University of Ghana http://ugspace.ug.edu.gh SELECTION OF COWPEA (Vigna unguiculata (L). Walp) FOR HIGH YIELD UNDER LOW SOIL PHOSPHORUS CONDITIONS By SOULEYMANE Abdou (10496581) THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE AWARD OF DOCTOR OF PHILOSOPHY DEGREE IN PLANT BREEDING WEST AFRICA CENTRE FOR CROP IMPROVEMENT COLLEGE OF BASIC AND APPLIED SCIENCES UNIVERSITY OF GHANA LEGON [April, 2018] University of Ghana http://ugspace.ug.edu.gh DECLARATION I hereby declare that except for references to works of other researchers, which have been duly cited, this work is my original research and that neither part nor whole has been presented elsewhere for the award of a degree …………………………………............. Souleymane Abdou Student ………………………………….............. Professor Kwadwo Ofori Supervisor …………………………………............. Professor Eric Y. Danquah Supervisor …………………………………............. Professor Frank. K. Kumaga Supervisor …………………………………............. Dr. Agyemang Danquah Supervisor …………………………………............. Dr. C. A. Fatokun Supervisor …………………………………............. Dr. Ousmane Boukar Supervisor i University of Ghana http://ugspace.ug.edu.gh ABSTRACT Nitrogen (N) and phosphorus (P) are among the main edaphic factors that limit crop production in Sub-Saharan Africa (SSA). As a nitrogen fixing legume, cowpea production is not significantly affected by N, but P deficiency in soils. Two hundred (200) cowpea genotypes composed of both landraces and improved varieties were screened in the field under low P soil conditions in order to (i) determine genetic variability and heritability of traits associated with low P tolerance, (ii) assess the genotype by environment interactions for yield and its components under low P conditions, (iii) identify superior cultivars for use as varieties and/or as parents in improvement programs and (iv) determine the effects of planting date on the cycle and yield of local cowpea varieties in Niger. A participatory rural appraisal (PRA) was conducted to obtain farmers‘ knowledge and perceptions on low soil fertility and preference for cowpea varieties. The PRA results revealed that 11, 49 and 40 % of respondents claimed that their farms are respectively very low, low and high in fertility. About 75 % of farmers indicated that their cowpea production was low. Drought and low soil fertility were identified as the major production constraints. In terms of varietal preferences, early maturity and high yield potential were the traits preferred by farmers. High diversity with regards to yield and yield related traits under low soil P conditions was detected in the evaluated germplasm. High estimates of heritability, genetic variance, phenotypic variance as well as GCV and PCV were obtained in rainy season in comparison to dry season. The mean number of days to flowering, days to 50 % maturity, plant height, fodder yield, harvest index and hundred seed weight were the most highly heritable traits with 92.51 %, 85.02 %, 70.56 %, 80.6 %, 80.72 % and 71.94 % broad-sense heritability estimates, respectively. Grain yield was positively correlated with shoot dry weight, plant height and shoot color ; it was however negatively correlated with days to flowering and days to 50 % maturity. Genotype x Environment interaction was only significant for days to flowering, days to 50 % maturity and ii University of Ghana http://ugspace.ug.edu.gh fodder yield. The contribution of genotypes to the variation was high regarding fodder yield and days to flowering, but medium with regard to grain yield. Genotypes with high grain and fodder yield under low P soil conditions were detected; nevertheless, only G150 was stable across the test environments. The largest contribution of genotypes to traits variation was observed in fodder yield (46.76 %) while the smallest was in shoot dry weight (1.5 %). Genotypes were also responsible for a significant variation with regard to days to flowering (21.57%) as well as grain yield (15.06 %). The environment was the main cause of the variability detected in shoot dry weight, plant height, days to 50 % maturity and grain yield with 82.45 %. 72.04 %. 58.69 % and 36.55 %, respectively. The GEI part of the variation was high in days to flowering (36.85 %). fodder yield (22.73 %) and days to 50 % maturity (18.66 %). The G x E interaction although not significant for grain yield, contributed considerably (15.85 %) to the yield variation. Planting date had a significant effect on yield and yield related traits of local cowpea cultivars in Niger. Significant reduction in days to flowering and days to 50 % maturity with subsequent decrease in grain and fodder yield was observed when varieties were planted late. The first planting dates recorded the highest mean grain yield in both 2016 and 2017 while the lowest belonged to last planting dates. G56 and G34 were the best and worst performing genotypes with respectively 40.47 and 1.78 g grain yield per plant across the test environments. Superior genotypes in terms of fodder yield were G35, G78 and G74 with 178.50, 173.63 and 167.37 g per plant, respectively. Cultivars tolerant to both low P soil conditions and striga were also identified.The population developed for recurrent selection and QTL detection is at F2 generation. Candidate Genotypes for release as variety were identified. . iii University of Ghana http://ugspace.ug.edu.gh DEDICATION I dedicate the present work to my father (Elhadj Souleymane Soli) and mother (Khadidjatou Sara) for their indefatigable effort in shaping me into the person I have become. iv University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENTS All praises be to ALLAH who has made the completion of the present work possible. I would like to express my deep gratitude to the Alliance for a Green Revolution in Africa (AGRA) for giving the grant in order to carry out this programme. I would also like to thank Dr Fatokun C. A., Prof. Ofori K., Prof. Danquah E. Y., Dr Boukar, O., Dr Agyemang D. and Prof. Kumaga F. for their guidance, supervision and encouragements in the present work. My thanks to the staff of the West Africa Centre for Crop improvement (WACCI) for their supports I am indebted to Dr Germaine Ibro for her support, guidance, patience and enthusiastic assistance with regards to the present work. I would like to extend my special thanks to my wife and children for their sacrifice throughout the period this research work. I am immensely grateful to INRAN, my institute for giving me the study leave as well facilities that allow me carry out this PhD programme. My special thanks to IITA for the genotypes. ICRISAT, Niger is also acknowledged for providing me germplasm, weather data and field for my trials. The supports of Dr Kapran I., Dr Tahirou A., and Dr Bokar M. are acknowledged. Let me extend my gratitude to Mr Issaka Boureima, Issoufou Banaou, Lawali Iro, Illia Habou, Abdoussalami Issoufou, Mahamadou Seydou, Saddi Nagari, Adamou Aoula, Maazou Boubacar and Hamadou for their valuable assistance in the field work. v University of Ghana http://ugspace.ug.edu.gh I am also grateful for Dr Garba Mamane, Mr Ibro Madougou, Dr Falalou Hamidou as well all the colleagues from INRAN. Finally, I wish to extend my thanks to my colleagues students of WACCI, especially Chiaka, D., Ismail K., Kanfani G. and Zakaria D. for their help in this work. vi University of Ghana http://ugspace.ug.edu.gh Table of contents DECLARATION .................................................................................................................... i ABSTRACT .......................................................................................................................... ii DEDICATION ...................................................................................................................... iv ACKNOWLEDGEMENTS ................................................................................................... v LIST OF TABLES ................................................................................................................ xi LIST OF FIGURES ............................................................................................................ xiii LIST OF ABBREVIATIONS ............................................................................................ xiv CHAPTER ONE ........................................................................................................................ 1 1.0. INTRODUCTION ................................................................................................... 1 CHAPTER TWO ........................................................................................................................ 6 2.0. LITERATURE REVIEW ........................................................................................ 6 2.1. Origin and domestication of cowpea .................................................................... 6 2.2. Classification of cowpea ...................................................................................... 7 2.3. Biology and agronomy of cowpea ....................................................................... 8 2.4. Distribution and production ............................................................................... 10 2.5. Constraints to production ................................................................................... 11 2.5.1. Soil fertility and mineral fertilizer utilization ............................................. 11 2.5.1.1. The soil phosphorus problem ...................................................................... 13 2.5.1.2. Forms of phosphorus in soils ...................................................................... 14 2.5.1.3. Phosphorus availability to plants ................................................................ 15 2.5.1.4. Phosphorus function in plants ..................................................................... 16 2.5.1.5. Phosphorus deficiency symptoms ............................................................... 17 2.5.2. Impact of phosphorus on crop performance ............................................... 17 2.5.2.1. Impact of phosphorus on legumes and cowpea .......................................... 19 2.5.3. Adaptation of cowpea to low P soil conditions .......................................... 20 2.5.4. Mechanisms of adaptation to low P soils conditions .................................. 23 2.5.4.1. Root induced processes ............................................................................... 24 2.5.4.2. Root morphology ........................................................................................ 25 2.5.4.3. Mycorrhizal dependency ............................................................................. 26 2.5.4.4. Rhizosphere processes ................................................................................ 27 CHAPTER THREE .................................................................................................................. 29 3.0. Evaluation of farmers knowledge, perceptions on low soil fertility and their cowpea varietal preferences ................................................................................................. 29 vii University of Ghana http://ugspace.ug.edu.gh 3.1. Introduction ........................................................................................................ 29 3.2. Materials and methods ....................................................................................... 32 3.2 1. Site selection .................................................................................................... 32 3.2.2 Sampling method ............................................................................................... 33 3. 2. 3 Interview guides and data collection ............................................................... 33 3. 2.4 Data analysis .................................................................................................... 34 3.3. Results ................................................................................................................ 34 3.3.1. Socio-economic characteristics of respondents ................................................ 34 3.3.2. Farms and soils ................................................................................................. 35 3.3.3. Cropping systems ............................................................................................. 37 3. 3. 4. Farm size ........................................................................................................ 38 3.3.5. Cowpea production and yield ........................................................................... 38 3.3.6. Production constraints ...................................................................................... 39 3.3.7. Management of production constraints by farmers .......................................... 41 3.3.8. Causes of soil fertility deterioration ................................................................. 43 3. 3.9. Farmers varietal preferences ........................................................................... 44 3.4. Discussions ......................................................................................................... 45 3.5. Conclusion .......................................................................................................... 49 CHAPTER FOUR .................................................................................................................... 50 4.0. Genetic variability and heritability of cowpea traits under low P soil conditions . 50 4.1. Introduction ........................................................................................................ 50 4.2. Materials and methods ....................................................................................... 52 4.2.1. Rainy season trial ............................................................................................. 52 4.2.2. Dry season experiment ..................................................................................... 59 4.3. Results ................................................................................................................ 60 4.3.1. Genetic variation .............................................................................................. 60 4.3.2. Variance components and heritability .............................................................. 66 4.3.3. Correlation between traits ................................................................................ 68 4.3.4. Principal component Analysis (PCA) .............................................................. 70 4.4. Discussion .......................................................................................................... 71 4.5. Conclusion .......................................................................................................... 75 CHAPTER FIVE ...................................................................................................................... 76 5.0. Assessment of genotype by environment interactions for yield and its components under low P conditions ........................................................................................................ 76 viii University of Ghana http://ugspace.ug.edu.gh 5.1. Introduction ........................................................................................................ 76 5.2. Materials and methods ....................................................................................... 78 5.2.1. Experimental sites ............................................................................................ 78 5.2.2. Germplasm ....................................................................................................... 80 5.2.3. Method ............................................................................................................. 82 5.2.4. Data collection .................................................................................................. 82 5.2.5. Data analysis .................................................................................................... 84 5.3. Results ................................................................................................................ 86 5.3.1. Days to flowering (DAF) ................................................................................. 89 5.3.2. Shoot dry weight (SDWt) ................................................................................. 92 5.3.3. Plant Height (PHt) ............................................................................................ 95 5.3.4. Days to 50 % maturity (MAT50) ..................................................................... 98 5.3.5. Grain Yield (Gyield) ...................................................................................... 101 5.3.6. Fodder yield (Fyield) ...................................................................................... 103 5.3.7. Pod yield (Pyield) ........................................................................................... 106 5.3.8. Seed number per pod (Snber) ......................................................................... 107 5.3.9. Pod length (Plength) ....................................................................................... 108 5.3.10. Harvest index (HI) ........................................................................................ 109 5.3.11. Threshing percentage (Thresh %) ................................................................ 110 5.4. Discussion ........................................................................................................ 111 5.5. Conclusion ........................................................................................................ 115 CHAPTER 6 ........................................................................................................................... 116 6.0. Effects of planting date on the cycle and yield of local cowpea varieties in Niger ... .............................................................................................................................. 116 6.1. Introduction ...................................................................................................... 116 6.2. Materials and methods ..................................................................................... 118 6.2.1. Sites ................................................................................................................ 118 6.2.2. Agro-climatic data .......................................................................................... 119 6.2.3. Germplasm ..................................................................................................... 120 6.2.4. Method ........................................................................................................... 121 6.3. Results .............................................................................................................. 125 6.3.1. Days to 50 % maturity (MAT50) ................................................................... 129 6.3.2. Grain yield (Gyield) ....................................................................................... 133 6.3.3. Shoot dry weight (SDWt) ............................................................................... 136 ix University of Ghana http://ugspace.ug.edu.gh 6.3.4. Plant height (PHt) ........................................................................................... 137 6.3.5. Pod length (Plength) ....................................................................................... 141 6.3.6. Seed number per pod (Snumber) .................................................................... 142 6.3.7 Fodder yield (Fyield) ....................................................................................... 143 6.3.8. Striga .............................................................................................................. 147 6.4. Discussion ........................................................................................................ 149 6.5. Conclusion ........................................................................................................ 152 CHAPTER SEVEN ................................................................................................................ 154 7.0. General conclusions and recommendations ......................................................... 154 7.1. General conclusions ......................................................................................... 154 7.2. Recommendations ............................................................................................ 156 REFERENCES ....................................................................................................................... 158 APPENDIX ............................................................................................................................ 183 x University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES Table 2.1 Top 20 world cowpea producing countries in 2014 ................................................. 10 Table 3.1 Number of respondents per region ........................................................................... 33 Table 3.2 Average age of respondents (years) ......................................................................... 34 Table 3.3 Formal Education levels of respondents .................................................................. 34 Table 3.4 Distance of farms from villages ............................................................................... 35 Table 3.5 Mode of farm acquisition ......................................................................................... 36 Table 3.6 Soil texture according to farmers in the study areas ................................................ 36 Table 3.7 Area (ha), production (kg), and yield (kg/ha) of cowpea ......................................... 39 Table 3.8 Farmers assessment of their year 2014 cowpea production ..................................... 39 Table 3.9 Ranking of cowpea production constraints by farmers during FGD ....................... 40 Table 3.10 Farmers reasons for low cowpea production.......................................................... 40 Table 3.11 Assessment of the fertility status of farms by respondents .................................... 41 Table 3.12 Erosion level of farms according to respondents ................................................... 41 Table 3.13 Use of agricultural inputs by farmers (Questionnaire) ........................................... 43 Table 3.14 Ranking of variety traits by farmers ....................................................................... 44 Table 4.1. Some chemical properties of plot C8 soil at ICRISAT Sadore ............................... 54 Table 4.2. Weather data of 2016 rainy season at ICRISAT Sadore, Niger .............................. 55 Table 4.3. Chemical soil characteristics of the minor season trial site .................................... 59 Table 4.4 Mean square and variance of genotypes and blocks for traits in the rainy season ... 61 Table 4.5 Summary statistics for traits in the rainy season ...................................................... 61 Table 4.6 Mean square and variance of genotypes and blocks for traits in the dry season...... 62 Table 4.7 Summary statistics for traits in the dry season ......................................................... 63 Table 4.8 Summary statistics for traits in rainy and dry seasons ............................................. 64 Table 4.9 Days to flowering of photosensitive varieties in both seasons................................. 65 Table 4.10 Superior genotypes with regards to grain yield in both seasons ............................ 66 Table 4.11 Estimates of variance components and broad-sense heritability in the main season .................................................................................................................................................. 67 Table 4.12 Estimates of variance components and broad-sense heritability in the dry season 68 Table 4.13 Genetic correlation coefficient among traits in the rainy season ........................... 69 Table 4.14 Genetic correlation between traits in the dry season .............................................. 69 Table 4.15 Principal components analysis for the rainy season trial ....................................... 70 Table 5.1 Cowpea varieties used for the trials at Sadore and Maradi during 2017 rainy season .................................................................................................................................................. 81 Table 5.2 Level of significance of factors for the traits under study (combined analysis) ...... 87 Table 5.3 Mean squares of factors for traits (AMMI analysis) ................................................ 88 Table 5.4 Partitioning of total AMMI sum of square into its different components ................ 89 Table 5.5 Mean days to flowering of genotypes over all the environments ............................ 91 Table 5.6 Mean shoot dry weight of genotypes across the environments ................................ 94 Table 5.7 Means plant height of genotypes across the environments ...................................... 97 Table 5.8 Mean days to 50 % maturity for genotypes across the environments .................... 100 Table 5.9 Mean grain yield of genotypes across the environments ....................................... 103 Table 5.10 Mean fodder yield of genotypes across the environments ................................... 105 xi University of Ghana http://ugspace.ug.edu.gh Table 5.11 Mean pod yield per plant of genotypes at E3 and E4. ......................................... 107 Table 5.12 Mean number of seeds per pod at E3 and E4 ....................................................... 108 Table 5.13 Mean pod length for genotypes at E3 and E4 ...................................................... 109 Table 5.14 Mean harvest index of genotypes at E3 and E4 ................................................... 110 Table 5.15 Mean threshing percentage of genotypes at E3 and E4 ....................................... 111 Table 6.1 Rainfall data of the experimental sites ................................................................... 119 Table 6.2 Weather data of Sadore in 2016 ............................................................................. 120 Table 6.3 List of cowpea genotypes used for the study ......................................................... 121 Table 6.4 Mean squares of traits (combined ANOVA using six planting date‘s data) .......... 125 Table 6.5 Mean squares of traits (AMMI analysis)................................................................ 126 Table 6.6 Mean squares of traits (combined ANOVA using 2016 planting dates‘ data) ...... 126 Table 6.7 Mean squares of traits (combined ANOVA using 2016 planting dates‘ data) ...... 127 Table 6.8 Contribution of the interaction components to G x E variation ............................. 128 Table 6.9 Partitioning of total AMMI sum of squares into its different components ............ 129 Table 6.10 Comparison of the 2016 mean days to flowering and days to 50 % maturity of the first and last planting dates ..................................................................................................... 130 Table 6.11 Mean days to 50 % maturity and stability values of genotypes across the environments .......................................................................................................................... 131 Table 6.12 Mean grain yield per plant and stability values of genotypes across test environments .......................................................................................................................... 134 Table 6.13 Mean shoot dry weight per plant and stability values of genotypes across test environments .......................................................................................................................... 137 Table 6.14 Mean plant height and stability values of genotypes across test environments ... 139 Table 6.15 Mean pod length of genotypes across the 2017 planting dates ............................ 142 Table 6.16 Mean number of seeds per pod of genotypes across 2017 planting dates............ 143 Table 6.17 Mean fodder yield per plant and stability values of genotypes across 2017 planting dates ........................................................................................................................................ 145 Table 6.18 Mean scores of genotypes for tolerance/susceptibility to striga across 2017 planting dates .......................................................................................................................... 148 xii University of Ghana http://ugspace.ug.edu.gh LIST OF FIGURES Figure 3-1 Farmers‘ reasons for growing cowpea in association with millet and other crops. 37 Figure 3-2 Management of production constraints by farmers (FGD) .................................... 42 Figure 3-3 Factors causing reduction of soil fertility according to farmers ............................. 43 Figure 4-1 Distribution of total P in the ICRISAT sahelian center Sadore 2011 ..................... 52 Figure 4-2. Total N distribution in the ICRISAT sahelian center Sadore 2011. ...................... 53 Figure 4-3. Distribution of sites where local cowpea varieties were collected ........................ 56 Figure 5-1 Weather data of ICRISAT Sadore for 2016 and 2017 rainy seasons ..................... 79 Figure 5-2 Variation of weather factors between rainy and dry season ................................... 80 Figure 5-3 Boxplots for means of days to flowering across environment ............................... 90 Figure 5-4 GGE biplots for days to flowering means .............................................................. 92 Figure 5-5 Boxplot for shoot dry weight means ...................................................................... 93 Figure 5-6. GGE biplots for shoot dry weight ......................................................................... 95 Figure 5-7 Boxplot for plant height means .............................................................................. 96 Figure 5-8. GGE biplots for plant height means ...................................................................... 98 Figure 5-9. Boxplot for days to 50 % maturity means ............................................................. 99 Figure 5-10. GGE biplots for days to 50 % maturity means ................................................. 101 Figure 5-11 Boxplot for grain yield means per plant ............................................................. 102 Figure 5-12. Boxplot for fodder yield means ......................................................................... 104 Figure 5-13. GGE biplots for fodder yield means .................................................................. 106 Figure 6-1 GGE biplots of days to 50 % maturity ................................................................. 132 Figure 6-2 Discriminating power and representativeness of the environments ..................... 133 Figure 6-3 GGE biplots of grain yield per plant .................................................................... 135 Figure 6-4 Discriminating power and representativeness of environments ........................... 136 Figure 6-5 Boxplot for plant height means ............................................................................ 138 Figure 6-6 GGE biplots for plant height ................................................................................ 140 Figure 6-7 Discriminating power and representativeness of environments ........................... 141 Figure 6-88 Boxplot for fodder means per plant .................................................................... 144 Figure 6-9 GGE biplots for striga infestation ......................................................................... 149 xiii University of Ghana http://ugspace.ug.edu.gh LIST OF ABBREVIATIONS ABU: Ahmad Bello University AGRA: Alliance for Green Revolution in Africa AMMI: Additive Main Effect and Multiplicative Interaction ANOVA: Analysis of Variance CV: Coefficient of Variation FAO: Food and Agricultural Organization GCV: Genetic Coefficient of Variation GE: Genotype by Environment GEI: Genotype by Environment Interactions GGE: Genotype and Genotype by Environment Interactions IAR: Institute of Agricultural Research ICRISAT: International Crop Research Institute for the Semi-Arid Tropics IITA: International Institute for Tropical Agriculture INERA: Institut de L‘Environment et de Recherches Agricoles INRAN: Institut National de la Recherche Agronomique du Niger PCV: Phenotypic Coefficient of Variation PRA: Participatory Rural Appraisal PPB: Participatory Plant Breeding PVS: Participatory Varietal Selection xiv University of Ghana http://ugspace.ug.edu.gh QTL: Quantitative Trait Locus USA: United States of America WACCI: West Africa Centre for Crop Improvement xv University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE 1.0. INTRODUCTION Cultivated cowpea, Vigna unguiculata [L. Walp] is an annual crop (2n = 22) that belongs to the family fabaceae. It is a seed-pod crop adapted to warm agro-ecology and susceptible to chilling. It can be cultivated in areas with annual rainfall ranging from 500 to 1200 mm, however extra-early and early varieties can be grown in zones having less than 500 mm (Dugje et al., 2009). As one of the most important leguminous crops in the dry savanna areas of Sub-Saharan Africa, cowpea provides food, fodder and cash to farmers. Its leaves and grains are of high nutritional value with a mean protein content of about 26 % (Ibro et al., 2011), making it the cheapest source of dietary protein to small scale farmers. The high amount of antioxidants in its grain contributes in lowering the risk of cancer (Mahamane et al., 2008). In the Sahel, cowpea is a key component of agricultural systems as it is mostly grown in association with cereals especially millet and sorghum. It is a source of fodder needed to sustain a crop-livestock integration system practiced by small scale farmers through out the semi-arid zone. In some areas, farmers living around the urban zones grow cowpea just for fodder due to its high price in cities as it is used for fattening sheep and goat. Cowpea plays a significant role in the improvement of soil fertility as it replenishes nitrogen of the soil. Studies have shown that the presence of legumes such as cowpea in cropping systems is a cheap alternative to improve soil fertility and subsequently yield. It is also used as a cover crop for the prevention of wind and water erosion. Cowpea is grown widely in Africa, Asia and America. World production amounts to 5,718,145 tons over an area of 11,316,145 ha (FAOSTAT, 2014). Niger Republic with an average production of 1,000,000 tons over the last five years is the second largest cowpea producing country in the world after Nigeria. In Niger, cowpea has become the second crop in 1 University of Ghana http://ugspace.ug.edu.gh terms of production after millet (FAOSTAT, 2014) and the most important legume as a result of the drastic decline in groundnut production. In the last decade, cowpea has become a major cash crop and occupies about 90 % of area allocated to leguminous crops thereby improving significantly the income level of small scale farmers (Ibro et al., 2007). About 70 % of its production is exported mainly to neighbouring countries; consequently, it is among the main sources of Niger‘s foreign exchange with 27 % of the total agricultural export. Crop production in general and cowpea in particular has been dwindling due to many limiting factors. These factors are the main causes of the observed low yield across cowpea producing areas in the semi-arid areas of the tropics. The mean grain yield is about 324 kg/ha across major world cowpea production countries (Singh et al., 2002). In Niger cowpea yield ranges from 68 to 388 kg/ha from 2000 to 2012 with a mean of 208 kg/ha (INS, 2013). This yield is low even though it is obtained from a mixed cropping system where cowpea shares the same field with millet or sorghum. The low grain yield is compounded by several factors such as insects, diseases and parasitic weeds as well as poor soil conditions, the use of unimproved varieties, inadequate and poor cultural practices (Rachie, 1985). Low soil fertility is among the most important abiotic factors that cause severe setbacks in cowpea production. Many studies have concluded that in the driest zones of the Sahel and other parts of West Africa, low soil fertility causes yield reduction more than rainfall (Payne, 1997). In Niger, only about 1/3 of the country‘s land area is suitable for agriculture. Agricultural soils, especially upland soils are generally poor, acidic (4.5 to 6.0 pH), have low content of organic matter (0.1 to 0.7 %) and deficient in P with a bray-1 P varying from 0.4 to 3.4 mg/kg (Nouri et al., 2016). The soils are predominantly sandy with low silt and clay content (Republique du Niger, 2001). However, soils with low clay content and kaolinite‘s low activity tend to contain high free iron oxide to clay fraction that make inherent P fixed or 2 University of Ghana http://ugspace.ug.edu.gh blocked by iron forms such as goethite (West et al., 1984). Apart from P, N is another deficient macroelement that limits crop production in Niger. Cowpea being a legume has the ability of fixing nitrogen, its production is not therefore affected by N deficiency in soils, however low P content in soils could hinder N fixation of leguminous crops (Krasilnikoff et al., 2003; Jemo et al., 2006; Qiao et al., 2007). P plays a major role in crops shoot and growth, respiration, photosynthesis, flower initiation and fruit development (Theodorou and Plaxton, 1993; Schachtman et al., 1998; Uchida, 2000). The low P nature of agricultural soils in Niger could therefore be a major edaphic factor responsible for the observed low cowpea yield in the country. Several methods of improving P levels of the soil and/or its availability to crops were reported in literature. These include application of compost (Nelson and Nikkelsen, 2008; Verma et al., 2013), poultry manure (Yu et al., 2013), crop residues (Fuller et al., 1956,), rock phosphate (Bationo et al., 2002; Saidou et al., 2011) as well as mineral P fertilizers (Saidou et al., 2007; Odundo et al., 2010; Oladiran et al., 2012; Haruna and Usman, 2013; Nkaa et al., 2014). Adoption of these technologies by farmers was however limited as a result of many socio- economic constraints. Compost and animal manure application are conditioned by their availability, labour as well as distance of farms from villages. Williams et al. (1999) reported labour requirement and availability, farm size, farm ownership and its distance from villages are prerequiste to organic residues utilization. The situation may become worse in future with the increased pressure on arable land due to the slow but steady spread of the desert, high population growth rate, exponential urbanization and overgrazing. In almost all crop growing areas of the country, the fields are now continously farmed until it is worn out, leading to complete depletion of soil nutrients and other forms of soil degradation. Crop residues that can help in resuscitating soil nutrients are completely collected from farms for livestock feeding, fuel and housing. Furthermore, erosion caused by wind and rainfall add to the 3 University of Ghana http://ugspace.ug.edu.gh problem by sweeping the top soils since the fields are left bare with no cover. Also P mineral fertilizer application is limited by low income of farmers, restricted access to credits, as well as low level of education (Abdoulaye and Sanders, 2005; Kelly, 2005; Ndjeunga and Batalian, 2005). Under low P soil conditions, adoption of tolerant varieties could be a good alternative for resource poor farmers and environmental preservation as well. Significant genotypic variations with regards to P uptake and P use efficiency have been found in rice, wheat, clover and maize (Trolove et al., 1996; Romer and Schenk 1998). Yan et al. (1995) reported that legumes vary in their P use efficiency with respect to growth and yield. Genotypic variability has been found in pigeon pea (Cajanus cajan) and common bean (Phaselous vulgaris) regarding their ability to acquire P (Subbarao et al., 1997; Bonser et al., 1996). In cowpea large varietal differences with respect to adaptation to P deficient soil conditions were reported (Mahamane et al., 2008; Vesterager et al., 2006; Krasilnikoff et al., 2003; Oladiran et al., 2012; Alkama et al., 2009; Jemo et al., 2006; Saidou et al., 2007; Ojo et al., 2007). Researchers have taken advantage of that variation and developed mapping populations in order to identify quantitative trait loci associated with low P soil tolerance. Rothe et al. (2014) reported three SSR markers namely CLM0269, 221/222, and CLM0298 linked to low P soil tolerance in cowpea. In addition, Ravelombola et al. (2017) screened 357 USDA cowpea lines for high performance on P deficient soil conditions and response to rock phosphate and found ten and eight QTLs linked to respectively low P tolerance and rock phosphate use efficiency. Despite the inherent low P nature of most agricultural soils in Niger, that is being aggravated by continuous cultivation of fields, farmers‘ socio-economic conditions and some of their practices, as well as low availability and high price of P fertilizers, very little attempts have been made to develop cowpea varieties adapted to P deficient soil conditions. The general 4 University of Ghana http://ugspace.ug.edu.gh objective of the present work was to select cowpea lines with tolerance to low P soils conditions. The specific objectives were to: i) Evaluate farmers perceptions on low soil fertility and determine their cowpea varietal preferences; ii) Determine genetic variability and heritability of traits associated with low P tolerance; iii) Assess the genotype by environment interactions for yield and its components under low P conditions and iv) Identify superior cultivars for use as varieties and/or as parents in improvement programs. 