL MINERAL BIOAVAILABILITY OF COWPEA MUTANTS AND NORMAL GENOTYPES ...................................................................................................... 84 4.1 Abstract ......................................................................................................................... 84 4.2 Introduction .................................................................................................................. 85 4.3 Materials and methods ................................................................................................ 86 4.3.1 Plant material and experimental sites ..................................................................... 86 4.3.2 Trial design and management ................................................................................. 86 4.3.3 Data collection and analysis ................................................................................... 87 4.3.4 Statistical analysis ................................................................................................... 88 4.4 Results ........................................................................................................................... 88 4.4.1 ANOVA for protein, selected minerals, phytic acid, potential mineral bioavailability, moisture and ash of cowpea mutants and normal genotypes ........................................... 88 4.4.2 Combined ANOVA for protein, selected minerals, phytic acid, potential mineral bioavailability, moisture and ash of cowpea mutants and normal genotypes .................. 89 4.4.3 Combined ANOVA for cowpea groupings of genotypes for all traits ..................... 89 4.4.4 Cowpea mutants and normal genotype performance for all characteristics across environments ..................................................................................................................... 93 4.4.5 Cowpea group performance for all characteristics across environments .............. 94 4.4.6 Phenotypic correlation ............................................................................................ 99 4.4.7 Principal component analysis (PCA) .................................................................... 100 4.4.8 Clustered heat map for protein, mineral elements, phytic acid, potential mineral bioavailability, moisture and ash in cowpea mutant and normal genotypes ................. 102 4.5 Discussion .................................................................................................................... 103 4.5.1 ANOVA for for protein, selected minerals, phytic acid, potential mineral bioavailability, moisture and ash of cowpea mutants and normal genotypes ................ 103 4.5.2 Combined ANOVA for cowpea groupings of genotypes for all characteristics .... 105 4.5.3 Cowpea mutants and normal genotype performance for all characteristics across environments ................................................................................................................... 105 4.5.4 Cowpea group performance for all characteristics across environments ............ 107 vi 4.5.5 Phenotypic correlation .......................................................................................... 107 4.5.6 Principal component analysis (PCA) .................................................................... 108 4.5.7 Clustered heat map for protein, mineral elements, phytic acid, potential mineral bioavailability, moisture and ash in cowpea mutant and normal genotypes ................. 108 4.6 Conclusions ................................................................................................................. 109 4.7 References ................................................................................................................... 110 CHAPTER 5 ......................................................................................................................... 115 EVALUATION OF GENOTYPE BY ENVIRONMENT INTERACTION AND STABILITY OF COWPEA MUTANTS AND NORMAL GENOTYPES FOR GRAIN YIELD ................................................................................................................................... 115 5.1 Abstract ....................................................................................................................... 115 5.2 Introduction ................................................................................................................ 116 5.3 Materials and methods .............................................................................................. 117 5.3.1 Plant material and experimental sites ................................................................... 117 5.3.2 Trial design and management ............................................................................... 117 5.3.3 Data collection ...................................................................................................... 118 5.3.4 Statistical analysis ................................................................................................. 118 5.4 Results ......................................................................................................................... 118 5.4.1 The ANOVA for AMMI analysis of cowpea grain yield ........................................ 118 5.4.2 Mean value performance and stability ranking of genotypes across environments ........................................................................................................................................ 119 5.4.3 Adaptability, additive main effects and multiplicative interaction (AMMI) biplot analysis ........................................................................................................................... 122 5.4.4 Genotype plus genotype by environment interactions (GGE) biplot analysis ....... 124 5.4.5 Ideal genotypes and environments for cowpea grain yield ................................... 126 5.5 Discussion .................................................................................................................... 128 5.5.1 The ANOVA for AMMI analysis of cowpea grain yield ........................................ 128 5.5.2 Mean value performance of genotypes across the environments .......................... 130 5.5.3 Adaptability, additive main effects and multiplicative interaction (AMMI) biplot analysis and stability ranking of genotypes .................................................................... 131 5.5.4 Genotype plus genotype by environment interactions (GGE) biplot analysis, ideal genotypes and environments for cowpea grain yield ..................................................... 131 vii 5.6 Conclusions ................................................................................................................. 133 5.7 References ................................................................................................................... 133 CHAPTER 6 ......................................................................................................................... 137 GENERAL DISCUSSION, CONCLUSION AND RECOMMENDATIONS FOR FUTURE RESEARCH ........................................................................................................ 137 6.1 General discussion ...................................................................................................... 137 6.2 Conclusions ................................................................................................................. 140 6.3 Recommendations for future research ..................................................................... 140 6.4 References ................................................................................................................... 141 viii LIST OF TABLES Table 2.1 African countries with the highest cowpea production (tons) in 2020…………………………………………………………………… 13 Table 2.2 Types of protein fractions found in cowpea, their solubility and profile bands…………………………………………………………………... 17 Table 2.3 Amino acids profile of cowpea seeds………………………………...... 18 Table 3.1 Source of collection and origin of cowpea genotypes used in the study... 52 Table 3.2 Characteristics of the five environments…….....……………………… 52 Table 3.3 Soil characteristics of the five environments during the 2021/2022 cropping season……………………………………………………….. 52 Table 3.4 Analysis of variance showing mean square values and broad-sense heritability (H2) for grain yield (GY) and yield components of cowpea at five environments................................................................................ 57 Table 3.5 Analysis of variance showing mean square values and broad-sense heritability (H2) for grain yield (GY) and yield components of cowpea across five environments......................................................................... 59 Table 3.6 Analysis of variance showing mean square values for grain yield (GY) and yield components of cowpea three groups based on their origins...... 60 Table 3.7 Mean values of grain yield (GY) and yield components of each of the 31 genotypes across five environments................................................... 63 Table 3.8 Mean values for grain yield (GY) and yield components of cowpea three groups based on their origin……………....................................... 67 Table 3.9 Pearson correlations for grain yield (GY) and yield components............ 68 Table 3.10 Principal component analysis (PCA) of grain yield and yield components in cowpea genotypes........................................................... 71 Table 4.1 Analysis of variance showing mean square values and broad-sense heritability (H2) for protein, selected mineral elements, phytic acid, potential mineral bioavailability, moisture and ash of cowpeas at five environments.......................................................................................... 91 Table 4.2 Analysis of variance showing mean square values and broad-sense heritability (H2) for protein, selected mineral elements, phytic acid, potential mineral bioavailability, moisture and ash of cowpeas across five environments................................................................................... 92 ix Table 4.3 Analysis of variance showing mean square values for protein, selected mineral elements, phytic acid, potential mineral bioavailability, moisture and ash of three cowpea groups based on their origins.............. 92 Table 4.4 Mean values of protein, selected mineral elements, phytic acid, potential mineral bioavailability, moisture and ash of each of 31 genotypes across five environments........................................................ 95 Table 4.5 Mean values of cowpea genotype groupings based on their origin…... 99 Table 4.6 Pearson correlations for protein, selected mineral elements, phytic acid, potential mineral bioavailability, moisture and ash.........................….... 99 Table 4.7 Principal component analysis (PCA) of 11 traits in cowpea genotypes... 101 Table 5.1 Additive main effects and multiplicative interaction (AMMI) analysis of variance (ANOVA) for grain yield of 31 genotypes..........………….. 119 Table 5.2 Mean values of all genotypes’ performance across environments for cowpea grain yield (kg ha-1)……...................………….....…………… 121 x LIST OF FIGURES Figure 2.1 World cowpea production.......................................................................... 12 Figure 2.2 Cowpea area harvested (ha), production (ton) and grain yield (kg ha-1) in South Africa from 2010 to 2020................................................................. 14 Figure 3.1 Mean value performance of Namibian mutants, IITA (International Institute of Tropical Agriculture) and SA (South Africa) genotypes for grain yield (A), number of pods per plant (B), number of seeds per plant (C) and number of seeds per pod (D)............................................................... 64 Figure 3.2 Mean value performance of Namibian mutants, IITA (International Institute of Tropical Agriculture) and SA (South Africa) genotypes for pod weight per plant (A), seed weight per plant (B), hundred seed weight (C) and harvest index (D)........................................................................... 65 Figure 3.3 Mean value performance of Namibian mutants, IITA (International Institute of Tropical Agriculture) and SA (South Africa) genotypes for pod width (A), pod length (B), number of branches per plant (C) and plant height (D)................................................................................................... 66 Figure 3.4 Heat map showing the phenotypic correlation between cowpea grain yield (GY) and yield components............................................................... 69 Figure 3.5 Principal component analysis (PCA) showing the associations among the genotypes with their traits.......................................................................... 72 Figure 3.6 Clustered heat map showing the groupings of genotypes based on the measured traits........................................................................................... 73 Figure 4.1 Mean value performance of Namibian mutants, IITA (International Institute of Tropical Agriculture) and SA (South Africa) genotypes for protein (A), iron (B), zinc (C) and copper (D)............................................. 96 Figure 4.2 Mean value performance of Namibian mutants, IITA (International Institute of Tropical Agriculture) and SA (South Africa) genotypes for manganese (A), boron (B), phytic acid (C) and phytic acid:iron (D).......... 97 Figure 4.3 Mean value performance of Namibian mutants, IITA (International Institute of Tropical Agriculture) and SA (South Africa) genotypes for phytic acid:zinc (A), moisture (B) and ash (C)........................................... 98 Figure 4.4 Heat map representing phenotypic correlation between measured traits for cowpea genotypes................................................................................. 100 xi Figure 4.5 Principal component analysis (PCA) displaying genotypes with their associated traits.......................................................................................... 102 Figure 4.6 Clustered heat map showing the associations between the genotypes with the measured traits...................................................................................... 103 Figure 5.1 Adaptability, additive main effects and multiplicative interaction (AMMI) biplot for grain yield of 31 genotypes across five environments.. 123 Figure 5.2 Adaptability, additive main effects and multiplicative interaction (AMMI) biplot for grain yield stability of 31 genotypes across five environments.............................................................................................. 124 Figure 5.3 Genotype plus genotype by environment interactions biplot highlighting the mega-environments for cowpea grain yield......................................... 125 Figure 5.4 Genotype plus genotype by environment interactions biplot ranking indicating the mean grain yield performance of 31 cowpea genotypes................................................................................................... 126 Figure 5.5 Genotype plus genotype by environment interactions biplot showing the ideal genotypes according to the average grain yield of cowpea across the five environments....................................................................................... 127 Figure 5.6 Genotype plus genotype by environment interactions biplot showing ideal environments for cowpea grain yield................................................. 128 xii LIST OF ABBREVIATIONS AND SI UNITS a Constant AEC Average environment coordination AFLP Amplified fragment length polymorphism AMMI Additive main effects and multiplicative interaction ANOVA Analysis of variance ARC Agricultural Research Council ASV Additive main effects and multiplicative interaction stability value AW Atomic weight B Boron b Slope Ca Calcium CEC Cation exchange capacity CRSP Collaborative research support programme Cu Copper CV Coefficient of variation DNA Deoxyribonucleic acid E Environment Fe Iron G Genotype GE Genotype by environment interaction GGE Genotype and genotype by environment interaction Gr Grouping Gr.E Grouping by environment interaction GY Grain yield H2 Broad-sense heritability h2 Narrow-sense heritability HCl Hydrochloric acid HI Harvest index IITA International Institute of Tropical Agriculture IPCA Interaction principal component axis ISRA Institute Senegalais de Recherches Agricoles K Potassium xiii LSD Least significant difference M2 Second mutation generation M4 Fourth mutation generation M9 Nineth mutation generation Masl Meters above sea level Max Maximum mc Moisture content Mg Magnesium Min Minimum Mn Manganese Mr Molar ratio 𝑀𝑆𝑒 Mean square error 𝑀𝑆𝑔 Mean square of genotype 𝑀𝑆𝑔𝑒 Mean square of genotype by environment interaction MW Molecular weight Na Sodium NBPP Number of branches per plant NIR Near-infrared spectroscopy NPP Number of pods per plant NSPPx Number of seeds per plant NSPPy Number of seeds per pod OECD Organisation for economic co-operation and development P Phosphorus PA Phytic acid PA:Fe Molar ratio of phytic acid with iron PA:Zn Molar ratio of phytic acid with zinc PC Principal component PCA Principal component analysis PH Plant height PL Pod length PW Pod width PWP Pods weight per plant QTL Quantitative trait locus R Replication xiv rASV Ranking of additive main effects and multiplicative interaction stability value Rep. (E) Replications within environments S Sulphur SA South Africa SSR Simple sequence repeat SWP Seeds weight per plant UCR University of California Riverside USA United State of America Zn Zinc 100 SW Hundred seeds weight % Percentage ∆A Change in absorbance 𝜎2𝑎 Additive varince 𝜎2𝑔 Genotypic variance 𝜎2𝑝 Phenotypic variance 𝜎2𝑔𝑒 Genotype by environment variance interaction 𝜎2𝑒 Error/residual variance cm Centimeter Da Dalton g Gram g 100 g-1 Gram per 100 gram ha Hectare kg ha-1 Kilogram per hectare mg 100 g-1 Milligram per 100 gram mg kg-1 Milligram per kilogram mm Millimeter ml Milliliter na Nanometer ppm Parts per million 1 ABSTRACT Cowpeas produce a substantial amount of grain, which is a significant source of vitamins, minerals and protein to disadvantaged people with limited access to adequate nutrients. Improved cowpea genotypes have been introduced through new breeding techniques such as mutagenesis to increase the phenotypic, genetic and nutritional diversity of the crop. The objectives of this study were: 1) to determine the phenotypic diversity and characterise cowpea mutants and normal genotypes for grain yield and yield components, to identify superior cowpea mutants and normal genotypes and to determine the correlation between all measured characteristics, 2) to evaluate the variability of cowpea mutants and normal genotypes for protein content, selected mineral elements, phytic acid and the potential bioavailability of iron (Fe) and zinc (Zn), to identify superior cowpea mutants and normal genotypes, and to determine the interrelationship between all measured characteristics, and 3) to determine genotype by environment (GE) interaction, to identify superior genotypes for grain yield and to determine the adaptability and stability of cowpea mutants and normal genotypes in South Africa. Thirty-one cowpea genotypes (16 Namibian mutants, seven International Institute of Tropical Agriculture (IITA) genotypes and eight South Africa genotypes) were planted in five different environments in South Africa during the 2021/2022 cropping season. Significant (P ≤ 0.05) genotype and GE interaction effects were observed for grain yield, yield components, protein content, minerals, phytic acid and potential mineral bioavailability. Broad-sense heritability (H2) values above 50% were observed for yield components, protein, mineral elements, phytic acid and potential mineral bioavailability, while low H2 values below 50% were observed for grain yield and Boron (B), indicating the complexity in selection and genetic improvement of these traits. Superior Namibian mutants (ShR10P12, ShR3P4, ShR4P1 and BrR11P2), IITA genotype (98K-476-8) and South African genotype (Enchore) for grain yield were identified. Superior Namibian mutants (ShL3P7-2, ShR3P4, ShR4P1, BrR11P11, BrR4P11, NKL9P7, NKR8P9, NKR9P9 ShR2P11, BrR11P2, ShL2P7, ShR3P4 and NKRuP5), IITA genotypes (98K-476-8, IT82E-18, IT93K-452-1, IT99K-573-2-1 and ITOOK 1263), and South African genotypes (Oloyin, Orelu, Pan 311, Bechuana White, Enchore and Glenda) for protein content, Fe and Zn concentration were also identified. The IITA genotype (IT93K-452- 1) and South African genotypes (Oloyin and Orelu) had a potential of good Fe bioavailability. All cowpea mutants and normal genotypes had a potential of poor Zn bioavailability. High yielding and stable Namibian mutants (NKL9P7, ShR10P12 and ShR2P11), IITA genotype (ITOOK 1263) and South African genotype (Agrinawa) were identified. Two mega- 2 environments, namely, 1) Taung and Mafikeng, and 2) Mafikeng, Bloemfontein, Polokwane and Potchefstroom were identified, indicating broad adaption of the genotypes. Potchefstroom and Taung were identified as ideal environments for evaluation of cowpea genotypes. Significant positive correlations between grain yield with almost all yield components were observed. Significant positive correlations of protein content with mineral elements and phytic acid were also observed, indicating the potential to simultaneously select these traits. Namibian