5 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO 2.0. LITERATURE REVIEW 2.1. Origin and domestication of cowpea The origin of cowpea has been controversial as different centers were suggested by many authors. The centers proposed include India (Faris, 1964), Persia (Wright, 1907), South America (Piper, 1913) and Africa (Ames, 1939). However, Fery (1990) stated that the majority of the evidence published recently indicates Africa as center of origin of cowpea. Wild Vigna species center of diversity is located in Southeastern Africa (Ng and Padulosi, 1988, Padulosi et al., 1997). With regards to cultivated cowpea, its center of highest diversity is the Nothern Guinea Savanna of West Africa. Steele (1972) believes that Vigna unguiculata originated from West Africa since the subspecies dekindtiana which is most cross-compatible with cultivated cowpea is found exclusively in that region. V. unguiculata was a perennial and predominantly outbreeding crop; however it has changed to annual and inbreeding growing habit through the process of evolution according to Ng (1995). Evolving from species dekindtiana, cultivated cowpea experienced the loss of seed dormancy, pod dehiscence, pod and seed size along with an increase in yield through the process of domestication, selection and cultivation. Although the evidences showed cowpea domestication took place in Africa, the precise location where it was domesticated in Africa is not certain. According to Ng and Maréchal, 1985, Ethiopia, Central Africa, Central and South Africa as well as West Africa were among the possible centers of domestication. Similarly, Padulosi and Ng (1997) postulated that cowpea could have been brought under cultivation in any part of Sub-Saharan Africa taking into account the fact that variety dekindtiana is widely spread in that region. 6 University of Ghana http://ugspace.ug.edu.gh Nevertheless, Ehlers and Hall (1997) stated that West Africa was the first center of cowpea domestication. Their claim corroborates with the findings of Ng (1995) who asserted that the greatest diversity of cultivated cowpea is found in the Northwest of Cameroon, part of Burkina Faso, Savanna of Nigeria, Northern Benin, Togo and Southern Niger. Based on archaeological findings from carbon dating of wild or cultivated cowpea of Kimtampo rock shelter, central Ghana is so far the eldest place where cowpea was discovered (Flight, 1976). Archaeological evidence showed that cattle existed in West Africa since 3000 BC (Clutton- Brock, 1989) and probably early African farmers used wild cowpea as feed for their cattle. Ng (1995) hypothesized that the domestication of the cowpea cultigroug unguiculata occurred in West Africa as far back as 2000 BC. Cowpea was undoubtedly brought the first time to India during the Neolithic period (Pant et al, 1982) and its large genetic variation existing in that region suggests that it is the secondary center of the crop diversity. Cowpea was brought to th Europe via North-eastern Africa about 300 BC, to tropical Americas in the 17 century and to th Southern USA in 18 century (Padulosi and Ng, 1997) 2.2. Classification of cowpea Cowpea, an herbaceous dicotyledonous plant, is a member of the order Fabales. It belongs to the family Fabaceae, subfamily Faboideae, tribe Phaseoleae, subtribe Phaseolineae, genus Vigna and section Catiang (Maréchal et al., 1978). There are many species in the genus Vigna, its numbers however differ among authors. One hundred and eighty four species were reported by Phillips (1951), 150, 154 and 84 by Verdcourt (1970), Steele (1972) and Marechal et al. (1978) respectively. Other members of the Vigna species apart from cowpea are blackgram (V. mungo), mungbean (V radiata) and adzuki bean (V. angularis). Cowpea has strong similarities with Phaseolus and Dolichos. Maréchal et al. (1978) asserted that genus Vigna was at first divided into many subgenera on the basis of morphological features, degree of genetic hybridization/reproductive seclusion and geographic dissemination of 7 University of Ghana http://ugspace.ug.edu.gh species. The subgenera Vigna and Haydonia from Africa, Ceratotrophis from Asia and Sigmoidotropis and Lasioprom from America are the main groups. However, among all the cowpea subgenera, only Vigna unguiculata which is subdivided in four cultigroups is under cultivation. The cultigroups are unguiculata, biflora, sesquipedalis and textilis (Ng and Maréchal, 1985). Apart from cultivated cowpea, wild cowpea relatives exist within the genus Vigna. Padulosi (1993), reported that more than 400 wild Vigna unguiculata were characterized at IITA; that work along with study of live materials in situ and samples from main European and African herbaria as well as cytological investigations brought about the discovery of new taxa and a modification of the nomenclature of some species (Ng, 1995). The species dekindtiana, tenuis and stenophylla were classified under the Vigna (Padulosi, 1993, Padulosi and Ng, 1997, Maréchal et al., 1978) while variety protacta and pubescens were placed into two distinct subspecies owing their hair possession in pods and other plant parts as well as leaves, flowers, pollen and grain morphology and their root systems. In this classification system, new varieties such as grandiflora, huillensis, ciliolata and congolensis, were added to subspecies dekindtiana, however varieties parviflora, oblonga and tenuis, were distinguished in subspecies tenuis while protrata, rhomboidea kgalagadiensis were recognized within the subspecies protracta. 2.3. Biology and agronomy of cowpea At germination, the two leaves of cowpea, above the cotyledon, are simple and opposite, but the subsequent trifoliate leaves are alternate. The shape of cowpea leaves is either globose, sub-globose, hastate/lanceolate or sub-hastate (http://genebank.iita.org). The plants are of various types and they are classified as erect, semi-erect, prostrate, semi-prostrate and climbing. Cowpea growth habit ranges from indeterminate to determinate with the indeterminate habit being predominant. Most varieties are tap rooted with few lateral roots. 8 University of Ghana http://ugspace.ug.edu.gh The peduncles stemming from the leaf axis support the raceme which in turns bears the flowers. The flowers are of white, yellow, red or purple color. There is a relationship between flower and coat color according to Saunders (1960). With regards to the pods form, it can be straight, curved or coiled. In general fresh pods are green in color, but some can be partially or completely purple or black as a result of anthocyanin content (Harland, 1919). Cowpea seeds have generally kidney form, however when it is limited in the pod, its shape turns to globular. Seeds are usually of medium size, but genotypes with small or large seed size do exist. With respect to seed coat, it can be smooth or wrinkled while its color may be white, red, cream, green, buff, brown, black , speckled, mottled or blotchy. Many are labelled as eyed (blackeye, pinkeye purple hull etc.) whereas the white colored helium is delimited by a different color (Davis et al., 2010). Cultivated cowpea differs considerably from its wild relatives but it shares some similarities in terms of morphology and growth habit. Among the wild cowpea, variety dekindtiana is the most closely related to cultivated cowpea landraces, however they are significantly different with regards to color, size and smoothness of the mature pods as mature pods of dekindtiana are small, black, scabrous, shatter and contain tiny, dark speckled or solid black seeds resembling those of other wild varieties (Padulosi and Ng, 1997). There is no dormancy in cultivated cowpea (Ezedinma et al., 1962) while its wild relative‘s seed have to be scarified to enhance germination as a result of some levels of dormancy (Stanton, 1961). In suitable moisture conditions, the germination of cultivated cowpea seeds occurs within 4 to 5 days after planting (DAP). Flower bud initiation depends on the variety growth habit as well as its reaction to photoperiodism. In general, determinate varieties are early maturing while the indeterminate ones mature late. Most determinate varieties start flowering from 4-5 weeks after planting, reach the peak at 7 weeks whereas flowering begins at 7 weeks in the case of indeterminate genotypes. With regards to cycle, cowpea varieties are classified as extra-early, 9 University of Ghana http://ugspace.ug.edu.gh early, medium and late maturing. Extra-early varieties reach maturity within 60-69, early around 70-79, medium in 80-89 and late in 90-120 DAP (Dugje et al., 2009). 2.4. Distribution and production Cowpea is a versatile legume crop that tolerates to a large extent harsh climatic conditions. It is cultivated worldwide, however, large number of producing countries do not store production data. In most developing countries, reliable data with regards to agriculture can only be obtained from FAOSTAT. In 2014, cowpea was produced in 42 countries and its world production amounted to 5, 602, 715 tons over a total area of 12, 6 million hectares (FAOSTAT, 2017). Africa, to which the seven top producing countries belong, is by far, the leading continent in terms of cowpea production and the bulk of the production comes from West Africa (Table 2.1.). Table 2.1 Top 20 world cowpea producing countries in 2014 Rank Country Production (tons) Area (ha) Yield (Kg/ha) 1 Nigeria 2. 137.900 3. 701.500 578 2 Niger 1. 593.166 5. 325.168 299 3 Burkina Faso 573.048 1. 205.162 475 4 United Republic of Tanzania 190.500 197.323 965 5 Cameroon 174.251 209.019 834 6 Mali 149.248 353.382 422 7 Kenya 138.673 281.877 492 8 Myanmar 115.200 132.000 873 9 Mozambique 103.837 377.900 275 10 Sudan 80.000 260.000 308 11 D R C 70.042 159.945 438 12 Senegal 64.088 153.142 418 13 Malawi 35.903 81.753 439 14 Haiti 29.895 41.525 720 15 United States of America 21.591 12.060 1.790 16 Peru 17.588 12.779 1.376 17 Serbia 16.189 4.777 3.389 18 Sri Lanka 15.281 11.519 1.327 19 China, mainland 13.500 13.000 1.038 20 Uganda 10.100 25.000 404 Source: FAOSTAT, 2017. 10 University of Ghana http://ugspace.ug.edu.gh Nigeria, Niger and Burkina with 2,137,900, 1,593,166 and 573,048 tons respectively, are the major producing countries and account for 77 % of the world production. 2.5. Constraints to production Cowpea production is subject to complex biotic and abiotic limiting factors that severely reduce its yield. The average yield in the world in 2014 is 444 kg/ha (FAOSTAT, 2017), it is very low compared to the potential which was reported to be more than 3000 kg/ha by Singh et al. (1997). The biotic constraints include insects, diseases and witch weeds. However insects appear to be the most common and serious biotic enemy of cowpea both on the field and in storage. Fatokun (2002) stated that at each stage of development, there are one or more main insects that attack cowpea and impede its development with subsequent negative effect on the yield. Aphis craccivora, for example attacks cowpea at seedling, flowering and pod stage and is believed to be one of the major pest of cowpea in Africa, Asia and Latin America (Jackai and Daoust, 1986). Drought, heat, low soil fertility and poor use of improved varieties are among the critical abiotic factors which badly dwindle cowpea production. In the semi- arid of the tropics, an area characterized by erratic rainfall, drought is a foremost cause of poor crop yield, however according to Payne et al (1997), poor soil fertility bring about yield loss more than drought in the driest parts of the Sahel and other areas of West Africa. 2.5.1. Soil fertility and mineral fertilizer utilization The bulk of cowpea production comes from the West African Semiarid Tropics (WASAT). This zone is characterized by inherent low soil fertility as a result of low levels of organic carbon (generally less than 0.3%), poor total and available phosphorus and nitrogen, as well as low effective cation exchange capacity (ECEC) (Bationo et al., 2002). Organic matter contributes significantly to soil fertility since it contains around 70 to 80 and 60 to 70% of nitrogen and phosphorus respectively. Soil of the WASAT not only tend to be sandy and low in pH, organic-matter content, N, P and K availability, water holding capacity, and 11 University of Ghana http://ugspace.ug.edu.gh cation-exchange capacity (CEC) but also its clay mineralogy is largely kaolinitic (Jones and Wild, 1975; Weil, 2000; Wong et al., 1991; Casenave and Valentin, 1992; Manu et al., 1996; Rockström et al., 1998; West et al., 1984). Soils with low clay content and kaolinite low activity tend to have a high ratio of free iron oxide to clay, causing much of the native P to be fixed or occluded by iron forms such as goethite (West et al., 1984). In the past three -1 -1 -1 decades, around 660 kg N ha , 75 kg P ha , and 450 kg K ha has been lost on average from approximatively 200 million hectare of farmland in 37 African countries. Soil N and P can be restored through the use of fertilizers and organic matter. However while N is mainly replenished via organic matter, P is mostly restored through the use of mineral fertilizers (Sanchez et al., 1997). However, the striking point is that both mineral and organic fertilizer are used in extremely small quantities resulting in very low amount applied per ha and slow increase of the rate of consumption (Kelly and Naseem, 2009). Most farmers cannot afford it due generally to its high cost and poor availability (Trolove et al., 2003), a situation that persisted for decades (Payne, 2006). With an average rate of 10 kg/ha, SubSaharan Africa is the world area where chemical fertilizer is less used (McIntire (1986); the whole continent consumption represents only 2 to 3% of that of the world (Kelly and Naseem, 2009). The situation is even worse in Niger where only 1 kg/ha were used (Pender et al., 2008). With regard to P fertilizers utilization, Africa accounts for less than the quarter of that the world (Lott et al., 2011). In Niger mineral P application is very low; NPK is the most popular P fertilizer, however its utilization is limited by many socio economic constraints. The world has a considerable reserve of rock phosphate (RP) that can be used as source of good quality as well as relatively low cost phosphorus fertilizer. The largest world deposit of RP is found in Morocco and Western Sahara (US Geological Survey, Mineral commodity summaries 2014). In West Africa there are ample sedimentary deposits of rock phosphate of varying 12 University of Ghana http://ugspace.ug.edu.gh quality (Pieri, 1990, Sanchez et al., 1997). Niger is blessed with large amount of natural rock phosphate reserve in Tahoua region. 2.5.1.1. The soil phosphorus problem Erosion, leaching, crop uptake, crop removal as well as runoff are the major causes of phosphorus lost from the soil. The concentration of P in the tissues of plants varies from 0.1 to 0.5 % on a dry weight basis (Mullins, 2009). When phosphorus is utilized by crop roots or lost via runoff, part of the sparingly soluble P becomes available through the solid soil phase (McDowell et al., 2001). This process is regulated by the equilibrium between the soil adsorption system, soil solution and precipitated P compounds (Sample et al., 1980). Applied P fertilizer changes to low soluble or insoluble forms through a process known as fixation. The fixation of phosphates by Fe, Al, and Ca is among the main factors that cause low phytoavailability (McBeath et al., 2005), because at least 70–90% of P that enters the soil is fixed, making it difficult for plants to absorb and use (Kou et al., 1999; Lei et al., 2004; Liu et al., 2000). Both dissolved and eroded soil particles forms of soil phosphorus may be carried away by water that moves across earth surface and could increase the concentration of bioavailable P in surface water namely rivers, streams, lakes and oceans (Mullins, 2009, Bomans et al., 2005). This will result in eutrophication which is waters overenrichment with mineral elements. The consequences are high production of autotrophs, such as cyanobacteria and algae (Bomans et al., 2005, Correl, 1998, Carpenter et al., 1998). The removal of P through leaching is viewed as less important compared surface runoff because soils, namely phosphorus-deficient subsoils held P very tightly. However, the potential of P leaching may increase in the case of soils with high water tables as well as phosphorus-saturated soils (Mullins, 2009). 13 University of Ghana http://ugspace.ug.edu.gh 2.5.1.2. Forms of phosphorus in soils Phosphorus is a mineral element that is highly immobile in the soil. It reacts with soil elements such as iron (Fe), aluminum (Al) and calcium (Ca) to form stable compounds. It therefore exists most often in non-available forms or in forms that become available only outside of the rhizosphere (Schachtman et al. 1998). Phosphate in the soil is either in organic or mineral forms. Inorganic P is exceedingly reactive. It is present in the lithosphere in more than 150 different mineral forms. These minerals are variable with regard to their solubility and ability to supply the soil solution with P. Mineral P solubility is affected by pH, the concentration of Al, Fe, Ca and magnesium (Mg), the behavior and surface area of soil particles and soil moisture content. On the other hand, organic P is composed of relatively labile phospholipids, nucleic acids, inositols, fulvic acids and humic acids (McDowell et al., 2001). Phosphorus is however, most often present in the organic form in the soil. Richardson (1994) reported that nearly 80 % of the phosphate found in the soil is in the organic form. Phytic acid (inositol hexaphosphate) is usually a principal component of organic phosphate; the other part is made up of inorganic fraction that contains 170 mineral forms of P. About 80 % of P in the soils turns out to be immobile and non-available for plant use as a result of precipitation, adsorption or conversion to the organic form (Holford, 1997). Plants take up P 2- from the soil solution in the form H2PO4- and HPO4 when soil pH is respectively below 7.2 and pH above 7.2. Labile P is available to plants while non-labile P is not. Immobile P is released into soil solution as a result of the decomposition of soil minerals through weathering, but the process is gradual. Acid soils contained iron and aluminum phosphates, while calcareous soils, octacalcium phosphate and apatite are found in neutral soils. P is low available in soil because it moves mainly through diffusion while minerals that are more available move through the soil via bulk flow and diffusion. P diffuses slowly at the rate of 14 University of Ghana http://ugspace.ug.edu.gh -12 -15 2 -1 10 to 10 m s , a zone depleted of P occurs around the roots when plant absorption rates are high (Schachtman et al. 1998). Decomposition of the organic matter by soil microorganisms releases a considerable amount of phosphorus to soil. The mineralization of P by microorganisms linked to organic matter is crucial in preserving P levels in soil solution. 2.5.1.3. Phosphorus availability to plants Available P refers to P pool readily available to plant (Holford, 1997). The form of P most readily available to plants is Pi (McDowell et al., 2000) and most of the times, its concentration do not go beyond 10 µM in soil solutions (Bieleski, 1973). Breeuwsma et al., 1995, cited by McDowell et al., 2000, reported that Dutch scientists extracting soil with acid ammonium oxalate found that the molar concentration of P was in correlation with the average concentration of Al and Fe on a molar basis, and multiplied by 100, gave a degree of P saturation (DPS). Sardans et al. (2006) studied the effect of warming and drought on soil phosphatase activity and soil P availability in a Mediterranean shrubland. Their results revealed that warming without significant changes in soil water content leads to an increase in mineralomass accumulation with subsequent high microbial activity, and the combine action of these two factors results in the reduction of the total and available soil-P contents. However under drought conditions, plant growth and P demand reduce resulting in a rise in Pi availability. Zahedifar et al. (2011) investigated the effect of phosphorus and organic matter on soil-plant phosphorus relationships in spinach and found out that soil P content is higher in all organic matter treatments particularly at 6th weeks after emergence. Furthermore, they realized that P uptake is greater in treatment that received organic matter without P than where P was applied. In some soils organic phosphorus can contribute 50 percent of the available phosphorus. Organic matter application on phosphorus adsorption of three soil parent materials namely basalt, granite, and river alluvial deposits, was studied 15 University of Ghana http://ugspace.ug.edu.gh (Yu et al. 2013), the result revealed that organic fertilizers available P was significantly higher in all treatment groups compared after the addition .The release of soil P to plant roots is influenced by many chemical and biological processes such as soil temperature, texture, aeration and compaction, moisture, pH, mineralogy, crop residues, organic matter, effect of plant roots, effect of mycorrhizae and interaction with other nutrients (Mullins, 2009, Armstrong et al., 1999). Low soil temperature, low pH, moisture stress, low organic matter, limited oxygen supply as well as low mycorrhizae population in the soil negatively affect phosphorus availability to plants. 2.5.1.4. Phosphorus function in plants There are 17 essential mineral elements for plant growth which roles cannot be accomplished by any other element. P is one of these nutrients and is considered as a major nutrient. It is needed by crops in relatively sizable quantities (Armstrong et al. 1999), however, P is often deficient for crop production. P is a major macronutrient for plant and accounts for about 0.2% of a plant‘s dry weight. It is reported to be a key component in molecules such as ATP, phospholipids and nucleic acids (Schachtman et al. 1998). As such, it plays an important role in the storage and transfer of energy in the form of ADP and ATP (adenosine di- and triphosphate) and DPN and TPN (di- and triphosphate) during photosynthesis and respiration. Pi is associated also in the control of key enzyme reactions and in the metabolic pathways regulation (Theodorou and Plaxton, 1993). Phosphorus is essential for plant development and it is part of every living plant cell (Schachtman et al. 1998). According to Grant et al. (2001), plants need an adequate supply of P from the early stage of development for optimum crop production. The highest concentration of P in a mature plant is found in seeds. Metabolism and cell division are high in young cells such as shoots and roots tips, hence it needs P in large amount. P helps in root, seeds and fruits 16 University of Ghana http://ugspace.ug.edu.gh development, it is also involved in flower initiation. P is reported to lower disease incidence and increase the quality of some crops (Uchida, 2000). 2.5.1.5. Phosphorus deficiency symptoms The description of the classic deficiency symptoms for phosphorus or other nutrients in crops can be readily done, but symptoms are not always so clear in the field. Sometimes, other conditions can complicate the appearance causing thereby difficulty in diagnosis. Interactions involving more than one nutrient may also affect appearance of symptoms (Armstrong et al., 1999). Phosphorus deficiency symptoms may vary from one crop to another, but the general symptom is slow, weak and stunted growth. For example while phosphorus deficiency may delay blooming and maturity in soybean, premature senescence of leaves may occur on P-deficient cotton plants later in the season. The appearance of reddish-purple color indicates P deficiency in plants, but sometimes it occurs as a result of sugars accumulation which causes the development of anthocyanin pigments. The development of reddish-purple color may also be a normal plant characteristic. Red-purple color induced by P deficiency can be noticed during early spring on low P sites, but usually this symptom persist only under extrem low P soil conditions. As a mobile element, P can be moved to sites of new growth, leading to the appearance of dark to blue-green coloration on older leaves of some plants. P deficiency may sometimes resemble N deficiency when plants are small, especially under cold temperatures that limit the extension of roots with subsequent reduction in P uptake. When the soil warms, deficiencies may disappear. Delayed maturity as well as poor seed and fruit development occur when crops are grown under low P soil conditions (Armstrong et al. 1999, Uchida, 2000). 2.5.2. Impact of phosphorus on crop performance Soil fertility is essential for optimum crop growth and development as well as the production of adequate food quantities. Undoubtedly, low fertile soils cause low crop productivity 17 University of Ghana http://ugspace.ug.edu.gh (Omotayo and Chukwuka, 2008). The poor crop yield in Africa is associated with low soil fertility in nitrogen as well as low soil availability of phosphorus. Low soil fertility is widely spread across the crop growing areas of sub-Saharan Africa. It is reported that reduction in soil fertility of smallholders‘ farms is the major biophysical cause of low food production in Africa (Sanchez et al., 1997). Low soil fertility especially in N and P has been identified by researchers as one of the main abiotic limiting factor to African agriculture (Sanchez et al., 1997). After N, P is the second most frequently limiting macronutrient for plant growth (Schachman et al., 1998). Phosphorus is the 11th most abundant element in the earth crust, its supply is however limited (Syers et al., 2011). It is among the macronutrients that play a major role in plant growth and development. Owing to P significant impact on crop yield, its availability to plants is crucial for food production. A soil with plant available P content less than 40 mg kg-1 (determined by Bray-1 method) is regarded as P deficient (Mourice and Tryphone, 2012). Adequate P supply leads to increase grain production, high quality crop, greater stalk strength, more root growth and early crop maturity (Beegle and Durst, 2002), thus it is essential for achieving and maintaining food security. Agriculture accounts for about 80 to 90 % of world demand of phosphorus and it is globally the main user of that macronutrient (Tirado and Michelle, 2012). Agricultural production is constrained in many parts of the world by the shortage of available P to plants (Frossard et al., 2009). As stated by Grant et al. (2005), crop production is restricted by limited P supply at the early season. Similar results have been reported on many crops species. On corn, enhanced early-season P supply leads to higher partitioning of dry matter to grain (Gavito and Miller 1998a), higher grain yield (Barry and Miller 1989; Lauzon and Miller 1997). The effect of P supply before 6 weeks of growth on wheat and barley grain yield is much greater than when it is applied later. In early-season, P deficiency results in growth reduction in radish and lettuce (Avnimelech and Scherzer 1971), broadleaved willow (Atkinson and Davidson 1971), 18 University of Ghana http://ugspace.ug.edu.gh wheatgrass (Boatwright and Viets 1966) and a variety of other crops (Crafts-Brandner 1992; Elliott et al. 1997). 2.5.2.1. Impact of phosphorus on legumes and cowpea Legume crops have the ability to fix N from the air, but its optimum development as well as nitrogen fixation depend strongly on soils adequate supply of phosphorus. Apart from P, a good balance of micronutrients (iron, sulfur, and molybdenum), appropriate pH, and good aeration is also necessary (Valenzuela and Smith, 2002). The production of legumes is not constrained to a large extent by nitrogen deficiency in the soil. Rock phosphate application on P deficient soils of West African humid zone has been reported to enhance biomass and N accumulation in C. micans, a root-nodulating N-fixing legume (Somado et al., 2006). Common bean yield is reduced by 60% when it is grown on soil deficient in plant available P (Acosta-Díaz et al., 2009). P deficiency in soil is among the most limiting soil fertility factors to cowpea production (Bationo et al., 2002). Nkaa et al. (2014) studied the effect of phosphorus fertilizer on the growth and yield of cowpea. They found out that P significantly influences cowpea yield and growth parameters. They reported improved plant height, leaf area, number of leaves and number of branches in all the weeks of measurement with P application. Similarly Odundo et al. (2010) while investigating the effect of phosphorus on survival, P accumulation and yield of cowpea varieties across a soil fertility gradient in western Kenya deduced that P application is essential for enhanced dry matter accumulation and yield. Phosphorus positively impacts on grain yield, number and dry weight of nodules, shoot and root dry weight, total biomass, number of pods per plant as well as the total N and P uptake in cowpea (Oladiran et al., 2011, Singh et al., 2011, Asuming et al., 2013). Similar results were reported by Audu et al. (2013) who reported that Sokoto Rock Phosphate (SRP) application raises the quantity of P available to plants in the soil resulting in increment in grain yield and fodder. They then suggested the application of SRP at the rate of 75 kg/ha if 19 University of Ghana http://ugspace.ug.edu.gh the soil pH is below 6.5. Greater amount of nitrogen is fixed by cowpea following P application (Fongi and Begoude, 2014). Saidou et al. (2011) reported that rock phosphate application increased AMF infection of cowpea roots. The application of P at the rate of 30 kg/ha increases dry matter production in cowpea by 74 % in comparison to the control (Odundo et al, 2010), they also noted significant rise in grain yield with 15 kg/ha P supply. Resource poor small scale farmers are now dealing with low crop yields as well as income and food scarcity as a result of low soil available P in most of SSA soils (Verde and Matusso, 2014). Adequate action should thus be taken to prevent food problem and poverty in the near future. 2.5.3. Adaptation of cowpea to low P soil conditions Cowpea is a legume crop that is adapted to harsh climatic conditions such as drought and low soil fertility. The ability of this crop to fix nitrogen from the air allows it to have good performance on low fertile soils. However the lack of available phosphorus in the soil negatively affects the nitrogen fixing capacity of cowpea. The combine effect of N and P deficiencies results in significant reduction of the crop grain and biomass yield. Improvement in yield can be achieved through phosphorus fertilizer application, unfortunately the released P is rapidly bound by Fe and Al oxide, and consequently it becomes unavailable for plant use. Several research findings have revealed variability among crop species and cultivars regarding specific nutrient elements uptake and use efficiency, phosphorus included, as a result of several morphological, physiological and biochemical mechanisms (Hoffland et al., 1989; Raghothama, 1999; Neumann and Römheld, 1999; Akhtar et al., 2007). Large genotypic difference associated with P acquisition and P use efficiency has been detected in many crops such as rice, wheat, clover and maize (Trolove et al., 1996; Romer and Schenk 1998). A major quantitative trait locus (QTL) named Pup1 for P use efficiency was 20 University of Ghana http://ugspace.ug.edu.gh identified in Kasalath, a rice variety (Wissuwa et al., 2002), and near isogenic lines to which this QTL was introgressed have significant yield in constrast to intolerant line nipponbare (Wissuwa, 2005). Yan et al. 1995 reported that legumes vary in their P use efficiency with respect to growth and yield. Genotypic variability has been found in common bean (Phaselous vulgaris) and pigeon pea (Cajanus cajan) regarding their ability to acquire P (Bonser et al., 1996; Subbarao et al., 1997). Evaluations of cowpea lines have been conducted to explore potential genotypic variation for P uptake and P use efficiency. Mahamane et al. (2008) who screened 696 US Core collection and IITA accessions for adaptation to low soil P environments and for response to rock phosphate application found out significant differences in the cowpea lines ability to adjust to P stress deficiency and P acquisition from rock phosphate. Saidou et al. (2011) evaluated more than 100 cowpea accessions obtained from IITA, IAR, Zaria and INRAN on field and greenhouse conditions and reported wide genotypic variability among cowpea genotypes with regards to grain yield, P use efficiency and AMF infection rate as well as shoot-to-root ratio. A study conducted to assess P use efficiency and P uptake variability of 21 pigeon pea varieties using cowpea as control has indicated that pigeon pea and cowpea differ in P uptake strategies (Vesterager et al., 2006). Their findings showed more accumulated P as well as more shoot and root in cowpea compared to pigeon pea. In another study, three groups of cowpea genotypes namely low, moderate and high, were obtained on the basis of Phosphorus Efficiency Index (PEI) (Kugblenu et al., 2014). Genotypes with high PEI are IT03K-351-1, IT00K-901-5, IT93K-452-1 and IT98K-1263 while IT97K-819- 118 has moderate and Soronko has the least. However, when the classification was made on the basis of PEI obtained from PCA and growth potential of shoot dry weight, nodule dry weight and P uptake efficiency at high P, IT93K-452-1, IT98K-1263 and IT03K-351-1 were detected as efficient and responders whereas IT97K-819-118 and Soronko were non 21 University of Ghana http://ugspace.ug.edu.gh efficient and poor responders to P application. Oladiran et al. (2012) screened 10 cowpea genotypes to determine their response to three levels of phosphorus and found out that the line IT97K-414-5 was the most efficient in P uptake. Similarly varietal variation with regard to P uptake from Gafsa rock phosphate was obtained when cowpea lines were assessed in a low P ultisol (Ankomah et al., 1995).Two cowpea lines with high P use efficiency were also reported by Jemo et al. (2006). Saidou et al. (2007) evaluated cowpea accessions both in greenhouse and field conditions on three phosphate levels namely low P, RP and SSP. They reported lines tolerant to low P soil conditions but their performance appeared to be influenced by the environment. Some studies have surprisingly detected that seed size has an influence on P use efficiency in common bean and cowpea. Yan et al. (1995) observed that large-seeded Phaseolous vulgaris lines performed better on low P soils compared to small- seeded varieties. Ojo et al. (2007) reported that cowpea with large grain size have high P uptake than those with small seed size. They crossed IT90K-277-2, a cowpea line with high RP use efficiency and IT89KD-288 which responds negatively to RP to study the genetic basis of RP use efficiency. They noticed that backcross to each parent produces larger seed with more seed P and these backcrosses are more efficient in P uptake than either parent. A study was conducted to examine the response of cowpea genotypes to sub-optimal P supply in alfisols of western Kenya by Gweyi-Onyango (2011). His findings showed that the cowpea line Elanda has greater P use efficiency than Inz since significant difference was detected in Elanda nodule weights but not in Inz under the P treatments (+ P and – P). However, the cowpea genotype Inzeku did not show any response to P application, its yield remains unchanged under + P and –P. Seven cowpea lines were evaluated for P uptake and –1 P use efficiency under three P treatments (0 kg P, 30 kg P ha as triple superphosphate –1 (TSP) and 90 kg P ha as Togo phosphate rock (PR) by Jemo et al. (2006) and significant variation was observed since nodule dry matter, shoot dry matter, grain yield, and P uptake 22 University of Ghana http://ugspace.ug.edu.gh differ significantly among the varieties tested. These authors also affirm that P application significantly augments N2 fixation in cowpea. Eight varieties were screened in greenhouse under no P, low P and normal P treatments (Rothe et al., 2014). Each variety was uprooted at the onset of flower buds, dried overnight at 75°C and roots and shoot biomass weight were recorded. Among the tested varieties, Aloka and Danila were considered as susceptible because their shoot biomass was significantly inferior under low P compared to normal treatment, and their P susceptibility index (PSI) were higher than that of the tolerant varieties Big John, CB46, Golden eye, IT98K-476-8, and IT97K-1069-6. A mapping population was developed using IT98K-476-8 (tolerant) and Aloka (susceptible) for the identification of QTL for low P tolerance (Rothe et al., 2014). 75 SSR markers that were polymorphic on the two parents were run on 120 F6 RILs progenies. The RIL shoot biomass data on low P treatment and the PSI obtained from both low and normal treatmebts were the phenotypic data used. Three SSR markers which are linked to low P tolerance in cowpea were identified from that study. 2.5.4. Mechanisms of adaptation to low P soils conditions Plants require suitable climatic and soils conditions for adequate growth and development. However these environmental factors have been changing over time becoming more harsh and unfavorable for crop production as a result of natural happenings as well as human activities. Consequently, the rainfall is becoming low and erratic and the soils turn out to be deficient in organic matter and some essential mineral nutrients such as nitrogen, phosphorus and potassium. P which is a non-renewable element is among the crucial mineral nutrients deficient in Sub-Saharan Africa agricultural soils. Owing to the role it plays in crop development and yield, P deficiency in agricultural soils is a serious threat to food production. As a way of survival, some plant genotypes have developed mechanisms of adaptation to such conditions. Cowpea varieties that adapt well to low P soil conditions have 23 University of Ghana http://ugspace.ug.edu.gh been reported in literature. Adaptation mechanisms used by plants include root induced processes, root characteristics, mycorrhizal dependency, P uptake kinetics parameters, and rhizosphere processes (Nwoke et al., 2005, Nielsen and Barber, 1978; Nielsen and Shjorring, 1983; Nye and Tinker, 1977, Nielsen, 1993). 2.5.4.1. Root induced processes Inorganic phosphorus reacts with iron (Fe), aluminum (Al) and calcium (Ca) to become fixed, leading to P deficiency in soils. The secretion of organic acids and/or phosphatases and proton release that triggers the change in pH are among the mechanisms used by some plants genotypes to adapt to such conditions. Organic acids contained in root exudates have the ability to dissolve Ca, Fe and Al bound soil P (Dakora and Phillips 2002). Otani et al. (1996) observed that pigeon pea crude root exudates have more power in melting Fe and Al bound P than those of other crops, particularly at 4 and 5 weeks following transfer to the nutrient solution. They further noticed that more malonic, oxalic, and piscidic acid were released from pigeon pea roots compared to those of other crops. The exudation of organic acid from cowpea accessions tested on low P soils was investigated by Pypers et al. (2006). They reported increased exudation amount on low P soils without enhance P uptake due to the low concentration of P in the soil. Likewise organic acid production was determined on soybean and cowpea grown on low P soil by Nwoke et al. (2008). Their findings suggested that root length may be important since only citric acid known to have insignificant correlation with high P uptake in cowpea was detected. Mahamane et al. (2008) assessed the organic acid release of cowpea cultivars grown in hydroponic culture under P-deficient and P-sufficient conditions. Their results indicate that a tricarboxylic acid was the main organic acid secreted by cowpea roots that can be a piscidic acid which Ae et al. (1990) and Ishikawa et al. (2002) reported as the principal root exudates of pigeonpea. The quantity of organic acid released with P application is greater than that obtained without P (Mahamane 24 University of Ghana http://ugspace.ug.edu.gh et al., 2008). This finding do not corroborate with the results of Rengel (2002), Dong et al. (2004), Hoffland (2006), and Pearse et al. (2007) who reported more organic acid secretion under P stress. Another study conducted on soybean by Liao et al. (2006) showed that citrate and oxalate exudation was stimulated by Al toxity and P deficiency respectively while malate secretion was triggered by both treatments. 2.5.4.2. Root morphology Plants partially depend on mineral elements contained in the soils for growth and development. However these elements are not only unevenly distributed in the soils but they have also been depleted as a result of continuous cultivation, wind and water erosion as well as other natural causes and human activities. In such conditions the volume of soils exploited by the plant roots and root hairs may largely determine the amount of nutrients the plant can reach. The exhaustion of phosphate modifies the root system architecture to a large extent leading to superficial root system with more and elongated lateral roots as well as congested root hairs (Peret et al., 2011). They further postulated that Arabidopsis thaliana copes with low P by decreasing primary root growth, improving lateral root formation and growth, and producing longer and more root hairs. Larger root system, root diameter and root hairs are very important root parameters that may play a paramount role in enhanced P uptake in plants. Larger root system results in more root-soil contact which is especially critical for P uptake (Gahoonia et al., 2004). Root diameter determines the soil volume that can be exploited by roots when a given amount of photosynthate was invested (Atkinson, 1991). The geometrical arrangement of root hairs as well as its capability to amplify effective root surface area allow it to improve its efficiency to exploit rhizosphere soil for P (Gahoonia et al., 1997). Root growth, root branching and root hair morphology are essential in order to take up P efficiently. Maize genotypes with high P efficiency have larger root systems, greater P uptake per unit root weight, more nodal roots, nodal root laterals, and 25 University of Ghana http://ugspace.ug.edu.gh greater root hair density of nodal root main axes and first-order laterals than did P inefficient accessions under P deficiency (Bayuelo-Jimenez et al., 2013). Miguel et al. (2013) carry out a study to assess the correlation between the root traits variability and P uptake efficiency in common bean genotypes using radioactive phosphorus 32P. Their results revealed that lines with longer roots have high P uptake than those having short roots. However the exploitation of larger soil volume alone by plants may not confer to them high P uptake. Danila and IT82D-716 have low P uptake despite the exploitation of a relatively high soil volume by their roots and root hairs while IT89KD-349 exploited low volume of soil but pulls out more P (Krasilnikoff et al., 2003). Similarly Kugblenu et al. (2014) found no correlation between cowpea root traits and grain yield per plant under low P in the field experiment. 