mutants (NKR1P3, BrR11P2, ShL2P7, ShR2P11 and ShR10P12), IITA genotype (98K-476- 8) and South African genotypes (Glenda, Dr Saunders, Enchore and Oloyin) were associated with high grain yield. Namibian mutants (ShR10P12, ShR3P4, ShR9P5, BrR11P11, BrR11P2, BrR4P11, NKR1P3, NKR9P9 and NKRuP5), IITA genotypes (IT07K-292-10, IT07K-318-33, IT82E-18 and IT99K-573-2-1), and South African genotypes (Agrinawa, Bechuana White and Dr Saunders) were associated with high protein content, manganese (Mn), phytic acid, Molar ratio of phytic acid with iron (PA:Fe), Molar ratio of phytic acid with zinc (PA:Zn), and ash content. These genotypes have a potential of long-term profitability to the agricultural production industry. Keywords: cowpeas, grain yield, protein, mineral elements, mutation, phenotypic diversity, bioavailability, broad-sense heritability, stability, genotype by environment interaction 3 CHAPTER 1 INTRODUCTION The Vigna unguiculata (L.) Walp also known as cowpea, is a self-pollinated perennial diploid of the Fabaceae family and subfamily Faboideae (Agbogidi 2010; Moussa et al. 2011). Given the world population's rapid increase, legumes (Fabaceae) such as cowpea (Vigna unguiculata), soybean (Glycine max) and chickpeas (Cicer arietinum) are regarded as one of the most crucial plant families for human nutrition (Gepts et al. 2005; Smykal et al. 2015). Africa is predominating with the largest area of cowpea production and consumption (Kareem and Taiwo 2007; Timko and Singh 2008). In 2018/2019, 15.05 million hectares of cowpea dried grain were produced worldwide, of which 8.90 million tons (96.79%) of the dried grains were produced in Africa (FAOSTATS 2020). In the past, cowpea was one of the undervalued and underutilised crops in South Africa, and as a result, agricultural improvement projects have paid little attention to it (Gerrano et al. 2020). Recently, the Agricultural Research Council (ARC) in South Africa has made significant progress in obtaining and selecting nutritious cowpea genotypes with high grain yield in an attempt to increase food security and reduce malnutrition. Globally, it was estimated that 38.91 million people were obese or overweight due to malnutrition, with 45 million people being wasted (low height for their age), 149 million children below the age of five being stunted (looking too young for their age) and 38.91 million being overweight (WHO 2021). Due to poverty and malnutrition, almost one-third of African children experience numerous physical and mental (stunted growth, impaired immune system and pneumonia) complications (White and Broadley 2011; Mohajan 2022). Therefore, legume crops such as cowpeas can potentially contribute to the eradication of malnutrition and hunger. The whole cowpea plant is edible for both humans and animals. The crop can be consumed either as fresh leaves, immature pods or dry grains and its fodder has good value for animal feeding. Cowpeas are relatively a good source of protein with values ranging from 25.00 to 35.00% for fresh leaves, 21.00 to 29.00% for immature pods and 19.00 to 25.00% for dry grains (Okonya and Maass 2014; Gerrano et al. 2017). The crop is also rich in essential minerals [Fe, Zn, Mn and calcium (Ca)], carbohydrates (> 60%), amino acids (lysine and tryptophan), vitamins (B and C), fat (1.50%) and fibre (Hall et al. 2003; Elhardallou et al. 2015; Xiong et al. 2016). Due to its high nutritional quality, the crop has the ability to improve dietary quality by lowering cholesterol (Frosta et al. 2014). 4 Cowpeas can withstand the ever-changing climate better than other grain legumes such as dry beans and soybeans (Omomowo and Babalola 2021). Cowpeas are bred to withstand heat stress, drought tolerant stress and have better nitrogen fixation than drybeans and soybeans (Muhammad et al. 2010; Hall 2012). The crop is well suited to dry regions of the savannah belt where various crops may perform poorly or fail to germinate because of infertile soils or water stress caused by shortage of rainfall (Boukar et al. 2018). The crop thrives in an ideal environment with annual precipitation between 500 and 1200 mm, but recent study revealed that cowpeas can withstand annual rainfall as less as 400 mm (Ukpene and Imade 2015; DAFF 2011). In addition, most wild relatives of cowpea have been selected and used in cowpea breeding programmes due to their ability to tolerate drought stress, diseases and pest infestations (Agbicodo et al. 2009; Boukar et al. 2020; Gerrano et al. 2020). The crop also has the capacity to fix nitrogen in the soil which promotes soil health and decreases the need for inorganic nitrogenous fertilisers (Horn and Shimelis 2020). Numerous research studies evaluated cowpea diversity using morphological, agronomic, molecular and nutritional characteristics (Ba et al. 2004; Lazaridi et al. 2017; Menssen et al. 2017). According to these studies, cowpea populations exhibit limited genetic variability, which may be related to the plant's self-pollinating reproductive mechanism, which prevents the introduction of new genetic variation. Generally, genetic variability in plant breeding can be created through cross-breeding and has proved to be a very important tool. However, cross breeding cowpea genotypes is challenging, costly and time consuming to achieve due to the self-incompatibility of the crop. Therefore, mutation breeding has been employed in cowpea breeding programmes (Horn et al. 2016). Mutation is a change that occurs in the deoxyribonucleic acid (DNA) sequence that induces genetic variation and has a chance of creating broader genetic diversity. Radiation, chemical mutagens and other appropriate biotechnology approaches have been used to induce variability in crop plants (Sikora et al. 2011; Kozgar et al. 2012; Holme et al. 2019). Cowpea mutant genotypes obtained from Namibia were gamma irradiated (Horn et al. 2016). The ARC in South Africa obtained the cowpea mutants from Namibia and normal genotypes from the IITA to compare their diversity with South African genotypes and determine their adaptability and stability with the aim of introducing new variation to the South African cowpea germplasm collection. Cowpea mutant genotypes requires characterisation and assessment for nutritional quality and grain yield to assist with identification of potential parental genotypes which will be used to develop and improve cowpea cultivars (Gerrano et al. 2019; Mbuma et al. 2022). Therefore, breeders will be able to identify and choose the best genotypes that can be used as parents for 5 hybridisation with the help of the knowledge on the present genetic variability among cowpea mutants and normal genotypes (Gerrano et al. 2015). Breeding populations with wide diversity should be tested in multiple environments for their adaptability and stability (Tresina and Mohan 2012; Takinami et al. 2016). Multi-environmental trials are employed to assess the performance of genotypes and to determine their adaptability and stability in various environments (Negash et al. 2013). The change in the genotypes’ performance across two or more environments is known as the GE interaction. Several statistical models have been used to quantify GE interaction, and genotype adaptability and stability before a cultivar is released and recommended for commercial production (Yan 2002; Asrat et al. 2009). These statistical methods include analysis of variance (ANOVA), genotype plus genotype by environment interaction (GGE) biplot analysis, principal component analysis (PCA) and the additive main effects and multiplicative interaction (AMMI). The ANOVA is an important technique for determining the variability among genotypes, environments and other sources of variation in experimental trials. The PCA is used to visualise the similarity and differences between genotypes and their trait association while phenotypic correlations are used to determine the association among traits. The AMMI analysis incorporates a significant amount of GE interaction sum of squares and distinguishes between main (genotypes and environments) and interaction effects, which is essential in agricultural research (Gauch 2006). Additionally, the adaptability and stability of genotypes in multiple environments is assessed using the AMMI analysis. However, the AMMI biplot lacks the classifying capacity required to evaluate mega-environments. Mega-environments refer to the grouping of environments that have a similar influence on the performance of genotypes. The GGE biplots have been proposed and used to effectively classify superior genotypes and mega- environments (Peprah et al. 2020). 1.1 Aim of the study The aim of this research was to quantify the variation and to characterise cowpea mutants and South African and IITA genotypes for grain yield and yield components as well as for nutritional quality traits. 6 1.2 Objectives 1. To determine the phenotypic diversity and characterise cowpea mutants and normal genotypes for grain yield and yield components, to identify superior cowpea mutants and normal genotypes and to determine the correlation between all measured characteristics. 2. To evaluate the variability of cowpea mutants and normal genotypes for protein content, selected mineral elements, phytic acid and the potential bioavailability of Fe and Zn, to identify superior cowpea mutants and normal genotypes and to determine the interrelationship between all measured characteristics. 3. To determine GE interaction, to identify superior genotypes for grain yield and to determine the adaptability and stability of cowpea mutants and normal genotypes in South Africa. 1.3 References Agbicodo E, Fatokun C, Muranaka S, Visser R, Van der Linden CG (2009) Breeding drought tolerant cowpea: constraints, accomplishments, and future prospects. Euphytica 167:353-370. https://doi:10.1007/s10681-009-9893-8 Agbogidi O (2010) Screening six cultivars of cowpea (Vignia unguiculata L. Walp) for adaptation to soil contaminated with spent engine oil. Journal of Environmental and Chemical Ecotoxicology 2:103-109 Asrat A, Fistum A, Fekadu G, Mulugeta A (2009) AMMI and SREG GGE biplot analysis for matching varieties onto soybean production environments in Ethiopia. Academica Journal 4:1322-1330 Ba FS, Pasquet RS, Gepts P (2004) Genetic diversity in cowpea [Vigna unguiculata (L.) Walp.] as revealed by RAPD markers. Genetic Resources and Crop Evolution 51:539-550. https://doi.org/10.1023/B:GRES.0000024158.83190.4e Boukar O, Belko N, Chamarthi S, Togola A, Batieno J, Owusu E, Haruna M, Diallo M, Lawan M, Olusoji M, Fatokun C (2018) Cowpea (Vigna unguiculata): genetics, genomics and breeding. Plant Breeding 138:415-424. https://doi.org/10.1111/pbr.12589 Boukar O, Abberton M, Oyatomi O, Togola A, Tripathi L, Fatokun C (2020) Introgression breeding in cowpea [Vigna unguiculata (L.) Walp.]. Frontiers in Plant Science 11:1439. https://doi.10.3389/fpls.2020.567425 https://doi:10.1007/s10681-009-9893-8 https://doi.org/10.1023/B:GRES.0000024158.83190.4e https://doi.org/10.1111/pbr.12589 https://doi.10.3389/fpls.2020.567425 7 Department of Agriculture, Forestry and Fisheries (DAFF) (2021) Directorate of plant production. Production guideline for Cowpea www.daff.gov.za Elhardallou SB, Khalid II, Gobouri AA, Abdel-Hafez SH (2015) Amino acid composition of cowpea (Vigna ungiculata L. Walp) flour and its protein isolates. Food and Nutrition Sciences 6:790-797. https://doi:10.4236/fns.2015.69082 FAOSTATS (Food and Agriculture Organisation of the United Nations) (2020) Statistical Database. Rome, Italy. https://www.fao.org/faostat/en/#data/QCL. Accessed on 22 August 2022 Frosta KMG, Filho RDS, Ribeiro VQ, Arêas JAG (2014) Cowpea protein reduces LDL- cholesterol and apolipoprotein B concentrations, but does not improve biomarkers of inflammation or endothelial dysfunction in adults with moderate hypercholesterolemia. Nutricion Hospitalaria 31:1611-1619. https://doi:10.3305/nh.2015.31.4.8457 Gauch HG (2006) Statistical analysis of yield trials by AMMI and GGE. Crop Science 46:1488-1500. https://doi.org/10.2135/cropsci2005.07-0193 Gepts P, Beavis WD, Brummer EC, Shoemaker RC, Stalker HT, Weeden NF, Young ND (2005) Legumes as a model plant family: genomics for food and feed report of the cross-legume advances through genomics conference. Plant Physiology 137:1228-1235. https://doi:10.1104/pp.105.060871 Gerrano AS, van Rensburg WSJ, Adebola PO (2017) Nutritional composition of immature pods in selected cowpea [Vigna unguiculata (L.) Walp.] genotypes in South Africa. Australian Journal of Crop Science 11:134-141. https://doi:10.21475/ajcs.17.11.02.p72 Gerrano AS, van Rensburg WSJ, Koto FR (2019) Agronomic evaluation and identification of potential cowpea (Vigna unguiculata L. Walp) genotypes in South Africa. Acta Agriculturae, Scandinavica, Section B - Soil and Plant Science 69:295-303. https://doi.org/10.1080/09064710.2018.1562564 Gerrano AS, Adebola PO, van Rensburg WSJ, Laurie SM (2015) Genetic variability in cowpea (Vigna unguiculata (L.) Walp.) genotypes. South African Journal of Plant and Soil 32:165- 174. https://doi.org/10.1080/02571862.2015.1014435 Gerrano AS, van Rensburg WSJ, Mathew I, Shayanowako AI, Bairu MW, Venter SL, Swart W, Mofokeng A, Mellem J, Labuschagne MT (2020) Genotype and genotype x environment http://www.daff.gov.za/ https://doi:10.4236/fns.2015.69082 https://www.fao.org/faostat/en/#data/QCL https://doi:10.3305/nh.2015.31.4.8457 https://doi.org/10.2135/cropsci2005.07-0193 https://doi:10.1104/pp.105.060871 https://doi:10.21475/ajcs.17.11.02.p72 https://doi.org/10.1080/09064710.2018.1562564 https://doi.org/10.1080/02571862.2015.1014435 8 interaction effects on the grain yield performance of cowpea genotypes in dryland farming system in South Africa. Euphytica 216:80. https://doi.10.1007/s10681-020-02611-z Hall AE (2012) Phenotyping cowpeas for adaptation to drought. Front Physiology 3:155. https://doi.org/10.3389/fphys.2012.00155 Hall EA, Cissé N, Thiaw S, Elawad HOA, Ehlers JD, Ismail AM, Fery RL, Roberts PA, Kitch LW, Murdock LL, Boukar O, Phillips RD, McWatters KH (2003) Development of cowpea cultivars and germplasm by the bean/cowpea CRSP. Field Crops Research 82:103-134. https://doi.org/10.1016/S0378-4290(03)00033-9 Holme IB, Gregersen PL, Brinch-Pedersen H (2019) Induced genetic variation in crop plants by random or targeted mutagenesis: convergence and differences. Frontiers in Plant Science 10:1468. https://doi.org/10.3389/fpls.2019.01468 Horn LN, Ghebrehiwot HM, Shimelis HA (2016) Selection of novel cowpea genotypes derived through gamma irradiation. Frontiers in Plant Science 7:262. https://doi.org/10.3389/fpls.2016.00262 Horn LN, Shimelis H (2020) Production constraints and breeding approaches for cowpea improvement for drought prone agro-ecologies in sub-Saharan Africa. Annals of Agricultural Sciences 65:83-91. https://doi.org/10.1016/j.aoas.2020.03.002 Kareem KT, Taiwo MA (2007) Interactions of viruses in cowpea: effects on growth and yield parameters. Virology Journal 4:15. https://doi.org/10.1186/1743-422X-4-15 Kozgar MI, Khan S, Wani MR (2012) Variability and correlations studies for total iron and manganese contents of chickpea (Cicer arietinum L.) high yielding mutants. American Journal of Food Technology 7:437-444. https://doi:10.3923/ajft.2012.437.444 Lazaridi E, Ntatsi G, Savvas D, Bebeli PJ (2017) Diversity in cowpea [Vigna unguiculata (L.) Walp.] local populations from Greece. Genetic Resources and Crop Evolution 64:1529-1551. https://doi.org/10.1007/s10722-016-0452-6 Mbuma NW, Gerrano AS, Lebaka N, Labuschagne M (2022) Interrelationship between grain yield components and nutritional quality traits in cowpea genotypes. South African Journal of Botany 150:1-10. https://doi:10.1016/j.sajb.2022.07.006 Menssen M, Marcus L, Omondi EO, Abukutsa-Onyango M, Dinssa FF, Winkelmann T (2017) Genetic and morphological diversity of cowpea [Vigna unguiculata (L.) Walp.] entries from East Africa. Scientia Horticulturae 226:268-276. https://doi.org/10.1016/j.scienta.2017.08.003 https://doi.10.1007/s10681-020-02611-z https://doi.org/10.3389/fphys.2012.00155 https://doi.org/10.1016/S0378-4290(03)00033-9 https://doi.org/10.3389/fpls.2019.01468 https://doi.org/10.3389/fpls.2016.00262 https://doi.org/10.1016/j.aoas.2020.03.002 https://doi.org/10.1186/1743-422X-4-15 https://doi:10.3923/ajft.2012.437.444 https://doi.org/10.1007/s10722-016-0452-6 https://doi:10.1016/j.sajb.2022.07.006 https://doi.org/10.1016/j.scienta.2017.08.003 9 Mohajan HK (2022) Food insecurity and malnutrition of Africa: a combined attempt can reduce them. Journal of Economic Development, Environment and People 11:2285-3642. http://dx.doi.org/10.26458/jedep.v1i1.716 Moussa B, Lowenberg-DeBoer J, Fulton J, Boys K (2011) The economic impact of cowpea research in west and central Africa: a regional impact assessment of improved cowpea storage technologies. Journal of Stored Products Research 47:147-156. https://doi.org/10.1016/j.jspr.2011.02.001 Muhammad A, Dikko AU, Audu M, Singh A (2010) Comparative effects of cowpea and soybean genotypes on N2 - fixation and N-balance in Sokoto dry sub-humid agro-ecological zone of Nigeria. Nigerian Journal of Basic and Applied Science 18:297-303 Negash K, Tumsa K, Gebeyehu S, Amsalu B (2013) Multi-environment evaluation of early maturing cowpea (Vigna unguiculata L.) varieties in the drought prone areas of Ethiopia. Ethiopian Journal of Crop Science 3:1-7 Okonya JS, Maass BL (2014) Protein and iron composition of cowpea leaves: an evaluation of six cowpea varieties grown in eastern Africa. African Journal of Food, Agriculture, Nutrition and Development 14:2129-2140. https://doi:10.18697/ajfand.65.13645 Omomowo OI, Babalola OO (2021) Constraints and prospects of improving cowpea productivity to ensure food, nutritional security and environmental sustainability. Frontiers in Plant Science 12:751731. https://doi.org/10.3389/fpls.2021.751731 Peprah BB, Parkes E, Manu-Aduening J, Kulakow P, van Biljon A, Labuschagne M (2020) Genetic variability, stability and heritability for quality and yield characteristics in provitamin A cassava varieties. Euphytica 216:31. https://doi.org/10.1007/s10681-020-2562-7 Sikora P, Chawade A, Larsson M, Olsson J, Olsson O (2011) Mutagenesis as a tool in plant genetics, functional genomics, and breeding. International Journal of Plant Genomics 2011:13. https://doi.org/10.1155/2011/314829 Smykal P, Coyne CJ, Ambrose JM, Maxted N, Schaefer H, Blair MW, Berger J, Greene SL, Nelson MN, Besharat N, Vymyslicky T, Tokerk C, Saxena RK, Roorkiwal M, Pandey MK, Hub J, Lim YH, Wang LX, Guom Y, Qium LJ, Reddenn RJ, Varshneyi RK (2015) Legume crops phylogeny and genetic diversity for science and breeding. Critical Reviews in Plant Sciences 34:43-104. https://doi:10.1080/07352689.2014.897904 http://dx.doi.org/10.26458/jedep.v1i1.716 https://doi.org/10.1016/j.jspr.2011.02.001 https://doi:10.18697/ajfand.65.13645 https://doi.org/10.3389/fpls.2021.751731 https://doi.org/10.1007/s10681-020-2562-7 https://doi.org/10.1155/2011/314829 https://doi:10.1080/07352689.2014.897904 10 Takinami PY, Uehara VB, Teixeira BS, del Mastro NL (2016) Radiation, plant proteins and sustainability. American Journal of Biological and Environmental Statistics 2:28-33. https://doi:10.11648/j.ajbes.20160204.11 Timko MP, Singh BB (2008) Cowpea, a multifunctional legume. In: Moore PH, Ming R (eds) Genomics of tropical crop plants. Plant Genetics and Genomics. Springer, New York, pp. 227- 258. https://doi.org/10.1007/978-0-387-71219-2_10 Tresina PS, Mohan VR (2012) Physico-chemical and antinutritional attributes of gamma irradiated Vigna unguiculata (L.) Walp. subsp. unguiculata seeds. International Food Research Journal 19:639-646 Ukpene AO, Imade FN (2015) Amino acid profiles of seven cowpea varieties grown in Agbor. Nigerian Annals of Natural Sciences 15:072-078 White PJ, Broadley MR (2011) Physiological limits to zinc biofortfication of edible crops. Frontiers in Plant Science 2:8. https://doi.org/10.3389/fpls.2011.00080 WHO (World Health Organisation) (2021) Malnutrion. https://www.who.int/news-room/fact- sheets/detail/malnutrition#:~:text=Globally%20in%202020%2C%20149%20million,age%20 are%20linked%20to%20undernutrition. Accessed on 25 August 2022 Xiong H, Shi A, Mou B, Qin J, Motes D, Lu W, Ma J, Weng Y, Yang W, Wu D (2016) Genetic diversity and population structure of cowpea (Vigna unguiculata L. Walp). PLoS ONE 11:8. https://doi.10.1371/journal.pone.0160941 Yan W (2002) Singular-value partitioning in biplot analysis of multi-environment trial data. Agronomy Journal 94:990-996. https://doi.org/10.2134/agronj2002.9900 https://doi:10.11648/j.ajbes.20160204.11 https://doi.org/10.1007/978-0-387-71219-2_10 https://doi.org/10.3389/fpls.2011.00080 https://www.who.int/news-room/fact-sheets/detail/malnutrition#:~:text=Globally%20in%202020%2C%20149%20million,age%20are%20linked%20to%20undernutrition https://www.who.int/news-room/fact-sheets/detail/malnutrition#:~:text=Globally%20in%202020%2C%20149%20million,age%20are%20linked%20to%20undernutrition https://www.who.int/news-room/fact-sheets/detail/malnutrition#:~:text=Globally%20in%202020%2C%20149%20million,age%20are%20linked%20to%20undernutrition https://doi.10.1371/journal.pone.0160941 https://doi.org/10.2134/agronj2002.9900 11 CHAPTER 2 LITERATURE REVIEW 2.1 Origin, history and production of cowpea crop Cowpea domestication was first proposed by Vavilov in 1926 with suggestions that Ethiopia was a major centre, with China and India as minor centres of domestication (Herniter et al. 2020). Other research studies revealed that wild cowpea relatives