2.5.4.3. Mycorrhizal dependency The term ‗mycorrhiza‘ is used to designate the relationship between certain non-pathogenic fungi and the roots of higher plants. Rhizobiums are in the soil, but they can only fixed N when they develop a symbiotic association with legumes. They enter the roots cortex by means of invagination and form an infection thread into the root hairs. They cause proliferation of cortical cells resulting in nodules formation on roots. The bacteria living in the nodules convert N into ammonia and make it available to plants for its growth. Dowdle et al. (1980) investigated the mycorrhiza inoculum levels in soils and quantified their contribution to the P nutrition of cowpea. Their results show that vesicular arbuscular mycorrhizae (VAM) improves P as well as Zn, S, K, and Sr uptake. It performs a crucial function in making P available to higher plants, especially in low P environment as it is in the tropics. Eighteen to 24 fold increase in pearl millet root and shoot dry matter was obtained through application of P without VAM treatment while the root and shoot dry matter of sorghum and cowpea relied on the interaction between P and VAM (Bagayoko et 26 University of Ghana http://ugspace.ug.edu.gh al., 2000). Ngwene et al. (2010) investigated the interaction of AMF inoculation and the N03-: NH4+ ratio in making rock phosphate (RP) available to cowpea, but their results failed to provide any evidence that the combination of AM fungi and NO3- at neutal pH helps in making more RP available to plant. Similarly, a study conducted by Ahiabor and Hirata (2003) revealed that mycorrhizae (Glomus etunicatum) treatment did not have any impact on cowpea shoot dry weight whereas it has improved that of groundnut and pigeon pea; but the combine effect of rock phosphate, soluble phosphate and mycorrhizae has enhanced the cowpea total shoot dry matter production and the P content of its vegetative parts. These authors postulated that although the three inputs have acted synergically in improving the cowpea shoot dry weight, mycorrhizae had played a major role since it has not only stimulated additive sporulation responses on cowpea, but it has also increased root ramification in the soils allowing the plant to pull out more sparingly soluble P. Nutrients uptake of two cowpea varieties was enhanced by AM application under drought condition (Neumann and George, 2009). 2.5.4.4. Rhizosphere processes Rhizosphere processes are among the important mechanisms that lead to mineral nutrients availability to plants especially in nutrient stressed environments. Roots secrete H+ or HCO3- (and CO2) that not only triggers pH change but also cause modifications in the redox potential. pH reduction promotes the melting of Ca bound phosphorus thus making more P available to plants. Proton release and organic acid exudation occur at the same time causing acidification of the rhizosphere to liberate sparingly soluble P for plant use (Mahamane et al. 2008). Latati et al. (2014) carried out a study to investigate if cowpea cultivation can make P more available in the rhizosphere to enhance the yield of cereals intercropped with legumes. Their results showed that cowpea cultivation as sole or intercropped with cereals improves the availability of P in the rhizosphere. Under low soil P 27 University of Ghana http://ugspace.ug.edu.gh level, increasing rhizoplane pH and citrate secretion in apical root zone of two cowpea varieties (Espace 10 and Santo Inacio) was observed to react to Al toxicity (Akinrinde et al., 2007). Exploring the kinetics and mechanisms of cowpea root reaction to changes in calcium solution, Blamey et al. (2014) reported that roots respond to variation in Ca solution via impacts on cell wall relaxation of the rhizodermis and outer cortex in the elongation zone. Variation in adaptation to low P soils of three cowpea lines was investigated by Alkama et al. (2009) and they observed that the proton efflux for P-efficient genotype 26- 73 was 43% and 60% greater than for the P-inefficient Danila in hydroaeroponics and in soil respectively, under P deficiency. They then inferred that the P use efficiency variation among the cowpea genotypes under inquiry depend on the nodulated-root proton efflux and respiration in the plant rhizosphere. 28 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE 3.0. Evaluation of farmers knowledge, perceptions on low soil fertility and their cowpea varietal preferences 3.1. Introduction Cowpea production in Niger is affected by many production constraints which include biotic and abiotic stresses. Low soil fertility was identified by farmers as one of the main abiotic factors causing serious threats to cowpea yield in the crop growing areas of Niger (SNV, 2008). The country is populated by over 17 million inhabitants composed up of about 90 % small scale farmers. Unfortunately only about 1/3 of the country‘s land area is suited to crop production; the arable land is therefore continuously cultivated without fallow since no new land is available. As a result, the soil nutrients are almost completely depleted as there is little or no organic matter since crop residues are used for feeding livestock as well as housing and nearly none is left on the field. In addition, chemical fertilizer‘s application is very low as the mean amount applied is about 1 kg/ha due to high cost, unavailability and poor income conditions of the farmers. Improper land use, overgrazing, deforestation, prolonged cultivation of the same field without suitable and required regulations bring about land degradation as well as soil nutrients depletion (Conant et al., 2003; Kebede et al., 2013; Bernoux et al., 1998). Farmers‘ actions contribute to a large extent in the loss of agricultural soils nutrients and acquaintance with their perceptions about soil fertility may be a key factor in designing agricultural soil management strategies. All the actors involved in the protection of soils used for cropping can gain from taking into account local knowledge and farmers‘ vision with respect to soil fertility (Marenya et al, 2008). Farmers‘ awareness of various soil types and its management as well as their ability to assist researchers in soils classification has been reported in many studies (Tabor, 1992; Pawluk et al., 1992; Hecht, 1990). 29 University of Ghana http://ugspace.ug.edu.gh Farmers use soil color, texture, depth, topograghy, drainage and farm distance from home to determine soil fertility (Corbeels et al., 2000; Yeshaneh, 2015). Other indicators of soil fertility according to growers are reduction in crop yield, weed infestation, rocky outcrop, early plant wilting in the growing cycle (Corbeels et al., 2000; Bizabih et al., 2016; Mahler et al., 2015, Marenya et al., 2008) and soil erosion (Kediro, 2015). The emergence of striga in a field is viewed as a sign of low soil fertility by farmers in some areas of Niger (Salifou et al., 2015). Also Zouera et al. (2016) associated infestation of a field by striga with poor soil fertility. Farmers‘ assessment of soil fertility is subjective as it is based on observations and does not involve any scientific test. The combination of farmers‘ assessments with the ones obtained from soil laboratories may give an insight on soil fertility status. Authors have called for finding the correlation between indigenous and scientific soil knowledge for sustainable management of agricultural land resources (WinklerPrins, 1999; Sherwood and Uphoff, 2000). An agreement was detected between the physicochemical analysis and farmers‘ assessment of soil fertility (Yeshaneh, 2015). Also, Marenya et al. (2008), using a univariate regression without controlling covariates, observed a strong significant positive correlation between farmers‘ soil fertility perception and percent soil carbon content; however when the yield is controlled, the relationship becomes non significant. Many technological packages with respect to cropping systems, integrated soil management and improved crop varieties have been developed and disseminated to farmers; however their adoption rate remains very low. In developing countries, small scale farmers, especially in rural areas do not generally readily adopt innovations. Some farmers believe that they are experts in agriculture due to the possession of indigenous knowledge. With regard to crop cultivars, farmers have selected their own varieties, the landraces, which are adapted to their social, economic and environmental conditions as well. Witcombe et al. (1996) stated that 30 University of Ghana http://ugspace.ug.edu.gh farmers of marginal areas in most developing countries keep using landraces since improved varieties which performed better than theirs were not widely disseminated to them. But low adoption may also occur if the new variety is not adapted to farmers‘ conditions and perception. Almekinders and Louwaars (1999) hypothesized that collaboration with farmers‘ right from the stage of goals‘ setting, up to the selection of varieties can foster adoption. However, the conventional breeding top-down decision-making and information flows, where the views of farmers and other stakeholders are not sufficiently considered in varietal development has caused low adoption rate of improved crop cultivars. The conventional package of new varieties and external inputs, although efficient under good environment, did not generally improve the livelihood of farmers in rural areas (Ceccarelli et al, 2009). Participatory plant breeding (PPT) and participatory varietal selection (PVS) have evolved to address the problem of low adoption rate of improved of crop varieties. Ashby (2009) stated that several investigations inferred that PPB and PVS increase farmers‘adoption of formally bred cultivars in marginal areas since these approaches take into account their needs in developing, testing and release of varieties. In cowpea, like any other crop, farmers‘ preferences with respect to varieties vary from one region to another. The important cowpea varietal traits include yield (grain and fodder), phenological cycle, tolerance to biotic and abiotic constraints, seed size, seed color and market value. Usually, farmers go for large white or brown seeds with rough coat in West Africa (Nielson et al., 1993). However, grain yield has always been among the important traits to farmers with respect to cowpea varieties, but nowadays, with the shortening of rainy seasons in Niger caused by climate change, cowpea growers living in the dry areas have abandoned their landraces which are late maturing, high yielding for both grain and fodder and started requesting for extra early varieties. Resource- poor farmers do not have enough means to invest in agricultural inputs such as fertilizers and insecticides, as a result they ask for varieties that can give acceptable yield under biotic or 31 University of Ghana http://ugspace.ug.edu.gh abiotic stresses with no or minimum added inputs. In Niger, cowpea is presently the most important cash legume and farmers have started paying attention to its market value. Preferences of cowpea growers in some localities of the country have been determined through PRA (Salifou et al, 2015) and it turns out that high yield potential, early maturity, white seed color and good taste are the most preferred traits. It is however necessary to determine farmers‘ cowpea varieties preferences in all major cowpea growing areas in order to make specific recommendations for each locality. Hence, the present study was carried out to: - assess farmers‘ knowledge and perception of low soil fertility and its effect on cowpea production; - determine farmers‘ preferences regarding cowpea varieties. 3.2. Materials and methods 3.2 1. Site selection The survey was conducted in the 3 major cowpea growing areas of the country which are Dosso, Maradi and Zinder. In each region, 3 departments were selected on the basis of the importance of cowpea production as well as soil fertility problem. A total of nine departments namely Birni Ngaoure, Loga, Doutchi in Dosso , Guidan Roumdji, Dakoro, Mayahi in Maradi, Matameye, Magaria and Goure in Zinder were selected for the present study. Three villages were sampled per department. However, only two villages were selected in Birnin Ngaoure as well as in Loga as a result of some constraints. The present survey was carried out in 25 villages which were selected on the basis of informations received from extension services. 32 University of Ghana http://ugspace.ug.edu.gh 3.2.2 Sampling method In each village, household heads were gathered at the chief‘s house. Random sampling method namely ballot papers were used to select respondents among household heads present at the meeting. Two hundred and fifty (250) respondents (Table 3.1), 10 in each village, were interviewed. Data on cowpea production, soil fertility knowledge and perception and other production constraints were collected per farm and every respondent was allowed to give information of a maximum of five farms. In each household, data were collected on female farm when available to take into account gender aspect. Table 3.1 Number of respondents per region Region Number Percentage (%) Dosso 70 28.00 Maradi 90 36.00 Zinder 90 36.00 Total 250 100.00 3. 2. 3 Interview guides and data collection Structured, semi-structured questionnaires as well as ranking were the materials used for data collection. While the structured questionnaire was administered to individual household heads, the semi-structured questionnaire and ranking were used in the focus group discussion (FGD). One FGD was organized per village; this approach was used to get general information about agriculture in each area and to list and rank cowpea production constraints. In the case of household questionnaire, data was collected electronically with the help of tablets on which CS Entry 6.1 software was installed, the questionnaire was than uploaded to the tablets. Data were collected on the socio-economic characteristics of respondents, soil nature and constraints to cowpea production. 33 University of Ghana http://ugspace.ug.edu.gh 3. 2.4 Data analysis The data analysis which consists of descriptive statistics and ranking was carried out using Stata 14 software. 3.3. Results 3.3.1. Socio-economic characteristics of respondents The age of respondents ranged from 17 to 90 years. The mean age of respondents across the study areas was 42.56 years. Household heads in Zinder were slightly younger compared to those of Dosso and Maradi. The mean age was 44.23 years for Dosso, 43.23 years for Maradi and 40.17 years for Zinder (Table 3.2). Table 3.2 Average age of respondents (years) Region Mean Range Dosso (n=70) 43.27 55 Maradi (n=90) 44.23 72 Zinder (n=90) 40.17 58 Total (n=250) 42.56 73 With regard to education, the results showed that 72.4% of interviewees had not been to school revealing the low level of formal education in these areas, 17.2 % of respondents had primary education while 9.2 % had secondary (Table 3.3). Formal education was slightly higher in Zinder compared to the other two regions. Most of the respondents who went to school dropped out from primary level. Table 3.3 Formal Education levels of respondents Region Dosso Maradi Zinder Total LEVEL OF EDUCATION N % N % N % N % None 53 75.71 67 74.44 61 67.78 181 72.40 Primary 14 20.00 13 14.44 16 17.78 43 17.20 Secondary 3 4.29 9 10.00 11 12.22 23 9.20 Tertiary 0 0.00 1 1.11 2 2.22 3 1.20 34 University of Ghana http://ugspace.ug.edu.gh The principal occupation of most of the respondents was agriculture. The most important secondary occupations were small business and livestock rearing. 3.3.2. Farms and soils The number of farms per household varied from one to more than five among respondents. Most farms, 71.33 % are located far from homes while 28.1 % of them were very close to villages. The number of backyard farms was significantly lower in Dosso compared to the two other regions (Table 3.4). Table 3.4 Distance of farms from villages Region Far ms Dosso (n=70) Maradi (n=90) Zinder (n=90) Total (n=250) Back yard 19.15 30.75 32.41 28.10 Far from villages 80.75 69.25 67.59 71.87 The mode of acquisition of farms include inheritance, purchasing, gift, borrowing and hiring. However the most important mode was inheritance as its proportion across the study area amounted to 82.23 %. Higher number of respondents inherited their farms in Dosso and Zinder compared to Maradi. The most important mode of farm acquisition after inheritance were purchasing and borrowing with 7.65 and 7.10 % respectively. The proportion of respondents who purchased their farms was greater in Maradi in comparison to Dosso and Zinder. Farm hiring was done mainly in Maradi region. Borrowing was however higher in Dosso and Zinder compared to Maradi. Very few of respondents received their farms from government (Table 3.5). 35 University of Ghana http://ugspace.ug.edu.gh Table 3.5 Mode of farm acquisition Region Mode of acquisition Total Dosso (n=70) Maradi (n=90) Zinder (n=90) (n=250) Inheritance 87.85 74.66 85.42 82.23 Purchase 1.36 15.10 5.09 7.65 Received from 0.92 0.56 1.11 0.86 Government Borrow 9.46 4.45 7.92 7.10 Hire 0.41 5.64 0.46 2.16 Farms usually belong to household heads. However, in every household, women possessed small plots of lands obtained from their families or borrowed/gifted by their husbands. The management of women farms was carried out mainly by men, who are most often their husbands, sons or brothers. The present study showed that only about 4 % of women farms were managed by women themselves. Farmers classified soils as sand, clay and sand-clay. Across the study areas, 68.22 % of respondents claimed that the soils of their farms are made up of sand while those who described their farms soil as sandy-clay and clay were 18.62 and 3.95 % respectively. According to the results, the proportion of sandy soils was higher in Zinder than Maradi and Dosso while that of sandy-clay was high in Dosso (Table 3.6). Table 3.6 Soil texture according to farmers in the study areas Region Soil type Dosso (n=70) Maradi (n=90) Zinder (n=90) Total (n=250) Sand 53.26 67.13 80.93 68.22 Clay 2.57 5.93 3.04 3.95 Sandy-clay 22.94 21.15 12.73 18.62 NA 20.23 5.78 3.29 9.21 NA = proportion of farmers who were able not to give information about the soil texture of their farms. 36 University of Ghana http://ugspace.ug.edu.gh 3.3.3. Cropping systems More than 20 different crops were grown by farmers of the surveyed areas. It includes millet, cowpea, sorghum, groundnut, Bambara nut, sorrel, okra, maize, cassava, sweet potato, tomato, tiger nut, sesame, watermelon. However, the most important and widely cultivated crops were millet, cowpea, sorghum and groundnut. Millet and cowpea were the only crops that adapt to almost all agro-ecological zones of the country. Cowpea was mainly grown in association with millet according the findings of the present study. During the focus group discussion (FGD), in 18 of the 25 villages where the PRA was conducted, farmers affirmed that cowpea was mostly cultivated in association with millet. According to them, mixed cropping is a way of adaptation to shortage of land but it also helps in reducing the risk of total production loss because of its diversification nature. Although mixed cropping was the main cropping system, improved varieties were cultivated in sole cropping since they do not give high grain yield in mixed cropping (Fig. 3.1). 20% 32% 48% Diversification Land shortage Low grain yield PRA survey, 2015) Figure 3-1 Farmers‘ reasons for growing cowpea in association with millet and other crops 37 University of Ghana http://ugspace.ug.edu.gh The results of the FGD showed that both local and improved varieties of cowpea were used in 88 % of the sampled villages while in 12 % of the villages, only local varieties were cultivated. Farmers who grow cowpea in sole cropping system were either improved seeds key distributors and/or those who participated in the demonstration programs or attended capacity building workshops organized by extension services and non-governmental organizations (NGO). 3. 3. 4. Farm size Wide range was observed in terms of farm size among respondents. Farm area per household varied from 0.5 to 53, 0.7 to 54 and 0.3 to 16 ha in Dosso, Maradi and Zinder respectively. The mean farm size per household is 5.58 ha varying from 3.23 to 7.24 across the study area. Our results showed that households in Zinder have significantly less farm land compared to Maradi and Dosso (Table 3.7). 3.3.5. Cowpea production and yield The production and the yield of cowpea reported in the present work concerned only the cropping season of year 2014. The mean production per household ranged from 0 to 2750 kg with a mean of 223.69 kg per household. The mean yield was 59.5 kg/ha and varied from 0 to 870 kg/ha. There were variation among the three regions with regard to both production and yield. Both the production and the yield were higher in Dosso in comparison to the two other regions. The lowest production was observed in Zinder while the lowest yield was recorded in Maradi. Farmers claimed that 2014 was a bad cropping season. They also asserted that the production of cowpea was generally low in recent years; however the production in a normal year is usually higher than that of the 2014. The results suggested that cowpea performed better in Dosso in comparison with the two other regions (Table 3.7) 38 University of Ghana http://ugspace.ug.edu.gh Table 3.7 Area (ha), production (kg), and yield (kg/ha) of cowpea Area (ha) Production (kg) Yield (kg/ha) Region Mean Range Mean Range Min Mean Range Dosso (n=70) 6.47 52.5 331.11 2750.00 79.82 870.00 Maradi (n=90) 7.24 53.3 234.97 2000.00 36.80 166.67 Zinder (n=90) 3.23 15.7 128.85 750.00 66.44 330.77 Mean (n=250) 5.58 53.7 223.69 2750.00 59.52 870.00 Farmers may have some difficulties to adequately estimate yield, they can however easily judge the production. According to majority of them (75.63%), the 2014 cowpea production was low, however 22.11 % affirmed that it was average while only 2.26 % viewed it as good (Table 3.8) Table 3.8 Farmers assessment of their year 2014 cowpea production Region Evaluation of the production Total Dosso (n=70) Maradi (n=90) Zinder (n=90) (n=250) Low 77.84 66.49 83.06 75.63 Medium 21.12 29.85 15.13 22.11 Good 1.04 3.66 1.81 2.26 3.3.6. Production constraints Many production constraints were faced by cowpea growers; the most serious ones which often cause economic yield loss were drought, insects, poor soil fertility, striga, diseases and poor seed quality. The rankings obtained from the focus group discussion showed that drought is the predominant constraint impeding cowpea production in the study areas; it is followed by low soil fertility. Taking into account the overall ranking, other major problems include insects, diseases, poor seed quality and striga (Table 3.9.). 39 University of Ghana http://ugspace.ug.edu.gh Table 3.9 Ranking of cowpea production constraints by farmers during FGD Constraints Score1 Score2 Score3 Score4 Score5 Drought 72.0 4.4 - 4.2 5.0 Low soil fertility 12.0 34.8 13.7 4.2 25.0 Insects 4.0 30.4 50.0 16.7 25.0 Poor seed quality 4.0 13.0 18.2 33.3 20.0 Striga 4.0 - 4.5 16.7 20.0 Diseases 4.0 - - - - Birds - - - 4.1 - Others - 17.4 13.6 20.8 - TOTAL 100 100 100 100 100 Like the FGDs, the results of household questionnaire revealed that farmers were aware of soil fertility problem. The results showed that low soil fertility was the major constraint to cowpea production as it recorded the highest mean percentage of 69.91 % across the survey areas. Other factors that cause sizeable reduction in cowpea production were low rainfall, insect pests and poor field management. The results showed that low rainfall impact on cowpea production was higher in Zinder compared to Maradi and Dosso (Table 3.10). Table 3.10 Farmers reasons for low cowpea production Region Reasons Dosso (n=63) Maradi (n=78) Zinder (n=85) Total (n=226) Low rainfall 9.52 28.21 48.24 30.53 Rainfall distribution 1.59 1.28 4.71 2.65 Flood 1.59 0.00 0.00 0.44 Wind 0.00 0.00 1.18 0.44 Bird 0.00 1.28 0.00 0.44 Insect 4.76 1.28 10.59 5.75 Rodents 0.00 1.28 1.18 0.88 Livestock 1.59 2.56 0.00 1.33 Late onset of rains 3.17 11.54 2.35 5.75 Late weeding 6.35 6.41 0.00 3.98 Poor Field management 9.52 3.85 3.53 5.31 Low soil fertility 87.30 73.08 54.12 69.91 Farm far from village 0.00 0.00 1.18 0.44 40 University of Ghana http://ugspace.ug.edu.gh While assessing the fertility status of their farms, 48.87 % of the respondents claimed that their field soils are poor, the farms are even very poor according to 11.29 % of interviewees. However 39.48 % of interviewed household heads believed that their fields are fertile. The proportion of poor soils was higher in Dosso and Zinder compared to Maradi (Table 3.11). Table 3.11 Assessment of the fertility status of farms by respondents Region Perception on soil Fertility Dosso (n=70) Maradi (n=90) Zinder (n=90) Total (n=250) fertile 32.61 43.97 41.33 39.84 Poor 52.67 43.38 51.41 48.87 Very poor 14.72 12.65 7.26 11.29 Soil erosion is among the factors that bring about the problem of soil fertility. Over 77% of the agricultural soils were eroded according to respondents; about 14% of respondents claimed their soils are even much eroded. Only 9 % of the respondents stated that the soils are not eroded (Table 3.12). Table 3.12 Erosion level of farms according to respondents Region Perception on soil Erosion Dosso (n=70) Maradi (n=90) Zinder (n=90) Total (n=250) No eroded 12.93 11.78 3.27 9.04 Eroded 75.41 70.57 85.48 77.29 Very eroded 11.67 17.65 11.25 13.67 3.3.7. Management of production constraints by farmers Farmers have devised their own ways of dealing with production constraints. To reduce the impact of drought, they used early maturing varieties although this alternative works only in the case of terminal drought. Some farmers planted trees, left crop residues on the field, or/and applied organic and inorganic fertilizer (usually NPK, Urea and DAP) to handle the problem of soil fertility deterioration (Fig. 3.2 and Table 3.13). The overall proportion of respondents that applied fertilizers were 40%; those who used manure, pesticides and 41 University of Ghana http://ugspace.ug.edu.gh improved seeds were 76.8 %, 10 % and 32 % respectively. Among the inputs, manure is the most utilized by respondents. In contrast the use of pesticides was very low. The utilization of improved varieties as revealed by the results was relatively high (32.80 %). The use of agricultural inputs was higher in Maradi in comparison to Dosso and Zinder. Striga was however manually uprooted. With respect to erosion some respondents said that they leave millet straw in order to protect the soil from such degradation. Other more efficient technologies such as wind breaks and stone bunds/lines transferred by extension services, development projects and NGOs were used by some farmers. 4% 4% 35% 57% Plantation of trees Uproot striga Use of early maturing varieties Use of organic and mineral ferti Source: PRA survey, 2015) Figure 3-2 Management of production constraints by farmers (FGD) 42 University of Ghana http://ugspace.ug.edu.gh Table 3.13 Use of agricultural inputs by farmers (Questionnaire) Region Inputs type Dosso (n=70) Maradi (n=90) Zinder (n=90) Total (n=250) Fertilizer 70.00 42.22 14.44 40.00 Manure 75.71 75.56 78.89 76.80 Pesticides 8.57 16.67 4.44 10.00 Cowpea improved variety 28.57 43.33 25.56 32.80 3.3.8. Causes of soil fertility deterioration According to the focus group discussion results, the factors that contributed to soil fertility deterioration include collection of crop residues from farms, lack of fallow or reduction in its duration, wind erosion and lack of fertilizer application. However, collection of crop residues from fields and lack of fallow were the main causes of reduction in soil fertility (Fig. 3.3). 8% 8% 42% 42% Crop residues collection No fallow No fertilizer application Wind erosion Source: PRA survey, 2015) Figure 3-3 Factors causing reduction of soil fertility according to farmers 43 University of Ghana http://ugspace.ug.edu.gh 3. 3.9. Farmers varietal preferences Early maturity and high grain yield potential were the traits preferred by farmers in terms of cowpea variety. Early maturity had a mean score of 2.35 while high grain yield recorded a mean of 2.40 across the study areas. Furthermore a median of 1 and 2 were recorded by early maturity and high grain yield respectively, meaning that at least 50 % of respondents ranked early maturity as first preferred traits and high yield as second (Table 3.14). By giving preference to high grain yield, farmers expressed their preferences for drought, low soil fertility and insect pest tolerance. The remaining traits that include seed size, seed weight, seed coat color, taste, cooking time and market value were of low importance to respondents according to the results of the present work. Table 3.14 Ranking of variety traits by farmers Region Dosso (n=70) Maradi (n=90) Zinder (n=90) Total (n=250) Traits Mean Median Mean Median Mean Median Mean Median Grain size 3.86 5.00 4.10 5.00 4.92 5.00 4.33 5.00 Grain color 4.56 5.00 4.68 5.00 4.91 5.00 4.73 5.00 Grain weight 4.49 5.00 4.80 5.00 4.81 5.00 4.72 5.00 Grain yield 2.86 3.00 2.02 2.00 2.43 2.00 2.40 2.00 Fodder yield 4.29 5.00 4.42 5.00 4.70 5.00 4.48 5.00 Drought tolerance 4.94 5.00 4.09 4.00 3.44 3.00 4.10 4.00 Field Insect tolerance 4.63 5.00 4.64 5.00 4.09 4.00 4.44 5.00 Striga tolerance 4.96 5.00 4.87 5.00 4.91 5.00 4.91 5.00 Poor soil tolerance 4.89 5.00 4.90 5.00 4.78 5.00 4.85 5.00 Early maturity 3.59 5.00 2.48 2.00 1.27 1.00 2.35 1.00 Dual purpose 4.91 5.00 4.61 5.00 4.91 5.00 4.80 5.00 Taste 3.57 4.00 4.90 5.00 4.96 5.00 4.55 5.00 Cooking time 4.97 5.00 5.00 5.00 5.00 5.00 4.99 5.00 Market value 4.64 5.00 4.60 5.00 4.96 5.00 4.74 5.00 Store insect tolerance 3.94 5.00 5.00 5.00 4.91 5.00 4.67 5.00 Early maturing varieties were preferred in Zinder and Maradi, however in Dosso at least 50 % of respondents gave it a rank of 5 which is the last rank. The preference of high yielding 44 University of Ghana http://ugspace.ug.edu.gh varieties cut across all the study areas. Respondents of Maradi and Zinder stated three reasons for preferring early maturing varieties: - Rainy seasons were becoming shorter from one year to another, rainfall hardly goes beyond 2.5 months, so they need varieties that can mature before the rains stopped; - Shortage of food occurred almost every year from the middle of the cropping season, farmers who planted early maturing varieties can get food from their farms earlier and this allowed them to stay on their farms and carry out remaining farm operations; - Towards the end of the rainy season, the price of feedstuffs, especially cowpea, is high on local markets, the earlier the harvest, the more likely to take advantage of the good price. 3.4. Discussions The low level of formal education among respondents reported in the present work corroborated with Republique du Niger, (2001) where it was reported that illiteracy rate among adults in rural areas is 83 %. Education rate in rural areas started rising only in the year 2000 when the government adopted mass education policies. The very poor level of education of respondents revealed by this study is partly due to the fact that young men, especially those with some level of formal education do not have any interest in agriculture and have therefore migrated from the rural to urban areas. The use of improved cowpea varieties along with local cultivars in 18 of the 25 surveyed villages may be attributed to the promotion of ―community seed growers‖ among the farmers by extension services; emergence of private seed companies as well as the establishment of input shops in rural areas. However, most of the released improved varieties were not adopted by farmers; only few improved varieties namely IT90K-372-1-2, KVX-30-309-6G, TN5-78, 45 University of Ghana http://ugspace.ug.edu.gh IT97K499-38, IT97K499-35 and IT99K-573-1-1 were widely used by growers from the eighteen improved varieties published in the national catalogue of plant species and varieties. The cultivation of cowpea in association with millet by majority of our respondents across the nine districts where the survey was conducted is in agreement with the findings of Salifou et al. (2015) who reported that cowpea was mainly intercropped with millet and sorghum in three districts of Niger. It also corroborated with Langyintuo et al. (2003) who found that cowpea was mostly planted in association with millet and sorghum in West and Central Africa. The mean cowpea yield of 59.5 kg/ha reported in the present work was an indication of vey low cowpea yield in farmers field. It was below the range reported in the national statistics where it is stated that cowpea yield in Niger ranged from 68 to 388 kg/ha with a mean of 208 kg/ha from 2000 to 2012 INS (2013). Although the yield remains low, there was an increase in cowpea production in Niger in recent years as a result of expansion of area under cultivation since the crop progressively occupied most of the areas allocated to cash crops, especially legumes. The very low production of cowpea in 2014 revealed by the current study was not consistent with FAOSTAT (2014) where it was reported that 2014 cowpea production in Niger was higher compared to 2013. However, the trends of both production and yield followed an uneven development (zigzag pattern) from one year to another depending on the incidence of one or more major production constraints. The identification of drought and low soil fertility as most important production constraints in the present work was consistent with previous findings in many cowpea producing countries. In Niger, drought, insects, striga, low soil fertility were among the top constraints to cowpea production (SNV, 2008). Similar results were reported by Asante et al. (2001) and Egbadzor et al. (2013) in Ghana, Tignegre, (2010) in Burkina and Adipala et al. (2000) in Uganda. The claim by farmers that striga was associated with low fertile soils is in agreement with Salifou et al. (2015). It also corroborated with Zouera et al. 46 University of Ghana http://ugspace.ug.edu.gh (2016) who declared that soils infested by striga were significantly poorer than those that were not. Hence the spread of striga in agricultural soils of Niger may support the hypothesis that low soil fertility is a widely spread problem in agricultural soils of Niger with its subsequent setbacks to crop production. Agricultural soils of Niger, especially those of upland are deficient in essential nutrients required for optimal plant growth. They contain on average 0.1 to 0.7% soil organic matter, a pH of 4.5 to 6.0, Bray-1 P of 0.4 to 3.4 mg/kg of soil and low available N (Nouri et al., 2017). Low P content in soils, causes stunted growth, delays in flowering and pod setting leading therefore to poor grain and fodder yield. In the present study, farmers were able to assess the fertility status of their farms although their assessment may be subjective since not based on any scientific test. According to Corbeels et al, (2000) and Yeshaneh, (2015), farmers determine soil fertility on the basis of soil color, texture, depth, topography, drainage and farm distance from home. Their findings agreed to some extent with the results of the current study since the respondents believed that back yard farms are richer than those located far from villages because the village waste was dumped on it. However with regards to soil texture, the interviewees could only identify sand, clay and sand-clay soil texture types. They believed that clay soils are more fertile than sandy soils as they grew crops that they think are more demanding in terms of nutrients such as sorghum on clay soils while millet and cowpea were mostly cultivated on upland sandy soils. The predominantly sandy texture of soils in the study areas revealed by the results was consistent with Republique du Niger, (2001) where it was reported that agricultural soils of Niger have a sand proportion varying from 80 to 90 %, 1 to 8 % of clay and 2 to 6 % of silt. The respondents‘ beliefs that fertile soils gave higher production compared to poor soils was consistent with Mahler et al., (2015) who reported that low crop yield is an indicator of poor soil fertility to farmers. 47 University of Ghana http://ugspace.ug.edu.gh Before the advent of crop research in the sub-Saharan African countries, farmers used to choose varieties through mass selection. In the selection of these cultivars called ―landraces‖, farmers make use of their own criteria that may not tally with those of formal plant breeders. Usually farmers select varieties that are adapted to their agro-ecological environments and to their socio-cultural needs and economic conditions. The results of the present study showed that the two most important traits preferred by farmers with respect to cowpea varieties were early maturity and high grain yield potential. Similar results were obtained by Salifou et al. (2015) in Niger, Ansah et al. (2014) in Ghana and Orawu (2007) in Uganda. Early maturating varieties were prioritized by farmers in order to avoid yield loss caused by terminal drought since the rainy season is becoming shorter as a result of climate change. However early maturity alone may not be enough to prevent the negative impact of drought on cowpea production, seedling drought tolerance is necessary as rain distribution within the season is most often erratic especially in the beginning of the season, early maturity should therefore be accompanied with tolerance to seedling drought in order to obtain good harvest. The strong correlation between seedling and terminal drought tolerance in cowpea reported by Watanabe et al. (1997) was crucial information to breeders for the development of drought tolerant cultivars of cowpea. The fact that seed size and seed coat color were not cited among the important traits by our respondents was not consistent with many findings as large seed size and white coat color were reported as farmers preferred traits by Kormawa et al. (2000) in Nigeria, Salifou et al. 2015 in Niger, Tignegre, (2010) in Burkina, Egbadzor et al. (2013) in Ghana and Langyintuo et al. (2003) in the whole West and Central Africa. Large seed size was preferred by consumers in Niger according to Ibro et al. (2011) who studied the effects of physical and chemical characteristics of cowpea on its price in Niger. They further stated that consumers in Niamey were ready to pay more to get brown seed cowpea. Cowpea is presently the most 48 University of Ghana http://ugspace.ug.edu.gh important cash legume crop in Niger; however, only the respondents of Gouré, a department located in the eastern part of Zinder region talked about its market value, they claimed that brown coat seeded cowpea have higher price on the market of Kano (Nigeria) than varieties with white seed color. Ibro et al. (2011) also reported that brown seeded cowpea produced in Zinder was exported to Nigeria; as a result, white seeds were dominant on markets of Zinder, as well as Maradi and Niamey. 3.5. Conclusion In the present work, farmers have identified drought and low soil fertility as the most critical constraints to cowpea production in their various communities. They are aware that their soils are poor and they know that low fertile soils impact negatively on cowpea production. Early maturity and high yield potential are the preferred traits with regard to cowpea varieties. However low soil fertility, especially low levels of P not only leads to low yield, but also delays flower and pod setting in cowpea. Therefore, farmers are left with only two options: it is either to apply fertilizer (inorganic or organic) or to grow tolerant varieties. Owing to the high price as well as low availability of fertilizers in rural markets and the extremely low income condition of small scale farmers in Niger, the best alternative is to use varieties that are tolerant to low P. With the objective of identifying and/or developing cowpea varieties tolerant to low P soil conditions, this thesis research is thus a welcome development. 49 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR 4.0. Genetic variability and heritability of cowpea traits under low P soil conditions 4.1. Introduction In Niger, poor soil fertility, especially low phosphorus, is one of the serious problems dwindling cowpea production and yield as well. The use of fertilizers and/or organic manure in order to solve the problem is impeded by farmers‘ socio-economic conditions as well as poor availability and high price of mineral fertilizers on local markets. The cheap and environmental friendly alternative is the development of cultivars with high grain and fodder yield under low soil P conditions. However the successful development of elite cultivars depends on the crop germplasm diversity (Baudoin and Maréchal 1985). Although cowpea is a self-pollinating crop species, there is a large variation in its wild as well as cultivated germsplam that can be explored in breeding through genetic manipulations (Ng and Maréchal, 1985) for the development of better performing varieties. Breeders have made considerable progress in cowpea breeding by exploiting the crop‘s genetic diversity through the use of both phenotypic and genetic markers in the identification of sources of genes for tolerance/resistance to various abiotic and biotic stresses. Significant genetic variability have been reported for traits associated with resistance/tolerance to drought (Fatokun et al., 2012, Mafakheri et al., 2017), striga (Kamara et al., 2008; Omoigui et al., 2007; Tchiagam et al., 2010), aphids (Souleymane et al., 2013) and low P (Mahamane et al., 2008; Vesterager et al., 2006; Krasilnikoff et al., 2003; Oladiran et al., 2012; Alkama et al., 2009; Jemo et al., 2006; Saidou et al., 2007; Ojo et al., 2007) . Selection of superior genotypes requires not only large variation in the germplasm with regard to the traits under consideration, but also high heritability. Omoigui et al. (2006) reported a moderate-to-large genetic variance and heritability in some cowpea reproductive 50 University of Ghana http://ugspace.ug.edu.gh traits. They also noted substantial genetic variability regarding shoot weight, 100-seed weight, days to flowering and plant height. The same authors noted high heritability estimates for 100-seed weight, duration of reproductive phase, days to first flower appearance, maturity, and harvest index. Similarly, Gerrano et al. (2015) collected data on 16 phenotypic traits while screening 25 cowpea cultivars in South Africa and observed large significant differences for all the traits; however they noticed that phenotypic variance is larger in comparison to genetic variance with respect to all quantitative traits. They also reported high heritability for all the traits except for the number of branches. Several other authors including Inuwa et al.