were found in the African continent, dominating other continents as the primary centre of origin (Coulibaly et al. 2002; Agbogidi 2010). The Vigna unguiculata ssp. dekindtiana var. spontanea, which can be found in Africa, is the wild origin of the commercial cowpea (Pasquet and Padulosi 2013). However, there are still contradicting opinions and reviews regarding the origin of wild cowpea relatives, which have high level of diversity (Ba et al. 2004; Xiong et al. 2017). The introduction of cowpea from Africa to the Indian subcontinent is speculated to be roughly 2000 to 3500 years ago (OECD 2016). During the 17th century AD, the crop was distributed to India by Spanish people via slave trade and distributed across Asia (Xiong et al. 2016). Furthermore, West African slave trade led to its cultivation reaching the southern United States of America (USA) during the early 18th century (Pakela 2006). Currently, cowpea is produced worldwide throughout the tropical and subtropical areas (Kebede and Bekeko 2020). Approximately 8.90 million tons of dried cowpea seeds on average were produced between 2019 and 2020 worldwide with Africa accounting for about 96.79% of the world cowpea production (FAOSTATS 2020) (Figure 2.1). Europe only produces an estimated 25 732 tons of dry cowpea seeds annually. Nigeria is the largest cowpea producer globally, producing an estimated 3 million tons of annual yield, followed by Niger (Table 2.1). The top 10 cowpea producing countries (from Nigeria to Sudan) are also leading in cowpea research, hence high crop production, harvest area and grain yield. In most African countries, abiotic and biotic factors contribute to low cowpea grain yield. 12 Figure 2.1 World cowpea production. Source: FAOSTATS (2020) 13 Table 2.1 African countries with the highest cowpea production (tons) in 2020 Rank Country Production (tons) Area harvested (ha) Yield (kg ha-1) 1 Nigeria 3 576 361 4 303 005 831 2 Niger 2 386 735 5 725 433 417 3 Burkina Faso 652 454 1 354 100 482 4 Ethiopia 374 332 220 037 1701 5 Kenya 246 870 298 120 828 6 Mali 215 436 454 274 474 7 Cameroon 215 016 244 058 881 8 Ghana 202 735 149 102 1360 9 Senegal 184 137 290 677 634 10 Sudan 161 000 339 780 474 11 Tanzania 127 884 112 657 1135 12 DRC 76 292 175 418 435 13 Mozambique 90 461 331 424 273 14 Malawi 41 656 97 825 426 15 Madagascar 31 069 34 122 911 16 Uganda 12 697 33 350 381 17 Mauritania 7 933 22 415 354 18 Egypt 7 180 1 853 3875 19 South Africa 4 801 109 90 437 20 Eswatini 715 2 056 348 Source: FAOSTATS (2020) 2.2 Origin, history and production of cowpea in South Africa Studies suggest that South Africa was a centre of cowpea wild species (Fang et al. 2007; Boukar et al. 2020). An additional view was that South Africa, in the Transvaal area, was the evolutionary core of V. unguiculata due to the existence of the most significant original wild variations in that area (Weng et al. 2017). Southern African wild cowpea species had smaller seeds than those found in West Africa, such as V. dekindtiana var. spontanea, which have somewhat larger seeds (Boukar et al. 2015). Literature that highlighted the history of cowpea revealed that due to the existence of the most primitive wild species (V. rhomboidea, V. protracta, V. tenuis and V. stenophylla), West and South Africa were the centre of V. unguiculata speciation (Pakela 2003; Boukar et al. 2020). Due to the minimal and lack of breeding programmes in the past, most wild species were eliminated in South Africa. People who were moving between countries where the Vigna pubescens subspecies emerged transferred the wild species across to neighbouring countries such as Mozambique, Zimbabwe, Lesotho, Botswana and Namibia (Horn et al. 2022). Later, the subspecies spread to East and West Africa. In the form of slave trade, the Spanish distributed wild cowpea species from southern African countries together with other wild 14 species such as millet, groundnut and soybean to tropical America in the 17th century (Sedivy et al. 2017; Sousa and Raizada 2020; Panzeri et al. 2022). Cowpeas are grown in dryland areas of South Africa, particularly in Taung and Potchefstroom in the North West. Small-scale farmers are the leading producers of cowpeas (Magloire 2005). In 2020, cowpea production in South Africa was estimated to be 4867 tons (Figure 2.2; FAOSTATS 2020). South Africa is the second lowest cowpea producing country with low yields. The area of production of the country is also very low as compared to Nigeria, which indicates that cowpea production in South Africa is given little research attention. Other factors contributing to low production and productivity are the lack of improved varieties and locally adapted cultivars, heterogeneous flowering and maturity, low yield potential and poor processing quality (Villa et al. 2005; Shiringani and Shimelis 2011). Significant progress has been made at the ARC in sourcing and selecting potential diverse cowpea population from IITA which are created for high yielding and good nutrition (Gerrano et al. 2017). However, the biggest challenge is that cowpea has narrow genetic diversity, which limits its genetic improvement (Mbuma et al. 2020, 2021, 2022). Figure 2.2 Cowpea area harvested (ha), production (ton) and grain yield (kg ha-1) in South Africa from 2010 to 2020. Source: (FAOSTATS 2020) 15 2.3 Factors affecting productivity of cowpea The cowpea crop productivity and improvement are known to be constrained by abiotic and biotic factors. Abiotic factors which affect plant growth and can be divided into imposed effects of stress, such as heat, cold and drought stress (Timko et al. 2007; Hatfield and Prueger 2015; Ravelombola et al. 2018). Environmental stress can cause damage to the cowpea plant which can appear as necrosis in all or a portion of the plant leaves. Cowpea plant damage can also be due to reduced rate of growth as a result of temperature, physiological malfunction, soil chemistry and water availability (Challinor et al. 2014; Iseki et al. 2021). The mentioned effects reduces the grain yield of the crop (Nunes et al. 2022). Therefore, the sustainability of crop production is threatened by environmental stress. Insect pests and diseases are the second key production barrier for cowpea production (Afutu et al. 2017). Insect pests and diseases cause major economic crop losses. The pod borer is regarded as the most destructive and economically important cowpea insect pest predominately in sub-Saharan Africa and causing yield losses ranging from 20 to 80% (Gianessi 2013). Cowpea aphid infestation is another major constrain that causes chlorosis, leaf curling and stunted growth limiting cowpea productivity and production (Kamphuis et al. 2012; Choudhary et al. 2017; MacWilliams et al. 2020). In Africa, aphid infestation has been linked to cowpea grain yield losses which ranges from 20 to 100% (Kataria and Kumar 2013). Diseases caused by viruses, bacteria, fungi and nematodes are a major biotic constraint in cowpea productivity and production (Oliveira et al. 2012; Mbeyagala et al. 2014; Odedara and Kumar 2017). However, over the past years, natural microbial-based treatments have been effectively used in agro-ecological production to reduce crop losses and increase productivity (Omomowo and Babalola 2021). Natural microbial-based treatments etails the use microbial inoculants that can either act as biocontrol agents, biofertilizers biopesticides and 2.4 Cowpea taxonomy, botany and genetics Cowpea [Vigna unguiculata (L.) Walp] is a dicotyledon which belongs to the Fabaceae family (Agbogidi 2010). In 1824, Savi introduced the genus Vigna, whose taxa had previously been referred to as Phaseolus (Pasquet and Padulosi 2013). The Vigna and Phaseolus are genera of the same legume tribe Phaseoleae with similar morphological characteristics, hence the initial confusion of referring them as the same crop. The Vigna taxonomy has been revised over the past years with significant breakthroughs using molecular technologies like DNA finger printing (Pasquet and Padulosi 2013; OECD 2016). It was discovered that the unguiculata 16 species comprises of four groups or subspecies (spp.) namely: unguiculata, biflora, sesquipedalis and textilis (Pasquet and Gepts 2021). The unguiculata is also subdivided into wild and cultivated ssp. in which the unguiculata is classified as a cultivated ssp., while spontanea and mensensis are classified as wild ssp. (Pasquet 2000; Ba et al. 2004). Additionally, the unguiculata ssp. is widely cultivated in Africa, India, the Middle East, and South America primarily for dry grain and fodder as it provides a source of protein for millions of poor people (Timko et al. 2007). Furthermore, the seeds of the wild spontanea ssp. are 10x smaller than cultivated unguiculata ssp., have small and dehiscent pods and have hard seed coats (Pasquet and Padulosi 2013; OECD 2016). Cowpea breeding programmes widely make use of the two ssp. (unguiculata and spontanea) for crop improvement. The ssp. sesquipedialis, commonly known as asparagus bean or “yard long bean”, is cultivated for its immature green pods in Eastern and Southern Asia (Xu et al. 2010). Cowpea is a warm-season herbaceous, annual crop that looks similar to common bean, but the leaves are less pubescent, shinier and darker green compared to the common bean (Timko et al. 2007). The growing patterns of many cowpea plants differ which include climbing, erect, bushy, trailing or indeterminate types under favourable conditions. Long bean varieties of cowpea with climbing and indeterminate growth characteristics involve regeneration of green pods throughout the growth period (Herniter et al. 2020). Their stems are slightly hairy or smooth and are rarely shaded purple (Pakela 2006). Their leaf form is either trifoliate or alternate (Pottorff et al. 2012). The inflorescence of cowpea contains two flowers. The flowers have short pedicels, are self-pollinating and a have corolla that can be in any of the following colours: purple, dirty yellow, white, light blue or violet (Ige et al. 2011). The flower opens in the morning and closes at about noon; they wilt and fall off after the flowering season. The seeds are accommodated in pods, which vary in texture, shape, size, and colour. Some plants are erect, coiled, or can be in the form of a crescent (Deshpande et al. 2018). When the pods mature, their colour could be yellow, purple or brown. Each pod contains about eight to 20 seeds, depending on the variety. Cowpeas are diploids with 2n = 22 chromosomes. Legume crops such as chickpea (Cicer arietinum) and common bean (Phaseolus vulgaris L.) have high genomic collinearity (a pair of loci that are shared by two species but are on different chromosomes) compared with cowpea (Vasconcelos et al. 2015; Merga and Haji 2019). Cowpea has a 613 megabase genome size and maternally inherited chloroplasts (OECD 2016). The ability of the crop to develop and perform successfully is determined by genetics and a range of environmental conditions. 