(2012); Animasaun et al. (2015), Ajayi et al. (2014); Damarany, (1994); Nwosu et al. (2013); Fonji et al. (2015); Dias et al. (2016); Ubi et al. (2001); Manggoel et al. (2012); Uguru, (1995) have reported significant genetic differences as well as high heritability estimates for cowpea traits under various environmental conditions. Significant genetic variability has been reported extensively in literature with respect to cowpea yield components under low P soil conditions. However, limited information is available on heritability. Ansah et al. (2014) screened cowpea accessions under two P treatments (low and high) and reported high heritability for total root length, lateral root numbers, root dry weight, nodule dry weight, shoot dry weight, total dry weight, shoot P content, P utilization efficiency. However, their conclusion was based on shoot and root biomass, no heritability estimates for the reproductive traits as well as yield and its components were determined in their work. A breeding program for the development of high grain and/or fodder yielding cowpea varieties under low P soil environment was initiated as P deficiency is wide spread in upland agricultural soils where cowpea is most often grown in Niger (Nouri et al., 2017). To this end we collected farmers‘ varieties from different cowpea growing areas of the country with the 51 University of Ghana http://ugspace.ug.edu.gh aim of quantifying the degree of their genetic diversity as well as estimating the heritability of yield related traits under low soil P conditions. 4.2. Materials and methods The experiments were carried out at the International Crops Research Institute for the Semi- Arid Tropics (ICRISAT), Sadore station, Niger. The research station lies on latitude 13°14'N and longitude 2°16'E and is located at an altitude of 235 m above sea level in Sahelian zone where the rainy season is between June and September. The mean annual rainfall between 1983 and 2015 is 556.92 mm with the peak being in August (ICRISAT unpublished). Sadore is situated 45 Km south of Niamey where the mean winter (December, January, and February) temperature is 25.3°C and mean spring (March, April and May) temperature is 32.7°C. The first trial was conducted in the rainy season while the second took place in the dry season. 4.2.1. Rainy season trial The trial was carried out on plot C8 which is low in total P and N according to the soil maps produced by the institute in 2011 (Figure 4.1 and 4.2). Figure 4-1 Distribution of total P in the ICRISAT sahelian center Sadore 2011 52 University of Ghana http://ugspace.ug.edu.gh (Source: ICRISAT Sadore farm unit) Figure 4-2. Total N distribution in the ICRISAT sahelian center Sadore 2011. (Source: ICRISAT Sadore farm unit) Plot C8 has been on fallow since 2013. Soil samples were collected from the plot to determine the P level and as well as some nutrients composition. It was divided into 3 blocks and soil samples were collected at 0-20 and 20-40 cm depths from each. Results of the soil analysis showed that the site was low in available P (6.42 ppm) although the level has increased during the fallow period. The soil was also acidic and poor in nitrogen (N), potassium (K), sodium (Na), organic matter (OM), magnesium (Mg) and exchangeable acidity (EA); the calcium (Ca) content was however medium (Table 4.1) 53 University of Ghana http://ugspace.ug.edu.gh Table 4.1. Some chemical properties of plot C8 soil at ICRISAT Sadore Soil properties 0-20 20-40 Mean pH 5.60 5.17 5.39 Ca 6.10 5.63 5.87 Mg 0.13 0.07 0.10 Na 0.14 0.13 0.14 K 0.07 0.17 0.12 AE 0.14 0.32 0.23 Available P 6.18 6.65 6.42 C 0.23 0.16 0.20 OM 0.39 0.26 0.33 N 0.02 0.02 0.02 The weather data covering the period of the trial were recorded by the ICRISAT Sadore weather station. Data include information about rainfall, temperature, relative humidity, solar radiation, soil temperature and wind (Table 4.2). The rains lasted for 4 months (June, July, August and September). The total rainfall was 572.4 mm. The total monthly rainfall ranged from 85 mm in June to 202.5 mm in August. The relative humidity (min and max) was lower at the beginning and end of the season and higher in the middle of the season with the peak in August. Although there was not much variation in air temperature between months, June and October had the highest means with 36.4 and 38.1°C respectively. The two months also had the highest evaporation. Solar radiation data was obtained for August, September and October only and the maximum value was recorded in October. The wind speed was high in the beginning of the season, decreased in the middle and rose towards the end. 54 University of Ghana http://ugspace.ug.edu.gh Table 4.2. Weather data of 2016 rainy season at ICRISAT Sadore, Niger Month Rain AirT_Min AirTC_Max RH_Max RH_Min SOLAR ST05-07 ST05-13 ST10-07 ST10-13 WIND Evap. June 85 26.1 36.4 73.3 47.7 29.4 44.5 30.5 37.7 4.6 6.98 July 186.6 24.4 33.4 87.9 63.3 27.2 39.5 28.1 34.4 3.3 4.5 Aug 202.5 23.5 32.3 92.5 67.8 19.8 26.1 37.3 27 32.8 2.8 4.2 Sept 98.3 23.4 33.8 90.5 59.5 19.6 26.9 41.8 28 36.2 2.1 4.2 Oct 0 23.4 38.1 75.7 34.4 21.3 27.2 50.1 30.8 38.6 2.4 6.3 Evap. = evaporation, Min TP = minimum temperature, Max TP = maximum temperature, HUM-07= min relative humidity, HUM-13 = maximum relative humidity, SOLAR = solar radiation, ST05-07 = min soil temperature at 5 cm depth, ST05-13 = max soil temperature at 5 cm depth, ST10-07 = min soil temperature at 5 cm depth, ST10-13 = max soil temperature at 10 cm depth. Units: rain (mm); AirT_Min, AirT_Max, AirT_Min, ST05-07, ST05-13, ST10-07, ST10-13 (°C); RH_Min and RH_Max (%); Solar (W/m^2); Wind (meters/second); Evap (mmday-1). 55 University of Ghana http://ugspace.ug.edu.gh The tested cowpea germplasm lines were made up of 150 accessions that were planted on th June 30 , 2016. The germplasm included 22 improved varieties, six wild relatives and 122 local cultivars. The improved lines were composed of 16 from IITA, four from ICRISAT and two from INRAN. The local accessions consisted of five from ICRISAT, Sadore genebank, three from Burkina, one from United States of America (USA) and 114 from farmers. The latter were collected during a survey in the cowpea growing areas of Niger Republic in 2015. The farmers ‗varieties collection survey was conducted in 52 villages of the 3 major cowpea production regions of the country which are Maradi, Dosso and Zinder. Seeds samples were collected from 15 villages in Maradi, 19 in Zinder and 18 in Dosso. These villages were visited for the collection based on information obtained from extension services as well as farmers. The sites of the collection are indicated on the Niger Republic map (Figure 4.3). Figure 4-3. Distribution of sites where local cowpea varieties were collected 56 University of Ghana http://ugspace.ug.edu.gh Before planting, the field was ploughed; small trees mostly Guera senegalensis were uprooted while big ones were slashed to avoid shading. The residues of both trees and grass were removed from the field to reduce soil organic matter content. The genotypes were planted in a 10 x 15 alpha lattice design with three replications. The blocks were 14 m long and 6 m wide, and aligned from 1 to 15 in each replication. The distance between blocks was 2 m. The replications were also separated from one another by a distance of 2 m. The lines were planted each in two rows of 6 m length with a spacing of 0.5 m intra and 1 m inter rows. Each plot contained 24 hills. Borders were planted 1 m apart from the first row on each side of the field. Seeds were sterilized with Thioral (25 % Hyperchlor, 25 % Thirame) prior to planting. Four seeds were planted per hill and thinned to 2 plants at 21 days after planting (DAP). After emergence, a guard was hired for three weeks to prevent birds and squirrels from uprooting the plants to feed on cotyledons. Only one weeding was done at three weeks after planting. Eight different types of insecticides were applied through14 treatments from 4 to 14 weeks DAP to control a variety of insects. The insects included grasshoppers, aphids, th Clavigralla spp, and caterpillar pod borer. Insecticide application was halted on October, 6 2016, when we realized that it was no more efficient. Despite all these insecticide treatments, two of the insects, namely Clavigralla tomentosicollis and pod borer were not controlled successfully. Destructive sampling method was used to collect plant samples at eight weeks after planting (WAP) for determination of shoot dry weight (SDWt). Plants were randomly harvested from six hills in each plot i.e. two hills at the beginning, two in the middle and two at the end. Plant samples were dried under room temperature and weighed using an electronic scale. Plant height was recorded also at the time of plant sample collection for SDWt determination. 57 University of Ghana http://ugspace.ug.edu.gh Clavigralla tomentosicollis attack started around eight WAP when early and medium maturing varieties started setting pods. An assessment of tolerance was carried out on 63 DAP using a one to four rating scale with 1= Resistant (No insect found on the plants and no pod was attacked or destroyed), 2 = Tolerant (the plant is invaded by insects, but the majority (above 70 %) of the pods were not destroyed), 3= Susceptible (the pods were destroyed, except few), 4 = very susceptible (All the pods were destroyed and dropped, the peduncle of the plants bore no pod at all). Data were also collected on days to flowering (DAF), days to 50% maturity (MAT50), and grain yield (Gyield) per plant. The collected data were subjected to analysis of variance using Breeding View 1.5.0.16982 statistical software. The software runs two analyses automatically. In the first analysis, genotypes were treated as random effects to estimate genetic variance components and heritability (Equation (1)) (random terms underlined): yij = µ + blockj + Genotypei + errorij, (1) 2 The model includes a fixed intercept µ, and the random effects of blockj ∼ N(0,σ b), 2 2 genotypei ∼ N(0,σ g)), and errorij ∼ N(0,σ ). The estimated variance components are used in the second analysis, where genotypes were moved to the fixed part of the model. This model with random genotypic effects was used to obtain Best Linear Unbiased Predictions (BLUP) of the genotypes. In the second analysis, genotypes were treated as fixed effects, so the model was changed to Equation (2) y ij = µ + blockj + Genotypei + errorij (2) 58 University of Ghana http://ugspace.ug.edu.gh In this model, the variance components other than the error were not estimated, but the ones obtained from model1 were used. They were used to obtain adjusted means or Best Linear Unbiased Estimates (BLUE), which will be used in GxE. The phenotypic variance (VP), the phenotypic coefficient of variation (PCV), the genetic coefficient of variation (GCV) were estimated using the formulae below 2 VP = σ g + MSe 2 GCV = (√σ g /X) 100 2 PCV = (√σ p /X) 100 2 Where σ g = genotypic variance; MSe = mean squares of error; X is the grand mean for the 2 measured traits, and σ p = phenotypic variance. Principal component analyses (PCA) and correlation among traits were also carried out. 4.2.2. Dry season experiment There were no irrigation facilities on the site that hosted the rainy season trial; the dry season test was hence conducted on the A3 experimental plot. According to previous soil tests, A3 was poor in P. The site had been on fallow for two years. Soil tests revealed that the field is acidic, low in available P, Carbon, Ca, K as well as Cation Exchange Capacity (CEC) (Table 4.3). Table 4.3. Chemical soil characteristics of the minor season trial site Soil properties 0-20cm 20-40cm Mean pH H2O (1/2.5) 4.49 4.49 4.49 pH KCl (1/2.5) 4.25 4.27 4.26 Carbon 0.23 0.18 0.24 P total (ppm) 10.53 9.47 10.00 P assimilable (ppm) 5.62 4.78 5.20 Ca (meq/100) 2.63 2.50 2.57 K (meq/100) 0.10 0.10 0.10 CEC (meq/100) 5.00 6.00 5.50 59 University of Ghana http://ugspace.ug.edu.gh The trial was conducted under irrigation. The germplasm was composed of two hundred (200) genotypes which consist of 135 already tested in the previous trial and 65 other accessions. The latter were made up of two lines from University of Maradi, one wild cowpea relative; eight most widely adapted improved varieties and 54 local lines collected from farmers in October 2016 from the regions of Tillaberi and Dosso in Niger. th Planting took place on February 4 , 2017 after the field was ploughed and harrowed. The experimental design was a 10 X 20 alpha lattice design with 3 replications. Seed treatment, sowing and thinning were carried out as in the above trial. Seedlings that failed to emerge were supplied 2 weeks after sowing. Each block measured 3 m long and 14 m wide. Every line was planted on two rows of 3 m with the same spacing as the rainy season experiment. The blocks were separated by 1.5 m while the replications were 4 m apart. Border rows were planted 1 m from replications. The plants were watered twice a week. The field was sprayed twice with insecticide to control aphids. The data collection and analysis were carried out the same as the first trial. However, in addition to the previous experiment, fodder yield (Fyield), harvest index (HI) and hundred seed weight (HSWt) were recorded. 4.3. Results 4.3.1. Genetic variation In the rainy season, highly significant differences were detected for days to flowering (DAF), days to 50 % maturity (MAT50), plant height (PHt) and grain yield (Gyield) at p < 0.001 while pod sucking bug (PSB) was significant at at p = 0.001. There were no significant differences in shoot dry weight (SDWt). The largest variance was observed for plant height while the smallest was in pod sucking bugs (PSB) damage score (Table 4.4). 60 University of Ghana http://ugspace.ug.edu.gh Table 4.4 Mean square and variance of genotypes and blocks for traits in the rainy season Mean square Traits Genotypes Block Variance DAF 146.45*** 2.38 174.00 MAT50 49.64*** 0.21 48.00 PHt 925.5*** 761.6** 2604.00 SDWt 14.2 79.3** 225.90 Gyield 78.9*** 37.5* 285.60 PSB 0.06*** 0.21*** 0.43 *significance at 0.05, **significant at 0.01 ***significance at ≤ 0.001 probability level, DAF =days to flowering, MAT50 = days to 50 % maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, PSB = pod sucking bugs. The mean, range and median for the traits are given for Table 4.5 below. Table 4.5 Summary statistics for traits in the rainy season Traits Mean Range Median DAF 57.00 64.00 52.00 MAT50 83.00 28.00 81.00 PH 150.8 279.00 152.00 SDWt 30.90 101.80 27.50 Gyield 29.77 103.70 28.01 PSB 3.42 2.00 4.00 DAF =days to flowering, MAT50 = days to 50 % maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, PSB = pod sucking bugs. The mean number of days to flowering was 57 varying from 36 to 100. The genotypes that flowered early were G5, G22 and G49 with a mean of 40.48, 41.23 and 41.99 days after planting, respectively. G139, G60, G55 which flowered respectively 96.89, 93.98 and 92.94 days after planting were the last genotypes to set flowers. Days to 50 % maturity ranged from 77 to 105 days after planting with an average of 83. G100, G3, G2 were among the earliest maturing varieties while G39, G50 and G41 were part of those that matured late. Plant height ranged from 10 to 289 cm with a mean of 150.8. The median plant height of 152 cm revealed that the germplasm evaluated in the present study were mostly made up of prostrate to semi- prostrate genotypes. The results showed also that there was substantial genetic diversity among the lines under investigation with regards to shoot dry weight per plant. The minimum shoot dry weight was 7.91g while the maximum was 109.7 g. The accessions with highest dry 61 University of Ghana http://ugspace.ug.edu.gh plant weight were G134 (63.51 g), G85 (58.48 g) and G99 (54.57 g). Similarly wide variation among tested lines was detected with respect to grain yield which ranged from 0.73 to 104.38 g per plant. The mean grain yield was 29.77 g and the superior genotypes were G118, G62, G150 and G116 with 67.62, 60.02, 56.64 and 55.46 g per plant respectively. The plants were attacked by pod sucking bugs (PSBs) throughout the podding stage . In an effort to control the insects, several treatments using a variety of insecticides were carried out; the applied chemicals were however able to slow down the insects development only for a few weeks. The insects population build up was very high towards the end of the season despites the treatments resulting in a complet yield loss of late maturing varieties. The scoring results showed that all the genotypes were susceptible to PSBs, except Dan Missira and Yawa 04 which expressed some moderate level of tolerance . In the results of the dry season experiment, high significant differences at p <0.001 were detected for all the traits except the shoot dry weight. The blocks were also significantly different for days to 50 % maturity, plant height, shoot dry weight and grain yield (Table 4.6). Table 4.6 Mean square and variance of genotypes and blocks for traits in the dry season Mean square Traits Genotypes Block Variance DAF 10.56*** 1.34 25.63 MAT50 6.34*** 1.18** 16.11 PHt 61.90*** 56.9*** 406.3 Shoot dry weight 0.21 11.59*** 43.13 Gyield 42.70*** 51.1*** 224.2 Fyield 1101.20*** 212.7 1866.00 HI 0.05*** 0.008 0.0909 HSWt 4.93*** 0.00 7.67 *significance at 0.05, **significant at 0.01 ***significance at ≤ 0.001 probability level, DAF =days to flowering, MAT50 = days to 50 maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, Fyield = fodder yield, HI = harvest index, HSWt = hundred seed weight. 62 University of Ghana http://ugspace.ug.edu.gh Traits with large total variance values were fodder yield, plant height and grain yield in descending order. In terms of coefficient of variation, fodder (62.63 %) and grain (54.5%) yield were also the traits where high estimates were observed. Other traits with high CV values were shoot dry weight and harvest index with 51.70 and 60.23 % respectively. Hundred seed weight got the lowest estimates of both CV and variance. The total variance revealed that largest variability among genotypes resides in the fodder and to some extent in grain yield. The median showed that at least half of the lines under investigation started flowering at 53 DAP. The number of days to flowering ranged from 42 to 65 days after planting and the mean was 52.59 (Table 4.7). The varieties that flowered early were G205, G49 and G20 while G201, G84 and G208 flowered late. Table 4.7 Summary statistics for traits in the dry season Trait Mean Range Median SDWt 12.70 51.83 11.86 DAF 52.59 23.00 53.00 MAT50 72.99 30.00 72.00 PHt 45.10 144.20 40.67 Gyield 27.46 122.30 25.82 Fyield 68.98 279.10 60.00 HI 0.50 1.65 0.48 HSWt 16.68 26.78 16.58 DAF =days to flowering, MAT50 = days to 50 maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, Fyield = fodder yield, HI = harvest index, HSWt = hundred seed weight Even though no significant differences were detected among accessions regarding SDWt, large variation was observed as it ranged from 1.38 to 53.20 g. Genotypes with higher shoot dry weight per plant were G211, G190 and G70 which were all local varieties. Taking into account the median of 72 days after planting for days to 50 % maturity, the results revealed that majority of the lines belong to the medium class of maturity. MAT50 ranged from 55 to 85 DAP, thus showing the presence of some early and late maturing varieties in the tested 63 University of Ghana http://ugspace.ug.edu.gh germplasm. Plant height ranged from 10.80 to 155.00 cm, however the mean was 45.10. The minimum grain yield per plant was 0.314 while the maximum was 122.30 g. The mean grain yield per plant was 27.46 g and the most promising accessions were G118, G108 and G150. Large variability was also detected in the case of fodder yield; it had the largest range of 279.10 and varied from 10.91 to 290.00 g per plant. Genotypes with high fodder yield include G164, G197, G118 and G27. Harvest index (HI), a parameter that helps to classify varieties as source of grain, fodder or dual purpose showed considerable variation in this study. It ranged from 0.003 to 1.65. Varieties with high HI were classified as grain source; those with intermediate were dual purpose while the ones having low HI were forage source. According to our results, G205, G146 and G20 can be used for grain production while G164, G197 and G202 could serve as excellent source of fodder. In terms of dual purpose cultivars, G118 had an HI of 0.61 and was among superior lines with respect to both grain and fodder, thus it was a well indicated candidate for such uses. The results regarding hundred seed weight showed that the germplasm under investigation was composed of small, medium and large seed size lines since it varied from 3.67 to 30.45 g. However, most of them were of medium size seeds as shown by a mean value of 16.68 g as well as a median value of 16.58 g. Comparison of the results obtained from the two studies revealed that lower means were recorded in the dry season as compared to the rainy season for all the traits (Table 4.8). Table 4.8 Summary statistics for traits in rainy and dry seasons MEAN MEDIAN RANGE VARIANCE CV(%) Traits RS DS RS DS RS DS RS DS RS DS DAF 57.00 52.59 52.00 53.00 64.00 23.00 174.00 25.63 23.00 9.635 MAT50 83.00 72.99 81.00 72.00 28.00 30.00 48.00 16.11 8.00 5.50 PHt 150.8 45.10 152.00 40.67 279.00 144.20 2604.00 406.3 33.84 44.69 SDWt 30.90 12.70 27.50 11.86 101.80 51.83 225.90 43.13 48.64 51.70 Gyield 29.77 27.46 28.01 25.82 103.70 122.30 285.60 224.2 56.76 54.5 RS = rainy season, DS = dry season 64 University of Ghana http://ugspace.ug.edu.gh The difference between the two seasons was larger with regards to mean plant height (150.8 cm in rainy season and 45.10 cm in dry season) and shoot dry weight (30.90 and 12.70 g for rainy and dry seasons, respectively). Three types of genotypes were detected regarding the mean days to flowering stability in the two seasons: - Genotypes whose number of days to flowering remained approximatively the same in both seasons. Almost all the early maturing varieties belonged to this group - Genotypes that flowered early in the rainy season and relatively late in the minor season; an example was G5 which had the mean DAF of 40.48 in the rainy season and 57.96 days in the dry season. - Genotypes that flowered late (≥ 69 DAP) in the rainy season but relatively early in the dry season (< 57 DAP); this group is composed of late maturing varieties. Examples are given in Table 4.9 below. The observed difference ranged from 18.03 for G34 to 43.42 days for G141. Table 4.9 Days to flowering of photosensitive varieties in both seasons CODE DESIGNATION DS RS DIF G34 Bougou Bougou 57.95 75.97 18.03 G74 Baadare Karanguiya 53.04 71.58 18.54 G41 Kwaama Maiki 49.92 69.40 19.49 G78 Lakade Maiki 53.53 80.14 26.60 G120 Lakade karanguiya 49.71 77.31 27.59 G27 Hannoun Mareni Shalla 49.73 78.39 28.66 G94 Sabon Kafi sababba sata 56.26 87.88 31.61 G83 Tanout marche 54.27 91.46 37.19 G141 Kolalla Dan arbain 48.39 91.81 43.42 DS = minor season; RS = rainy season; DIF = difference between the 2 mean days to flowering In terms of MAT50, the difference between means of rainy season and dry season was narrow because in the rainy season, all the late maturing varieties were not able to reach 50 % maturity as a result of PSB attack. 65 University of Ghana http://ugspace.ug.edu.gh Some genotypes consistently produced high grain yield in both environments (Table 4.10). G131 and G11 performed even better in the minor season. These superior and stable cultivars can therefore be used as varieties and/or sources of genes in breeding programs. Table 4.10 Superior genotypes with regards to grain yield in both seasons CODE DESIGNATION DRY SEASON RAINY SEASON G118 Botsotsuwa 57.22 67.62 G150 Jiraini Oubanzaoura 53.50 56.64 G149 Waken chaibou 42.31 52.97 G131 Dan Saley 45.67 34.08 G11 Local chical 43.66 34.68 4.3.2. Variance components and heritability The estimates of genetic variance, phenotypic variance, genetic coefficient of variation (GCV), phenotypic coefficient of variation (PCV) and heritability for the rainy season are shown in Table 4.11. The residual term variance was greater than the genetic variance in plant height, shoot dry weight, grain yield and pod sucking bugs. In contrast genetic variance was higher than that of the error in days to flowering and days to 50 % maturity. The highest values of genetic variance were recorded in plant height and days to flowering. The GCV and PCV ranged from 1.30 to 24.77 and 2.50 to 35.40 respectively. Heritability estimates varied from 92.51 % in DAF to 21.05 % in SDWt. MAT50, PHt, and Gyield recorded 85.02, 70.56 and 53.05 % respectively as heritability values. 66 University of Ghana http://ugspace.ug.edu.gh Table 4.11 Estimates of variance components and broad-sense heritability in the main season 2 2 2 Traits σ2e σ g σ p GCV PCV h DAF 25.75 146.45 172.2 16.03 17.38 92.51 MAT50 8.333 49.64 57.97 7.73 8.36 85.02 PHt 964.7 925.5 1890.2 24.77 35.4 70.56 SDWt 131.9 14.2 146.1 6.78 8.9 21.05 Gyield 165.4 78.9 244.3 16.28 28.65 53.05 PSB 0.156 0.06 0.21 1.30 2.50 31.59 DAF =days to flowering, MAT50 = days to 50 % maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, PSB = pod sucking bugs. In the dry season, the partition of variance into its different components showed that the residual term variance was considerably greater than the genetic variance for DAF, MAT50, PHt, Gyield and SDWt traits. This resulted in a significantly high phenotype variance observed in these traits (Table 4.12). In contrast, the genetic variance was substantially higher than that of the error term in the case of fodder yield and in hundred seed weight. Harvest index, fodder yield, hundred seed weight, days to 50 % maturity and days to flowering recorded broad-sense heritability estimates of 80.72, 80.6, 71.94, 70.88 and 70.75 % respectively. GCV and PCV were also calculated and the results are given in Table 4.12. GCV values ranged from 1.29 for dry shoot weight to 39.96 % in fodder yield while the estimates of PCV varied from 4.13 % for harvest index to 49.76 % for fodder yield. Grain yield recorded 12.46 as GCV and 23.78 as PCV. 67 University of Ghana http://ugspace.ug.edu.gh Table 4.12 Estimates of variance components and broad-sense heritability in the dry season σ 2 2 2Traits 2e σ g σ p GCV PCV h DAF 12.15 10.56 22.71 4.48 6.57 70.75 MAT50 6.715 6.34 13.05 2.95 4.23 70.88 PHt 229.6 61.90 291.50 11.72 25.42 38.18 SDWt 24.26 0.21 24.47 1.29 13.88 1.88 Gyield 112.6 42.70 155.30 12.46 23.78 48.53 Fyield 607.00 1101.20 1708.20 39.96 49.76 80.6 HI 0.0324 0.05 0.09 3.26 4.13 80.72 HSWt 3.05 4.93 7.98 5.44 6.92 71.94 DAF =days to flowering, MAT50 = days to 50 % maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, PSB = pod sucking bugs, Fyield = fodder yield, HI = harvest index, HSWt = hundred seed weight. Taking into account, the results of the two seasons, higher estimates of heritability, genetic variance, phenotypic variance as well as GCV and PCV were obtained in the rainy season. 4.3.3. Correlation between traits The results of correlation analysis for the rainy season experiment revealed the existence of both positive and negative relationships among traits (Table 4.13). Grain yield was positively correlated with shoot dry weight, plant height and shoot color ; it was however negatively correlated with days to flowering and days to 50 % maturity. Shoot dry weight had the highest positive correlation with grain yield. It was however surprising to notice that SDWt was inversely related to plant height. The most high positive correlation was found between days to flowering and days to 50% maturity traits confirming that early flowering leads to early maturity. Similarly a positive correlation was observed between MAT50 and PSB indicating that early maturing varieties were likely to escape PSB attack compared to late maturing because they may reach maturity before the build up of the insect population. 68 University of Ghana http://ugspace.ug.edu.gh Table 4.13 Genetic correlation coefficient among traits in the rainy season Biomass 1 - COLOR 2 –0.01 - DAF 3 0.04 –0.14 - Height 4 –0.02 –0.04 –0.016 - MAT50 5 –0.07 –0.17 0.60 0.00 - PSB 6 0.02 0.04 –0.04 0.01 0.01 - Yield 7 0.23 0.13 –0.26 0.17 –0.35 0.10 - 1 2 3 4 5 6 7 DAF =days to flowering, MAT50 = days to 50 maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, PSB = pod sucking bugs. The dry season trial correlation results showed that grain yield was positively correlated with all the traits except hundred seed weight. Its positive relationship was only strong with SDWt (0.57) and moderate with harvest index (0.35) (Table 4.14). The most important positive correlation was observed between DAF and MAT50 (0.72). Fodder yield was positively associated with SDWt, plant height and MAT50, the relationship was however higher with SDWt, MAT50 and plant height in ascending order. On the other hand strong negative correlation was detected between fodder yield and harvest index. Table 4.14 Genetic correlation between traits in the dry season SDWt 1 - DAF 2 –0.12 - Fyield 3 0.14 0.18 - HSDWt 4 0.06 –0.33 –0.10 HI 5 0.24 –0.46 –0.69 0.22 MAT50 6 –0.10 0.72 0.30 –0.23 –0.40 - PHt 7 0.48 0.04 0.40 0.03 –0.32 –0.12 - Gyield 8 0.57 0.10 0.10 –0.11 0.354 0.11 0.16 - 1 2 3 4 5 6 7 8 DAF =days to flowering, MAT50 = days to 50 maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, Fyield = fodder yield, HI = harvest index, HSWt = hundred seed weight. 69 University of Ghana http://ugspace.ug.edu.gh 4.3.4. Principal component Analysis (PCA) In the rainy season trial, the two first components of the PCA biplot explained 92.35 % of the observed variation with PC-1 accounting for 82.74 % and PC-2 for 9.61 %. The first component was highly associated with plant height while the second was related to grain yield and shoot dry weight (Table 4.15). Table 4.15 Principal components analysis for the rainy season trial Traits 1 2 3 4 5 SDWt 0.12 0.37 0.92 -0.01 0.05 DAF -0.04 -0.16 0.05 0.77 0.61 Height 0.99 -0.12 -0.08 0.02 0.02 MAT50 0.00 -0.15 0.11 0.59 -0.79 Gyield 0.08 0.89 -0.36 0.25 -0.04 PCA 82.74 9.61 5.10 2.19 0.36 Eigenvalues 2695.6 313.1 166.0 71.4 11.7 DAF =days to flowering, MAT50 = days to 50 maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, PCA = principal component axis In the dry season trial, 86.71 % of the variation was explained by the two first axis of PCA with 69.77 % for PC-1 and 16.94 % for PC-2. PC-1 had a strong positive correlation with fodder yield whereas the PC-2 was highly associated with plant height and grain yield (Table 4.16). Table 4.16 Principal components analysis for the dry season trial Traits 1 2 3 4 5 SDWT 0.06 0.24 -0.08 0.01 0.86 DAF 0.00 -0.07 0.00 0.78 -0.16 Fyield 0.96 -0.29 0.03 -0.05 0.00 HI 0.00 0.01 -0.01 -0.01 0.00 HSWt -0.01 0.02 0.02 -0.17 0.28 MAT50 0.03 -0.07 0.02 0.60 0.32 PHt 0.23 0.80 0.53 0.07 -0.16 Gyield 0.17 0.46 -0.85 0.06 -0.16 PCA 69.77 16.94 9.35 2.03 0.87 Eigenvalues 1140.2 276.8 152.8 33.3 14.3 DAF =days to flowering, MAT50 = days to 50 maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, Fyield = fodder yield, HI = harvest index, HSWt = hundred seed weight, PCA = principal component axis. 70 University of Ghana http://ugspace.ug.edu.gh 4.4. Discussion According to Hazelton and Murphy (2007), P content in soil less than 12 ppm are low. With a mean P level of 6.42 and 5.20 in the rainy and dry seasons respectively, the sites used in current study were accordingly very low in P; thus qualifying it for hosting the trials to determine genetic variability and heritability of cowpea traits under low P soil conditions. Despite the different environmental conditions, high significant difference was detected in both seasons with regards to days to flowering, days to 50 % maturity, plant height and grain yield. The present findings corroborated with Omoigui et al. (2005), Ansah et al. (2014) and Gerrano et al. (2015) who reported significant variability in all measured traits. However the present study results, disagreed with theirs with respect to shoot dry weight no significant differences were observed among genotypes in both seasons concerning that trait. SDWt data were recorded at 8 WAP for all the genotypes, Rothe et al. (2014) collected SDWt data at flowering time of each cultivar and detected highly significant difference among lines; this is an indication that their method may be more appropriate for investigating genotype divesrsity with regards to that trait. Mean days to flowering of 52.59 in the dry season and 57 days in the rainy season were significantly higher than the range of 36 to 42 for days to 50 % flowering reported by Ishiyaku and Singh (2003) in Nigeria. Lower means were also reported by Gerrano et al. (2015) in South Africa, Nagalakshmi et al. (2010) in India, Cabbinah et al. (2011) in Ghana and Ehlers and Hall (1997) in USA. The higher values observed in the current study could be attributed to the presence of large number of local varieties in the screened germplasm which were known to be generally highly photosensitive and/or late maturing. It can also be due to the effect of the environmental factors since Hadley et al. (1983) stated that temperature and photoperiodism have an influence on time to flowering. With an overall mean of 78 days after planting for days to 50 % maturity, findings of the present study revealed that most of the 71 University of Ghana http://ugspace.ug.edu.gh genotypes tested belonged to the medium class of maturity according to Dugje et al. (2009) maturity classification. The mean plant height of 150.8 cm obtained in the rainy season was in agreement with Nwosu et al. (2013) who reported a range of 134.29 cm at Ibadan and 159.73 cm at Keffi in Nigeria. However significantly lower plant height (41.5 cm) was recorded in the dry season and this could be a sign of genotype by environment interaction. Similar plant height values were revealed by the study of Animasaun et al. (2015) who stated that plant height ranged from 40.5 to 68.59 cm among 10 cowpea cultivars. Large variation was observed between the means of the two seasons with regards to shoot dry weight per plant as considerably higher values were obtained in the rainy season when environmental conditions seemed to be more favorable for legume crops production. Despite the low available P of the sites, some genotypes were able to produce very high grain yield per plant, expressing therefore high level of tolerance. The best cultivar recorded a mean yield of 67.62 g in the rainy season and 57.22 g in the dry season. Slightly lower values were reported by Nwosu et al. (2013) while evaluating the genetic variability, heritability and genetic advance in cowpea cultivars. However our results were not in agreement with Ojo et al. (2007) who got 29.73 g as mean grain yield per plant for the P tolerant parent when studying the genetics of P utilization in tropical cowpea. Similarly Gerrano et al. (2015) stated that the best genotype in their study yielded 29.9 g grain per plant. From the results, we observed that genotypes with high fodder yielding potential during the dry season existed within the germplasm under evaluation. A very narrow variation was observed regarding the mean grain yield per plant between the two seasons. Taking into account the non-significant difference between grain yield mean per plant in the two seasons, we can conclude that the minor season also provides a condusive environment for cowpea production provided irrigation facilities are available. Despites the weather conditions that made prostrate and semi-prostrate to become semi-erect and semi-erect to become erect, some genotypes were not affected and produced high amount 72 University of Ghana http://ugspace.ug.edu.gh of fodder. These genotypes which included G164, G197 and G27 can be used for forage production under irrigation or via the flood recession cropping to solve the problem of feed shortage that occurs most often during the dry season. The partition of variance into its different components showed that the variabilty due to the genetic constitution of genotypes varied among seasons as well as traits. Despite the fact that more accessions were involved in the minor season experiment, lower genetic variance was observed in that season compared to the main season for all the traits, thus the expression of the genetic diversity of genotypes may be restricted by the weather conditions prevailing in the dry season. For example despites the addition of 50 lines in the dry season to the 150 evaluated during the major season, the genetic variance detected in the rainy season with respect to days to flowering and plant height was respectively 14 and 15 times greater than that observed in the dry season. The greater residual term variance in comparison to genetic variance observed in plant height, shoot dry weight and grain yield in both seasons indicated that environmental factors contribution to the variability detected in the present work with regard to these traits was sizeable. These results corroborated with Ajayi et al. (2014) and Inuwa et al. (2012). However, Uguru (1995) obtained a slightly greater genetic to residual variance for grain yield while investigating the association between heritability and variability of yield and yield components in vegetable cowpea. Similarly higher genetic variance than error variance was reported by Manggoel et al. (2012) with respect to days to 50 % flowering and grain yield. Their findings contrasted with those of the current study regarding grain yield but agreed with it concerning days to flowering and hundred seed weight. Genetic constitution of the tested genotypes has played a major role in the variability revealed by the present study with regard to fodder yield, hundred seed weight and harvest index since the genetic variance was greater than that of the error term. 73 University of Ghana http://ugspace.ug.edu.gh GCV and PCV estimates were considered as high, medium or low when their values are respectively > 20, between 10 and 20 and < 10 according to Deshmukh et al. (1986). Accordingly, our results showed that only plant height in the rainy season and fodder yield in the dry season recorded large GCV while the traits with high PCV include days to flowering, plant height and grain yield. Both large GVC and PCV were reported by Manggoel et al. (2012), Gerrano et al. (2015) with regard to grain yield, however in the present work only large PCV was detected for that very trait. Sufficient variation with respect to plant height, days to flowering, fodder and grain yield warranting selection existed in the tested genotypes taking into account their GCV and PCV values obtained in the present work. Heritability estimates for a trait determine the extent to which it can be transmitted from parents to offspring. According to Burton and Devane (1953) the progress that can be achieved through selection is determined jointly by heritability and GCV estimates. Pramoda and Gangaprasad (2007) classified heritability estimates ≥ 80 % as very high, 60 to 79 % as moderately high, 40 to 59 % as medium and < 40 as low. According to their classification, the findings of the present study indicated that very large heritability values were obtained for days to flowering, fodder yield and harvest index; moderately high estimates were recorded for hundred seed weight, days to 50 % maturity while moderate values were observed for grain yield. Similar results were reported by Idahosa et al. (2010), Omoigui et al. (2006), Gerrano et al. (2015) and Manggoel et al. (2012). The present findings were however not in agreement with Gerrano et al. (2015) regarding grain yield who reported moderately high heritability. In contrast, Inuwa et al. (2012) reported very low heritability with respect to days to 50 % flowering and days to 90 % maturity. From the current results, fodder yield combined both large GCV as well as very high heritability; it could therefore be improved easily and efficiently through selection. Selection of superior genotypes with regard to grain yield is also possible since the trait recorded medium GCV and heritability values. 74 University of Ghana http://ugspace.ug.edu.gh 4.5. Conclusion The results of the present study revealed that high diversity with regard to yield and yield related traits exists in the evaluated germplasm, hence selection could be carried successfully. High to medium heritability was also detected for yield and its components. However both the diversity and heritability were higher in the rainy season compared to the dry season. In the rainy season, grain yield was positively correlated with shoot dry weight and plant height; it was however negatively correlated with days to flowering and days to 50 % maturity. The dry season trial correlation results showed that grain yield was positively correlated with all the traits except hundred seed weight. Cowpea cultivars with high grain, fodder and/or both were identified and can be used as varieties or as parents in improvement programs. 75 University of Ghana http://ugspace.ug.edu.gh CHAPTER FIVE 5.0. Assessment of genotype by environment interactions for yield and its components under low P conditions 5.1. Introduction The increasing cowpea production observed in recent years in Niger was mainly due to expansion of cultivation area rather than yield improvement. Now, new cropping land is scarce as population growth rate is high (Pender et al., 2008). Consequently, the mean cropping area per household has been reducing resulting in the reduction of the fallow period. Resource poor farmers are left with no choice than using marginal lands and/or continuous cultivation of the fields which soils are predominantly sandy and deficient in organic matter, nitrogen (N) and phosphorus (P), leading to a steady depletion of the meager soil nutrients. About 15 kg/ha of N, 2 kg/ha of P and 11 kg/ha of potassium (K) were mined from agricultural soils every year in Niger (Buerkert and Hiernaux, 1998). However, fertilizer application country wide is very low; on average basis, less than 1 kg/ha of mineral fertilizers was applied by farmers in Niger (Pender et al., 2008). As a nitrogen fixing legume, cowpea can overcome N deficiency in soils; however Traore (1974) stated that crops respond minimally to N when P requirements are not met. In some soils, P may exist but not in the form that it is available to crops. However some genotypes have the ability of making sparingly soluble P available in its rooting zone through mechanisms such as root induced processes, root characteristics, mycorrhizal dependency, P uptake kinetics parameters, and rhizosphere processes (Nielsen and Barber, 1978; Nielsen and Shjorring, 1983; Nye and Tinker, 1977). Cowpea cultivars with high P use efficiency or tolerant to low P soil conditions have been reported extensively in literature. However most of these reports were based on studies conducted under screen house conditions. As a result, the yield stability of P efficient or tolerant genotypes remains largely under investigated. The success of a breeding program is most often dependent on genotype by environment interaction. 