17 2.5 Nutritional value/importance and functional compounds 2.5.1 Proteins and amino acids Cowpea is one of the most nutritious legume crops that provide populations in tropical regions with proteins and minerals (Dakora and Belane 2019). Previous research has shown that cowpea crude protein content varies with different varieties (Ravelombola et al. 2016). Protein content of cowpea genotypes was reported to be ranging from 21 to 25% (Itatat et al. 2013). Another study reported that cowpea has a comparatively higher protein content (17 to 30%) than cereals (7 to 13%) (El-Niely 2007). There is wide variation in the protein content of cowpea leaves, immature pods and the seeds. Studies have reported cowpea protein in leaves to be ranging from 25 to 32% (Okonya and Maass 2014), while for immature pods it ranges from 21 to 29% (Gerrano et al. 2017). Cowpea leaves, immature pods and seeds are all edible and mostly consumed in African countries. Cowpea has a complex profile of proteins. The relative proportions of each protein fraction have a significant impact on the nutritional quality of the crude protein in cowpea seeds (Teka et al. 2020). The major protein fractions (classified according to their solubility in different solvents) in cowpea are glutelins (alkali) and albumins (water) followed by globulins (salt solution) and prolamins (alcohol) (Vasconcelos et al. 2010; Santos et al. 2012) (Table 2.2). Because of their protein structural features, globulin fractions are naturally resistant to digestion (Phillips et al. 2003). Albumin content in cowpea seeds range from 8.20 to 11.90% (Tchiagam et al. 2011). Globulins and albumins are graded as metabolic and enzymatic proteins including lectins, protease inhibitors and lipoxygenase (Park et al. 2010). Prolamin occurs in the lowest concentration ranging from 2.30 to 5% (Gupta et al. 2010). Table 2.2 Types of protein fractions found in cowpea, their solubility and profile bands Protein profile Solubility Protein profile bands Glutelin alkali 21 Albumin water 20 Globulin salt solution 16 Prolamin alcohol 1 The ability of a food protein to meet the metabolic needs of humans for amino acids can also be defined as its nutritional quality (Frota et al. 2017). A protein's ability to satisfy both the physiological needs of the organism and the requirements for essential amino acids is measured 18 by its protein quality (Kurpad 2013). Cowpea amino acid profiles vary between genotypes (Table 2.3). There are 17 basic amino acids in cowpea protein, most of which are daily required in human diet (Elhardallou et al. 2015). Isoleucine, threonine, leucine and lysine are some of the basic amino acids found in cowpea protein (Gonçalves et al. 2016). As a result, it can supplement lysine-deficient proteins in staple cereals, tubers and starchy roots (Teka et al. 2020). In comparison, mature seeds have a lower concentration of free amino acids than those in immature seeds (Jayathilake et al. 2018). High concentrations of amino acids can also indicate reduced anti-nutritional factors, and enhanced digestibility and mineral bioavailability (Sarwar et al. 2012). The total amino acid content of cowpea seeds range from 27.50 to 33.43 g 100 g-1 (Gupta et al. 2010). Table 2.3 Amino acids profile of cowpea seeds Amino acids g 100 g-1 Cysteine 0.84 – 1.08 Tryptophan 1.00 – 1.33 Methionine 1.28 – 2.06 Histidine 1.85 – 2.47 Threonine 3.89 – 5.12 Isoleucine 4.17 – 5.46 Leucine 6.45 – 8.50 Lysine 7.30 – 8.74 2.5.2 Mineral concentration About 232 million people in Africa suffer from trace element deficiency, a problem that can be solved by legumes which have the ability of accumulating micronutrients in organs and boost brain development (Polak et al. 2015). Legumes have edible leaves and their seeds contain significant levels of dietary minerals required for human nutrition and health (Dakora and Belane 2019). Over 2 billion people are affected by insufficient Fe intake (the most widespread dietary risk factor in the world), while 49% of people worldwide are at risk from inadequate Zn diet (Brown et al. 2001). Plant breeders are focusing on the genetic improvements needed to support breeding schemes with the objective of increasing the accumulation of seed minerals in elite cowpea genotypes (Beaver and Osorno 2009). Micronutrients play crucial physiological roles in processes including haemoglobin formation (Fe), protein synthesis and stabilization of plasma membranes (Zn), bone health (Ca), Fe metabolism [copper (Cu)], promotion of blood clotting and fat metabolism (Mn) and stimulation of hormonal levels (B) (Mogobe et al. 2015; Dimkpa and Bindraban 2016; 19 Alejandro et al. 2020). Even though cowpeas contain a significant amount of Fe and Zn, its bioavailability is low. However, if taken with meals that are high in vitamin C, Fe and Zn, absorption is improved (Messina 2016). Large variation of mineral composition in different cowpea collections has been reported previously (Belane and Dakora 2011; Gonçalves et al. 2016; Gerrano et al. 2017; Mbuma et al. 2021). For example, a study that was conducted in South Africa reported large variation in cowpea immature pods for mineral concentrations which ranged from 60.05 to 97.78 mg kg-1 for Fe, 32.53 to 56.25 mg kg-1 for Zn, 21.05 to 47.72 mg kg-1 for Mn, 4.84 to 9.54 mg kg-1 for Cu and 21.29 to 40.31 mg kg-1 for B (Gerrano et al. 2017). When compared to cowpea grains, immature pods were observed to have a similar mineral concentration. For example, previous research has reported that mineral concentrations in cowpea grain ranged from 60 to 99 mg kg- 1 for Fe, 44 to 65 mg kg-1 for Zn, 5 to 32 mg kg-1 for Mn, 8.30 to 14.70 mg kg-1 for Cu and 10 to 22 mg kg-1 for B (Belane and Dakora 2011). A recent study that was conducted in South Africa reported mineral concentration in cowpea seeds to be ranging from 1.80 to 257.30 mg kg-1 for Fe, 19.40 to 55.30 mg kg-1 for Zn, 1.70 to 21.70 mg kg-1 for Mn, 3.80 to 35.70 mg kg- 1 for Cu (Mbuma et al. 2021). In Brazil, ranges of 58.50 to 69 mg kg-1 for Fe, 46.75 to 64.25 mg kg-1 for Zn, 8.25 to 16 mg kg-1 for Mn, 4.25 to 7 mg kg-1 for Cu were reported in cowpea grain (Gonçalves et al. 2020). While the immature pods and grains have similar mineral concentrations, higher values have been reported in cowpea leaves, varying from 42 to 55 mg kg-1 for B, 8.60 to 19.70 mg kg-1 for Cu, 49 to 104 mg kg-1 for Zn, 196 to 394 mg kg-1 for Mn, and 142 to 626 mg kg-1 for Fe (Belane and Dakora 2011). These studies have shown that there is a large variation of mineral concentration within and between collections. Furthermore, this variation may be explained by the collections' various genetic background and their growing conditions. 2.5.3 Starch, fibre and oligosaccharides Cowpea is the most important starch-protein grain legume in the West African sub-region, with a wider range of applications than other legumes (Atuobi et al. 2011). Starches are macromolecules produced by plant tissues that are used in both food and non-food products. Cowpea seeds have a carbohydrate content that ranges from 53 to 66%, with starch accounting for most of the carbohydrate content (Huang et al. 2007; Ashogbon and Akintayo 2013). Cowpea starch is gaining popularity in research due to its resilience to shear thinning and its high amylose content (39 to 42%) (Hoover et al. 2010; Ratnaningsih et al. 2016). Due to its high amylose concentration and resistance to enzyme hydrolysis, cowpea starch could be used 20 to make thickeners, textural modifiers and gelling agents in food formulations, and could be included in food products (Adebooye and Singh 2008; Ratnaningsih et al. 2016). Both soluble and insoluble dietary fibre are abundant in cowpeas (Liyanage et al. 2014; Jayathilake et al. 2018). Water can dissolve soluble fibre, which stabilizes cholesterol and blood glucose level. The colon and large intestine absorb water through insoluble fibre, which does not dissolve in water, to keep waste items wet and moving smoothly (Eashwarage et al. 2017). Increased intake of fibre content is somehow associated with cardiovascular disease, diabetes, obesity, cancer and many other chronic syndromes according to numerous epidemiological studies (Lattimer and Haub 2010; Threapleton et al. 2013; Barber et al. 2020). Cowpeas contains about 12 to 14.80 g 100 g-1 of total dietary fibre (Kirse and Karklina 2015; Eashwarage et al. 2017). Flatulence or gas production in the intestine is caused by the presence of oligosaccharides in legumes, which causes discomfort such as bloating and constipation (Afiukwa et al. 2012). The oligosaccharides stachyose, raffinose and verbascose are primarily responsible for this reaction to legumes, which varies depending on gender, age, intestinal microbiota composition and other factors (OECD 2016). Cooking of cowpea grains inactivate or reduce their anti-nutrients including oligosaccharides, thereby improving their nutritional quality (Wang et al. 2008; Tresina and Mohan 2012). 