76 University of Ghana http://ugspace.ug.edu.gh Multilocational evalutation allows an unbiased estimation of the yield and also helps in predicting the performance of a cultivar under predictable and unpredictable weather conditions, thereby allowing the identification and selection of stable and high yielding genotypes (Kamdi, 2001). Yield is a trait highly influenced by environment, thus varieties with high yield over a large number and various environments are scarce. Adewale et al. (2010) asserted that breeders recommendation of varieties to different locations is influenced by G x E interactions, therefore when the interaction is large, specific recommendations are made to environments. Stability of yield and its components in cowpea were assessed under twelve environments in Egypt (El-Shaieny et al., 2015). They detected significant G x E for all the characters except in pod length, hundred seed weight and pod weight per plant. Nevertheless, they were able to identify some stable genotypes such as Dokii 331 and Cream 12. G x E interaction with regard to flowering date, pod maturity and grain yield in cowpea was also reported by Aliyu and Makinde (2016) when phenotyping 21 cowpea breeding lines for seed yield and yield components in the southern Guinea Savanna of Nigeria. However, a significant environmental effect on soybean genotypes with respect to seed yield per plant in the humid southeast parts of Nigeria was detected by Nwofia et al. (2016). But they noticed some level of G x E in TGx 1448-2E and TGx 1485-1D genotypes with respect to number of pods per plant. Amara and Suale (1996) evaluated cowpea cultivars in multi-environments for P use efficiency in Sierra Leone and observed large G x E interaction in terms of yield. They identified two lines (IT86D-1010 and IT86D-716) with high performance, however these varieties were not stable over all the test environments as they produced poor yield in low rainfall areas. Also some genotypes with high P use efficiency were selected for use in 77 University of Ghana http://ugspace.ug.edu.gh improvement programs by Abaidoo et al., 2017 following a two years evaluation in the Northern Guinea savanna of Nigeria. Soils in Niger, across the agro-ecological zones, are low and variable in P content, in such situation, organic and mineral fertilizer application may help in obtaining high yield, but development of cultivars with high P use efficiency or tolerant to deficient P soils may undoubtedly provide a better alternative to resource poor farmers. In the present study, local cowpea varieties were evaluated for tolerance to low P soil conditions in four environments with the aims of: - assessing of G x E interactions for yield and its components - evaluating and selecting stable and high yielding cultivars for recommendations to growers - identifying superior genotypes for use in varietal improvement programs 5.2. Materials and methods 5.2.1. Experimental sites The trials were conducted at the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Sadore and Institut National de la Recherche Agronomique du Niger (INRAN), Maradi, Niger. Sadore lies on latitude 13°14'N and longitude 2°16'E at an altitude of 235 m above sea level in Sahelian zone where the rainy season is between June and September. The mean annual rainfall between 1983 and 2015 is 556.92 mm with the peak being in August (ICRISAT unpublished). Sadore is located 45 km south of Niamey where the mean winter (December, January, and February) temperature is 25.3°C and mean spring (March, April and May) temperature is 32.7°C. 78 University of Ghana http://ugspace.ug.edu.gh The ICRISAT‘s experimental fields used were C8 and A3. These sites were low in P and N according to the 2011 soil map of the statition (see Fig. 4.1 and 4.2 in Chapter four). Also the analysis of soil samples collected at 0 – 20 and 20 – 40 cm from the two sites confirmed the low P content of their soils (Table 4.1 and 4.4 in Chapter four). Data on rainfall, temperature, relative humidity and solar radiation were obtained from Sadore ICRISAT‘s weather station. The environmental factors varied only slightly between 2016 and 2017 rainy seasons (Figure 5.1). 140 120 1 RS_2016 100 RS_2017 80 DS_2017 60 1= monthly rain, 40 rain duration is 4 months 20 0 Figure 5-1 Weather data of ICRISAT Sadore for 2016 and 2017 rainy seasons However, relative humidity, solar radiation and rainfall differed significantly between the rainy (June, July, August and September) and dry (February, March, April and May) seasons (Figure 5. 2). Rainfall and relative humidity were also variable within the rainy season; they were low at the beginning of the season, reached the peak towards the middle and started dropping from the end of August. 79 University of Ghana http://ugspace.ug.edu.gh 250 120 200 100 80 150 60 100 40 50 20 0 0 Feb, March, April, May = Minor season, June, July, Aug, Sept= Main season Rain_mm_Tot AirTC_Max AirTC_Min RH_Max RH_Min Figure 5-2 Variation of weather factors between rainy and dry season Maradi is located in the southern part of Niger republic at latitude 15°26‘ N, longitude 8°33‘ E and altitude of 383 m above sea level. Its Soil is 90 % sand, 8 % clay, and 2 % silt with a pH of 6.5. No confirmed information was available about soil P content of the study site. However it is P deficient according to the station management. Mean rainfall from 2010 to 2016 was 517.18 mm. Minimal and maximal mean temperatures were 21.6 and 35.8 from 2010 to 2014. 5.2.2. Germplasm The study was conducted in four different trials during 2016 and 2017. The number of genotypes tested varied from one trial to another. The first trial (E1) took place in the rainy season of 2016 when 150 lines were evaluated. The second trial (E2) was carried using 200 genotypes in the dry season of 2017. The third (E3) and fourth (E4) trials were conducted during the 2017 rainy season at Sadore and Maradi respectively. In the last two trials, 28 80 Rainfall RH, Temp University of Ghana http://ugspace.ug.edu.gh genotypes made up of promising genotypes selected from the results of the two first experiments together with some widely adopted improved varieties by farmers (Table 5. 1). Fiften promising genotypes that cut across the 4 trials were used in the assessment of genotype by environment interaction. Table 5.1 Cowpea varieties used for the trials at Sadore and Maradi during 2017 rainy season No ENTRY_CODE VARIETY NAME TYPE ORIGIN 1 206 KVX30-309 Improved INERA (Burkina Faso) 2 209 TN5-78 Improved (INRAN) Niger 3 187 NELOC-187 Local Niger 4 165 NELOC-165 Local Niger 5 33 Dan batsatsa Local Niger 6 1 Gorom Local Local Burkina Faso 7 131 Dan Saley Kornaka Local Niger 8 43 Farin wake Kantari Local Niger 9 149 Waken chaibou Local Niger 10 62 Dan Baouchi Dan Saga Local Niger 11 173 NELOC-173 Local Niger 12 108 Tamalalo Araouraye Local Niger 13 48 Dan Gawouna Gourgouzou Local Niger 14 176 NELOC-176 Local Niger 15 146 Jan Pass Ismihou Local Niger 16 96 Dan Raha Local Niger 17 11 Local chical Local Niger 18 32 Matarka Nawa Issufuri Local Niger 19 118 Botsotsuwa Ishirnawa Local Niger 20 150 Jiraini Ouban zaoura Jaja Local Niger 21 116 Yagaou Dan saga Local Niger 22 98 NDIAMBOUR Local Niger 23 197 NELOC-197 Local Niger 24 115 Bahaouche Mai Tsakoni Local Niger 25 205 IT90K-372-1-2 Improved IITA 26 101 Karadua Garam Tsofua Local Niger 27 164 NELOC-164 Local Niger 28 125 Bonbohi Fabirdji Local Niger 81 University of Ghana http://ugspace.ug.edu.gh 5.2.3. Method th The first experiment (E1) was planted on 30 June in a 10 x 15 alpha lattice design with three replications. The blocks were 14 m long and 6 m wide. The distance between blocks was 2 m. The replications were also separated from one another by a distance of 2 m. The lines were planted each in two rows of 6 m length with a spacing of 0.5 m intra and 1 m inter rows. Each plot contained 24 hills. The second experiment (E2) was carried out under irrigation. The planting took place on th February 4 , 2017. The experimental design was a 10 X 20 alpha lattice design with 3 replications. Supply of missing hill was carried out two weeks later. Each block measured 3 m long and 14 m wide. Each line was planted on two rows of 3 m with the same spacing as the first experiment. The blocks were separated by 1.5 m while the replications were 4 m apart. th th The third (E3) and fourth (E4) trials were planted on 4 July Sadore and 20 June at Maradi in the 2017 rainy season. The design was a 4 X 7 alpha lattice with 3 replications. The plot 2 area was 16 m . The spacing consisted of 0.8 and 0.5 m inter row and intra plants respectively. For all the trials, seeds were sterilized with Thioral (25 % Hyperchlor, 25 % Thirame) prior to planting. Four seeds were planted per hill. At 21 days after planting (DAP), the plants were thinned to 2 per hill for E1 and E2 and 1 per hill for E3 and E4. Border lines were planted 1m apart from the first row on each side of the experiment. Weeds and insects were controlled as necessary. 5.2.4. Data collection Data were collected on both agronomic, yield and yield related traits which included the following: 82 University of Ghana http://ugspace.ug.edu.gh - shoot dry weight (SDWt) : Plants were uprooted from four hills in each plot at 8 weeks after planting, dried at room temperature and weighted ; - plant height (PHt) : The tallest two plants selected randomly per plot were measured using a tape measure at 8 weeks after planting; - days to flowering (DAF): number of days from planting to flowering; - days to 50 % maturity (Mat50): number of days from planting to 50 % maturity; - grain yield per plant (Gyield): dry seed weight per plot divided by the number of plants per plot; - fodder yield per plant (Fyield) : dry fodder weight per plot divided by the number of plants per plot; - pod yield (Pyield) : dry pod weight per plot divided by the number of plants per plot; - harvest index (HI):Grain yield/biological yield x 100; - hundred seed weight (HSWt), - pod length (Plength), the length of two to three normal dry pods per plot was measured with a ruler; - number of seeds per pod (NSPd): the number of seeds contained in two to three normal dry pods per plot was determined; - Threshing percentage (Thresh): the weight of dry grains per plot over the weight of dry pods per plot x 100; - The field at Maradi was highly infested by striga. The genotypes were scored for tolerance/susceptibility using a rating scale adapted from Singh and Emechebe (1997). 83 University of Ghana http://ugspace.ug.edu.gh The striga density (DS) which is expressed as the number of emerged striga shoots per plot over the surface area of the plot was calculated. The data on fodder yield and hundred seed weight were collected at three environments (E2, E3 and E4) while that on pod yield, pod length, seed number per pod and threshing percentage were measured at two environments ( E3 and E4). 5.2.5. Data analysis Data were subjected to the combined analysis of variance using SAS 9.4. AMMI which has on one hand, an additive component accounting for genotype and environment main effects and on the other, a multiplicative terms that models G x E was performed to study the patterns of G x E and to determine differential adaptation of genotypes to environments. The model is as follow: Y𝑖𝑗 = 𝜇 +𝐺𝑖 𝐸𝑗 k + +∑ 𝝀k=1 𝒌Y + 𝜖𝑖𝑗 GGE (G plus GxE) biplots which is similar to AMMI except that the multiplicative terms include the genotypic main effect in addition to GE was run to: - directly identify the best performing genotypes, - determine the mega environments (group of environment with same best performing genotypes). AMMI analysis was not performed for traits that were measured at only two environments. Combined anova and AMMI analysis of variance tables for each trait were attached to the appendix. 84 University of Ghana http://ugspace.ug.edu.gh Percentage contribution of IPCA1, IPCA2 and the residual to G x E variation was calculated by dividing the sums of square of each of them by that of the interaction times 100. It was however calculated only when both IPCAs were significant. AMMI sum of squares was partitioned into its different components in order to determine the contribution of genotypes, environment, G x E and the error to the total variation observed in each trait. The contribution of each factor was determined by dividing its sum of squares by the total sum of squares and then multiplies the result by 100. AMMI stability values were calculated using the formula below: ASV(i thGenotype)  (((SSPCA1/ SSPCA2)*PCA1score )2  (PCA2score )2 )1/ 2i i ASV: AMMI stability value SSPCA1: sum of squares of PCA1 SSPCA1: sum of squares of PCA2 th PCA1score: PCA1 score for the i genotype th PCA1score: PCA1 score for the i genotype th PCA2score: PCA2 score for the i genotype GGE biplot analysis was carried out to determine the mega-environments as well as the winning genotypes per mega-environment using Genstat 12 statistical software. The winning genotype for a mega-environment is the vertex cultivar at the intersection of the two polygon sides whose perpendicular lines form the boundary of that environment. 85 University of Ghana http://ugspace.ug.edu.gh Test environment with longer vectors are the ones that discriminate the genotypes better; if the environment has a short vector, it means the performance of all the genotypes is similar within that environment; the difference between cultivars cannot therefore be detected. The representativeness of the environments is determined according to the angle formed by the environment and the average environment coordination (AEC), tests environments that have small angle with AEC are more representative than the ones having large angle. The results from the data collected on striga infestation were not shown since the attack occurred only at one site. 5.3. Results Combined anova as well AMMI analysis results showed significant differences (p < 0.01, p < 0.001) among genotypes for days to flowering, plant height, days to 50 % maturity, pod length, pod, grain and fodder yield and harvest index. No significant differences were however detected for shoot dry weight and seed number per pod. Environments were also significantly different (p < 0.01, p < 0.001, p = 0.05) for all the traits except for pod length and seed number per pod. Moreover, significant (p < 0.01, p < 0.001) genotype by environment interactions (G x E) were detected for days to flowering, days to 50 % maturity, pod yield per plant, fodder yield per plant, harvest index and threshing percentage (Tables 5.2 and 5.3). However AMMI PCA components (IPCA1 and IPCA2) were significant only for days to flowering, days to 50 % maturity and fodder yield. Cumulative percentage of PCA axis accounted for 93.66 % and 84.51% of the total GEI variation in days to flowering and days to 50 % maturity respectively. But, in the case of fodder yield, the whole GEI variation was due to the PCA components with the residual having zero contribution. 86 University of Ghana http://ugspace.ug.edu.gh . Table 5.2 Level of significance of factors for the traits under study (combined analysis) SOV DAF MAT50 PHt SDWt Gyield Snber Plength Pyield Fyield HI Thresh% ** *** *** ** *** Rep(Env) 34.99 29.41 7358.22 749.59 824.82 3.32 3.44 1641.60 973.34 0.039 86.20 ** * Block(Rep) 3.02 17.21 2854.71 291.78 367.33 3.53 2.12 122 1261.63 0.013 48.92 *** *** *** *** *** * *** Env 690.45 1650.48 113156.36 35446.00 5959.83 2.76 0.3 607.84 6006.90 0.036 48.46 *** *** ** *** ** *** *** *** Genotype 113.18 56.81 1832.53 131.36 722.86 2.86 7.65 756.53 6641.93 0.086 134.72 *** *** ** *** ** * Env*Genotype 70.17 48.86 984.13 163.00 291.94 4.58 3.06 472.73 1932.27 0.054 174.37 Error 10.21 15.04 742.03 142.14 44.68 2.28 2.58 161.93 577.66 0.024 83.37 *significance at 0.05, **significant at 0.01 ***significance at ≤ 0.001 probability level, DAF =days to flowering, MAT50 = days to 50 % maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, Fyield = fodder yield, HI = harvest index,Snber = number of seeds per pod, Plength = pod length, Pyield = pod yield, Thresh% = threshing percentage 87 University of Ghana http://ugspace.ug.edu.gh Table 5.3 Mean squares of factors for traits (AMMI analysis) SOV DAF MAT50 PHt SDWt Gyield Fyield *** *** *** *** *** *** Treatments 117.00 158.70 7492.00 2035.00 801.00 3902.00 *** *** ** *** *** Genotypes 127.30 65.00 2096.00 151.00 861.00 7689.00 *** *** *** *** *** ** Environments 691.00 2121.40 124134.00 37097.00 7500.00 6007.00 ** *** *** ** Block 34.10 26.50 5763.00 616.00 847.00 851.00 *** *** *** Interactions 72.50 49.80 959.00 158.00 302.00 1869.00 *** *** * *** IPCA 143.00 86.70 1232.00 264.00 399.00 2393.00 *** * * IPCA 40.30 27.20 1130.00 124.00 242.00 1269.00 Residuals 16.10 27.00 395.00 58.00 243.00 Error 9.80 15.20 859.00 151.00 255.00 646.00 Total 46.20 62.60 3246.00 787.00 447.00 1725.00 *significance at 0.05. **significant at 0.01 ***significance at ≤ 0.001 probability level, DAF =days to flowering, MAT50 = days to 50 % maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yieldFyield = fodder yield. 88 University of Ghana http://ugspace.ug.edu.gh The partitioning of the AMMI sums of squares into its different components revealed the contribution of every treatment factor as well as the error term to the variation observed in each measured traits (Table 5.4). The contribution of genotype to the variation varied from one trait to the other. The largest contribution of the genotypes was observed for fodder yield (46.76 %) while the smallest was for shoot dry weight (1.5 %). The genotypes were also responsible for a significant variation with regards to days to flowering (21.57%) as well as grain yield (15.06 %). The environment was the main cause of the variability detected in shoot dry weight, plant height, days to 50 % maturity and grain yield with 82.45 %, 72.04 %, 58.69 % and 36.55 % respectively. The GEI part of the variation was high for days to flowering (36.85 %), fodder yield (22.73 %) and days to 50 % maturity (18.66 %). The GE interaction, although not significant for grain yield, contributed considerably (15.85 %) to the yield variation. The error has a great contribution in the variation observed in grain yield (32.54%). Table 5.4 Partitioning of total AMMI sum of square into its different components SOV DAF SDWt PHt MAT50 Fyield Gyield Genotype 21.57 1.50 5.05 8.12 46.76 15.06 Environment 28.39 82.45 72.04 58.69 6.94 36.55 GE 36.85 4.72 6.93 18.66 22.73 15.85 Error 13.19 11.33 15.98 14.54 23.57 32.54 Total 100.00 100.00 100.00 100.00 100.00 100.00 DAF =days to flowering, MAT50 = days to 50 maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yieldFyield = fodder yield. 5.3.1. Days to flowering (DAF) A relatively narrow variation was recorded among environmental means regarding days to flowering. Early flowering occurred at E3 compared to E2 and E4. E1 was comparatively 89 University of Ghana http://ugspace.ug.edu.gh intermediate in terms of days to flowering. The highest heterogeneity of variance was obtained at E4 (Figure 5.3). Figure 5-3 Boxplots for means of days to flowering across environment The overall mean number of days to flowering ranged from 44.54 days after planting in G98 to 56.75 in G32. G108 was the most stable genotype with an AMMI stability value (ASV) of 1.61 (Table 5.5). 90 University of Ghana http://ugspace.ug.edu.gh Table 5.5 Mean days to flowering of genotypes over all the environments Code Genotype E1 E2 E3 E4 Mean ASV G108 Tamalalo 48.00 59.67 46.67 54.67 52.25 1.61 G149 Waken chaibou 52.00 56.00 48.00 58.00 53.50 1.84 G116 Gnagaou 57.00 51.67 49.33 53.67 52.92 1.9 G1 Gorom ocal 48.67 54.33 43.33 49.67 49.00 1.93 G125 Bonbohi 56.00 50.67 49.67 53.67 52.50 2.05 G146 Jan pass 45.67 49.33 40.67 44.33 45.00 2.62 G62 Dan baouchi 51.33 58.00 46.00 51.67 51.75 2.77 G150 Jiraini 50.67 53.67 46.67 59.33 52.58 3.93 G101 Karadua 51.00 55.67 45.33 47.33 49.83 4.45 G118 Botsotsuwa 46.00 52.67 48.00 44.00 47.67 4.69 G43 Farin wake 45.00 53.67 45.00 61.00 51.17 6.13 G96 Dan raha 47.67 57.00 43.67 44.00 48.08 6.45 G11 Local chical 48.67 51.33 41.33 62.67 51.00 7.66 G98 Ndiambour 45.50 57.33 36.00 39.33 44.54 8.89 G32 Matarka nawa 52.00 51.00 51.67 72.33 56.75 12.72 Mean 49.68 54.13 45.42 53.04 50.57 E1 = Sadore rainy season 2016, E2 = Sadore dry season 2017, E3 = Sadore rainy season 2017, E4 = Maradi rainy season 2017, ASV = AMMI stability value. The results showed that the mean days to flowering at E3 was lowest compared to other environments. Environments were grouped into 2 mega environments with regard to days to flowering. GGE biplot analysis found E1, E3 and E4 as similar environments and grouped them into one mega- environment. Genotypes with higher number of days to flowering were considered as superior cultivars by the software, so E1, E3 and E4 were put in the same mega environment because they shared the same winning line which was G32. E2 formed a different mega environment. The best performing cultivars in E2 was G108 (Figure 5.4). 91 University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 77.37%) E2 G108 G62 E1 G149 G96 G101 E3 G116 G98 G150 G125 G1 E4 G118 G32 G43 G11 G146 PC1 - 55.90% Genotype scores Environment scores Convex hull Sectors of convex hull Mega-Environments Figure 5-4 GGE biplots for days to flowering means 5.3.2. Shoot dry weight (SDWt) The variation between environments was very large with regard to this trait as it can be seen in the boxplot for means of shoot dry weight (Figure 5.5). E4 had the highest mean and variance; it was therefore the most discriminating as well as most favorable environment for shoot development. The experiments of E2 and E3 were carried out on the same field and same year but different season with E2 being the minor season and E3 the main season of 2017. However the shoot dry weight per plant at E3 is more than double of that of E2 suggesting a significant effect of the season on the trait. On the other hand, E3 and E4 are the same year but different location and planting date, the results showed that the mean of E4 was more than double of that of E3. 92 PC2 - 21.47% University of Ghana http://ugspace.ug.edu.gh Figure 5-5 Boxplot for shoot dry weight means The overall mean shoot dry weight ranged from 37.72 to 48.84 g per plant with G118 having the highest value and G96 recorded the lowest across locations (Table 5.6.). All the genotypes performed significantly better at E4 in comparison to the other environments. Despite being the line with the lowest shoot dry weight, G96 was the most stable as it recorded the smallest AMMI stability value (1.08). 93 University of Ghana http://ugspace.ug.edu.gh Table 5.6 Mean shoot dry weight of genotypes across the environments Code Genotype E1 E2 E3 E4 Mean ASV G96 Dan raha 27.56 9.11 33.55 80.67 37.72 1.08 G108 Tamalalo 40.04 13.86 35.62 88.33 44.46 1.81 G43 Farin wake 34.41 11.49 31.48 84.58 40.49 1.81 G32 Matarka Nawa 36.14 12.56 32.03 75.83 39.14 1.82 G125 Bonbohi 35.81 17.42 37.37 89.92 45.13 1.95 G101 Karadua 32.54 16.84 50.09 85.00 46.12 2.65 G98 Ndiambour 30.70 13.26 35.30 72.67 37.98 2.68 G149 Waken chaibou 32.63 20.26 32.75 73.17 39.70 2.83 G62 Dan baouchi 45.76 10.84 32.18 78.00 41.69 2.90 G118 Botsotsuwa 32.01 26.71 42.05 94.58 48.84 3.38 G11 Local chical 46.39 14.10 30.90 93.67 46.26 4.21 G146 Jan pass 33.33 19.24 29.53 97.50 44.90 5.97 G116 Gnagaou 38.81 11.51 38.63 101.83 47.70 6.12 G1 Gorom local 33.40 14.01 44.93 67.25 39.90 6.66 G150 Jiraini 42.01 17.32 42.35 66.50 42.05 7.40 Mean 36.10 15.24 36.58 83.30 42.81 E1 = Sadore rainy season 2016. E2 = Sadore dry season 2017. E3 = Sadore rainy season 2017. E4 = Maradi rainy season 2017, ASV = AMMI stability value. Three mega environments were formed, E2 and E3 were grouped into the same mega- environment as they shared the same winning genotypes namely G1. E1 and E4 formed a mega- environment each. The best performing genotype at E1 was G62 while G116 won at E4 (Figure 5.6). 94 University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 79.46%) G118 G1 G101 E3 E2 E4 G116 G146 G98 G125 G1G51049 G96G14038 G11 G32 E1 G62 PC1 - 58.43% Genotype scores Environment scores Convex hull Sectors of convex hull Mega-Environments Figure 5-6. GGE biplots for shoot dry weight 5.3.3. Plant Height (PHt) The variation of the mean plant height per environment is shown in the figure below. 95 PC2 - 21.02% University of Ghana http://ugspace.ug.edu.gh Figure 5-7 Boxplot for plant height means The highest mean plant height was obtained at E1 while the largest variance was observed at E4. E2 had the lowest performance as well as the narrowest variation. Large variation was detected among the tested genotypes regarding mean plant height (Table 5.7). It ranged from 100.68 to 145. 34 cm. G101 and G149 had the highest and lowest mean plant height respectively. The most and least stable genotypes were G125 and G150. 96 University of Ghana http://ugspace.ug.edu.gh Table 5.7 Means plant height of genotypes across the environments Code Genotype E1 E2 E3 E4 Mean ASV G125 Bonbohi 187.89 44.91 122.67 129.00 121.12 1.46 G1 Gorom local 164.00 44.35 122.33 112.83 110.88 1.89 G108 Tamalalo 186.22 48.11 118.67 119.83 118.21 2.11 G62 Dan baouchi 180.11 47.21 105.33 113.50 111.54 2.20 G118 Botsotsuwa 192.17 65.29 109.67 162.17 132.32 2.98 G11 Local 159.33 51.88 124.33 144.00 119.89 3.03 G101 Karadua 182.56 82.47 152.17 164.17 145.34 3.16 G149 Waken chaibou 135.67 53.56 102.00 111.50 100.68 3.60 G98 Ndiambour 163.50 43.09 111.50 89.33 101.86 3.63 G146 Jan pass 203.89 50.34 97.67 142.67 123.64 3.85 G116 Gnagaou 141.22 44.60 129.17 130.83 111.45 4.35 G96 Dan raha 208.44 38.75 140.50 182.33 142.51 4.52 G43 Farin wake 156.56 28.90 90.50 153.67 107.40 4.83 G32 Matarka nawa 205.67 49.19 111.00 111.67 119.38 4.98 G150 Jiraini 203.44 46.55 116.50 105.67 118.04 5.20 Mean 178.04 49.28 116.93 131.54 118.95 E1 = Sadore rainy season 2016. E2 = Sadore dry season 2017. E3 = Sadore rainy season 2017. E4 = Maradi rainy season 2017, ASV = AMMI stability value. Only 1 mega environment was formed by the GGE as there was no re-ranking of genotypes according to environments. The superior genotypes were G96 and G101 (Figure 5.8). 97 University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 77.42%) G116 G11 E4 G43 G149 G101 G1 G11E83 G98 E2 G96 G108G125 G62 G146 G150 G32 E1 PC1 - 50.02% Genotype scores Environment scores Convex hull Sectors of convex hull Mega-Environments Figure 5-8 GGE biplots for plant height means 5.3.4. Days to 50 % maturity (MAT50) The variation of the mean number of days to 50 % maturity according to environments is shown in the figure 5.9. It illustrated the heterogeneity of variance of the trait among environments. The largest variability was observed at E4. The highest mean was recorded at E4 while E3 got the lowest making it the best environment for early maturity. 98 PC2 - 27.41% University of Ghana http://ugspace.ug.edu.gh Figure 5-9 Boxplot for days to 50 % maturity means The most stable genotype across the environments was G146 with an overall mean of 73.04 days after planting and an ASV of 0.96 (Table 5.8). General mean indicated that G96 was the first genotype to reach 50 % maturity. G98 exhibited a wide variability among environments; for example a difference of 24 days was observed between its means of E1 and E3. It is the most instable genotype with an ASV of 12.06 according to the results. 99 University of Ghana http://ugspace.ug.edu.gh Table 5.8 Mean days to 50 % maturity for genotypes across the environments Genotype Designation E1 E2 E3 E4 Mean ASV 146 Jan pass 79.18 69.66 62.90 80.42 73.04 0.96 1 Gorom local 76.51 72.21 67.44 81.58 74.43 1.13 118 Botsotua 76.68 75.33 68.79 80.73 75.38 1.19 11 Local chical 81.75 70.83 62.76 84.39 74.93 1.23 43 Farin wake 79.48 71.68 69.85 79.50 75.13 1.28 116 Gnagaou 85.77 72.88 70.23 86.45 78.83 1.38 125 Bonbohi 83.39 70.55 71.23 85.51 77.67 1.75 150 Jiraini 80.13 75.38 68.12 90.28 78.48 3.30 62 Dan baouchi 77.00 76.19 67.37 89.75 77.58 3.77 149 Waken chaibou 78.05 75.84 68.67 88.28 77.71 3.86 108 Tamalalo 78.19 77.58 68.44 90.04 78.56 3.99 101 Karadua 80.72 72.76 68.70 76.53 74.68 4.16 96 Dan raha 76.72 74.92 64.00 73.10 72.18 5.18 32 Matarka nawa 78.10 69.68 75.42 90.61 78.45 6.36 98 Ndiambour 86.09 73.46 61.13 68.52 72.30 12.06 Mean 79.85 73.26 67.67 83.05 75.96 E1 = Sadore rainy season 2016, E2 = Sadore dry season 2017, E3 = Sadore rainy season 2017, E4 = Maradi rainy season 2017, ASV = AMMI Stability Value. The GGE biplots grouped the environments into 3 mega environments (Figure 5.10). It considered genotypes with highest number of days to maturity as the best ones. E3 and E4 belonged to the same mega-environment with G32 as the winning genotype. The early maturing variety at E3 and E4 was G98. E1 and E2 formed each a mega environment. The winning cultivars at E1 were G116 and G125 while G62 and G96 won at E2. 100 University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 83.34%) G116 G125 E1 G32 G98 E3 G101G4G311 E4 G146 G150 G1 E2 G149 G118 G108 G62 G96 PC1 - 69.04% Genotype scores Environment scores Convex hull Sectors of convex hull Mega-Environments Figure 5-10 GGE biplots for days to 50 % maturity means 5.3.5. Grain Yield (Gyield) A wide variability was observed among environments in terms of grain yield per plant (Figure. 5.11). E1 and E2 were the best environments since they recorded the highest mean grain yield per plant. Comparatively, the yield at E3 and E4 was very low. The largest yield variability was recorded at E1 followed by E4. 101 PC2 - 14.30% University of Ghana http://ugspace.ug.edu.gh Figure 5-11 Boxplot for grain yield means per plant With the general means of 48.33 and 20.35 g per plant, G118 and G98 were respectively the best and less performing cultivars (Table 5.9). G x E was not significant for grain yield, so the superior genotypes across the environments were G118, G150, G149, G108 and G62 with 48.23 g, 45.23 g, 40.32 g, 40.44 g and 39.03 g mean grain yield per plant. G118 was the winning cultivar at E1 and E2; however its performance decreased at E3 and E4. G150 was the most stable genotype as its yield was high at all the environments. 102 University of Ghana http://ugspace.ug.edu.gh Table 5.9 Mean grain yield of genotypes across the environments Code Designation E1 E2 E3 E4 Mean G150 Jiraini 52.55 53.74 35.65 38.97 45.23 G125 Bonbohi 31.66 36.08 10.70 27.23 26.42 G11 Local chical 48.13 44.61 21.25 18.96 33.24 G96 Dan raha 31.77 24.74 21.33 16.67 23.63 G149 Waken chaibou 51.85 46.43 24.96 38.02 40.32 G116 Gnagaou 55.56 36.91 17.60 22.35 33.11 G98 Ndiambour 36.19 35.87 11.75 -2.42 20.35 G62 Dan baouchi 59.57 36.60 19.45 40.49 39.03 G101 Karadua 17.83 40.09 15.21 16.15 22.32 G146 Jan pass 33.24 43.29 32.64 15.43 31.15 G32 Matarka nawa 59.93 33.38 11.38 8.56 28.31 G1 Gorom local 59.59 33.35 34.53 7.67 33.78 G43 Farin wake 48.87 16.61 20.08 17.60 25.79 G118 Botsotua 71.60 76.52 27.69 17.11 48.23 G108 Tamalalo 38.70 57.06 31.99 34.00 40.44 Mean 46.47 41.02 22.41 21.12 32.75 E1 = Sadore rainy season 2016, E2 = Sadore dry season 2017, E3 = Sadore rainy season 2017, E4 = Maradi rainy season 2017. 5.3.6. Fodder yield (Fyield) The boxplot for fodder yield means showed that E2 and E3 were the best environments (Figure 5.12). Very low fodder yield mean was observed at E4 due to severe striga infestation. The boxplot showed that the genotypes expressed largest variability at E2. 103 University of Ghana http://ugspace.ug.edu.gh Figure 5-12 Boxplot for fodder yield means Mean fodder yield of culivars across the environments varied from 31.24 to 121.42 g per plant. The superior genotypes were G197, G118 and G164 with respectively 121.42, 108.79 and 104.59 g per plant (Table 5.10). However, all the high fodder yielding varieties showed differential adaptation to environments. The widely adopted improved varieties were highly instable and low in fodder yield. 104 University of Ghana http://ugspace.ug.edu.gh Table 5.10 Mean fodder yield of genotypes across the environments Code Genotype E2 E3 E4 Mean ASV G149 Waken chaibou 68.45 70.78 58.58 65.94 1.45 G108 Tamalalo 90.17 84.83 79.45 84.82 1.46 G96 Dan raha 40.61 47.57 44.10 44.10 2.22 G32 Matarka nawa 40.22 72.02 33.82 48.69 2.78 G62 Dan baouchi 80.65 45.04 65.52 63.74 3.30 G150 Jiraini 88.85 68.33 60.88 72.69 3.50 G118 Botsotsuwa 123.72 89.65 112.99 108.79 3.63 G1 Gorom local 46.27 76.68 32.00 51.65 3.82 G101 Karadua 81.25 111.68 69.00 87.31 4.20 G98 Ndiambour 83.36 47.16 30.52 53.68 4.22 G43 Farin wake 20.61 36.07 42.33 33.00 4.70 G206 KVX 37.78 27.62 28.32 31.24 5.01 G209 TN-5-78 34.49 61.52 30.35 42.12 6.32 G197 NELOC-197 173.19 92.38 98.70 121.42 8.43 G205 IT90K-372-1-2 5.70 62.64 42.95 37.10 8.60 G164 NELOC-164 171.65 123.55 18.58 104.59 16.31 Mean 74.19 69.85 53.01 65.68 E2 = Sadore dry season 2017, E3 = Sadore rainy season 2017, E4 = Maradi rainy season 2017, ASV = AMMI Stability Value. Environments were grouped into 2 mega-environments with E2 and E3 in one and E4 in the other. G164 and G197 were the winning genotypes at E2 and E3 while G118 was the best at E4 (Figure 5.13). 105 University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 91.70%) G164 G1 G98 GG20392 E3 E2 G206 G205G96 G1 G50101G149 G43 G62 G108 G197 G118 E4 PC1 - 77.62% Genotype scores Environment scores Convex hull Sectors of convex hull Mega-Environments Figure 5-13 GGE biplots for fodder yield means 5.3.7. Pod yield (Pyield) Pod yield per plant was higher at E3 in comparison to E4. Large variability among genotypes was however observed at E4 compared to E3. Mean pod yield per plant varied from 20.67 to 51.29 g per plant at E3 and from 2.45 to 71.76 g per plant at E4. The overall best genotypes were G62, G150 and G149 with 53.42, 49.38 and 45.84 g per plant, respectively. The least performing cultivars were G32, G98 and G101 as they recorded mean pod yield of 12.08, 16.93 and 19.68 g per plant, respectively (Table 5.11). 106 PC2 - 14.07% University of Ghana http://ugspace.ug.edu.gh Table 5.11 Mean pod yield per plant of genotypes at E3 and E4. Code Genotype E3 E4 Mean G62 Dan Baouchi 35.08 71.76 53.42 G150 Jiraini 47.01 51.74 49.38 G149 Waken chaibou 43.07 48.62 45.84 G118 Botsotua 49.09 31.90 40.50 G108 Tamalalo 45.75 34.81 40.28 G1 Gorom local 51.29 24.47 37.88 G11 Local chical 30.25 43.96 37.10 G43 Farin Wake 39.19 32.87 36.03 G146 Jan Pass 48.43 16.37 32.40 G116 Gnagaou 28.16 35.01 31.59 G125 Bonbohi 21.32 35.70 28.51 G96 Dan Raha 39.30 17.19 28.24 G101 Karadua 21.59 17.77 19.68 G98 Ndiambour 31.42 2.45 16.93 G32 Matarka Nawa 20.67 3.48 12.08 Mean 36.78 31.21 E3 = Sadore rainy season 2017. E4 = Maradi rainy season 2017. 5.3.8. Seed number per pod (Snber) There was a narrow variation among the environments with respect to number of seeds per pod. The mean per environment was 11.75 and 11.38 seeds per pod for E3 and E4, respectively. General mean number of seeds per pod varied from 9.43 to 12.48 with G62 having the highest and G98 the lowest mean. Variation among genotypes was larger at E4 than E3 (Table 5.12). 107 University of Ghana http://ugspace.ug.edu.gh Table 5.12 Mean number of seeds per pod at E3 and E4 Code Genotype E3 E4 Mean G62 Dan Baouchi 12.65 12.31 12.48 G149 Waken chaibou 11.34 13.03 12.19 G11 Local chical 12.22 12.14 12.18 G32 Matarka Nawa 12.36 11.79 12.08 G118 Botsotua 10.50 13.44 11.97 G116 Gnagaou 11.54 12.08 11.81 G146 Jan Pass 11.94 11.65 11.79 G96 Dan Raha 12.34 11.08 11.71 G150 Jiraini 11.33 12.02 11.67 G43 Farin Wake 12.04 11.13 11.58 G125 Bonbohi 10.81 12.17 11.49 G108 Tamalalo 12.30 10.59 11.44 G1 Gorom local 12.25 10.21 11.23 G101 Karadua 10.93 9.94 10.44 G98 Ndiambour 11.76 7.09 9.43 Mean 11.75 11.38 E3 = Sadore rainy season 2017, E4 = Maradi rainy season 2017. 5.3.9. Pod length (Plength) The two environments were similar in terms of pod length. E4 seemed to discriminate the cultivars better than E3. Genotypes with longest pods were G11 and G62. The pods of G98 were significantly shorter in comparison to those of the remaining genotypes (Table 5.13). 108 University of Ghana http://ugspace.ug.edu.gh Table 5.13 Mean pod length for genotypes at E3 and E4 Code Genotype E3 E4 Mean G11 Local chical 14.52 15.61 15.06 G62 Dan Baouchi 15.04 15.05 15.05 G149 Waken chaibou 14.46 15.38 14.92 G125 Bonbohi 14.98 14.34 14.66 G150 Jiraini 14.41 14.87 14.64 G108 Tamalalo 15.03 14.14 14.59 G43 Farin Wake 14.24 14.69 14.47 G118 Botsotua 13.37 15.37 14.37 G116 Gnagaou 14.18 14.51 14.34 G1 Gorom local 14.77 12.72 13.74 G96 Dan Raha 14.00 13.12 13.56 G146 Jan Pass 13.31 13.82 13.56 G101 Karadua 13.22 12.79 13.00 G32 Matarka Nawa 11.41 12.79 12.10 G98 Ndiambour 12.90 8.78 10.84 Mean 13.99 13.87 13.93 E3 = Sadore rainy season 2017. E4 = Maradi rainy season 2017. 5.3.10. Harvest index (HI) Mean harvest index was higher at E4 than E3 meaning that the cultivars produced more grain than fodder at the latter than the former. The overall mean HI was 0.41. The mean per genotype ranged from 0.12 to 0.49 with G149 having the highest mean and G101 the lowest. Genotypes that were best in grain and fodder production were G149, G150, G43 and G62. Those suited for fodder production were G101, G32 and G116 (Table 5.14). 109 University of Ghana http://ugspace.ug.edu.gh Table 5.14 Mean harvest index of genotypes at E3 and E4 Code Designation E3 E4 Mean G149 Waken chaibou 0.66 0.32 0.49 G150 Jiraini 0.48 0.45 0.46 G43 Farin wake 0.40 0.50 0.45 G62 Dan baouchi 0.49 0.40 0.44 G11 Local chical 0.18 0.56 0.37 G96 Dan raha 0.21 0.51 0.36 G108 Tamalalo 0.41 0.31 0.36 G146 Jan pass 0.14 0.53 0.33 G118 Botsotsuwa 0.27 0.28 0.27 G1 Gorom local 0.19 0.32 0.26 G116 Gnagaou 0.24 0.17 0.20 G98 Ndiambour 0.10 0.29 0.20 G32 Matarka nawa 0.08 0.20 0.14 G101 Karadua 0.15 0.08 0.12 Mean 0.37 0.45 0.41 E3 = Sadore rainy season 2017, E4 = Maradi rainy season 2017. 5.3.11. Threshing percentage (Thresh %) The two environments did not differ from one another in terms of threshing percentage. The mean across genotypes was 61.9 %. It varied from 56.39 % in G101 to 70.53 % G149 (Table 5.15). 110 University of Ghana http://ugspace.ug.edu.gh Table 5.15 Mean threshing percentage of genotypes at E3 and E4 Code Designation E3 E4 Mean G149 Waken chaibou 63.72 77.33 70.53 G150 Jiraini 68.57 65.27 66.92 G118 Botsotsuwa 61.20 71.37 66.29 G11 Local chical 71.03 56.59 63.81 G62 Dan baouchi 66.66 58.85 62.76 G43 Farin wake 59.82 65.19 62.50 G146 Jan pass 68.54 54.96 61.75 G108 Tamalalo 55.06 67.55 61.31 G96 Dan raha 70.19 50.88 60.53 G32 Matarka nawa 60.87 59.42 60.15 G116 Gnagaou 53.81 63.11 58.46 G1 Gorom local 65.30 51.42 58.36 G98 Ndiambour 55.36 58.34 56.85 G101 Karadua 54.54 58.23 56.39 Mean 62.48 61.32 61.90 E3 = Sadore rainy season 2017, E4 = Maradi rainy season 2017. 5.4. Discussion The highly significant differences detected among genotypes for almost all studied traits indicated the existence of large genetic diversity in the germplasm under investigation. Similar findings in cowpea were reported by El-shaieny et al. (2015) and Teixeira et al. (2007). The significant variation observed between environments implied that the weather factors differed widely according to year and location. Major climatic determinants for plant productivity are moisture, temperature, light, carbon dioxide and daylength (White and Howden 2010). Available weather data for E1, E2 and E3 suggested relative humidity and solar radiation as the main causes of variation observed between environments. The mean available phosphorus of 6.42 ppm and 5.20 ppm for E1 on one hand and E2 and E3 on the other showed that these sites did not vary significantly in terms of soil available P content. In the case of E4, the rainfall was the only available weather data and it was similar to that of the other sites. Genotype by 111 University of Ghana http://ugspace.ug.edu.gh environment interaction occurs when environments are not correlated or the genotypic variance is heterogeneous or both conditions are present (Crossa et al., 1993; 2004). The detection of significant G x E for days to flowering, days to 50 % maturity, fodder yield, pod yield, harvest index and threshing percentage in the present work was an indication of the occurrence of at least one of the conditions cited above. Significant G x E interactions for days to flowering, days to maturity, pod and seed weight per plant was observed by Aliyu and Makinde (2016).The lack of significant genotype by environment interactions in terms of grain yield per plant revealed by the current study was not consistent with Santos et al. (2015) who reported significant G x E interactions for grain yield while studying the adaptability and stability of cowpea cultivars to Midwest Brazil areas. Significnat G x E for grain yield were also observed by Omoigui et al. (2017) who used biplots analysis to assess the performance of cowpea varieties under striga infestation in the dry agro-ecology of Nigeria. Similar results were reported by Santos et al. (2017) following three years stability evaluation of erect cultivars in Mato Grosso do Sul state of Brazil. Furthermore, Nwofia et al. (2016) found both significant year by genotype and planting date by genotype interactions regarding seed yield per ha in soybean, but they detected only year by genotype interactions in terms of number of seed per plant. The medium to high contribution of genotypes to the variation in days to flowering and fodder yield implied that selection for these traits can be carried out successfully. However, the gain in selection may be low, especially regarding days to flowering since the environmental contribution to its variation was also high. G98 was the best genotype for early flowering; however it was not stable and can therefore be recommended for use in rainy seasons at Sadore and Maradi where it performed well. The results showed that the large variation recorded by shoot dry weight has no genetic basis; it was mainly affected by the environment. Shoot dry 112 University of Ghana http://ugspace.ug.edu.gh weight was high in rainy seasons and very low in dry season, the only environmental factors that varied significantly between rainy and dry season were the relative humidity and solar radiation according to the weather data, these climatic factors may therefore be the cause of the observed slow seedlings growth that resulted in low shoot dry weight in the dry season. No information was available to explain the very high dry shoot weight obtained at Maradi compared to the other environments. That site could be relatively higher in available P content compared to the remaining test environments and according to Ansah et al. (2014) shoot dry weight was the most sensitive trait to P deficiency. The yield is usually known to be highly influenced by environmental conditions. The high environmental influence on yield (pod, grain and fodder) detected in the present work was in agreement the findings of Aliyu and Makinde (2016). It was also consistent with Omoigui et al. (2017) and Santos et al. (2015) who respectively reported that the environment was responsible for 35.01and 78.27 % of the variation observed in grain yield of cowpea. However dissimilar results were detected by Baraki et al. (2014) who working on sesame found that the genotypes contribution (37.3 %) to yield variation was greater than that of the environments (29.5 %). The environment contribution to the variation was higher in grain yield than fodder in the present study. However, in both traits, the genotypes accounted for sizeable part of the variation giving thereby room for selection of superior genotypes. The dimension of experimental plot could have affected grain yield as the present results showed that the grain yield per plant was higher in trials with small plots compared to those with large plots area. The similarity detected among test environments by the GGE biplots analysis in terms of grain yield per plant could be the reason why significant G x E interaction was not detected for that trait. However despites the absence of GEI, a change in ranking of genotypes was observed as 113 University of Ghana http://ugspace.ug.edu.gh G118 which was best at E1 and E2 was not the leading cultivar at E3 and E4, revealing differential adaptation of some genotypes to environments. P deficiency in soils caused stunted growth and low yield, therefore genotypes with high grain and/or fodder yield on such conditions could be considered as tolerant. In the present work, G150 and G149 recorded high grain yield, G101 produced high fodder while G118 and G108 were good in both grain and fodder yield. These cultivars were therefore considered as adapted to P deficient soil conditions. The overall mean harvest index of 41% obtained in this study implied that most of the lines under study were good in both grain and fodder. It was however greater in comparison to the value of 34.5 % reported by Rajput. 