2.5.4 Bioactive polyphenols and antioxidants Cowpea is high in essential bioactive compounds that can support human health in various ways, in addition to its nutritional benefits (Awika and Duodu 2017). These essential bioactive compounds are polyphenols, which include phenolic acid and flavonoids. Phenolic acid plays a role in the antioxidant properties of legumes (Xu and Chang 2009; Singh et al. 2017; Liu et al. 2020). Concentration of phenolic acids in cowpea cultivars varies greatly depending on the seed colour (Cai et al. 2003). Red cowpea phenotypes have significantly more phenolic acids than other phenotypes, while white varieties have the least (Ojwang et al. 2012). Flavonoids play a crucial role in plant defence (Nassourou et al. 2016). Flavonoids are highly concentrated in the seed coat of cowpea and other legumes (soybeans, black beans, and kidney beans) and they have a significant influence on the choice of cowpea varieties for food consumption in most countries (Ojwang et al. 2012). Flavonoids have been studied extensively for their anti-stress and anti-inflammatory benefits (González et al. 2011; Yang et al. 2014). Previous research in South Africa reported a range of 0.36 to 0.95 mg 100 g-1 (Salawu et al. 21 2014) and 0.40 to 6.91 mg 100 g-1 (Mbuma et al. 2021) for total flavonoid content. In contrast to the values observed in South Africa, a higher range (0.21 to 23.95 mg 100 g-1) of total flavonoids was reported in Burkina Faso (Sombié et al. 2018). Common beans showed comparative differences of total flavonoids for leaves which ranged from 2.15 to 44.59 mg 100 g-1, for pods it ranged from 0.87 to 3.64 mg 100 g-1 and for seeds it ranged from 1.65 to 9.29 mg 100 g-1 (Pham et al. 2020). These studies have also indicated that flavonoid content varies among the cowpea genotypes depending on the seed colour, the darker seed coated cultivars have higher flavonoid content compared to the cream white seed coated cultivars. Antioxidants are substances that can prevent the onset or progression of chain oxidation reactions, thereby reducing or inhibiting oxidative damage (Moreira-Araújo et al. 2017). Due to their potential function in disease prevention and health promotion, there is increasing interest in the antioxidant activity of phenolics and tannin of cowpea (Zia-Ul-Haq et al. 2013). Depending on the cultivar, the phenolic content composition and bioactive properties of cowpea can differ significantly (Sombié et al. 2018). Consuming food products that are rich in phenolic antioxidants can reduce the chances of having terminal illnesses which include diabetes, cancer and cardiovascular disease (Zhao et al. 2014). Polymeric tannins can directly interact with proteins and carbohydrates, limiting their bioavailability, digestion and transport (King et al. 2000; Awika and Rooney 2004). Tannins can also reduce mineral availability by preventing them from being absorbed (Delimont et al. 2017). Tannis attracted interest as a potential natural supplement to reduce the calorie content of foods due to their capacity to bond carbohydrates with proteins (Dunn et al. 2015; Amoako and Awika 2016). Cowpea has a reduced tannin concentration that may be beneficial for health promotion (Awika and Duodu 2017). 2.6 Anti-nutritional compounds Anti-nutritional factors are secondary metabolites present in plants that inhibit the absorption of nutrients in food (Soetan 2008). Trypsin inhibitors, phytic acid, and oxalic acid are the main limitations present in cowpeas that reduce nutrients’ bioavailability (Preet and Punia 2000; Parmar et al. 2017; Jayathilake et al. 2018). Phytic acid and oxalic acid are the main anti- nutritional factors in legume crops (Akande et al. 2010). The quantity of anti-nutrients in cowpea depends on factors such as environmental effects, genotypical variations, and use of high-phosphate fertilizers, which is highly variable (Simion 2018). 22 Mineral bioavailability is reduced by the presence of phytic acid and oxalic acids as they bind with the minerals (Shukkur et al. 2006). Phytate inhibits the absorption of Mg, Ca, Zn and Fe (Akond et al. 2011). On the positive side, the ability of phytic acid to bind minerals reduces cell damage in the epithelium of the digestive tract which lowers the risk of colon cancer (Vucenik and Shamsuddin 2003). Due to lack of digestive enzymes (phytase), humans and non-ruminant animals cannot absorb phytate molecules, but ruminants can easily digest phytate because their rumen bacteria produce phytase (Klopfenstein et al. 2002). Rumen bacteria can convert dietary oxalic acid into carbon dioxide and formic acid in ruminants (Simion 2018). 2.7 History of cowpea breeding in the world Conventional cowpea breeding began at National de Recherches Agronomiques in Senegal, and the USA in 1961. Only a few national agricultural research programmes (Nigeria, Uganda, Tanzania, Senegal, India and the USA) had a form of improvement programmes and they were maintaining various collections of cowpea germplasm (Simion 2018). Collaborative attempts to develop cowpea cultivars for Africa started before the bean/cowpea Collaborative Research Support Programme (CRSP) was established (Hall et al. 2003). The purpose of the collaboration between the University of California Riverside (UCR) and the Institute Senegalais de Recherches Agricoles (ISRA) was to produce cowpea cultivars for specific production zones in Senegal and California. Dr. L.W. Kitch of Purdue University initiated hybridizations in Cameroon in 1990, and the Purdue project later partnered with Dr. J.D. Ehlers who had joined the UCR CRSP project as a cowpea breeder in producing improved cultivars (Hall et al. 2003). The Ein El Gazal genotype was developed from a cross between the California cultivar CB5 and the Senegalese breeding line Bambey 23 for use as a dry grain cultivar in the Sahelian zone (Elawad and Hall 2002). Many consumers of cowpeas in the USA and Africa value the grain. Cowpea breeding has been going on for years throughout the world's primary cowpea-growing regions, namely, the USA (Singh 2005), Asia (Xu et al. 2017), Africa (Horn and Shimelis 2020) and Latin America (Xiong et al. 2016). In the last three decades, international and national research organizations have achieved significant progress in breeding for pest and disease resistance, yield and nutritional value through conventional breeding (Boukar et al. 2016). Using simple sequence repeat (SSR) and amplified fragment length polymorphism (AFLP) marker cowpea linkage maps, quantitative trait loci for resistance to floral bud thrips were discovered (Omo-Ikerodah et al. 2008). The use of SSR markers to evaluate the genetic diversity across improved cowpea breeding lines and varieties obtained from the IITA breeding nursery demonstrated that improved cowpea 23 varieties had a small genetic base in general (Li et al. 2001). Breeders continue to develop segregated populations in commercial breeding programmes by using superior lines as parental lines for crossings. 2.8 History of cowpea breeding in South Africa South African wild cowpea relatives have the highest genetic diversity and most prehistoric features like bearded stigma, pod shattering, hairiness, outbreeding, and small seeds and pods (Ng and Singh 2009). The cultivated cowpea (ssp. unguiculata) originated from selection and domestication of perennial wild cowpea relatives, according to a review on the background of Vigna species (var. dekindtiana) (Panzeri et al. 2022). Seed dormancy and pod dehiscence were lost throughout the domestication process and when the species was domesticated through selection, it resulted in an increase in seed and pod size (Soltani et al. 2021). Various incompatibilities and linkage drag among the cultivated and wild species made it difficult to exploit genetic variability in South African wild species for crop improvement (Sharma et al. 2013). Pre-breeding presents a valuable technique for enhancing the use of genetic variability found in wild and cultivated type germplasm in such circumstances. Cowpea is a self-pollinator, so even though some gene banks have a considerable collection of cowpea germplasm, the crop's genetic base is still limited (Boukar et al. 2020). It is vital to use alien germplasm collection, particularly from cross compatible wild relatives to broaden the genetic base of the crop (Pratap et al. 2021). Pre-breeding is used to identify desirable characteristics in susceptible genotypes/materials that cannot be utilised to breed wild species independently and introducing those qualities into genetic backgrounds that are well-adapted (Sharma et al. 2013). Currently, only pre-breeding of cowpea is practised in South Africa. At the ARC, a cowpea pre-breeding programme has been initiated with the aim of introducing novel genes or traits into the breeding programme leading to full-scale crop improvement. The ARC in South Africa has a large cowpea germplasm collection from IITA Ibadan, Nigeria and Namibia for evaluating yield potential, nutritional quality, resistance to pests and diseases, and adaptation to South African conditions. 2.9 Selection methods in cowpea breeding Cowpea breeding programmes around the world have focused on improving qualitative and quantitative traits to improve crop productivity. Pure-line selection, pedigree, mass selection, backcross and mutation breeding are some of the methods that have been commonly and 24 successfully used to improve cowpea (Horn and Shimelis 2020). The primary target for almost all breeding programmes is drought tolerance (Ravelombola et al. 2020), pest and disease resistance (Boukar et al. 2020), lodging tolerance and growth habit (erect) (Kim et al. 2018), light sensitivity (photo-insensitive) (Nuhu and Mukhtar 2014), higher grain yield (Gerrano et al. 2019; Mbuma et al. 2020) and nutritional value (Boukar et al. 2018; Mbuma et al. 2021, 2022). 2.9.1 Pure-line breeding Pure-line selection is done in heterogeneous populations such as introduced germplasm collections, landraces and materials from mass selection in self-pollinated species to isolate superior genotypes. Pure-line selection is the oldest method of crop enhancement (Breseghello and Coelho 2013). This method is effective in self-fertilizing crops like sorghum (Diallo et al. 2019), pea (Yan et al. 2017), wheat (Agorastos and Goulas 2005) and cowpea (Boukar et al. 2016). It is also useful for breeding inbred lines used to create hybrids in self-pollinated crops. The main objective of the pure-line programme is to generate cowpea genotypes with improved yield, to generate genotypes that are adaptable to diverse environments and have desired characteristics such as seed type, growth habit and days to maturity (Simion 2018). Pure-line selection has been used to identify potential cowpea lines from segregating populations after application of induced mutagenesis (Horn and Shimelis 2020). 2.9.2 Pedigree selection In pedigree selection, individual cowpea plants are selected from second filial generation and subsequent generations for progeny testing (Horn and Shimelis 2020). It is a valuable technique in a transgressive breeding programme for selecting genotypes with distinct characteristics including maturity, plant height and disease tolerance. The advantage of using pedigree breeding in cowpea breeding programmes is that it is well suited for simply inherited characters such as shape, colour, height and pod size (Nkoana 2018). Pedigree selection can provide information on the genetic components and combining ability of cowpea for morphological traits (Ayo-Vaughan et al. 2013). Such data can be utilised to create superior cowpea genotypes. 25 2.9.3 Backcross selection Backcross breeding aims to increase the genetic value of a locally adapted cultivar with genetic defects such as disease susceptibility, low yield, and poor nutritional quality (Mutlu et al. 2005). It involves repeatedly backcrossing the hybrid with the recurrent parent with the desired trait, to recover the genetic background of the recurrent parent. The backcross breeding approach is frequently used to incorporate simple hereditary traits into varieties as a means of rectifying some of the deficiencies that may exist in current varieties (Collard and Mackill 2008). Additionally, the approach has been used to develop cowpea cultivars with improved productivity in hot environments (Chamarthi et al. 2019). 2.9.4 Mutation breeding Genetic variability in plant breeding can also be obtained through mutation. Mutation breeding is a cost-effective, fast and consistent method for developing and screening crop genotypes with improved and unique agronomic traits (Raina et al. 2020). Plant propagules such as seeds can be exposed to physical and chemical mutagens, which will cause mutations (Mba et al. 2010). Finding the proper radiation dose for the target genotypes is necessary before undertaking large-scale mutagenesis (Tshilenge-Lukanda et al. 2012). Several plant traits, including disease resistance, plant height, oil quantity and quality, maturity and the size of cowpea starch granules have been successfully modified by induced mutagenesis (Goyal and Khan 2010; Singh et al. 2013). A study was conducted in South Africa on generating drought-tolerant mutants by screening cowpea genotypes from second mutation (M2) to fourth mutation (M4) generations (De Ronde and Spreeth 2007). In Nigeria, gamma irradiation was used to create early maturing cowpea mutants with tendrils on the leaflets, thick leaves and light green pods (Adekola and Oluleye 2007). Therefore, mutation breeding has been crucial in the creation of superior cowpea genotypes. 2.10 Genetic variance, heritability and predicted selection gains The basic concept underlying variance component approaches is to break down the overall variance of a phenotype into specific causes. The phenotypic variation is the result of genetic variation and environmental variation, while genetic variance is a measure of the variation that exists in the genetic makeup of individuals within a population (Vogt 2020). Plant breeding relies on genetic variability and new variation is critical for introducing new features into 26 breeding programmes (Holme et al. 2019). Cowpea genetic variability must be understood to develop and improve conventional breeding programmes (Gerrano et al. 2015). Genetic variability is required to ensure long-term selective genetic gain and reduce crop vulnerability to abiotic and biotic stress (Magloire 2005). External conditions have a bigger influence on cowpea genotypes than genetic factors during the growth season (Gerrano et al. 2015). Reliable and precise assessments of the degree of phenotypic variability and heritability of yield and yield components are necessary to improve selection efficiency (Shimelis and Shiringani 2010). Researchers can utilize the total amount of genetic variation responsible for a trait in calculations of the trait's heritability once they have determined the total amount of genetic variation responsible for the trait (Govindaraj et al. 2015). Heritability is the proportion of observable variation (the phenotype) in a population that can be attributed to genotypic variances. This could be the broad sense-heritability (H2), which includes epistatic and dominance interaction effects, or narrow sense-heritability (h2) which is the quantity of the additive genetic effects over the phenotypic variance (Raudonius 2017). Plant breeders commonly use heritability to evaluate the accuracy of single or multiple field trials since it affects the sensitivity of selection and is a useful statistic in quantitative breeding (Piepho and Möhring 2007). Cowpea breeders frequently prefer to evaluate the performance of all varieties on a plot mean basis since cowpea is a self-pollinated plant and thus plot mean- based broad-sense heritability is more effective (Xu et al. 2009). The high heritability for morphological and nutritional quality traits highlights the efficiency of conventional breeding for cowpea. High H2 values of above 50% obtained from a study recommended that selection within each preliminary set for grain yield, days to 50% flowering and hundred seed weight would be desirable, given the moderate magnitude of environmental influence (Ongom et al. 2021). Estimates of heritability and genetic gain are used to assess effective genotype selection for desirable characteristics (Patel et. 2016). Genetic gain is a valid approach for increasing crop selection efficiency (Daetwyler et al. 2014; Krchov and Bernardo 2015). Genetic gain can be defined as the measure of performance improvement obtained each year from artificial selection. The genetic gain achieved using the relevant selection procedures, including selection gain can be used to measure improvement or response to selection (Xu et al. 2020). In order to accelerate genetic gain in various legumes such as cowpeas, several modern breeding techniques have been developed, including marker- assisted breeding and genome engineering (Huynh et al. 2018; Olatoye et al. 2019). Improvements of genetic gains have been made to cowpea yield-related traits such hundred seed weight, pods per plant, harvest index, and total dry matter (Kamara et al. 2011). 27 2.11 Correlations and principal component analysis The linear relationship between traits can be determined using correlation analysis. There are two types of correlations, namely genetic and phenotypic correlations. A quantitative genetic measure that indicates the genetic relationship between two characteristics is known as genetic correlation (Adetiloye et al. 2017; Sodini et al. 2018). Phenotypic correlation is the measure of the phenotypic relationship between two characters due to genetic and environmental factors. Genetic correlation is a key measurement that gives insight on the pleiotropy and linkage background as well as the selection response of complex traits (Zhang et al. 2021). Genetic correlation measures the breeding values of the the sets of parents for different traits (Hadfield et al. 2007). When two characters have a significant correlation, it is possible to select a trait of interest by focusing on the correlated trait that is easier to measure, especially when the primary trait, such as grain yield, has a complex hereditary (Ceballos 2004). Corresponding heritability, genetic advance, genotypic and phenotypic coefficients of variation will aid in the selection of superior genotypes, which will be equivalent to the amount of genetic variation available and the degree to which the characteristics are inherited (Patel et al. 2016). The use of correlation coefficients to study the relationship between traits is essential for early plant or inbred line selection, as well as simultaneous selection when more than one trait is needed (Silva et al. 2016). This is because when compared to direct selection, indirect selection based on correlated response might result in more effective genetic development. Previous studies (Adetiloye et al. 2017; Walle et al. 2018) on genotypic and phenotypic correlation among cowpea genotypes have reported higher genotypic correlations compared with the corresponding phenotypic correlations. Their results indicated that the environmental factors had little concealing influence on the yield components. Significant positive genetic and phenotypic correlations between mineral elements (Mbuma et al. 2021), morphological traits (Mbuma et al. 2021; Owusu et al. 2021), and between morphology and nutritional quality traits combined were found (Mbuma et al. 2022), indicating a likelihood of simultaneous selection of the traits. Positive correlations are of utmost importance for traits of interest since they indicated that selection of one trait will simultaneously increase the other correlated traits. However, negative correlations are not preferred on traits of interest since they indicate that an increase in one trait will result in decreased values of the other correlated traits. The similarity and differences between the genotypes and their association with a trait can be visualised using the principal component analysis (PCA). PCA is a statistical technique that enables you to condense the information in large data tables by using a smaller set of data that 28 can be more easily viewed and studied (Clark and Ma’ayan 2011). Therefore, solving an eigenvalue and eigenvector problem results in the discovery of new variables known as the main principal components (PCs) and the total variation observed in the data set (Jolliffe and Cadima 2016). A study by Siwale et al. (2022) on Bambara discovered a minimum threshold eigenvalue of one with 59.36% of total variation observed. Additionaly A study on cowpea also found minimum threshold eigenvalue of one that explained 78.22% of the total variation observed in the cowpea genotypes (Mbuma et al. 2021). On a recent study by Mbuma et al. (2022), the PCA distingished four groups of cowpea genotypes with their trait association. The first group of genotypes was associated with pod width, number of seeds per pod, pod length, zinc, iron and 100 seed weight, second group was associated with calcium, manganese and copper, third group was associated with Magnesium, potassium, phosphorus and Sodium, and the fourth group was associated with flavonoids, total phenolics, number of branches and grain yield. 2.12 Genotype by environment interactions and stability analysis The GE interaction is the change in the relative performance of genotypes when tested in multiple environments. Multi-environmental trials are helpful in evaluating the relative performance of genotypes in different environments, making it possible for breeders to give recommendations on the stability and adaptability of genotypes (Negash et al. 2013). The analysis of GE interaction provides information on specific and broad adaptation as well as the stability of genotypes (Gurmu et al. 2009; Mohammed et al. 2016). Various statistical models have been presented to evaluate the impact of GE interaction in various legume crops (Yan 2002; Asrat et al. 2009). These methods include ANOVA, AMMI and GGE biplot analysis. The ANOVA is an important technique for determining the variability among genotypes, its limitation is that it does not determine the adaptability and stability of genotypes. Statistical analysis such as AMMI and GGE have been used mainly to quantify GE interactions and to determine the stability of tested genotypes. The advantages of utilizing the AMMI approach are that the analysis distinguishes between main and interaction effects, which is crucial for the majority of agricultural research, and integrates a significant amount of the GE interaction sum of squares (Gauch 2006). The GGE biplot was proposed to effectively classify superior genotypes under mega-environments (a collection of environments that share adapted genotypes) because the AMMI biplot lacks the classification capacity needed to evaluate mega-environments (Peprah et al. 2020). 29 To give suggestions for broad or targeted adaptation, the GGE analysis is utilised to assess the proportional yield potential and stability of enhanced cowpea genoty