1994. High striga infestation occurred at E4 and it significantly affected both grain and fodder yield of susceptible varieties. E4 recorded the highest mean shoot dry weight however it had the lowest mean fodder yield; this is an indication of striga damage on some varieties. Sampling for shoot dry weight determination was made at 8 weeks DAP, the fact that striga negative impact was not detected on that trait indicated that the parasite damage on susceptible varieties occurred after 8 weeks DAP. Therefore, scoring for striga tolerance at 8 weeks DAP or earlier may not give reliable results and this observation was in agreement with Singh and Emechebe (1997) who recommended scoring at 9 weeks DAP. According to these findings, the yield of cowpea extra- early cultivars may not be significantly affected by striga. Some of the tested genotypes had expressed high level of tolerance to striga. Positive association between striga infestation and field fertility status was reported. Zouera et al. (2016) stated that low fertile soils are more prone to striga infestation in comparison to fertile fields. Since farmers‘ fields, especially in Niger are predominantly poor, cowpea varieties that tolerate both striga and low P soil conditions could be the most adequate alternative for yield improvement. According to our results, G150, G118, 114 University of Ghana http://ugspace.ug.edu.gh G108, G149 and G62 had some good level of tolerance to both low P soil and striga could be therefore used as varieties or sources of genes for improvement of elite varieties. Apart from tolerating both striga and low P soils, G150 had a stability of 0.6, thereby cumulating high performance and stability. This finding did not corroborate with Amara and Suale (1996) who identified IT86D-1010 and IT86D-716 as most P efficient cowpea varieties but only adapted to high rainfall areas of Sierra Leone. However, high seed yield per plant and stability were detected in Cream 7, Azmerly, Blackeye crowder, Dokii and Black crowder genotypes by El- Shaieny et al. (2015). The overall mean seed number per pod of 11.57 obtained in this work was higher than the mean of 6.33 reported by Inuwa et al. (2012) but lower than that of 13.2 found by Rajput, (1994). The non-significant G x E interaction revealed by our results regarding seeds number per pod and pod length was not in agreement with Gerano et al. (2015) who detected significant GEI in these traits. 5.5. Conclusion The results showed that both the genotypes and environments were significantly different for most of the traits under study. However, G x E interaction was only significant for days to flowering, days to 50 % maturity, fodder yield harvest index, pod yield and threshing percentage. The contribution of genotypes to the variation was high regarding fodder yield and days to flowering, but medium in terms of grain yield. Genotypes with high grain and fodder yield under low P soil conditions were detected, nevertheless, only G150 was highly stable across the test environments. Cultivars tolerant to both low P soil conditions and striga were also identified. Candidate Genotypes for release as variety were identified. 115 University of Ghana http://ugspace.ug.edu.gh CHAPTER 6 6.0. Effects of planting date on the cycle and yield of local cowpea varieties in Niger 6.1. Introduction Genotype interaction with temperature, photoperiod and other factors of the environment affects the duration of vegetative and reproductive growth phases of grain legumes causing a reduction in yield (Summerfield. 1980; Hadley et al., 1983). High temperatures during long days cause two or more weeks delay in flowering in many cowpea genotypes (Warrag and Hall. 1983). Heat susceptible genotypes loose the five first flowers on its main stem following two weeks or more of successive or interrupted hot nights in the first four weeks after germination (Ahmed et al., 1992). However the negative impact of heat on the yield of many crops is affected by photoperiod (Hall, 1992) while temperature in turn moderates sensitivity to photoperiod (Vince- Prue, 1975). The effects of both heat and photoperiod on grain legume crops may be influenced by planting date. According to Azari and Khajepour, (2003), planting date is among the main factors that affect the growth and yield of leguminous crops. The same authors asserted that planting date is responsible for the positive correlation occuring between plant phenological stage and climatic factors. In soybean many studies were conducted to find out the effects of planting date on grain yield and its components. Alghamdi (2004) tested five soybean genotypes in six planting dates during 2000 and 2001 years and observed significant sowing date by genotype interactions with regard to seed weight per plant and grain yield per hectare. Similar findings were reported by Nwofia ( 2016); ElHerty et al. (2010); Kandil et al. (2012); Ngalamu et al. (2013) and Bello et al. (1996). Also May et al. (1989) affirmed that the yield advantage of early maturing soybean cultivars over late maturing ones was influenced by planting date. The use of unsuitable planting dates was cited among the factors that bring about low yield in cowpea 116 University of Ghana http://ugspace.ug.edu.gh (Coetzee, 1990; Singh et al., 2002; Ishiyaku et al., 2005). In an attempt to determine the appropriate sowing date of cowpea in Northwest and Limpopo provinces (South Africa), Shiringani et al. (2007) tested 10 cowpea genotypes in three planting dates. They reported th November 8 as best date for grain yield across the locations. In contrast, using the same cowpea genotypes, the same planting dates and the same locations, Shiringani and Shimelis, (2011) th identified December 6 as most adequate planting date for high grain yield. Dhital et al. (1997) stated that early planting gave higher seed yield compared to late sowing. In Ghana, six cowpea varieties were evaluated in six planting dates (Ansoba et al., 2013), the results showed that appropriate sowing date varied with variety, the authors however recommend early planting to take advantage of more suitable weather conditions and less insect damage. Planting date determines the time that the plant will flower. Late planting of early maturing varieties prevents farmers from getting the benefit of early harvests (Fatokun et al., 2002). Cowpea varieties have been classified into different maturity groups namely extra-early, early, medium and late maturing, but it should be noted that the cycle of many cultivars is subject to the environmental influence. Day length interaction with genotype may prolong or shorten the phenological cycle of photoperiod sensitive varieties with subsequent effect on its yield. In Niger, the adoption of improved cowpea varieties is very low; majority of farmers are still growing local varieties which are known to be mostly sensitive to photoperiod. Day length varies from 12h 55 min in June (the beginning of rainy season) to 11h 50 min in October (the end of the season) in Niger (https://planificateur.a-contresens.net/heure/NE-niger.html). Other environmental factors such rainfall, temperatures and relative humidity vary continuously throughout the cropping season. Under these conditions, planting date may considerably affect the cycle, yield and yield components of photosensitive cultivars. The identification of adequate planting dates of cowpea, 117 University of Ghana http://ugspace.ug.edu.gh especially local farmer varieties may significantly improve cowpea yield in small scale farmers‘ field in Niger. The objectives of the present work were to: - assess the effect of planting dates on the cycle, grain and fodder yield of some local varieties in Niger; - determine the best planting dates for high grain and/or fodder yield of farmers‘ cowpea varieties; - identify cultivars adapted to early and late planting conditions. 6.2. Materials and methods 6.2.1. Sites The experiments were conducted at ICRISAT Sadore and Ndounga experimental station of INRAN. All the locations are located around Niamey and share approximately the same climatic conditions. Niamey is situated in the Sahelian zone where the mean winter (December, January, and February) temperature is 25.3°C and mean spring (March,April and May) temperature is 32.7°C. Sadore is located at 45 km in the south of Niamey and lies on latitude 13°14'N and longitude 2°16'E and is located at an altitude of 235 m above sea level. Mean annual rainfall between 1983 and 2015 was 556.92 mm with the peak in August (ICRISAT unpublished). Ndounga is at 20 km in the south east of Niamey. It is located on latitude 13°22' N and longitude 2°14' E at an altitude of 217 m. Annual rainfall ranged from 495.5 to 610.8 mm between 2013 and 2017 with a mean of 561.6 mm. 118 University of Ghana http://ugspace.ug.edu.gh 6.2.2. Agro-climatic data Large difference was observed between Kollo and Sadore in terms of the amount of rainfall. The overall rainfall quantities received was 482.1 and 572.4 mm respectively at Kollo and Sadore. The amount of rainfall was higher at Sadore than Kollo in all the wet season months except September (Table 6.1) Table 6.1 Rainfall data of the experimental sites Month Kollo Sadore June Decade1 1.0 0 16 .9 Decade2 58.5 18.6 Decade3 17.5 49.5 July Decade1 70. 00 78 .1 Decade2 34.5 67.9 Decade3 59 40.6 August Decade1 74. 10 76 .8 Decade2 13 85.7 Decade3 62.7 40.0 September Decade1 22. 80 24 .8 Decade2 40 27.4 Decade3 29 46.1 Total 482.1 572.4 The only available climatic data at Kollo was rainfall, however for Sadore; data were obtained on air and soil temperature, relative humidity, solar radiation, evaporation and wind speed (Table 6.2). Mean evaporation and wind speed were considerably higher in June compared to other months. Mean minimum and maximum air and soil temperatures were also slightly higher in June. However mean minimum and maximum relative humidity were lower in June, reached the peak in August and dropped in September. 119 University of Ghana http://ugspace.ug.edu.gh Table 6.2 Weather data of Sadore in 2016 Month Evap. Min TP Max TP HUM-07 HUM-13 SOLAR ST05-07 ST05-13 ST10-07 ST10-13 WIND June Dec1 8.2 26.4 38.2 69 40 19.0 29.8 47.7 31.4 39.3 5.0 Dec2 7.4 26.9 37.1 71 43 - 29.5 45.9 30.3 38.7 5.5 Dec3 5.3 25.1 34.0 80 60 - 29.0 40.0 29.8 35.2 3.3 Mean 7.0 26.1 36.4 73 48 - 29.4 44.5 30.5 37.7 4.6 July Dec1 5.2 25.1 34.7 84 57 - 28.1 41.1 28.9 36.1 3.8 Dec2 4.3 23.8 32.9 90 67 - 26.9 38.6 27.7 33.9 3.4 Dec3 3.9 24.3 32.6 90 66 20.9 26.6 38.7 27.6 33.2 2.7 Mean 4.5 24.4 33.4 88 63 - 27.2 39.5 28.1 34.4 3.3 August Dec1 4.5 24.1 32.0 89 70 19.5 26.6 35.7 27.4 31.3 3.1 Dec2 3.6 22.9 31.8 95 68 19.3 25.5 35.8 26.1 31.8 2.3 Dec3 4.4 23.4 33.0 94 66 20.7 26.3 40.2 27.4 35.2 3.0 Mean 4.2 23.5 32.3 93 68 19.8 26.1 37.2 27.0 32.8 2.8 Sept. Dec1 3.7 23.1 32.4 92 64 16.8 26.6 40.3 27.8 34.7 2.0 Dec2 4.6 23.8 34.6 89 56 21.9 27.3 43.3 28.3 37.0 2.5 Dec3 4.3 23.3 34.4 91 59 20.1 26.7 41.8 27.9 36.9 1.9 Mean 4.2 23.4 33.8 90 60 19.6 26.9 41.8 28.0 36.2 2.1 Key: Dec = decade. Evap. = evaporation; Min TP = minimum temperature. Max TP = maximum temperature. HUM-07= min relative humidity. HUM-13 = maximum relative humidity. SOLAR = solar radiation. ST05-07 = min soil temperature at 5 cm depth. ST05-13 = max soil temperature at 5 cm depth. ST10-07 = min soil temperature at 5 cm depth. ST10-13 = max soil temperature at 10 cm depth 6.2.3. Germplasm A total of twenty one local cowpea varieties collected from the crop major growing areas in Niger were used for the study (Table 6.3) 120 University of Ghana http://ugspace.ug.edu.gh Table 6.3 List of cowpea genotypes used for the study Maturity ENTRY_NO NAME TYPE Village Region (Farmers) G16 Dan Illa INRAN Improved INRAN G33 Dan Batsatsa Local Kornaka Mar adi 80 G34 Bougou Bougou Local Zinder NA G35 Dan Dam Local D an Saga Maradi 100 G36 Tcheto Local Tcheto Dosso 60 G37 Tessa Local Tessa Dosso 90 G38 Bagagi 21 ans Local Bagagi Dosso NA G45 Gouma Local Gouma Dosso >90 G46 Dan Bawada Daji Local Bawada Dadji Dosso 90 G47 El Danke Local Maitsakoni Maradi 80 G51 Farin wake Local Matankari Dosso early G54 Salga Sabongari Local Salga Dosso medium G56 Dan Mangaza Local Wangarawa Maradi NA G59 Darey kiota Local Darey Dosso NA G63 Arnen wake Local Matankari Dosso NA G69 Gnagaou Local Lido Dosso NA G71 Hwarin wake Local Kornaka Maradi 90 G74 Baadare Local Karanguiya Maradi 75 G78 Lakade Local Maiki Maradi NA G80 Dan Baourota Local Kornaka Maradi 80-90 G58 Dan Bekori Local Bekori Zinder NA 6.2.4. Method The trials were carried out in 2016 and 2017. A total of six planting dates were used. th th rd The first, second and third 2016 experiments were planted on 30 June, 19 July and 3 August 2016 at ICRISAT Sadore. Seeds were sterilized with Thioral (25 % Hyperchlor, 25 % Thirame) prior to planting. Four seeds were planted per hill and thinned to 2 plants per hill at 21 days after planting (DAP). Alpha lattice design with 2 replications was used. Each genotype was planted each in two rows of 6 m long with a spacing of 1m inter and 0.5 m intra. Borders lines were planted 1m apart from the first row on each side of the field. Only one weeding was done at 3 121 University of Ghana http://ugspace.ug.edu.gh weeks DAP, later weeds were controlled when necessary by uprooting manually. Insects were controlled by chemical application; however no insecticide was effective with regard to Clavigralla tomentosicollis. The 2017 trials were conducted at INRAN experimental station of Ndounga. The first and th nd second plantings were carried out on 7 and 22 July respectively while the last one was sown th on 5 August. Experimental design was alpha lattice with 2 replications. Seeds were treated with fungicide before planting. Four seeds were sown and thinned to 1 plant per hill at 3 weeks after planting. The plot was 2.6 m long and 2 m wide. Within the plot, the spacing was 0.5 m intra and 0.8 m inter rows. Distance between plots was 1.5 m. Borders lines were planted 1.5 m apart the first rows. Weeding was carried out twice. The field was treated three times with insecticide. Data were collected on agronomic and yield parameters as follow: - shoot dry weight (SDWt): Plants were uprooted from four hills in each plot at 8 weeks after planting, dried at room temperature and weighed; - plant height (PHt): The tallest two plants selected randomly per plot were measured using a tape measure at 8 week after planting; - days to flowering (DAF) : number of days from planting to first flower appearance; - days to 50 % maturity (MAT50): number of days from planting to 50 % maturity; - grain yield per plant (Gyield): dry seeds weight per plot divided by the number of plants per plot. For 2017 trial, additional data were collected on the following parameters: - fodder yield per plant (Fyield) : dry fodder weight per plot divided by the number of plants per plot; 122 University of Ghana http://ugspace.ug.edu.gh - pod yield (Pyield) : dry pods weight per plot divided by the number of plants per plot; - hundred seed weight (HSWt). - pod length (Plength). the lenght of three normal dry pods per plot was measured with a ruler; - number of seeds per pod (NSPd): the number of seeds contained in three normal dry pods per plot was determined; The field was infested by striga; the genotypes were therefore scored for tolerance to the parasite according to a modified rating scale of Singh and Emechebe (1997). The scale was as follow: - 1. Resistant: no Striga emergence on the plot and no Striga symptoms observed on plants. - 2. Tolerant: Several Striga emergences but no significant yield reduction. - 3. Moderately resistant: few Striga emergence (2-3) per plot but no Striga symptoms observed on the plants. - 4. Scusceptible: Five to several Striga plants emerged per plot and plants showed severe Striga symptoms. - 5. Several Striga plants emerged per plot and most of the plants have died. The data were subjected to combined analysis of variance using SAS 9.4 sofware. AMMI which has on one hand, an additive component accounting for genotype and environment main effects and on the other, a multiplicative terms that models G x E was performed to study the patterns of G x E and to determine differential adaptation of genotypes to environments. The model is as follow: Y𝑖𝑗 = 𝜇 +𝐺𝑖 +𝐸𝑗 k +∑ 𝝀k=1 𝒌Y + 𝜖𝑖𝑗 123 University of Ghana http://ugspace.ug.edu.gh The GGE (G plus GxE) biplots which is similar to AMMI except that the multiplicative terms include the genotypic main effect in addition to GE was run to: - directly identify the best performing genotypes - determine the mega environments (group of environment with common best performing genotypes). The winning genotype for a mega-environment is the vertex cultivar at the intersection of the two polygon sides whose perpendicular lines form the boundary of that environment. Test environment with longer vectors are the ones that discriminate the genotypes better; if the environment has a short vector, it means the performance of all the genotypes is similar within that environment; the difference between cultivars cannot therefore be detected. The representativeness of the environments is determined according to the angle formed by the environment and the average environment coordination (AEC), tests environments that have small angle with AEC are more representative than the ones having large angle. Both AMMI and GGE biplots analysis were carried out using Genstat 12 statistical software. Percentage contribution of IPCA1, IPCA2 and the residual to the G x E variation as well as the contribution of genotypes, environment, G x E and the error to the total variation observed in each trait were determined as in chapter five. AMMI sum of squares was partitioned in its different components only for traits where the two IPCA were significant. GGE biplots analysis was not done for shoot dry weight because there were no significant interaction between genotypes and environments with regard to that trait. 124 University of Ghana http://ugspace.ug.edu.gh The data of days to flowering were not analyzed because of missing data. For days to 50 % maturity, only the data of five planting dates instead of six were used in the analysis. In the last planting date of 2017 many genotypes were not able to reach 50 % maturity because of drought and Clavigralla tomentosicollis damage. 6.3. Results Signficant differences were detected among genotypes with regard to grain yield, plant height, days to 50 % maturity and fodder yield. Genotypes were not significantly different for shoot dry weight and number of seeds per pod. However, environments were significantly different for all the traits. Significant genotype by environment interaction was observed in grain yield, fodder yield, plant height and days to 50 % maturity (Table 6.4 and 6.5). Table 6.4 Mean squares of traits (combined ANOVA using six planting date‘s data) Mean square Source Gyield PHt SDWt MAT50 *** *** *** *** Env 11302.79 17902.97 25167.94 2386.49 *** *** *** Genotype 1573.24 3371.55 375.88 205.72 *** *** Rep(Env) 185.21 20238.46 1803.61 7.17 * * BLOCK(Rep) 129.72 2641.48 729.55 12.41 *** *** Env*Genotype 352.04 1330.73 271.42 24.40 Error 122.12 1054.70 274.09 4.72 DAF =days to flowering, MAT50 = days to 50 maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield. 125 University of Ghana http://ugspace.ug.edu.gh Table 6.5 Mean squares of traits (AMMI analysis) Mean square SOV Gyield Fyield Height SDWt MAT50 Plength Snumber *** *** *** Genotypes 1870 8666 3948 449ns 261.5*** 13.81*** 12.52** *** *** *** Environments 11778 87472 18059 27232 2480.2*** 76.07** 292.89*** *** *** *** Block 192ns 7524 18075 1635 9.3ns 9.91* 19.11* *** ** * Interactions 391 2396 1625 291ns 27.4*** 5.24ns 4.55ns Error 123 1160 1173 309 5.1 3.49 5.01 DAF =days to flowering, MAT50 = days to 50 maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, ns = non significant. Planting dates were significantly different for all the traits both in 2016 and 2017. But in 2017, no significant difference was detected for striga infestation. Genotypes were also significantly different for traits within each of the two years except for shoot dry weight and number of seeds per pod (Table 6.6 and 6.7). Planting date by genotype interaction was detected for days to flowering, days to 50 % maturity, grain yield, fodder yield and striga infestation (Table 6.7). Table 6.6 Mean squares of traits (combined ANOVA using 2016 planting dates‘ data) Mean square SOV DAF PHt Gyield MAT50 SDWt *** * *** *** *** Planting date 2597.73 6299.84 1348.67 4908.38 2091.51 *** * *** *** Genotype 165.56 2774.95 1348.67 128.28 119.00ns *** *** Rep(Env) 18.01ns 41698.42 27.87ns 5.91ns 1028.84 * Block(Rep) 16.45ns 3271.52 55.77ns 3.99ns 196.15ns * * *** Env*Genotype 21.23 1111.24 110.94 23.15 76.28ns Error 12.62 1278.81 59.23 3.28 121.78 DAF =days to flowering, MAT50 = days to 50 maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, ns = non significant. 126 University of Ghana http://ugspace.ug.edu.gh Table 6.7 Mean squares of traits (combined ANOVA using 2016 planting dates‘ data) Mean square SOV PHt Gyield MAT50 SDWt Striga Plength Snber Fyield *** *** *** *** *** *** *** Planting date 16850.82 24208.73 129.81 6400.33 0.21ns 65.15 264.34 87472.12 * *** *** *** ** *** Genotype 2614.02 1284.52 128.00 602.21ns 4.36 9.42 8.57ns 6637.03 * * * Rep(Env) 1817.39ns 500.42ns 11.16ns 1878.17 0.26ns 12.60 20.07 7113.24ns * ** Block(Rep) 1566.16ns 128.13ns 13.11ns 718.36ns 0.52 2.89ns 1.15ns 1906.61 *** * * * Env*Genotype 1697.32ns 508.49 15.45 418.28ns 0.37 4.86ns 3.59ns 1936.48 Error 1168.59 190.43 6.01 418.28 0.21 3.63 5.89 1038.45 DAF =days to flowering, MAT50 = days to 50 maturity, PHt = plant height, SDWt = shoot dry weight, Gyield = grain yield, striga = striga damage scores, Plength= pod length, Snber = number of seeds per pod, Fyield = Fodder yield, ns = non significant. 127 University of Ghana http://ugspace.ug.edu.gh The interaction components namely IPCA1 and IPCA2 were significant for grain yield, plant height, days to 50 % maturity and fodder yield. They accounted for 86.11, 65.72, 87.04 and 100 % of the G x E variation observed in grain yield, plant height, days to 50 % maturity and fodder yield, respectively. In all these traits, IPCA1 explained the major part of the G x E variation (Table 6.8). The residual contribution to GEI was only high (34 %) with respect to plant height. Table 6.8 Contribution of the interaction components to G x E variation PCA components Gyield PHt Fyield MAT50 IPCA1 70.59 37.99 57.63 60.93 IPCA2 15.52 27.74 42.37 26.11 Residuals 13.89 34.28 0.00 12.95 Total 100.00 100.00 100.00 100.00 MAT50 = days to 50 maturity, PHt = plant height, Gyield = grain yield, Fyield = fodder yield The partitioning of the AMMI sums of square into its different components indicated that genotypes had a medium contribution to the total variation detected in yield and yield components. Genotypes contribution was low with regard to agronomic traits except days to 50 % maturity. However the environments were responsible for the largest part of the variation observed in all the traits. The error term contribution was high in plant height, pod length and number of seeds per pod (Table 6.9). 128 University of Ghana http://ugspace.ug.edu.gh Table 6.9 Partitioning of total AMMI sum of squares into its different components SOV Gyield PHt SDWt Fyield MAT50 Plenght Snber Genotypes 23.89 13.12 3.81 31.70 26.67 32.32 18.06 Environments 42.61 36.70 68.81 38.03 59.78 23.63 51.54 Interactions 24.96 27.00 12.35 17.53 11.16 24.52 13.14 Error 8.54 23.18 15.03 12.73 2.39 19.51 17.25 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 Genotypes contributed 42.83 and 24.69 % to the total variation observed in grain yield in 2016 and 2017, respectively. Planting date contribution to grain yield variation amounted to 16.92% in 2016 and 45.85 % in 2017. The variation in days to 50 % maturity in 2016 was mainly caused by planting date (72.98 %) while genotypes accounted for 18.74 %. Genotype by planting date interactions accounted for 28.5 % and 20.35 % to grain yield variation in 2016 and 2017, respectively. The interaction contributed 6.96 % to days to 50 % maturity variation in 2016. 6.3.1. Days to 50 % maturity (MAT50) A significant reduction in the phenological cycle was observed in all the genotypes following the comparison of the number of days to flowering and days to 50 % maturity of the 2016 first and last planting dates. All the varieties took more days to start flowering or to reach 50 % maturity when planted at the beginning of the season compared to when planting was carried out late. The mean number of days to flowering of the first sowing overpassed that of the third sowing by 16.08 days. The mean difference between the two planting dates with regard to days to flowering ranged from 11.43 in G59 to 30.21 days in G34. The difference between the two planting dates was even larger with respect to days to 50 % maturity. The mean days to 50 % maturity of the third sowing date was 21.18 days lower than that of first sowing. It varied from 15.41 in G80 to 33.86 days in G34 (Table 6.10). 129 University of Ghana http://ugspace.ug.edu.gh Table 6.10 Comparison of the 2016 mean days to flowering and days to 50 % maturity of the first and last planting dates Code Designation DAF1 DAF2 DIFF MAT1 MAT2 DIFF 16 Dan Illa INRAN 54.52 39.72 14.80 87.39 63.66 23.72 33 Dan Batsatsa Kornaka 51.28 39.60 11.68 80.89 61.40 19.48 34 Bougou Bougou 75.55 45.35 30.21 98.64 64.78 33.86 35 Dan Dam Ali neino 60.85 45.35 15.49 98.71 66.93 31.77 36 Tcheto Dan kassari 53.05 41.28 11.76 82.99 66.39 16.60 37 Tessa Abdoulaye mai goshi 62.08 43.76 18.33 86.54 65.98 20.56 38 Bagagi 21 ans 51.26 37.10 14.16 79.48 61.90 17.58 45 Gouma Sami Godje 53.30 40.56 12.74 79.31 64.78 14.53 46 Dan Bawada Daji 58.85 42.42 16.43 81.78 65.07 16.71 47 El Danke Maitsakoni 73.92 50.60 23.32 - - - 51 Farin wake Matankari 53.30 43.35 9.95 78.31 62.21 16.10 54 Salga Sabongari 54.72 38.80 15.92 79.29 62.81 16.48 56 Dan Mangaza Wangarawa 55.60 42.57 13.03 78.40 62.39 16.02 59 Darey kiota est 54.28 42.86 11.43 84.89 65.87 19.02 63 Arnen wake Matankari 54.08 39.42 14.66 83.54 63.57 19.98 69 Gnagaou Lido 55.52 41.08 14.44 85.39 66.04 19.34 71 Hwarin wake kornaka 52.85 40.85 12.01 84.43 61.71 22.73 74 Baadare Karanguiya 68.02 44.56 23.46 100.46 67.78 32.68 78 Lakade Maiki 74.43 48.44 25.98 100.73 70.97 29.76 80 Dan Baourota kornaka 51.06 39.36 11.70 78.78 63.37 15.41 Mean 58.42 42.35 16.08 85.79 64.61 21.18 DAF1 = days to flowering of the first planting date. DAF2 = days to flowering of the last planting date. MAT1 = days to 50 % maturity of the first planting date. MAT2 = days to 50 % maturity of the first planting date. DIFF = mean of the first planting date minus that of the last planting date. Mean days to 50 % maturity was 72.77 and ranged from 64.56 to 86.06 days after planting. The highest and the lowest means were observed at E1 and E3 respectively. The mean number of days to 50 % was 86.06, 67.40 and 64.56 for first, second and third planting dates in 2016. In 2017, the means were 74.11 and 71.73 for the first and second planting dates respectively. The earliest and latest maturing cultivars were G51 and G78 in 2016 while in 2017, the first genotype to reach maturity was G80 and the last was G74. Across the environments, G80 was the earliest maturing genotype with a mean of 67.27 days to 50 % maturity. Late maturing cultivars were 130 University of Ghana http://ugspace.ug.edu.gh G74, G34, G78 and G35 with the means of 84.13, 82.15, 80.06 and 79.98 days to 50% maturity, respectively. The most stable and early maturing genotype was G63 which recorded an overal mean of 70.50 days to 50 % maturity and a stability value of 0.1 (Table 6.11). Table 6.11 Mean days to 50 % maturity and stability values of genotypes across the environments Code Genotype E1 E2 E3 E4 E5 Mean ASV G63 Arnen wake 83.51 65.26 62.47 73.01 68.26 70.50 0.1 G71 Hwarin wake 84.26 63.26 62.76 73.76 67.26 70.26 0.9 G38 Bagagi 79.15 63.51 62.65 71.15 69.01 69.10 1.1 G46 Bawada 81.26 65.27 63.97 71.26 70.27 70.41 1.3 G51 Farin wake 77.97 63.27 63.26 73.47 68.27 69.25 1.3 G56 Dan mangaza 78.97 64.14 63.51 70.47 67.27 68.87 1.3 G16 Danila 87.59 61.97 63.47 73.09 69.97 71.22 1.5 G33 Dan Batsatsa 80.77 65.97 61.59 68.27 68.97 69.11 1.5 G54 Salga 78.33 64.77 62.47 68.83 70.27 68.94 1.6 G59 Darey 84.77 67.51 66.01 69.27 70.51 71.62 1.8 G69 Gnagaou 85.33 68.64 65.27 69.33 72.76 72.27 1.9 G36 Tcheto 83.90 67.97 66.27 71.90 74.97 73.00 2.0 G80 Dan baourota 78.52 63.02 63.26 65.52 66.02 67.27 2.4 G37 Tessa 85.77 69.91 65.65 68.77 68.65 71.75 2.5 G34 Bougou bougou 97.65 79.47 65.76 87.15 80.69 82.15 3.1 G78 Lakade 100.90 79.64 70.09 75.90 73.76 80.06 3.6 G35 Dan dam 100.01 67.91 66.02 86.51 79.47 79.98 5.3 G74 Baadare 100.33 71.64 67.52 96.33 84.83 84.13 7.6 Mean 86.06 67.40 64.56 74.11 71.73 72.77 E1 = first planting date in 2016. E2 = second planting date in 2016. E3 = last planting date in 2016. E4 = first planting in 2017. E5 = second planting date in 2017. E6 = last planting date in 2017, ASV = AMMI Stability Value. The test environments fell into two mega-environments with different winning genotypes. G78 was the best cultivar in the first mega-environment which was composed of E2 and E3 while G74 won in the second, which is made up of E1, E4 and E5 (Figure 6.1). 131 University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 94.60%) G78 E2 G37 GG5969 E1 E3 G34 G80 G33 G36 G54 G56G63 E5 G16 G38G71 G35 G51 E4 G74 PC1 - 83.50% Genotype scores Environment scores Convex hull Sectors of convex hull Mega-Environments Figure 6-1 GGE biplots of days to 50 % maturity E1 and E4 had similar power in discriminating genotypes as their vectors had approximatively the same length. However, E1 was most representative since it had smaller angle with PCA1 which explained the largest part of the variation (83.5 %) (Figure 6.2). 132 PC2 - 11.10% University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 94.60%) G78 E2 G37 E1 GG5969 E3 G34 G80 G33 G36 G54 G56 G63 G16 E5 G38 G71 G35 G51 G74 E4 PC1 - 83.50% Genotype scores Environment scores Vectors Figure 6-2 Discriminating power and representativeness of the environments 6.3.2. Grain yield (Gyield) Table 6.12 shows the results for grain yield per plant over all test environments. Mean grain yield varied from 11.67 to 58.15 g per plant with the highest recorded at E4 and the lowest at E6. The mean grain yield was 23.35, 15.94 and 12.67 g per plant for the first, second and third planting date in 2016 respectively. It was 58.15, 19.75 and 11.67 g per plant for planting date1, 2 and 3 in 2017. The best performing genotypes in 2016 were G63 and G33 while in 2017; the winning genotypes were G54, G59 and G56. G63, G33 and G71 produced highest grain yield for planting date1, 2 and 3 respectively in 2016. In 2017, G54 won in sowing date1 whereas G59 and G56 were the superior cultivars in date2 and 3 respectively. All the genotypes except G34, G35, G36 and 37 produced higher grain yield at E4 in comparison to the other environments. The first planting dates recorded the highest mean grain yield in both 2016 and 2017 while the 133 PC2 - 11.10% University of Ghana http://ugspace.ug.edu.gh lowest belonged to last planting dates. G56 and G34 were the best and worse performing genotypes, respectively with 40.47 and 1.78 g grain yield per plant across the test environments. Other lines with high grain yield were G33, G59, G54, G63 and G38. Genotypes G34, G35, G78, G74 and G47 had lowest performance across the environments. Higher grain yield was obtained in 2017 compared to 2016. All the high yielding cultivars were not stable. The most stable genotype was G46 which had an ASV of 0.2 and a mean of 24.04 g per plant. Table 6.12 Mean grain yield per plant and stability values of genotypes across test environments Code Genotype E1 E2 E3 E4 E5 E6 Mean ASV G56 Dan mangaza 32.78 28.95 26.67 95.65 30.77 28.02 40.47 8.5 G33 Dan Batsatsa 32.60 45.67 13.57 79.33 44.24 16.36 38.63 5.6 G59 Darey 26.62 19.58 24.09 104.17 34.90 20.21 38.26 15.8 G54 Salga 35.55 7.99 20.41 120.21 16.36 22.36 37.15 21.2 G63 Arnen wake 63.88 23.67 19.53 64.77 24.27 15.18 35.22 4.8 G38 Bagagi 37.34 17.97 14.74 84.42 33.61 16.83 34.15 8.4 G71 Hwarin wake 24.96 29.38 38.35 62.86 24.54 14.23 32.39 1.6 G80 Dan baourota 22.70 27.49 9.99 93.05 6.01 16.30 29.26 14.6 G51 Farin wake 32.84 24.73 19.49 69.55 3.94 12.51 27.18 2.4 G16 Danila 40.63 17.59 8.11 44.37 34.96 3.87 24.92 6.8 G46 Bawada 23.12 23.74 11.98 58.45 11.49 15.49 24.04 0.2 G69 Gnagaou 20.63 7.96 11.94 53.12 26.42 5.56 20.94 2.2 G47 Eldanke 3.31 -0.23 0.89 78.31 18.53 10.89 18.62 10.6 G36 Tcheto 29.84 12.79 3.84 17.33 9.23 -1.39 11.94 13.7 G78 Lakade 2.57 2.49 2.43 31.96 9.79 9.46 9.78 6.3 G74 Baadare 3.10 3.48 2.90 26.04 13.15 6.48 9.19 9.1 G37 Tessa 8.12 6.57 6.16 10.40 13.64 1.30 7.70 15.4 G35 Dan dam 4.90 0.83 3.51 10.46 13.33 6.46 6.58 15.9 G34 Bougou bougou -1.85 2.22 2.17 0.49 6.06 1.57 1.78 16.7 Mean 23.35 15.94 12.67 58.15 19.75 11.67 23.59 E1 = first planting date in 2016. E2 = second planting date in 2016. E3 = last planting date in 2016. E4 = first planting in 2017. E5 = second planting date in 2017. E6 = last planting date in 2017, ASV = AMMI Stability Value. 134 University of Ghana http://ugspace.ug.edu.gh The results showed that two mega-environments were formed. The first was composed of E1, E2, E3 and E5 and had G63 as the winning genotype. The second consisted E4 and E6 and the best performing cultivar was G54 (Figure 6.3). Scatter plot (Total - 88.74%) G63 G16 E1 G33 G36 G71 E2 E5 GE38 G37 G51 G56 G6G946 E6 G35 G34 G59 G74 E4 G78 G80 G54 G47 PC1 - 76.17% Genotype scores Environment scores Convex hull Sectors of convex hull Mega-Environments Figure 6-3 GGE biplots of grain yield per plant E4 was the most discriminating environment as it had the longest vector along the PC1 which accounted for 76.17 % of the yield variability; it also had the smallest angle with that principal component axis, hence most representative among test environments (Figure 6.4). 135 PC2 - 12.57% University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 88.74%) E1 G63 G16 E2 G36 E5 G33 G71E3 G38 G37 G51 G6G9 46 G56 G35 E6 G34 G74 G59 G78 G80 G54 E4 G47 PC1 - 76.17% Genotype scores Environment scores Vectors Figure 6-4 Discriminating power and representativeness of environments 6.3.3. Shoot dry weight (SDWt) Lower shoot dry weight was obtained in 2016 compared to 2017. It varied from 26.20 to 86.57 g per plant across the environments. In 2017, planting date1 and 2 recorded significantly higher shoot dry weight in comparison to the last planting date. In contrast, the last planting date recorded higher shoot dry weight compared to the two first sowing dates in 2016. The overall mean was 54.99 per plant. The best performance was observed in G56 (66.40 g) while G59 had the least (40.84 g). None of the lines was stable as the stability values ranged from 1.3 to 7.8 (Table 6.13). 136 PC2 - 12.57% University of Ghana http://ugspace.ug.edu.gh Table 6.13 Mean shoot dry weight per plant and stability values of genotypes across test environments Code Genotype E1 E2 E3 E4 E5 E6 Mean ASV G56 Dan mangaza 27.37 18.44 41.39 130.24 102.43 78.55 66.40 6.7 G47 Eldanke 22.12 36.31 47.21 109.29 87.25 78.36 63.42 3.8 G51 Farin wake 16.93 37.66 42.51 127.39 93.76 56.31 62.43 7.8 G34 Bougou bougou 41.09 47.74 50.01 97.12 75.10 57.45 61.42 3.9 G74 Baadare 29.63 27.71 52.24 69.37 112.75 67.68 59.90 7.4 G69 Gnagaou 29.50 30.11 53.06 73.91 92.86 63.87 57.22 5.0 G78 Lakade 30.30 27.21 32.30 111.35 80.67 59.93 56.96 4.6 G36 Tcheto 31.94 29.54 31.63 84.01 91.22 63.79 55.36 1.9 G35 Dan dam 28.43 35.53 26.46 86.00 75.86 77.15 54.90 3.1 G71 Hwarin wake 14.27 36.78 62.33 74.97 92.55 45.69 54.43 2.1 G54 Salga 12.49 32.30 36.22 92.17 92.05 60.09 54.22 2.8 G63 Arnen wake 38.18 23.22 44.01 60.92 85.02 71.89 53.87 5.9 G33 Dan Batsatsa 28.45 41.12 30.07 61.36 82.66 76.49 53.36 4.7 G46 Bawada 23.07 28.65 36.22 100.76 70.55 59.39 53.11 4.2 G37 Tessa 25.25 33.00 45.19 74.54 78.18 57.62 52.30 1.3 G38 Bagagi 33.03 29.59 35.52 63.20 70.17 61.49 48.83 4.7 G16 Danila 17.22 30.20 41.36 73.72 67.26 58.30 48.01 1.6 G80 Dan baourota 21.52 19.85 25.92 75.93 92.80 51.01 47.84 2.7 G59 Darey 27.05 21.59 28.92 78.52 46.70 42.26 40.84 2.8 Mean 26.20 30.87 40.14 86.57 83.67 62.49 54.99 E1 = first planting date in 2016. E2 = second planting date in 2016. E3 = last planting date in 2016. E4 = first planting in 2017. E5 = second planting date in 2017. E6 = last planting date in 2017, ASV = AMMI Stability Value. 6.3.4. Plant height (PHt) The environments of 2017 recorded higher plant height compared to those of 2016 (Figure 6.5). The means varied from 137.49 to 216.82 cm across the environments. Higher plant height was observed in all the genotypes at E5 except G33 and G35. Within the years, the second planting date recorded highest plant height. G34 was the cultivar with the highest mean plant height across the environments while G16 had the lowest. There was re-ranking of genotypes among environments resulting in low stability (Table 6.14). 137 University of Ghana http://ugspace.ug.edu.gh Figure 6-5 Boxplot for plant height means 138 University of Ghana http://ugspace.ug.edu.gh Table 6.14 Mean plant height and stability values of genotypes across test environments Code Genotype E1 E2 E3 E4 E5 E6 Mean ASV 34 Bougou bougou 195.59 224.04 229.87 199.67 243.04 208.70 216.82 5.8 54 Salga 135.77 148.63 135.21 186.35 262.63 260.71 188.21 8.8 36 Tcheto 169.10 186.46 163.46 138.35 204.71 212.63 179.12 3.4 37 Tessa 184.06 200.34 148.41 173.81 192.84 168.74 178.03 3.9 51 Farin wake 139.54 179.88 134.68 207.96 208.88 179.34 175.05 4.6 78 Lakade 180.28 131.20 172.74 178.45 195.20 173.74 171.94 2.0 47 Eldanke 111.93 148.68 167.15 198.10 202.68 201.90 171.74 5.4 38 Bagagi 157.82 172.78 140.81 159.09 215.03 180.15 170.95 1.4 74 Baadare 165.23 139.71 161.35 173.15 213.96 152.60 167.66 2.4 46 Bawada 148.20 176.06 125.10 187.20 194.81 174.60 167.66 2.4 59 Darey 178.13 181.42 157.42 177.13 184.67 106.42 164.20 4.8 56 Dan mangaza 100.31 142.13 100.15 211.10 228.79 191.49 162.33 9.7 69 Gnagaou 158.87 126.34 153.90 140.04 213.84 175.31 161.39 1.2 63 Arnen wake 180.40 138.42 120.27 141.81 192.42 185.60 159.82 3.8 71 Hwarin wake 170.78 187.20 100.34 152.03 190.20 150.09 158.44 3.4 35 Dan dam 157.18 154.04 141.94 133.09 140.04 176.28 150.43 6.5 33 Dan Batsatsa 182.96 126.85 80.98 142.13 174.35 192.90 150.03 5.7 80 Dan baourota 164.18 118.10 73.59 157.35 224.60 143.92 146.95 5.0 16 Danila 124.87 149.35 117.54 126.79 173.60 132.79 137.49 1.7 Mean 158.17 159.56 138.15 167.56 202.96 177.26 167.28 E1 = first planting date in 2016. E2 = second planting date in 2016. E3 = last planting date in 2016. E4 = first planting in 2017. E5 = second planting date in 2017. E6 = last planting date in 2017, ASV = AMMI Stability Value. Three meg-environments were formed with the first containing E1, the second containing E2 and E3 and the last was composed of E4, E5 and E6. G59 was the winning cultivar at E1, G34 won at E2 and E3 while the winning genotypes in the last mega-environment were G54 and G56 (Figure.6.6). 139 University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 65.80%) E1 G59 E3 E2 G34 G37 G35 G71 G16 G78G74 G36 G63 G38G6G9 46 G51 G33 G80 G47 E4 E5 G56 E6 G54 PC1 - 39.06% Genotype scores Environment scores Convex hull Sectors of convex hull Mega-Environments Figure 6-6 GGE biplots for plant height The scatter biplot explained only 69.8 % of the plant height variation. E3, E5 and E6 discriminated genotypes better than the remaining environments in terms of plant height; however E3 was the most discriminating representative among test environments as it had the longest vector and formed the smallest angle with PC1 axis (Figure 6.7). 140 PC2 - 26.75% University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 65.80%) E1 G59 E3 E2 G34 G37 G35 G71 G16 G78G74 G36 G63 G38G69G46 G51 G33 G80 G47 E4 E5 G56 E6 G54 PC1 - 39.06% Genotype scores Environment scores Vectors Figure 6-7 Discriminating power and representativeness of environments 6.3.5. Pod length (Plength) Table 6.15 shows the mean pod length per genotype. The pod length showed low variability between planting dates as it ranged from 9.55 (E6) to 12.26 cm (E4). The mean across the first, second and third sowing dates was 10.76 cm. The longest and shortest pods were observed in G59 and G34 with respectively 12.96 and 6.12 cm. G59 was the winning genotype at all the test environments since G x E was not significant. 141 PC2 - 26.75% University of Ghana http://ugspace.ug.edu.gh Table 6.15 Mean pod length of genotypes across the 2017 planting dates Code Genotype E4 E5 E6 Mean 59 Darey 16.04 13.85 8.99 12.96 46 Bawada 13.26 10.20 13.93 12.46 33 Dan batsatsa 13.76 12.25 10.83 12.28 78 Lakade 12.73 11.79 11.66 12.06 74 Baadare 15.03 7.89 11.60 11.51 71 Hwarin wake 13.19 11.14 9.71 11.35 56 Dan mangaza 14.35 10.95 8.61 11.30 80 Dan baraouta 13.40 12.83 7.37 11.20 58 Dan bekori 10.91 12.52 10.09 11.17 38 Bagagi21 14.34 9.12 9.30 10.92 37 Tessa 12.03 10.46 10.19 10.89 51 Farin wake 11.69 12.06 8.88 10.87 54 Salga 14.01 9.32 8.98 10.77 35 Dan dam 12.64 9.91 9.51 10.69 63 Arnen wake 11.03 9.63 8.30 9.65 36 Tcheto 11.09 9.67 7.90 9.55 69 Gnagaou 9.99 10.36 8.11 9.49 16 Danila 9.35 7.93 10.17 9.15 34 Bougou bougou 4.00 6.98 7.39 6.12 Mean 12.26 10.47 9.55 10.76 E4 = first planting, E5 = second planting date, E6 = last planting date. 6.3.6. Seed number per pod (Snumber) Large variability was observed among planting dates regarding number of seeds per pod. Significantly higher seeds per pod were obtained at the first planting date than the two other environments. The number of seeds per pod at sowing date1 (8.80) was more than two folds that of sowing date3 (3.63). The overall mean number of seeds per pod was 5.62. G59 recorded the highest mean while G34 had the lowest (Table 6.16). 142 University of Ghana http://ugspace.ug.edu.gh Table 6.16 Mean number of seeds per pod of genotypes across 2017 planting dates Code Genotype E4 E5 E6 Mean 59 Darey 10.97 8.38 4.25 7.87 33 Dan batsatsa 10.22 7.12 4.25 7.20 71 Hwarin wake 10.69 6.42 4.06 7.06 54 Salga 10.28 4.64 5.35 6.76 80 Dan baraouta 9.82 5.32 5.08 6.74 58 Dan bekori 8.25 6.46 5.24 6.65 56 Dan mangaza 10.61 6.26 2.66 6.51 38 Bagagi21 9.53 4.23 5.25 6.34 37 Tessa 9.84 5.11 3.20 6.05 16 Danila 7.27 4.12 5.34 5.58 51 Farin wake 8.94 4.14 3.28 5.45 74 Baadare 11.38 0.73 3.40 5.17 78 Lakade 8.71 2.20 4.23 5.05 46 Bawada 8.75 4.72 1.48 4.98 63 Arnen wake 7.51 4.25 3.11 4.95 35 Dan dam 8.39 2.01 2.36 4.25 36 Tcheto 7.97 2.27 1.72 3.99 69 Gnagaou 5.52 3.68 2.68 3.96 34 Bougou bougou 2.55 2.09 2.09 2.24 Mean 8.80 4.43 3.63 5.62 E4 = first planting in 2017, E5 = second planting date in 2017, E6 = last planting date in 2017. 6.3.7 Fodder yield (Fyield) The first and second planting dates were the most favorable environments for fodder production as they recorded significantly higher means than the third date (Figure 6.8). The fodder means were respectively 149.36, 139.95 and 64.07 g per plant for sowing date2, date1 and date3. The mean fodder yield across the environments was 117.19 g per plant. Superior genotypes were G35, G78 and G74 with 178.50, 173.63 and 167.37 g per plant. The cultivars with least performance were G58, G34 and G38 with a yield of 49.34, 73.84 and 74.72 g respectively. All 143 University of Ghana http://ugspace.ug.edu.gh the high yielding genotypes were not stable. G36 was a medium yielding and most stable genotype (ASV = 0.8) (Table 6.17). Figure 6-88 Boxplot for fodder means per plant 144 University of Ghana http://ugspace.ug.edu.gh Table 6.17 Mean fodder yield per plant and stability values of genotypes across 2017 planting dates Code Genotype E4 E4 E6 Mean ASV 35 Dan dam 251.75 185.44 98.30 178.50 7.9 78 Lakade 222.27 215.09 83.52 173.63 7.1 74 Baadare 229.90 197.24 74.98 167.37 5.0 37 Tessa 175.53 214.23 66.10 151.95 5.6 51 Farin wake 183.35 180.49 52.04 138.63 2.5 46 Bawada 156.88 165.53 88.19 136.87 3.8 47 Eldanke 206.00 102.06 80.51 129.52 5.6 63 Arnen wake 119.59 191.43 76.32 129.11 4.1 36 Tcheto 152.69 159.44 59.09 123.74 0.8 69 Gnagaou 129.24 166.42 59.44 118.36 3.3 71 Hwarin wake 134.23 168.60 43.76 115.53 3.5 56 Dan mangaza 155.85 136.89 52.55 115.10 2.0 54 Salga 110.82 150.65 74.60 112.02 3.2 80 Dan baourota 114.35 127.63 51.90 97.96 0.5 16 Danila 72.05 113.66 107.54 97.75 7.5 59 Darey 115.18 101.90 45.34 87.47 2.0 33 Dan Batsatsa 72.06 111.32 70.04 84.47 5.0 38 Bagagi 61.10 108.48 54.57 74.72 4.3 34 Bougou bougou 83.17 130.66 7.66 73.83 3.7 58 Dan bekori 53.01 60.04 34.98 49.34 6.8 Mean 139.95 149.36 64.07 117.79 E4 = first planting. E5 = second planting date. E6 = last planting date, ASV = AMMI Stability Value. The environments were grouped into two mega-environments. The first contained E5 and had G78 as winning cultivars. The second was made up of E4 and E6 and G35 was the best performing line (Figure.6. 9). 145 University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 92.58%) G47 G58 G35 E4G59 G56 G80 E6 G74 GG3136 GG3646 G38 G51 G78 G54 G34 G6791 G37 G63 E5 PC1 - 80.03% Genotype scores Environment scores Convex hull Sectors of convex hull Mega-Environments Figure 6-9 GGE biplots for fodder means per plant E4 provided much information about the differences among genotypes in terms of fodder yield since its vector was considerably longer compared to that of other environments and it was along the biplots axis which accounted for 80.03 % of the trait variation. E4 also formed the smallest angle with PC1 and was therefore most representative of the mega-environment to which it belonged. E6 had shortest vector which means that the performance of all the genotypes was similar in that environment (Figure 6.10). 146 PC2 - 12.55% University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 92.58%) G47 E4 G58 G35G59 G56 E6 G80 G74 GG3316 G3G646 G38 G51 G78 G54 G34 G6791 G37 G63 E5 PC1 - 80.03% Genotype scores Environment scores Vectors Figure 6-10 Discriminating power and representativeness of planting dates 6.3.8. Striga Most of the genotypes were either resistant or tolerant to striga. The only cultivars that expressed susceptibility to striga were G34 and G16 (Table 6.18). G34 was however the most susceptible as in many plots, its plants died before reaching maturity. It had been the most susceptibe cultivar across the planting dates. 147 PC2 - 12.55% University of Ghana http://ugspace.ug.edu.gh Table 6.18 Mean scores of genotypes for tolerance/susceptibility to striga across 2017 planting dates Code Genotype E4 E5 E6 Mean G47 Eldanke 0.52 1.48 0.83 0.94 G33 Dan batsatsa 1.08 0.52 1.33 0.98 G58 Dan bekori 0.83 1.08 1.13 1.01 G35 Dan dam 1.29 0.51 1.28 1.03 G51 Farin wake 0.84 1.08 1.29 1.07 G56 Dan mangaza 1.02 1.04 1.22 1.10 G63 Arnen wake 1.23 1.79 0.52 1.18 G74 Baadare 1.03 1.54 1.13 1.24 G38 Bagagi21 1.93 1.58 0.83 1.45 G78 Lakade 1.58 1.28 1.49 1.45 G71 Hwarin wake 1.78 1.28 1.29 1.45 G80 Dan baourota 1.13 2.13 1.29 1.52 G59 Darey 1.08 2.29 1.29 1.55 G54 Salga 1.05 2.08 1.84 1.66 G69 Gnagaou 2.04 1.79 2.23 2.02 G37 Tessa 2.23 1.93 2.11 2.09 G46 Bawada 2.28 2.73 1.52 2.18 G36 Tcheto 2.76 1.84 2.08 2.23 G16 Danila 4.04 2.02 2.01 2.69 G34 Bougou bougou 4.76 4.51 5.28 4.85 Mean 1.73 1.73 1.60 1.68 E4 = first planting. E5 = second planting date. E6 = last planting date. Only one mega-environment was formed as there was no re-ranking of genotypes across the planting dates, in other words, the tolerant as well as the susceptible cultivars did not vary with environments (Fig. 6.11). 148 University of Ghana http://ugspace.ug.edu.gh Scatter plot (Total - 93.81%) E5 GG5954 G80 G47 G34 G46 GG6734 E6 G5581 G56 GG6397 G78 G3781 G33 G36 G35 E4 G16 PC1 - 84.09% Genotype scores Environment scores Convex hull Sectors of convex hull Mega-Environments Figure 6-9 GGE biplots for striga infestation 6.4. Discussion The significant genotype by planting date interaction detected for days to flowering, days to 50 % maturity, grain and fodder was an indication of planting date effect on these agronomic and yield parameters. This finding was consistent with Azari and Khajepour (2003) who asserted that planting date was part of the major factors affecting growth and yield of leguminous crops. It further agreed with Morakinyo and Ajibade (1998) who said that the quantity as well as the quality of cowpea seeds was influenced by rainfall amount and distribution which in turn was affected by planting time. Although, rainfall amount was greater in 2016 compared to 2017, significantly higher grain yield was recorded in the latter than the former. This may be because 149 PC2 - 9.72% University of Ghana http://ugspace.ug.edu.gh the soil of the 2016 experimental site was low in fertility, especially P in comparison to that of 2017, this support the claim of Payne (1997) who stated that low soil fertility causes yield reduction more than rainfall in the driest agro-ecology of the Sahel and other areas of West Africa. The wide reduction in number of days to first flower appearance and days to 50 % maturity exhibited by all the genotypes in the last planting compared to first planting especially in 2016 was a signal of the cultivars sensitivity to photoperiod since the two periods were different in terms of daylength. Similar findings were reported by Akande et al. (2012) who noticed that cowpea varieties took long to reach 50 % maturity when planted in August compared to when they were planted in September. Similarly, a wide variation in days to 50 % flowering was exhibited by two late maturing improved varieties (IT99K-7-21-2-2 and IT99K-216-24-2) between July and August planting dates at Minjibir, an IITA station located in northern Guinea savanna of Nigeria (Sanda and Maina, 2013). Ehlers and Hall (1996) classified cowpea varieties as day neutral, quantitative short day and obligate short day based on their response to photoperiod and temperature. According to their classification, the lines tested in the present study exhibited a quantitative short day response as they expressed an increase in the number of days to appearance of flower buds under long days in comparison to short days. Furthermore, G34, G74, G78 and G47 seemed to be more sensitive to photoperiod as they started flowering towards the end of the season when the short days usually start in Niger. The sensitivity to photoperiod detected in these farmers‘ cultivars was in agreement with Wien and Summerfield (1980) who while evaluating 24 cowpea accessions collected from farmers‘ fields in Niger and Nigeria for adaptation to photoperiod, stated that flower initiation in local varieties of cowpea has been assigned to photoperiod control. 150 University of Ghana http://ugspace.ug.edu.gh There were differential adaptation of genotypes to planting dates with regards to yield and yield components within and across the two years. Delay in planting resulted in significant reduction in grain yield. As a result, the planting dates of June 30 and July 7 (early plantings) recorded the highest grain yield, pod length and number of seeds per pod in 2016 and 2017. These findings were consistent with Mirzaienasab and Mojaddam (2014) and Dhital et al. (1997). It also corroborated with Ezeaku et al. (2015) who reported that early planting resulted in significantly higher yield and yield components compared to late sowing in derived savanna ecology of southeastern Nigeria. However, Akande et al. (2012) used four planting dates (June, July, August and September) and identified August sowing date as best adapted for grain yield in southern Guinea savanna of Nigeria. Also Mbong et al. (2010) stated that grains of early sowing were low in quantity and quality compared to those of late planting as a result of scab infection, hence in areas prone to scab disease late planting may be better than early planting. The higher mean grain yield obtained from first planting dates in the present work was due to high performance of early maturing (less photosensitive) cultivars since late maturing highly photosensitive varieties (G34, G74, G78, G47 and G35) produced very low yield because the onset of its flowering occurred towards the end of the season regardless of planting date. In 2016 and 2017, the end of the rainy season was characterized by moisture as well as insect stresses and it negatively affected the yield of late maturing cultivars and late planted early maturing varieties. The combined effect of drought and Clavigralla tomentosicollis was the main reason why first sowing date (July) recorded considerably higher number of seeds per pod compared to last planting date (August). Innocent et al. (2013) found insignificant yield loss in early season planting while the loss was high and even total for late maturing varieties when planted late. The same authors noticed that population density of bruchids, maruca, pod sucking bugs and thrips 151 University of Ghana http://ugspace.ug.edu.gh was higher in late than early season. Similar findings were reported by Asante et al. (2001) who stated that the flowering and pod setting phases of late planted cultivars occurred at the time when the population of the three major post-flowering insects was at its peak leading to an economic yield loss. Ansobah et al. (2013) recommend early planting in Ghana to take advantage of more suitable weather conditions and less insects‘ damage. From the present results, high grain yield can be obtained in late planted cowpea when varieties used are tolerant to both terminal drought and pod sucking bugs, especially Clavigralla tomentosicollis. Sources of tolerance to drought (Fatokun et al., 2012; Watanabe et al., 1997) and to Clavigralla tomentosicollis (Koona et al., 2002; Dabire-Binso et al., 2010) had been reported, so development of varieties tolerant to both stresses is theoretically possible. The present studies revealed that biological yield of cowpea was affected by planting time as very low fodder yield was recorded in the last planting date compared to the two first sowing. Mojaddam and Nouri (2014) detected significant planting date by genotype interaction; however they observed that fodder yield of July 23 sowing was 6.7 % greater than that of July 6. Sanda et al. (2013) also found wide fodder yield variation among improved cowpea varieties according to planting dates; they noticed that September sowing gave significantly lower yield in comparison to July and August planting dates. 6.5. Conclusion Planting date had a significant effect on yield and yield related traits of local cowpea cultivars in Niger according to the results of the present work. Significant reduction in days to flowering and days to 50 % maturity with subsequent decrease in grain and fodder yield was observed when th varieties were planted late. Therefore, planting should be done from the end of June to July 10 in order to obtain optimum yield of farmers cowpea varieties in Niger, however highly 152 University of Ghana http://ugspace.ug.edu.gh photosensitive cultivars should be avoided regardless of the planting date especially for grain production. No cultivar had been consistently the best across the planting dates; however G56, G59, G54, G33, G63 and G71 showed relatively good performance and further studies are needed to determine which of them to recommend for all or each sowing date. 153 University of Ghana http://ugspace.ug.edu.gh CHAPTER SEVEN 7.0. General conclusions and recommendations 7.1. General conclusions The most widely crops cultivated across the study areas were millet, cowpea, sorghum and groundnut. However millet and cowpea were the only ones well adapted to almost all the surveyed areas. Cowpea was mainly grown in association with millet. Farms were generally inherited as more than 82 % of respondents affirmed obtaining their farms from their parents. The mean farm size per household was 5.58 ha and varied from 3.23 to 7.24 across the study areas. The soil texture of farms was sandy according to the majority of respondents and most of the farms were located far from homes. About 60 % of respondents claimed that their farms were low in fertility, 75 % stated that their cowpea production was low. The results of both focus group discussion and household questionnaire revealed that drought and low soil fertility were the top constraints impeding cowpea production and yield. The major causes of soil fertility deterioration were collection of crop residues from farms and lack of fallow. Early maturity and high yielding potential were the farmers‘ preferences in terms of cowpea varieties. There was low level of formal education among farmers as more than 72 % of respondents have not been to school. High genetic diversity was observed among genotypes with regard to adaptation to low soil P conditions. Lower means were recorded in dry season compared to rainy season for all the traits; however the difference was much larger in terms of plant height and shoot dry weight. There was differential adaptation of genotypes to the two seasons with respect to grain yield, but cultivars such as G118, G150, G149, G131 and G11 produced consistently high yield in both seasons. 154 University of Ghana http://ugspace.ug.edu.gh G131 and G11 recorded even higher grain in dry compared to rainy season. Higher estimates of heritability, genetic variance, phenotypic variance as well as GCV and PCV were obtained in the rainy season. Grain yield was positively correlated with shoot dry weight, plant height and shoot color ; it was however negatively correlated with days to flowering and days to 50 % maturity. The highest positive correlation was detected between days to flowering and days to 50 % maturity. Significant (p < 0.01. p < 0.001) genotype by environment interaction (G x E) was detected for all the parameters except grain yield, plant height, shoot dry weight, number of seeds per pod and pod length. The cumulative percentage of the PCA axis accounted for 93.66 and 84.51% of the total GEI variation in days to flowering and days to 50 % maturity respectively. However in the case of fodder yield, the whole GEI variation was due to the PCA components with the residual having zero contribution. The largest contribution of the genotypes to traits variation was observed in fodder yield (46.76 %) while the smallest was in shoot dry weight (1.5 %). Genotypes were also responsible for a significant variation with regard to days to flowering (21.57%) as well as grain yield (15.06 %). The environment was the main cause of the variability detected in shoot dry weight, plant height, days to 50 % maturity and grain yield with 82.45, 72.04, 58.69 and 36.55 % respectively. The GEI part of the variation was high in days to flowering (36.85 %), fodder yield (22.73 %) and days to 50 % maturity (18.66 %). The G x E interaction, although not significant for grain yield, contributed considerably (15.85 %) to the yield variation. The error had a great contribution in the variation observed in grain yield (32.54%). G118 and G98 were the best and worst performing cultivars with mean grain yield of 48.33 and 20.35 g per plant respectively. However none of them was stable according the AMMI stability value. G150 had a mean grain yield of 45.23 g per plant together with a stability value 155 University of Ghana http://ugspace.ug.edu.gh of 0.6, making it to be the most stable and high yielding cultivar. F2 seeds were obtained from the crosses between G150, G118 (tolerant) on one hand and G5 (susceptible to P deficient soils) on the other. The cultivars means fodder yield across the environments varied from 31.24 to 121.42 g per plant. The superior genotypes were G197, G118 and G164 with respectively 121.42, 108.79 and 104.59 g per plant. However, all the high fodder yielding varieties showed differential adaptation to environments. The widely adopted improved varieties were highly instable and low in fodder yield. The significant genotype by planting date interaction detected for days to flowering, days to 50 % maturity, grain and fodder was an indication of planting date effect on these agronomic and yield parameters. First planting dates recorded the highest mean grain yield in both 2016 and 2017 while the lowest belonged to last planting dates. Mean grain yield was 23.35, 15.94 and 12.67 g per plant for the first, second and third planting date in 2016 respectively. It was 58.15, 19.75 and 11.67 g per plant for planting date1, 2 and 3 in 2017. The best performing genotypes in 2016 were G63 and G33 while in 2017, the winning genotypes were G54, G59 and G56. G63, G33 and G71 produced highest grain yield for planting date1, 2 and 3 respectively in 2016. In 2017, G54 won in sowing date1 whereas G59 and G56 were the superior cultivars in date2 and 3 respectively. 7.2. Recommendations Extension workers should encourage farmers to not only leave crop residues on the farm but also to apply manure to reduce soil fertility deterioration. Cowpea breeding program in Niger should take into account early maturity, high yielding potential, tolerance to drought and low soil P conditions. Varieties that tolerant to P deficient soils should also be tolerant to striga as the two stresses seemed to occur together on the field. Most of the high performing cultivars with regard 156 University of Ghana http://ugspace.ug.edu.gh to both grain and fodder were not stable, so specific recommendation will be made to environments. Genotypes with high grain and/or fodder yield in the dry season were detected, however further studies are necessary for selection of varieties for dry season cowpea growers. Farmers should plant cowpea from the end of June to the first decade of July in order to get optimum grain and former yield. However the use of highly photosensitive varieties should be avoided regardless of the planting dates except if they are tolerant to both terminal drought and pod sucking bugs especially Clavigralla tomentosicollis. No cultivar had been consistently the best across the planting dates; however G56, G59, G54, G33, G63 and G71 showed relatively good performance and further studies are needed to determine which of them to recommend for all or each sowing date. 157 University of Ghana http://ugspace.ug.edu.gh REFERENCES Abaidoo, R., Dare. M. O. and Killani, S., 2017. Evaluation of early maturing cowpea (Vigna unguiculata) germplasm for variation in phosphorus use efficiency and biological nitrogen fixation potential. Journal of Agricultural Science 155(1):102–116. Abdoulaye, T. and Sanders, J. H., 2005. Stages and determinants of fertilizer use in semi-arid African agriculture : the Niger experience. Agricultural Economics 32: 167-179. Acosta-Díaz, E., Acosta-Gallegos, A. J., Trejo-López, C., Padilla-Ramírez, S. J. & Amador- Ramírez, D. M., 2009. Adaptation traits in dry bean cultivars grown under drouhgt stress. Agricultura Técnica en México. 35(4). 416-425. Adewale, B. D., Okonji, C., Oyekanmi, A. A., Akintobi, D. A. C. and Aremu, C. O., 2010. Genotypic variability and stability of some grain yield components of Cowpea. African Journal of Agricultural Research Vol. 5(9). pp. 874-880. DOI: 10.5897/AJAR09.481. Adipala, E., Nampala, P., Karungi, J. and Isubikalu, P., 2000. A review on options for management of cowpea pests: experiences from Uganda. Integrated Pest Management Reviews. 5. 185-196. Ae, N., Arihara, J,. Okada, K., Yoshihara, T. and Johansen, C ., 1990. Phosphorus uptake by pigeon pea and its role in cropping systems of the Indian subcontinent. Science 248. 477– 480. Afiukwa, C. A., Ubi, B. E., Kunert, K. J., Emmanuel, T. F. and Akusu, J. O., 2013. Seed protein content variation in cowpea genotypes. World Journal of Agricultural Sciences Vol. 1 (3) 094-099. Ahiabor, B. D. and Hirata, H., 2003. Associative influence of soluble phosphate. rock phosphate and arbuscular mycorrhizal fungus on plant growth and phosphorus uptake of three tropical legumes. West African Journal of Applied Ecology. vol. 4: 75-90. Ahmed, F. E., Hall, A. E. and DeMason, D. A., 1992. Heat injury during floral bud development in cowpea (Vigna unguiculata. Fabaceae). America Journal of Botany 79: 784-791. Agbicodo, E. M., Fatokun, C. A., Bandyopadhyay, R., Wydra, K., Diop, N.N ., Muchero, W., 2010. Identification of markers associated with bacterial blight resistance loci in cowpea [Vigna unguiculata (L.) Walp.]. Euphytica 175:215-226. Ajayi, A. T., Adekola, M. O., Taiwo, B. H. and Azuh ,V. O., 2014. Character Expression and Differences in Yield Potential of Ten Genotypes of Cowpea (Vigna unguiculata L. Walp). International Journal of Plant Research. 4(3): 63-71 Akande, S. R., Olakojo, S. A.. Ajayi, S. A., Owolade, O. F., Adetumbi, J. A. and Adeniyan, O. N., 2012. Planting date affects cowpea seed yield and quality at southern guinea savanna. Nigeria. Seed Technology 34: 51-60. 158 University of Ghana http://ugspace.ug.edu.gh Akhtar, M.S. Y., Oki, T., Adachi, Y., Murata, M. D. and Khan, H.R., 2007. Relative phosphorus utilization efficiency. growth response. and phosphorus uptake kinetics of Brassica cultivars under a phosphorus stress environment. Communication in Soil Science and Plant Analysis 38:1061–1085. Akarinde, E. A., Iroh, L., Obigbesan, G. O., Hilger, T., Romheld, V. and Neumann., G..,2007. Differential Expression of Aluminium Tolerance Mechanisms in Cowpea Genotypes under Phosphorus Limitation. Journal of Applied Science 6 (4): 854-859. Alghamdi, S. S., 2004. Yield stability of some soybean genotypes across diverse environments. Pakistan Journal of Biological Sciences 7 (12):2109-2114. Aliyu, O. M. and Makinde, B. O., 2016. Phenotypic Analysis of Seed Yield and Yield Components in Cowpea (Vigna unguiculata L.. Walp). Plant Breeding and Biotechnology 4(2):2526261. Alkama, N. E., Bolou, B. B., Vailhe, H., Roger, L., Ounane, J. M. and Drevon, J. J., 2009. Genotypic variability in P use efficiency for symbiotic nitrogen fixation is associated with variation in proton efflux in cowpea rhizosphere. Soil Biology and Biochemestry 41(9): 1823-1824. Almekinders, C. J. M. and Louwaars, N. P., 1999. Farmers‗seed production: New approaches and practices. Intermediate Technology Publication. Cromwell Press Trowbridge. U.K. 391pp. Amara, D. S. and Suale, D.S., 1996. Genotypic differences in yield formation. phosphorus utilization and nitrogen fixation by cowpeas in Sierra Leone. In: Isotope studies on plant productivity. IAEA pp.39-52. Ames, O., 1939. Economic Annals and Human Culture. Botanical Museum of Havard. Mass. (1939). 153pp. Andargie, M., Pasquet, R. S., Gowda, B. S., Muluvi, G. M., Timko, M. P., 2011. Construction of a SSR-based genetic map and identification of QTL for domestication traits using recombinant inbred lines from a cross between wild and cultivated cowpea (V. unguiculata (L.) Walp.). Molecular Breeding 28:413-420. Animasaun, D. A., Oyedeji, S., Azeez.,Y. K., Mustapha, O. T. and Azeez, M. A., 2015. Genetic Variability Study Among Ten Cultivars of Cowpea (Vigna unguiculata L. Walp) Using Morpho-agronomic Traits and Nutritional Composition. The Journal of Agricultural Sciences Vol. 10. No. 2: 119 – 130. Ankomah, A. B., Zapata, F., Danso, S. K. A. and Axmann, H., 1995. Cowpea varietal differences in uptake of phosphorus from Gafsa phosphate rock in a low-P Ultisol; Fertilizer Research 41:219–225. 159 University of Ghana http://ugspace.ug.edu.gh Ansah, S. G., Kumaga, F. K., Gracen, V., Adu-dapaah, H. and Nartey. E. K., 2014. Breeding cowpea (vigna unguiculata (l). walp) for phosphorus use efficiency in ghana. PhD thesis. West Africa Centre for Crop Improvement (WACCI). School of Agriculture College of Basic and Applied Sciences. University of Ghana. Legon. 225p. Ansoba, Y. E., Amoatey, C. A. and Kanton, R. A. L., 2013. Effect of planting date and variety of cowpea [Vigna unguiculata (l.)walp.] on green pod production. MPHIL thesis. Department of Crop Science. School of Agriculture College of Basic and Applied Sciences. University of Ghana. Legon 120 p. Asante, S., Tamo, M. and Jackai, L., 2001. Integrated management of cowpea insect pests using elite cultivars. date of planting and minimum insecticide application. African Crop Science Journal. 9. 655-665. Asare, A. T., Gowda, B. S., Galyuon, K. A., Aboagye, L. L., Takrama, J. F., Timko, M. P., 2010. Assessment of the genetic diversity in cowpea (Vigna unguiculata L. Walp.) germplasm from Ghana using simple sequence repeat markers. Plant Genetic Resources 8 (2) 142-150. Ashby, J. A., 2009. The impact of participatory plant breeding. In Plant Breeding and Farmer Participation. Ceccarelli. S.. Guimaraes. E. P. and Weltien. E. eds. FAO. Rome. Asuming-Brempong ,S., Wiafe ,Y. and Aggrey, M. K., 2013. Nodulation of cowpea (Vignaunguiculat [L]walp) at different levels of phosphorus in a Typic Kandiustalf. Agricultural Science Research Journal 3(12); pp. 387- 394. Armstrong, L. D., 1999. Functions of Phosphorus in Plants. Better crops with plant food. Vol. LXXXIII (83). No. 1. 3-39. Atkinson, D. and Davidson.,A. W., 1971. The effects of phosphorus deficiency on the growth of Epilobium montanum L. New Phytol. 70: 789–797. Audu, M., Kefas, M., Abdullahi.,A. A., Gambo, A.U., Dikko, B.A., Noma, S.S. and Muhammad, A., 2013. Influence of Sokoto Phosphate Rock on Some Soil Properties and the Growth and Yield of Cowpea [Vigna Unguiculata (L)Walp.] Varieties in Sudan Savanna of Nigeria. Journal of Biology. Agriculture and Healthcare. Vol.3. No.18: 75- 80. Avnimelech, Y. and Scherzer, S., 1971. The effect on yield of phosphorus uptake by young plants. Pages 365–384 in R. M. Samish. eds. Recent advances in plant nutrition. Proceedings of the 6th International Colloquium on Plant Analysis and Fertilizer Problems. Tel Aviv. March 1970. Gordon and Breach Science Publishers. London. UK. Vol. 2. Azari, A. and Khajepour, M. R., 2003. Effect of planting pattern on growth. development. grain yield and yield components in sunflower cv. Kooseh Isfahan in spring planting. Journal of Science and Technology of Agriculture and Natural Resources 7: 155-167. 160 University of Ghana http://ugspace.ug.edu.gh Badiane, F. A., Gowda, B. S., Cisse, N., Diouf. D., Sadio, O. and Timko, M. P., 2012. Genetic relationship of cowpea (Vigna unguiculata) varieties from Senegal based on SSR markers. Genetics And Molecular Research 11 (1) (2012) 292-304. Bagayoko, M., George, E., Römheld, V., Buerkert, A., 2000. Effects of mycorrhizae and phosphorus on growth and nutrient uptake of millet. cowpea and sorghum on a West African soil. Journal of Agricultural Science. 135: 399–407. Balemi, T and Negisho, K., 2012. Management of soil phosphorus and plant adaptation mechanisms to phosphorus stress for sustainable crop production: a review. Journal of Soil Science and Plant Nutrition. 12 (3). 547-562. Baraki, F., Tsehaye, Y. and Abay, F., 2014. AMMI analysis of genotype by environment interaction and stability of sesame genotypes in Northen Ethiopia. Asian Journal of Plant Science 13 (4-8): 178-183. Barry, D. A. J. and Miller, M. H., 1989. Phosphorus nutritional requirement of maize seedlings for maximum yield. Agronomy Journal 81: 95–99. Bationo, A. B. R., Ntare. S. A., Tarawali and Tabo, R., 2002. Soil fertility management and cowpea production in the semiarid tropics. Pp 301-318. In C.A. Fatokun. S.A. Tarawali. B.B. Singh. P.M. Kormawa and M. Tamo (eds) Challenges and opportunities for enhancing sustainable cowpea production. Proceedings of the world cowpea conference III held at the International Institute Tropical Agriculture (IITA). Ibadan. Nigeria. 4-8 September 2000. IITA. Ibadan. Nigeria. Baudoin, J. P. and Marechal, R., 1985. Genetic Diversity in Vigna. In: Cowpea Research. Production and Utilization. S.R.Singh and K.O. Rachie (eds.) 1985 John Wiley and Sons Ltd. Pp 3-10. Bayuelo-Jiménez, J. S., Craig, R. and Lynch, J.P., 2002. Salinity tolerance of Phaseolus species during germination and early seedling growth. Crop Science 42:1584–1594. Boatwright, G. O. and Viets, F. G. J., 1966. Phosphorus absorption during various growth stages of spring wheat and intermediate wheatgrass. Agronomy Journal 58: 185–188. Beegle, D. B. and Philip T. D., 2002. Managing Phosphorus for Crop Production. Produced by Ag Communications and Marketing. The Pennsylvania State University. Code UC055 05/14pod. Bello, L. L., Ojo, A. A., Adeyemo, M. O. and Omojor, Y. M., 1996. Effect of planting date on yield components and seed viability in soybean in southern Guinea Savanna. Nigeria. African Crop Science Journal 4: 393-397. Bernoux, M., Cerri, C. C., Neill, C. and de Moraes, J. F., 1998. The use of stable carbon isotopes for estimating soil organic matter turnover rates. Geoderma 82(1):43-58. 161 University of Ghana http://ugspace.ug.edu.gh Bezabih, B., Iemenih, M., and Regassa, A., 2016. Farmers‘ perception on soil fertility status of small- scale farming system in southwestern Ethiopia. Journal of Soil Science Environmental Management Vol. 7(9). pp. 143-153. Bieleski, R L., 1973. Phosphate pools. phosphate transport. and phosphate availability. Annual Review of Plant Physiology 24:225–252. Blamey, F. P. C., Wehr, J. B., Wang, P., Menzies, N. W. and Kopittke, P. M., 2014. Kinetics and mechanisms of cowpea root adaptation to changes in solution calcium. Plant and Soil 379. 301–314. Bomans, E., Fransen , K., Gobin, A., Mertens, J., Michiels, P., Vandendriessche, H., Vogels, N., 2005. Addressing phosphorus related problems in farm practice Final report to the European Commission. European Commission by the Soil Service of Belgium. W. de Croylaan 48. 3001 Heverlee. Belgium. Bonser, A. M., Lynch, J. P. and Snapp, S., 1996. Effect of phosphorus deficiency on growth. angle of basal roots in Phaseolus vulgaris. New Phytologist 132: 281–288. Breeuwsma , A., Reijerink, J.G . A and Shoumans, O. F., 1995. Impact of manure on accumulation and leaching of phosphate in areas of intensive livestock farming. In: Steele (ed) Animal waste and the land-water interface. pp 239–251. Lewis Publ.-CRC. NY. Burton, G. W. and Devane, E. H., 1953. Estimation of heritability in tall festca (Festuca arundinacea) from replicated clonal materials. Agronomy Journal. 45: 478–481. Buerkert, A. and Hiernaux, P., 1998. Nutrients in the West African Sudano-Sahelian zone: Losses. transfers and the role of external inputs. Z. Pflanzen-ernahe. Bodenk 161: 365– 383. Carpenter, S. R., Caraco, N. F., Correll, D. L., Howarth, R. W., Sharpley, A. N.and Smith, V. H., 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecological Applications 8:559-568. Casenave, A. and Valentin, C., 1992. A runoff classification system based on surface features criteria in semi-arid areas of West Africa. Journal of Hydrology 130:231-249. Ceccarelli, S., Guimaraes, E. P. and Weltien, E., 2009. Foreword. In Plant Breeding and Farmer Participation. Ceccarelli, S., Guimaraes. E. P. and Weltien, E. eds. FAO. Rome. Clutton-Brock, J., 1989. Cattle in ancient North Africa. The walking larder: patterns of domestication. pastoralism. and predation. J. Clutton-Brock. Unwin Hyman. London. UK. 200-206. Cobbinah, F. A., Addo-quaye, M. O.and Asante, I. K., 2011. Characterization. evaluation and selection of cowpea (Vigna unguiculata (L.) Walp) accessions with desirable traits from eight regions of Ghana. Journal of Agriculture and Biological Science 6: 21–31. 162 University of Ghana http://ugspace.ug.edu.gh Conant, R. T. J. and Paustian, K ., 2003. Land use effects on soil carbon fractions in the south eastern United States management incentive versus extensive grazing. Biology and fertility of soils. 38:386-392. Corbeels, M., Shiferaw, A. and Mitiku, H. M., 2000. Farmers‘ knowledge of soil fertility and local management strategies in Tigray. Ethiopia. Printed by Russell Press. Nottingham. Coetzee, J. J., 1990. Cultural practices for cowpea. Farming in South Africa. Leaflet. AL. Directorate: Agricultural Information. Pretoria. Correll, D. L., 1998. The Role of Phosphorus in the Eutrophication of Receiving Waters: A Review. Journal of Environmental Quality 27:261-266. Crafts-Brandner, S. J., 1992. Significance of leaf phosphorus remobilization in yield potential in soybean. Crop Science 32: 420–424. Crossa, J., Cornelius, P. L., Seyedsadr, M. and Byrne, P., 1993. A shifted multiplicative model cluster analysis for grouping environments without genotypic rank change. Theoritical and Applied Genetics 85(5): 577–586. Crossa, J., Yang, R. C. and Cornelius, P. L., 2004. Studying crossover genotype × environment interaction using linear–bilinear models and mixed models. Journal of Agricultural Biology and Environmental Statistics 9(3):362–380. Dabire-Binso, C. L., Malick, B. N., Sanon, A., Drabo, I. and Bi, K. F., 2010. Resistance mechanism to the pod-sucking bug Clavigralla tomentosicollis (Hemiptera: Coreidae) in the cowpea IT86D-716 variety. International Journal of Tropical Insect Science 30( 4):192–199. Dakora, F. D. and Phillips, D. A., (2002). Root exudates as mediators of mineral acquisition in low-nutrient environments. Plant and Soil 245:35–47. Daramany, A. M., 1994. Estimates of genotypic and phenotypic correlation. heritability and potency of gene sets in cowpea (Vigna unguiculata (L.). Assuit Journal of Agricultural Science 25. 1-8. Davis, D. W., Oelke, E. A., Oplinger, E. S., Doll, J. D., Hanson, C. V. and Putnam, D. H., 2010. Cowpea. Field Crop Manual. Department of horticultural science. agronomy and plant genetics. University of Wisconsi-Extension. Cooperative Extension University of Minnesota: Center for Alternative Plant and Animal Products and the Minnesota Extension Service. St. Paul. MN 55108 In web site. Deshmukh, S. N., Basu, M. S. and Reddy, P. S., 1986. Genetic variability. character association and path analysis of quantitative traits in Virginia bunch varieties of ground nut. Indian Journal of Agriculture Science. 56: 816–821. 163 University of Ghana http://ugspace.ug.edu.gh Dias, F. T. C., Bertini, C. H. C. M. and Filho, F. R. F., 2016. Genetic effects and potential parents in cowpea . Crop Breeding and Applied Biotechnology 16: 315-320. Brazilian Society of Plant Breeding. Printed in Brazil. Dhital, B. K., Harding, A. H. and Subedi, M., 1997. Effect of planting date on phenology. yield components and grain yield of determinate cowpea varieties in the low hills of Nepal. An electronic networking for sustainable development in Nepal. Dong, D., Peng, X. and Yan, X., 2004. Organic acid exudation induced by phosphorus deficiency and/or aluminum toxicity in two contrasting soybean genotypes. Physiologia Plantarum 122: 190–199. Dugje, I. Y., Omoigui, L. O., Ekeleme, F., Kamara, A. Y. and Ajeigbe, H., 2009. Farmers‘ Guide to Cowpea Production in West Africa. IITA. Ibadan. Nigeria. 20pp. Ehlers, J. D. and Hall, A. E., 1997. Cowpea [Vigna unguiculata (L.) Walp.]. Field Crops Research 53: 187–204. Ehlers, J. D. and Hall, A. E., 1996. Genotypic classification of cowpea based on responses to heat and photoperiod. Crop Science 36:673-679. El-Herty, E. H., Rizk, A. M. A., Gendy, E. K. and Abd El-Aal, H. T., 2010. Performance of soybean genotypes under four sowing dates at middle Egypt. Egyptian Journal of Plant Breeding 14: pp 283–293. Elias, A. A., Kelly, R. R., Doerge, R. W. and Mitchell, R. T., 2016. A Century of studying Genotype ´ Environment Interactions in Plant Breeding Experiments. Crop Science. 56:2090–2105 (2016). Elliott, D. E., Reuter, D. J., Reddy, G. D. and Abbott, R. J., 1997. Phosphorus nutrition of spring wheat (Triticum aestivum L.). 1. Effect of phosphorus supply on plant symptoms. yield. components of yield. and plant phosphorus uptake. Australian Journal of Agriultural Research 48: 869–881. El-Shaieny, A. A. H., Abdel-Ati, Y. Y., El-Damarany, A. M. and Rashwan, A. M., 2015. Stability analysis of component characters in cowpea. Journal of Agriculture and Forestry. Vol 7 (2) 24-35. Egbadzor, K., Yeboah, M., Offei, S., Ofori, K. and Danquah, E., 2013. Farmers' key production constraints and traits desired in cowpea in Ghana. Journal of Agricultural Extension and Rural Development. 5. 14-20. Ezeaku, I. E., Mbah, B. N. and Baiyeri, K. P., 2015. Planting date and cultivar effects on growth and yield performance of cowpea (Vigna unguiculata (L.) Walp). African Journal of Plant Science 9(11):439-448. 164 University of Ghana http://ugspace.ug.edu.gh Ezedinma, F. O. C. and Efekodo, J., 1962. Effect of alternating wetting and drying on germination of cowpea seeds. Nigeria Federal Department Agricultural Research Memorandum No. 56. FAOSTAT (http://faostat.fao.org/site/339/default.aspx). Faris, D. G., 1964. The chromosome number of Vigna sinensis (L.) Savi. Canadian Journal of Genetics and Cytology (6): 255-258. Fatokun, C. A., Boukar. O. and Muranaka. S., 2012. Evaluation of cowpea (Vigna unguiculata (L.) Walp.) germplasm lines for tolerance to drought. Plant Genetic Resources: Characterization and Utilization 10(3):171–176. Fatokun, C. A., 2002. Breeding cowpea for resistance to insect pests : attempted crosses between cowpea and Vigna vexillata. Challenges and Opportunities for Enhancing Sustainable Cowpea Production. Proceedings of the World Cowpea Conference III. 4-8th September 2000. C. A. Fatokun. S. A.. Tarawali. B. B.. Singh. P. M.. Kormawa. and M. Tamo. Eds. International Institute of Tropical Agriculture (IITA). Ibadan. Nigeria. 52-61. Fery, R. L., 1990. The cowpea: production. utilization and research in the United States. Horticultural Review (12): 197-222. Flight, C., 1976. The Kintampo culture and its place in the economic prehistory of West Africa. Origins of African plant domestication. J. R. Harlan. De Wet. J .M. J. and Stemler. A. B. L. Mouton. The Hague. Netherlands. 212-221. Fonji, A. M., Tongoona, P., Kumaga, F. K., and Boukar, O., 2015. Genetic analysis of traits related to biological nitrogen fixation in cowpea [Vigna unguiculata (L.) WALP] under low soil phosphorus. PhD thesis. West Africa Centre for Crop Improvement (WACCI). School of Agriculture College of Basic and Applied Sciences. University of Ghana. Legon. 205p. Fonji, A. M. and Begoude, B. A. D., 2014. Assessment of indigenious bradirhizobia spp. Population levels in low phosphorus soils of southern Cameroon using cowpea as trap crop. Journal of Ecology and Natural Environment. Vol. 6 (8): 280-286. Frank, M. P., Roland, J. B., Anne-Marie, N. I., Mokwunye, A. U., Fred, R. K., Cyrus G. N. and Paul, L. W..,1997. Soil Fertility Replenishment in Africa: An Investment in Natural Resource Capital. Replenishing Soil Fertility in Africa Proceedings of an international symposium cosponsored by Divisions A-6 (International Agronomy) and S-4 (Soil Fertility and Plant Nutrition). and the International Center for Research in Agroforestry. held at the 88th Annual Meetings of the American Society of Agronomy and the Soil Science Society of America. Indianapolis. Indiana. 6 November 1996. Frossard, D. E., Achat, L., Stefano, M. B., Else, K. B., Jean-Claude, F., Jan, J., Christian, M., Lilia, R., Lalajaona, R., Sokrat, S., Federica, T. and Astrid, O., 2010. The Use of Tracers 165 University of Ghana http://ugspace.ug.edu.gh to Investigate Phosphate Cycling in Soil–Plant Systems. In E.K. B€unemann et al. (eds.). Phosphorus in Action. Soil Biology 26. Fuller, W. H., Nielsen, D. R. and Miller, R. W., 1956. Some factors influencing the utilization of phosphorus from crop residues. Soil Science Society of America proceedings 20: 218- 224. Gahoonia, T. S. and Nielsen, N. E., 2004. Root traits as tools for creating phosphorus efficient crop varieties Plant and Soil 260: 47–57. Gauch, H.G., 1992. Statistical analysis of regional yield trials: AMMI analysis of factorial designs. Elsevier. Amsterdam; New York. Gavito, M. E. and Miller, M. H., 1998b. Changes in mycorrhiza development in maize induced by crop management practices. Plant Soil 198: 185–192. Gerrano, A. S., Patrick, O. A., Willem, S. J. V. R. and Sunette, M. L., 2015. Genetic variability in cowpea (Vigna unguiculata (L.) Walp.) genotypes. South African Journal of Plant and Soil 2015. 32(3): 165–174. Grant,C. A., Flaten, D. N., Tomasiewicz, D. J. and Sheppard. S. C., 2001. Importance of Early Season Phosphorus Nutrition. Better Crops/Vol. 85 No. 2. Grant, C., Bittman, S., Montreal, M., Plenchette, C. and Morel, C., 2005. Soil and fertilizer phosphorus: Effects on plant P supply and mycorrhizal development. Canadian Journal of Plant Science 85: 3–14. Gweyi-onyango, J. P., Akwee, P., Onyango, C., and Tesfamariam, T., 2011. Genotypic Responses of Cowpea (Vigna unguiculata L. Walp ) to Sub-Optimal Phosphorus Supply in Alfsols of Western Kenya : A Comparative Analysis of Legumes. Journal of Agricultural Science. 2(1). 1–8. Hadley, P., Roberts, E. H., Summerfield, R. J., Minchin, F. R., 1983. A quantitative model of reproductive development in cowpea [Vigna unguiculata (L.) Walp.] in relation to photoperiod and temperature and implications for screening germplasm. Annals of Botany 51: 531–543. Hall, A.E., 1992. Breeding for heat tolerance. Plant Breeding Review. 10: 129168. Harland, S. C., 1919. Inheritance of certain characters in the cowpea. Vigna sinensis. Journal of Genetics (8): 107-132. Haruna, I. M. and Usman, A., 2013. Agronomic efficiency of cowpea varieties (Vigna unguiculata (L.) Walp) under varying phosphorus rates in Lafia. Nassarawa state. Nigeria. Asian Journal of Crop Science. 5: 209-215. 166 University of Ghana http://ugspace.ug.edu.gh Hazelton, P. and Murphy, B., 2007. Interpreting soil test results: what do all the numbers mean? [2nd ed.]. CSIRO PUBLISHING. 150 Oxford Street (PO Box 1139) Collingwood VIC 3066. Australia). Hecht, S., 1990. ‗‗Indigenous soil management in the Latin American tropics: neglected Knowledge of native peoples.‘‘ In: M.A. Altieri and S.B. Hecht. ed. Agroecology and Small. Farm Development. CRC Press: Boca Raton. FL. Hiler, E. A., van Bavel, C. H. M.. Hossain, M.M. and Jordan, W.R.,1972. Sensitivity of Southern peas to plant water deficit at three growth stages. Agronomy Journal 64:60-64. Hoffland, E., Wei, C., and Wissuwa, M., 2006. Organic anion exudation by lowland rice (Oryza sativa L.) at zinc and phosphorus deficiency. Plant and Soil. 283(1-2). 155-162. Hoffland, E., Findenegg, G. R. and Nelemans, J.A., 1989. Solubilization phosphate by rape II: Local root exudation of organic acid as a response to Pstarvation. Plant and Soil. 113:161-165. Holford, I. C. R., 1997. Soil phosphorus: its measurement. and its uptake by plants. Australian Journal of Soil Research 35: 227–239. Ibro, G,. Bokar, M., Miriam, O., J. F. J. and James. L. D. B., 2007. ―Niébé: Importance au Niger.‖ Bean/Cowpea CRSP/INRAN publication. September Ibro, G., Thiombiano, T., Lowemberg-Deboer, J., Bernadette, K. D., Souleymane, S. and Rafiatou, A., 2011. Analyse de l‘effet des caractéristiques physiques et chimiques du niébé sur le prix au consommateur au Niger. Thèse de doctorat unique. Unité de Formation et de Recherche en Sciences Economiques et de Gestion (UFR/SEG). Université de Ouagadougou. Ouagadougou. Burkina Faso. 172p. Idahosa, D.O., Alika, J.E. and Omoregie, A.U., 2010. Genetic variability. heritability and expected genetic advance as indices for yield and yield components selection in cowpea (Vigna unguiculata L. Walp). Academia Arena. 2(5): 22 – 26. I. N. S. (Institut National de Statistique). 2013. Rapport National de Synthèse. Evaluation des Récoltes de la Campagne Agricole d‟Hivernage 2012 et Résultats Définitifs 2012-2013. 39p. Inocent, E. E., Mbah, B.N., Baiyeri, K. P. and Uguru, M. I., 2013. Productivity of grain cowpea (vigna unguiculata (l.) walp.) as influenced by season. genotype. insect pest management and cropping system in southeastern Nigeria. PhD thesis Department of Crop Science. University of Nigeria 270p. Inuwa, A. H., Ajeigbe, H. A., Muhammed, S. G. and Mustapha, Y., 2012. genetic variability and heritability of some selected of cowpea (vigna unguiculata (l) walp) lines. Agricultural society of nigeria. Proceedings of 46th annual conference held from 5th to 9th November 167 University of Ghana http://ugspace.ug.edu.gh 2012 kano EDITED BY J.M. Jibrin. M.A. Hussaini. B.M. Auwalu. E.U. Essiet. I.R. Muhammad. S.G. Mohammed. M.I. Daneji. Y. Garba and S.A. Pantami 624-628. Ishikawa, S., Adu-Gyamfi, J. J., Nakamura, T., Yoshihara, T., Watanabe, T., Wagatsuma, T., 2002. Genotypic variability in phosphorus solubilizing activity of root exudates by pigeonpea grown in low-nutrient environments. Plant Soil 245: 71-81. Ishiyaku, M. F. and Singh, B. B., 2003. Genetics of juvenile phase in cowpea [Vigna unguiculata (L.) Walp.]. Journal of Food. Agriculture and Environment 1: 133–136. Ishiyaku, M. F., Singh, B. B. and Craufurd, P. Q., 2005. Inheritance of time to flowering in cowpea (Vigna unguiculata (L.) Walp.). Euphytica. 142: 291-300. Ivica, G. D., Ivana, V. M., Rudolf, R. K.and Miodrag, Z. J., 2010. Mechanisms of adaptation of small grains to soil acidity. Proc. Nat. Sci. Matica Srpska Novi Sad. 118. 107—120. Jackai, L. E. N. and Daoust, R. A., 1986. Insect pests of cowpea. Annual Review of Entomology (31): 95-119. Jemo, M., Abaidoo, R. C., Nolte, C. and Horst, W. J., 2006. Genotypic variation for phosphorus uptake and dinitrogen fixation in cowpea on low-phosphorus soils of southern Cameroon. Journal of Plant Nutrition and Soil Science-zeitschrift fur pflanzenernahrung und bodenkunde. 169(6): 816-825. Jones, M.J., and Wild, A., 1975. Soils of the West African Savanna. Technical Communication No. 55. Commonwealth Bureau of Soils. Commonwealth Agricultural Bureaux. Farnham Royal. UK. Kakamega, K., 2011. Genotypic Responses of Cowpea (Vigna unguiculata) to Sub-Optimal Phosphorus Supply in Alfsols of Western Kenya: A Comparative Analysis of Legumes. Kamara, A.Y., Chikoye, D., Ekeleme, F., Omoigui, L. O. and Dugje, I.Y., 2008. Field performance of improved cowpea varieties under conditions of natural infestation by the parasitic weed Striga gesnerioides. International Journal of Pest Management Vol. 54. No. 3. 189–195. Kamdi, R. E., 2001. Relative stability. performance and superiority of crop genotypes across environments. Journal of Agricultural Biology and Environmental Statistics 6: 449-460. Kandil, A. A., Sharief, A. E., Morsy, A. R. and Manar, E. A. I., 2012. Performance of some promising genotypes of soybean under different planting dates using biplots analysis. Journal of Basic and Applied Science 8: 379-385. Kang, M. S. and Gorman, D. P., 1989. Genotype × environment interaction in maize. Agronomy Journal 81(4):662–664. 168 University of Ghana http://ugspace.ug.edu.gh Kebede, W., Awdenegest, M. and Fantaw, Y., 2013. Farmers perception of the effects of soil and water conservation structures on crop production: The case of Bokole watershed. Southern Ethiopia. Vol. 7(11). pp. 990-1000. Kediro, E., 2015. Farmers Perception on Soil Erosion and Effect of Conservation Structures on Soil Properties and Wheat Yield in Fulaso Watershed Sebeta Oromia. MSc thesis. Addis Ababa University. Addis Ababa. Ethiopia. 76p Kelly, V., 2005. Farmers demand for fertilizers in Sub-Saharan Africa. Department of Agricultural Economics. Michigan state university. East Lansing. MI 48:824-1039. Kitch, L., Boukar, O., Endondo, C. and Murdock, L., 1998. Farmer acceptability criteria in breeding cowpea. Experimental Agriculture. 34(4). 475-486. Kongjaimun, A., Kaga, A., Tomooka, N., Somta, P., Shimizu, T. and Shu, Y., 2012. An SSR- based linkage map of yardlong bean (Vigna unguiculata (L.) Walp. subsp. Unguiculata Sesquipedalis Group) and QTL analysis of pod length. Genome 55 (2) 81-92. Koona, P., Osisanya, E. O., Jackai, L. E. N., Tamo, M. and Markham, R . H., 2002. Resistance in accessions of cowpea to the coreid pod-bug Clavigralla tomentosicollis (Hemiptera: Coreidae). J Econ Entomol 95(6):1281-8. Kormawa, P. M., Chianu, J. N. and Manyong, V. M., 2000. Cowpea Demand and Supply Pattern in West Africa: The case of Nigeria.” In Fatokun. C. A.. Tarawali. S.. Singh. B. B.. Kormawa. P.. and Tamo. M. (Eds.) 2002. Challenges and opportunities for enhancing sustainable cowpea production. Proceedings of the World Cowpea Conference III held at the International Institute of Tropical Agriculture (IITA). Ibadan. Nigeria. 4–8 September 2000. IITA. Ibadan. Nigeria. pp 376 – 386. Krasilnikoff, G., Gahoonia, T. and Nielsen, N. E., 2003. Variation in phosphorus uptake efficiency by cowpea (Vigna unguiculata) due to differences in root and root hair length and induced rhizosphere processes. Plant and Soil. 251: 83-91. Kou, C., Wang, Q., Ren, L., 1999. Study on the variety of wheat and peanut in utilization in different phosphorus compound. China Journal of Soil Science 30. 181- 184. Kugblenu, Y. O., Kumaga, F. K., Ofori, K., and Adu-Gyamfi, J. J., 2014. Evaluation of cowpea genotypes for phosphorus use efficiency. Journal of Agricultural and Crop Research. 2(10). 202-210. Langyintuo, A., Lowenberg-Deboer, J., Faye, M., Lambert, D., Ibro, G., Moussa, B., Kergna, A., Kushwaha, S., Musa, S. and Ntoukam, G., 2003. Cowpea supply and demand in West and Central Africa. Field Crops Research. 82. 215-231. Latati, M., Blavet, D., Alkama, N., Laoufi, H., Drevon, J. J., Gérard, F., Pansu, M. and Ounane, S. M., 2014. The intercropping cowpea-maize improves soil phosphorus availability and maize yields in an alkaline soil. Plant and Soil 385:181–191. 169 University of Ghana http://ugspace.ug.edu.gh Lauzon, J. D. and Miller, M. H., 1997. Comparative response of corn and soybean to seed-placed phosphorus over a range of soil test phosphorus. Communication in Soil Science and Plant Analysis 28: 205–215. Lei, H., Zhu, C. and Liu, X., 2004. Phosphorus adsorption-desorption characteristics in acid soils under amendment. Acta Pedologica Sinica. 41. 636-640. Leiser, W. L., Rattunde, H. F. W., Weltzien, E., & Haussmann, B. I. (2014). Phosphorus uptake and use efficiency of diverse West and Central African sorghum genotypes under field conditions in Mali. Plant and soil, 377(1-2), 383-394. Liao, H., Wan, H. Y., Shaff, J., Wang, X. R., Yan, X. L. and Kochian, L. V., 2006. Phosphorus and aluminum interactions in soybean in relation to aluminum tolerance: exudation of specific organic acids from different regions of the intact root system. Plant Physiology 141:674–684. Liu, J. and Zhang, F., 2000. Dynamics of soil P pool in a long-term fertilizing experiment of wheat-maize rotation II. Dynamics of soil Olsen-P and inorganic P. China Journal of Applied Ecology 11. 360-364. Lott, J. N. A., Jurek, K., Graeme, D. B. and Lindsay, C. C., 2011.The critical role of phosphorus in world production of cereal grains and legume seeds Food Sec. 3:451–462 Mahamane, S., William, A. P., Richard, H. L., Creighton, J. M. and David, W. R., 2008. Evaluation of cowpea (Vigna unguicula L. Walp) genotypes for adaptation to low soil phosphorus conditions and to rock phosphate application. A Dissertation submitted to the Office of Graduate Studies of Texas A&M University in partial fulfillment of the requirements for the degree of doctor of philosophy 127p. Mahler, B., 2014. The importance of soil fertility in crop production. REACCH Annual Report. Year 4. pp24- 25. Mafakheri, K., Mohammad, R. B. and Ali, R. A., 2017. Assessment of genetic diversity in cowpea (Vigna unguiculata L.) germplasm using morphological and molecular characterisation Cogent Food & Agriculture 3: 1327092 Manggoel, W., Uguru, M. I., Ndam, O. N. and Dasbak, M. A., 2012. Genetic variability. correlation and path coefficient analysis of some yield components of ten cowpea [Vigna unguiculata (L.) Walp.] accessions. Journal of Plant Breeding and Crop Science 4: 80– 86. Mojaddam, M. and Nouri, A., 2014. The Effect of sowing date and plant density on yield and yield components of cowpea. Bulletin of Environment. Pharmacology and Life Sciences 3[8]:31-36. Marenya, P., Barrett, C. B. and Gulick, A. T., 2008. Farmers perception of soil fertility and fertilizer yield response I Kenya. http://ssrn.com/abstract=1845546. 170 University of Ghana http://ugspace.ug.edu.gh Marechal, R., Mascherpa, J. M. and Stainier, F., 1978. Etude taxonomique d‘un groupe d‘ especes des genres Phaseolus et Vigna (Papilionaceae) sur la base de donnees morphologiques et polliniques. traitees pour l‘analyse informatique. Boisiera (28): 1-273. McBeath, T. M., Armstrong, R. D., Lombi, E., Mclaughlin, M.J. and Holloway, R.E., 2005. Responsiveness of wheat (Triticum aestivum) to liquid and granular phosphorus fertilizers in southern Australian soils. Australian Journal of Soil Research 43. 203-212. McIntire, J., 1986.Constraints to Fertilizer Use in sub-Saharan Africa. In Management of Nitrogen and Phosphorus Fertilizers in sub-Saharan Africa. Eds A U. Mokwunye and P. L. G. Vlek Proceedings of a symposium. Martinus Nijhoff Publisher. McDowell, R.W., Sharpley, A. N., Condron, L. M., Haygarth, P.M.and Brookes, P.C., 2001. Processes controlling soil phosphorus release to runoff and implications for agricultural management. Nutrient Cycling in Agroecosystems 59: 269–284. 2001. Manu, A., fordresher, A. A. P., Geiger, S. C., Wilding, L.P. and Hossner, L. R., 1996. Soil parameters related to crop growth variability in Western Niger. West Africa. S.S.SA Journal. 60:283-288. Margaret, N., John, S. T., Erasmus, O., Drake, N. M. and Twaha, A. B., 2014. Development of Common bean (Phaseolus Vulgaris L.) Production under low Soil phosphorus and drought in Sub-Saharan Africa: A Review. Journal of Sustainable Development; Vol. 7. No. 5; 2014 ISSN 1913-9063 E-ISSN 1913-9071. May, M. L., Caviness, C. E. and Eldrge, I. L.,1989. Soybean response to early planting in northeast. Arkansas. Arkansas Farm Research. 385-388. Mbong, G.A., Akem, C. N., Alabi, O., Emechebe, A. M. and Alegbejo, M. D., 2010. Effect of Sowing Dates on Yield and Yield Components of Cowpea Infected with Scab. Asian Journal of Agricultural Sciences 2(2): 57-62. Miguel, M. A., Jochua, C. and Lynch, J. P., 2013. Evaluation and Selection of Common Bean (Phaseolus Vulgaris L.) Genotypes for Root Traits Associated with Phosphorus (P) Acquisition Efficiency and the Use of 32 P Isotope in Studies on P Uptake by Root Hairs. IAEA TECDOC SERIES. 93. Mirzaienasab, M. and Mojaddam, M.,. 2014. The effect of planting date on yield and yield components of two red bean cultivars in azna weather conditions. Indian Journal of Fundamental and Applied Life Sciences 4 (3):417-422. Mojaddam, M. and Nouri, A., 2014. The Effect of Sowing Date and Plant Density on Yield and Yield Components of Cowpea. Bulletin of Environment, Pharmacology and Life Sciences 3(8) 31-36. 171 University of Ghana http://ugspace.ug.edu.gh Morakinyo, J. A. and Ajibade, S. R., 1998. Effect of seasons and genotype x season interaction on vegetative and yield parameters of cowpea (Vigna unguiculata L. Walp). Nigerian Journal of Science 32(1): 21–25. Mourice, S. K. and Tryphone, G. M., 2012. Evaluation of Common Bean (Phaseolus vulgaris L.) Genotypes for Adaptation to Low Phosphorus. ISRN Agronomy. 1-9. Muchero, W., Ehlers, J. D., Close, T. J. and Roberts, P. T., 2009. Mapping QTL for drought stress-induced prematuren senescence and maturity in cowpea [Vigna unguiculata (L.) Walp.]. Theoretical and Applied Genetics 118 (5) (2009) 849-863. Muchero, W., Ehlers, J. D. and Roberts, P. T., 2010. QTL analysis for resistance to foliar damage caused by Thrips tabaci and Frankliniella schultzei (Thysanoptera: Thripidae) feeding in cowpea [Vigna unguiculata (L.) Walp.]. Molecular Breeding 25 (1) 47-56. Mullins, G., 2009. Phosphorus. agriculture and environment Produced by Communications and Marketing. College of Agriculture and Life Sciences. Virginia Polytechnic Institute and State University. Nagalakshmi, R. M., Usha, K. R. and Boranayaka, M. B., 2010. Assessment of genetic diversity in cowpea (Vigna unguiculata). Electronic Journal of Plant Breeding 1: 453–461. Ngalamu, T., Ashraf, M. and Meseka, S., 2013. Soybean (Glycine max L) genotype and environment interaction effect on yield and other related traits. American Journal of Experimental Agriculture 3 (4):977-987. Ngwene, B., George, E., Claussen, W. and Neumann, E., 2010. Phosphorus uptake by cowpea plants from sparingly available or soluble sources as affected by N-form and arbuscular– mycorrhiza–fungal inoculation. Journal of Plant Nutrition and Soil Science 173:353–359. Ndjeunga, J. and Batilan, M. S. C., 2005. Uptake of improved technologies in the semi-arid tropics of West Africa : why is agricultural transformation lagging behind ? E-Journal of Agricultural and Development Economics. 2(1) : 85-102. Nelson, N. and Mikkelsen, R., 2008. Meeting the phosphorus requirement on organic farms. Better Crops 92(1):12–14. Neumann, G. and Römheld, V., 1999. Root excretion of carboxylic acids and protons in phosphorus-deficient Plants. Plant and Soil 211:121–130. Ng, N. Q and Marechal, R., 1985. Cowpea taxonomy. origin and germplasm. Cowpea research, production and utilization. S. R. Sing and K. O. Rachie. Eds. John Wiley and Sons. Chichester. UK. 11-21. Ng, N. Q and Padulosi, S., 1988. Cowpea genepool distribution and crop improvement. Crop Genetic Resources of Africa 2: 161-174. 172 University of Ghana http://ugspace.ug.edu.gh Ng, N. Q., 1995. Cowpea Vigna unguiculata (Leguminosae-Papilionoideae). In evolution of crop plants. 2nd edition. Eds. Smartt. J. and Simmonds. N.W. Longman. Harlow. UK. 326332pp. Ng, N.Q. and Marechal, R., 1985. Cowpea Taxonomy. Origin and Germplasm. In: Cowpea Research, Production and Utilization. S.R.Singh and K.O. Rachie (eds.) 1985 John Wiley and Sons Ltd. pp11-22. Nielsen, S. S., Brandt, W. E. and Singh, B. B., 1993. Genetic variability for nutritional composition and cooking time of improved cowpea lines. Crop Science 33:469-472. Nielsen, N. E. and Barber, S.A., 1978. Differences among genotypes of corn in the kinetics of P uptake. Agronomy Journal 70: 695-698; Nielsen, N. E. and Schjørring, J. K., 1983. Efficiency and kinetics of phosphorus uptake from soil by various barley genotypes. Plant and Soil 72. 225–230. Nkaa, F.A., Nwokeocha, O.W. and Ihuoma, O., 2014. Effect of Phosphorus fertilizer on growth and yield of cowpea (Vigna unguiculata). Journal of Pharmacy and Biological Sciences Volume 9. Issue 5 Ver. IV. Nouri, M., Garba, M. and Wortmann, C. S., 2017. Optimizing fertilizer use within the context of integrated soil fertility management in Niger. In Fertiliier Use Optimization in Sub- Saharan Africa. Charles. S. Wortmann and Keith Sonest. eds. published by CABI. Nwofia, G. E., Edugbo, R. E. and Mbah, E. U., 2016. Interaction of genotype x sowing date on yield and associated traits of soybean [Glycine max (L.) Merrill] over two cropping seasons in a humid agro-ecological zone of south-eastern Nigeria. The Journal of Agricultural Sciences 11 (3) 164-177. Nwoke, O. C., Diels, J., Abaidoo, R., Nziguheba, G. and Merckx, R., 2008. Organic acids in the rhizosphere androot characteristics of soybean (Glycine max) and cowpea (Vigna unguiculata) in relation to phosphorus uptake in poor savanna soils. African Journal of Biotechnology 7. 3620-3627. Nwosu, D. J., Olatunbosun, B. D. and Adetiloye, I. S., 2013. Genetic variability. heritability and genetic advance in cowpea genotypes in two agro-ecological environments. Greener Journal of Biological Sciences 3: 202–207. Niu, Y. F., Ru, S. C., Gu, L. J., Huan, W., Cai, X. T. and Yong, S. Z., 2013. Responses of root architecture development to low phosphorus availability: a review. Annales Botanici Fennici 112 (2): 391-408 Nye, P. H. and Tinker, P. B., 1977. Solute movement in the soilroot system. Blackwell Scientific pub., Oxford, Reino Unido. 173 University of Ghana http://ugspace.ug.edu.gh Odion, E.C., Asiribo, O.E. Ogunlela, V.B., Singh, B.B. and Tarawali, S.A., 2007. Strategies to improve and sustain food production capacity in the savanna: The role of leguminous fodder crops in maintaining soil fertility and health. Journal of Food, Agriculture and Environment. 5:338-344. Odundo, S., Ojiem, O., Okalebo, J., Othieno, C., Lauren, J., Medvecky, B., 2010. Effect of phosphorus on survival. nodulation and yield of cowpea (Vigna unguiculata) varieties across a soil fertility degradation gradient in Western Kenya. World cowpea 2010 conference. 27th September to 1st October 2010. Saly. Senegal. Ogunkanmi, L. G., Ogundipe, O. T., Ng, N.Q., Fatokun, C. A., 2080. Genetic diversity in wild relatives of cowpea (Vigna unguiculata) as revealed by simple sequence repeats (SSR) markers. Journal of Food, Agriculture & Environment 6 (2008) 263-268. Ojo, D. K., Ogunbayo, S. A., Idehen, E. O., Akinwale, A. F., Oduwaye, O. A., 2007. Genetics evaluation of phosphorus utilization in tropical cowpea (Vigna unguiculata (L.) Walp.). Journal America Science, 3 (1). 77- 83. Oladiran, O., Fagbola, O., Abaidoo, R. C., Ikeorah, N., 2012. Phosphorus response efficiency in cowpea genotypes. Journal of Agricultural Science: Vol. 4. No. 1. Omoigui , L. O., Kamara, A. Y., Hakeem, A. A., Richard, O. A., Michael, P. T., Muhyideen, O. and Lateef, L; B., 2017. Performance of cowpea varieties under Striga gesnerioides (Willd.) Vatke infestation using biplot analysis. Euphytica 213:244. Omoigui, L. O., Kamara, A. Y., Massawe, F.S., Ishiyaku, M. F., Boukar, O., Alabi, S.O. and Ekeleme, F., 2007. Evaluation of cowpea genotypes for their reactions to Striga gesnerioides in the dry savanna of northern Nigeria. African Crop Science Conference Proceedings. 8: 273-278. Omoigui, L.O., Ishiyaku, M.F., Kamara, A.Y., Alabi, S.O. and Mohammed, S.G., 2006. Genetic variability and heritability studies of some reproductive traits in cowpea (Vigna unguiculata L. Walp). African Journal of Biotechnology. 5 (13): 1191 – 1195. Omotayo, O. E. and Chukwuka, K. S., 2008. Soil fertility restoration techniques in sub-Saharan Africa using organic resources. African Journal of Agricultural Research Vol. 4 (3). pp. 144-150. Orawu, M., 2007. Occurrence of cowpea aphid-borne mosaic virus and prospects of improving resistance in local cowpea landraces in Uganda. Ph.D thesis. School of Biochemistry. Genetics. Plant pathology and Microbiology. University of Kwazulu-Natal. South Africa, University of KwaZulu-Natal. 171p. Otani, T., Noriharu, Ae, and Hidenori, T., 1996 . Phosphorus (P) Uptake Mechanisms of Crops Grown in Soils with Low P Status significance of organic acids in root Exudates of Pigeonpea Soil ScL Plant Nutrition. 42 (3) 553-560. 553.60. 174 University of Ghana http://ugspace.ug.edu.gh Padulosi, S., 1993. Genetic diversity. taxonomy. and ecogeographic survey of the wild relatives cowpea(Vigna unguiculata [L.] Walpers). PhD thesis. Universite Catholique de Louvain- la-Neuve. Belgique. Padulosi, S. and Ng. N. Q., 1997. Origin. taxonomy. and morphology of Vigna unguiculata (L.) Walp. Advances in cowpea research. B. B. Singh. D. R. Mohan Raj. K. E. Danshiell. and L. E. N. Jackai. Copublication of IITA and JIRCAS. IITA. Ibadan. Nigeria. 1-11. Pandey, R. K., Maranville, J. W. and Admou, A., 2001. Tropical wheat response to irrigation and nitrogen in a Sahelian environment. I. Grain yield. yield components and water use efficiency. European Journal of Agronomy. Amsterdam. v. 15. n. 3. p. 93-105. 2001. Pant, K., Chandel, K. and Joshi, B., 1982. Analysis of diversity in Indian cowpea genetic resources. SABRAO Journal 14:103–111. Pawluk, R. R., Sandor, J. A and Tabor, J. A., 1992. ‗‗The role of indigenous soil knowledge in agricultural development.‘‘ Journal of Soil and Water Conservation 47 (4). 298-302. Payne, W.A., 2006. Dryland cropping systems of semiarid West and East Africa. p.733-768. In Gary A. Peterson et al. (ed.) Dryland agriculture. Agronomy Monograph No23. Second Edition 2006. ASSA, CSSA. and SSSA. Madison WI. Payne, W. A., 1997. Managing yield and soil water use of pearl millet in the Sahel. Agronomy Journal 89:481-490. Payne, W. A., Wendt, C.W. and Lascano, R.J., 1990. Root zone water balances of three low- input millet fields in Niger. West Africa. Agronomy Journal 82:813-819. Pearse, S. J., Venaklaas, E. J., Cawthray, G., Bolland, M. D. A. and Lambers, H., 2007. Carboxylate composition of root exudates does not relate consistently to a crop species‘ ability to use phosphorus from aluminium. iron or calcium phosphate sources. New Phytologist 173. 181–190. Pender, J., Abdoulaye, T. and Ndjeunga, J., 2008. Impacts of Inventory Credit. Input Supply Shops. and Fertilizer Microdosing in the Drylands of Niger. Published by INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE 2033 K Street. NW Washington. DC 20006-1002 USA. www.ifpri.org. Péret, B., Clément, M., Nussaume, L., and Desnos, T., 2011. Root developmental adaptation to phosphate starvation: better safe than sorry.Trends in plant science. 16(8). 442-450. Pieri, C., 1990. Les bases agronomiques de l'amélioration et du maintien de la fertilité des terres de savanes au sud du Sahara. p. 43-74. In C. Pieri (ed.). Savanes d'Afrique. terres fertiles. Comment produire plus et de façon durable en zone de savanes au sud du Sahara. Actes des Rencontres Internationales. 10-14 décembre 1990. Montpellier. France. 175 University of Ghana http://ugspace.ug.edu.gh Phillips, E. P., 1951. The genera of South African flowering plants. Botanical survey of the South Africa Memoir 25. 2nd edition Government Printer. Pretoria. South Africa. 25: 526-531 Piper, C. V., 1913. The wild prototype of the cowpea. Circular US Department of Agriculture and Bureau of Plant Industry (1913). No. 124. Pramoda, H. P. and Gangaprasad, S., 2007. Biometrical basis of handling segregation population for improving productivity in onion (Allium cepa L.). Asian Journal of Horticulture, 3 (4): 278-280. Pypers, P., Van Loon, L., Diels, J., Abaidoo, R., Smolders, E. and Merckx, R., 2006. Plant- available P for maize and cowpea in P-deficient soils from the Nigerian Northern Guinea Savanna – comparison of Eand L-values. Plant and Soil 283:251–264. Qiao, Y., Tang, C., Han, X. and Miao, S., 2007. Phosphorus deficiency delays the onset of nodule function in soybean. Journal of Plant Nutrition 30:1341–1353. Rachie, K. O., 1985. Introduction. Cowpea Research. Production and Utilization: S. R. Singh and K. O. Rachie. Eds. John Wiley and Sons. New York. i-iv. Raghothama, K.G., 1999. Phosphate acquisition. Annual Review Plant Physiology and Plant Mol. iol. 50:665-693. Rajput, A. L., 1994. Response of cowpea to rhizobium inoculation. date of sowing and phosphorus. Indian Journal of Agronomy 39 (4): 584-587. Ramaswamy, S. and Sanders, J. H., 1992. Population pressure. land degradation and sustainable agricultural technologies in the Sahel. Agricultural Systems 40: 361–378. Ravelombola, W., Qin, J., Shi, A., Lu, W., Weng, Y., Xiong, H., Yang, W., Bhattarai, G., Mahamane, S., Payne, W. A., MillerJr, J. C. and Scheuring, D., 2017. Cowpea Vigna unguiculata Low phosphorus conditions Rock phosphate Single nucleotide polymorphism (SNP) Association mapping. Euphytica 213: 183. Rengel, Z., 2002. Genetic control of root exudation. Plant and Soil 245:59–70. Richardson, A. E., 1994. Soil microorganisms and phosphorus availability. Soil Biota 50–62. Rockstrom, J., Jansson, P. E. and Barron, J., 1998. Seasonal rainfall portioning under runon and runoff conditions on sandy soil in Niger. On-farm measurements and water balance modeling. Journal of Hydrology 210:68-92. REPUBLIQUE DU NIGER., 2001. Initiative pour la fertilité des sols et la collecte des eaux de ruissellement (IFS/CER). Plan d’action national. 39p. Römer, W. and Schenk, H., 1998. Influence of genotype on phosphate uptake and utilization efficiencies in spring barley. European Journal of Agronomy; 8:215–224. 176 University of Ghana http://ugspace.ug.edu.gh Rothe, J. C., Hays, D. B., Singh, B. B., Jessup, R., Provin, T., Versaw, W. and Baltensperger, D., 2014. Breeding for tolerance of cowpea to low phosphorus soil conditions through physiological and genetic studies. A Dissertation submitted to the Office of Graduate Studies of Texas A&M University in partial fulfillment of the requirements for the degree of doctor of philosophy pp 110p. Saidou, A. K., Abaidoo, R. C., Singh, B. B., Iwuafor, E. N. O. and Sanginga, N., 2011. Genotypic variation in cowpea for rock phosphate use in low P soils of dru sudan and sahel savannah of Niger and Nigeria. West Africa. American Eurasian Journal of Agriculture And Environment Science 11(1) 62-70. Saidou, A. K., Abaidoo, R., Singh, B. B., Iwuafor, E. and Sanginga, N., 2007. Variability of cowpea breeding lines to low phosphorus tolerance and response to external application of Phosphorus. In: Bationo A., Waswa B., Kihara, J., Kimetu, J., (eds) Advances in Integrated Soil Fertility Management in sub-Saharan Africa: Challenges and Opportunities. Springer. Dordrecht. Salifou, M., Tongoona, P., Offei, S. K., Ofor, K. and Tignegre, J. B. D. L. S ., 2015. Breeding cowpea [Vigna unguiculata (L.) Walp.] for striga [Striga gesnerioides (wild.) vatke] resistance using marker assisted selection in Niger. PhD. thesis. West Africa Centre for Crop Improvement (WACCI). School of Agriculture College of Basic and Applied Sciences. University of Ghana. Legon. 156p. Sample, E. C., Soper, R. I. and Racz, G. J., 1980. Reactions of phosphate fertilizers in soils. In: Khasawneh FE. Sample EC & Kamprath EJ (eds) The Role of Phosphorus in Agriculture. pp 263–310. Madison. WI: ASA-CSSA-SSSA. Sanda, A. R. and Maina, M. I., 2013. Effect of drought on the yields of different cowpea cultivars and their response to time of planting in Kano State. Nigeria. International Journal of environment and bioenergy 6(3): 171-176. Sanchez, P. A., Shepherd, K. D., Soule, M. J., Place, F. M., Buresh, R. J., Izac, A. N., Mokwunye, A. U., Kwesiga, F. R., Ndiritu, C. G. and Woomer, P. L., 1997. in: Buresh. R. J., Sanchez. P. A. and Calhoun. F. (Eds). Replenishing soil fertility in Africa. Soil Science Society of America. Madison.Wisconsin. USA. p. 1–46. Santos, A., Ceccon, G., Teodoro, E. P., Correa, A. M., Alvarez, R. C. F., Silva, J. F. and Alves, V. B., 2016. Adaptability and stability of erect cowpea genotypes via REML/BLUP and GGE Biplot. Bragantia, Campinas, 75 (3): 299-306. Santos, A., Ceccon, G., Teodoro, E. P., Correa, A. M., Alvarez, R. C. F., Silva, J. F. and Alves, V. B., 2015. Adaptability and stability of erect cowpea genotypes via REML/BLUP and GGE Biplot. Bragantia, Campinas, 75 (3): 299-306. Sardans. J. Penuelas and M. Estiarte. 2006.Warming and drought alter soil phosphatase activity and soil P availability in a Mediterranean shrubland. Plant and Soil 289:227–238. 177 University of Ghana http://ugspace.ug.edu.gh Saunders, A. R., 1960. Inheritance in the cowpea. (Vigna sinensis Endb.) 2: Seed coat colour pattern. flower. and pod colour. South African Journal of Agricultural Science (3):141- 161. Sawadogo, M., Ouedraogo, J. T., Gowda, B. S.,Timko. M. P.. 2010. Genetic diversity of cowpea (Vigna unguiculata L. Walp.) cultivars in Burkina Faso resistant to Striga gesnerioides. African Journal of Biotechnology 9 (48) 8146-8153. Schachtman, D. P., Robert, J. R. and Ayling, S.M., 1998. Phosphorus Uptake by Plants: From Soil to Cell. Plant Physiology. 116: 447–453. Sherwood, S. and Uphoff, N., 2000. ‗‗Soil health: research. practice and policy for a more regenerative agriculture.‘‘ Applied Soil Ecology 15. 85-97. Shiringani, R. P., Shimeles, H. A. and Merwe, P. V., 2007. Effects of planting date and location on phenology. yield and yield components among selected cowpea varieties. MSc. Thesis. School of Agricultural and Environmental Sciences. Faculty of Science and Agriculture. University of Limpopo. South Africa 73pp. Shiringani, R. P. and Shimeles, H. A., 2011. Yield response and stability among cowpea genotypes at three planting dates and test environments. African Journal of Agricultural Science 6(4): 3259-3263. Singh, A., Baoule, A. L., Ahmed, H. G., Dikko, A. U., Aliyu, U. Sokoto,M. B., Alhassan, J., Musa, M. and Haliru, B., 2011 Influence of phosphorus on the performance of cowpea (Vigna unguiculata (L) Walp.) varieties in the Sudan savanna of Nigeria. Agricultural Science Vol.2. No.3. 313-317. Singh, B. B., Ehlers, J. D., Sharma, B. and Freire,F. F. R., 2002. Recent progress in cowpea Breeding. Challenges and Opportunities for Sustainable Cowpea Production. Proceedings of the World Conference III. 4-8th September 2000. C. A. Fatokun. S. A.. Tarawali. B. B.. Singh. P. M.. Kormawa. and M. Tamo. Eds. International Institute of Tropical Agriculture (IITA). Ibadan. Nigeria. 22-40. Singh, B. B. and Ishiyaku, M. F., 2000. Genetics of rough seed coat texture in cowpea. Journal of Heredity. 91: 170-174. Singh, B. B and Ajeigbe, H.A., 2000. Improving cowpea-cereals based cropping systems in the dry savannas of West Africa. In: Challenges and opportunities for enhancing sustainable cowpea production. Eds. C.A. Fatokun.. S.A. Tarawali.. B.B. Singh. P.M.. Kormawa and M. Tamo. Proceedings of the World Cowpea Conference III held at the International Institute of Tropical Agriculture. Ibadan. Nigeria. 4-8 September 2000 278-286pp. Singh, B. B. and Emechebe, A. M., 1997. Advances in research on cowpea Striga and Alectra. Advances in Cowpea Research. 215-224. SNV (2008) Etude de la filière niébé dans la région de Maradi. Juillet 2008. 59p. 178 University of Ghana http://ugspace.ug.edu.gh Somado, E. A., Kanwar, L. S. and Ronald, F. K., 2006. Rock phosphate-P enhances biomass and nitrogen accumulation by legumes in upland crop production systems in humid West Africa. Biology and Fertility of Soils (2006) 43: 124–130. Souleymane, A., Aken‘Ova, M. E., Fatokun, C. A. and Alabi, O. Y., 2013. Screening for resistance to cowpea aphid (Aphis craccivora Koch) in wild and cultivated cowpea (Vigna unguiculata L. Walp.) accessions. International Journal of Science and Environmental Technology 2. 611-621. Stanton, W. R., 1961. Unpublished report of research progress. Institute of Agricultural Research. Ahmadu Bello University. Samaru. Zaria. Nigeria. Steele, W. M., 1972. Cowpeas in Nigeria. Thesis submitted for the doctorate degree of the University of Reading. 258pp. Subbarao, G.V., Renard, C.R., Payne, W.A. and Bationo, A.B., 2000. Long-term effects of tillage. P fertilisation and crop rotation on pearl millet/cowspea productivity in West Africa. Experimental. Agriculture 36: 243-264. Subbarao, G. V., Ae, N. and Otani. T., 1997. Genetic variation in acquisition. and utilization of phosphorus from iron-bound phosphorus in pigeonpea. Soil Science and Plant Nutrition. 43 : 511–519. Summerfield, R. J., 1980. Effects of air temperature on seed growth and maturation in cowpea (Vigna unguiculata). Annals of Applied Biology 88: 437-446. Syers, K., Bekunda, M., Cordell, D., Corman, J., Johnston, J., Rosemarin, A. and Lougheed, T., 2011. Phosphorus and food production. UNEP Year Book. 34-45. Tabor, J.A., 1992. ‗‗Ethnopedological surveys—soil surveys that incorporate local systems of land classification.’’ Soil Survey Horizons Spring.1-5. Tchiagam, J. B. N., Bell, J. M., Birwe, S. G., Gonne, S. and Youmbi, E., 2010. Varietal response of cowpea (Vigna unguiculata (L.) Walp.) to Striga gesnerioides (Willd.) Vatke race SG5 infestation. Notulae Botanicae Horti Agrobotanici ClujNapoca. 38. 33-41. Teixeira, N. J. P., de Machado, C. F., Filho, F. R. F., de Rocha, M. M. and Gomes, R. L. F., 2007. Grain yield. yield components and their interrelationship in genotypes of cowpea of determinate growth habit. Editoria UFV. Univ. Fed. Vicosa Rev. Ceres 54: 375-383. Tignegre, J. B. D. L. S., 2010. Genetic study of cowpea (Vigna unguiculata (L.) Walp) resistance to Striga gesnerioides (Willd.) Vatke in Burkina Faso. Ph.D. thesis. School of agricultural sciences and agribusiness. University of Kwazulu-Natal. South Africa. 167p. 179 University of Ghana http://ugspace.ug.edu.gh Tirado, R. and Michelle, A., 2012. Phosphorus in agriculture Problems and solutions. Published in by Greenpeace International Ottho Heldringstraat 5 1066 AZ Amsterdam The Netherlands. Theodorou, M. E. and Plaxton, W. C., 1993. Metabolic adaptations of plant respiration to nutritional phosphate deprivation. Plant Physiology. 101: 339–344. Traore, F., 1974. Etude de la fumure azotée intensive des cereals et du role specifique de la matiere organique dans la fertilite des sols du Mali. Agronomie Tropicale 29: 567–586. Trolove, S. N., Hedley, M.J., Kirk, G.J.D., Bolan, N.S. and Loganathan, P., 2003. Progress in selected areas of rhizosphere research on P acquisition. Australian Journal of Soil Research 41:471- 499. Trolove, S. N., Hedley, M. J., Caradus, J. R. and Mackay, A. D., 1996. Uptake of phosphorus from different sources by Lotus pedunculatus and three genotypes of Trifolium repens. 1. Plant yield and phosphate efficiency. Australian Journal of Soil Research 34: 1015– 1026. Turk, K. J., Hall, A. E. and Asbell, C.W., 1980. Drought adaptation of cowpea. I. Influence of drought on seed yield. Agronomy Journal 72:413:420. Ubi, E., Mignouna, H. and Obigbesan, G., 2001. Segregation of seed weight. pod length and days to flowering. African Journal of Crop Science 9: 463–470. Uchida, R., 2000. Essential nutrients for plant growth: nutrient functions and deficiency symptoms. In Plant Nutrient Management in Hawaii’s soils. Approaches for Tropical and Subtropical Agriculture. J. A.. Silva and R. Uchida. eds. College of Tropical Agriculture and Human Resources. University of Hawaii at Monoa. Uguru, M. I., 1995. Heritable relationships and variability of yield and yield components in vegetable Cowpea. African Journal of Crop Science 3: 23–28. Valenzuela, H. and Smith, J., 2002. Cowpea: Cooperative Extension Service. College of Tropical Agriculture and Human Resources. University of Hawai at Manoa. Verdcourt, B., 1970. Studies of the leguminosae-papilionoidae for the flora of east Tropical Africa IV. Kew Bulletin (24): 507-569. Verde, B. and Matusso, J., 2014. Phosphorus in Sub-Sahara African Soils - Strategies and Options for improving available Soil Phosphorus in Smallholder Farming Systems: A Review. Academic Research Journal of Agricultural Science and Research. Vol 2(1) pp 1-5. Verma, S. L., Penfold, C. and Marschner, P., 2013. Effect of compost on soil phosphorus availability. A thesis submitted to the University of Adelaide in the fulfillment for the requirements for the degree of doctor of philosophy pp 142p. 180 University of Ghana http://ugspace.ug.edu.gh Vesterager, J. M., Nielsen, N. E., Høgh-Jensen, H., 2006. Variation in phosphorus uptake and use efficiencies between pigeonpea genotypes and cowpea. Journal of Plant Nutrition. 29: 1869–1888. Vince-Prue, D., 1975. Photoperiodism in plants. London Mc-Graw- Hill. pp. 263–291. Warrag, M.O.A. and Hall, A. E.,1983. Reproductive responses of cowpea to heat stress: genotype differences in tolerance of heat flowering. Crop Science. 23: 1088-1092. Watanabe, I., Hakoyama, S., Terao, T., and Singh, B. B., 1997. Evaluation methods for drought tolerance of cowpea. In Singh. B. B.. Mohan-Raj. D. R.. Dashiell. K. E.. and Jackai L. E. N. (Eds.) 1997. Advances in cowpea research. Co-publication of International Institute of Tropical Agriculture (IITA) and Japan International Research Center for Agricultural Sciences (JIRCAS). IITA. Ibadan. Nigeria. Pp 141-146. Weil, R. R., 2000. Soil and plant influence on crop response to two African phosphate rocks. Agronomy Journal. 92:1167-1175. Wien, H. C. and Summerfield, R.J., 1980. Adaptation of cowpeas in West Afica: effects of photoperiod and temperature responses in cultivars of diverse origin. Summerfield. R.J and A.H. Bunting.(eds) In Advances In Legume Science. Pp 405-407. West, L. T., 1984. Soil survey of the ICRISAT Sahelian Center. Niger. West Africa. Soil and Crop Sciences Department/Tropsoils. Texas A&M University. College Station.TX. 66p. White, D. H. and Howden, S.M., 2010. Climate and Its Effects on Crop Productivity and Management. Verheye. W.H. (ed). Soils. Plant Growth and Crop Production. I. EOLSS Publishers CoLtd. pp 44-78. Widders, I. E., 2005. The changing edible legume landscape: Adapting to new driving forces. agricultural research agendas. and funding opportunities. p. 17-21. In Proceedings of the First International Edible Legume Conference in Conjunction with the Fifth World Cowpea Congress. Durban. South Africa. Williams, O. T.,1999. Factors influencing manure application by farmers in semi-arid West Africa. Nutrient Cycling in Agroecosystem 55 : 15 – 22. WinklerPrins, A., 1999. ‗‗Local soil knowledge: a tool for sustainable land management.‘‘ Society and Natural Resources 12 (2). 151-161. Wissuwa, M., 2005. Combining a modeling with a genetic approach in establishing associations between genetic and physiological effects in relation to phosphorus uptake. Plant and Soil 269:57–68 Wissuwa, M., Wegner, J., Ae, N. and Yano, M., 2002. Substitution mapping of Pup1: a major QTL increasing phosphorus uptake of rice from a phosphorus deficient soil. Theoritical and Applied Genetics 105:890–897. 181 University of Ghana http://ugspace.ug.edu.gh Wissuwa, N. and Ae, N., 2001. Genotypic variation for tolerance to phosphorus deficiency in rice and the potential for its exploitation in rice improvement. Plant Breeding. 120:43-48. Witcombe, J. and Joshi A., 1996. Farmer Participatory Crop Improvement. II. Participatory Varietal Selection. a Case Study in India. Experimental Agriculture. 32(4). 461-477. Wong, M.T.F., Wild, A. and Mokwunye, A.U., 1991. Overcoming soil nutrient constraints to crop production in West Africa: Importance of fertilizers and priorities in soil fertility research. p. 105-115. In A.U. Mokwunye (ed.) Alleviating soil fertility constraints to increased crop production in West Africa. Kluwer Academic Publishers. Dordrecht. the Netherlands. Wright, W. F., 1907. The history of the cowpea and its introduction into America. Bulletin of the United States Department Bureau Plant Industry (102): 43-59. Yan, X., Lynch, J. P. and Beebe, S. E., 1995. Genetic variation for phosphorus efficiency of common bean in contrasting soil types: I. Vegetative response. Crop Science. 35. 1086– 1093. Yeshaneh, G. T., 2015. Assessment of Farmers‘ Perceptions about Soil Fertility with Different Management Practices in Small Holder Farms of Abuhoy Gara Catchemnt. Gidan District. North Wollo. American Journal of Environmental Protection. Vol. 3. No. 4. 137- 144. Yu, W., Ding, X., Xue, S., Li, S., Liao, X. and Wang, R., 2013. Effects of organic-matter application on phosphorus adsorption of three soil parent materials. Journal of Soil Science and Plant Nutrition vol.13no.4. Zahedifar, M., Karimian, N. A., Ronaghi, A., Yasrebi, J. and Emam Y., 2011. Effect of phosphorus and organic matter on soil-plant phosphorus relationships in spinach. Journal of Science & Technology Greenhouse Culture. Vol. 1. No. 4. Winter 2011. Isfahan Univ. Technol. Isf. Iran. Zannou, A., Kossou, D. K., Ahanchede, A., Zoundjihekpon, J. E., Agbicodo, J. E. and Struik, P. C., 2008. Genetic variability of cultivated cowpea in Benin assessed by random amplified polymorphic DNA. African Journal of Biotechnology 7 (24) 4407-4414. Zouera, S. B. G., Garba, M., Ayoade, J. O., Taiwo, O. J. and Mahaman, S., 2016. Striga Impacts on Farmers‘ Field and Control Methods Under Low-Input Agriculture in NigerStriga and Control Methods Under Smallholder Farming in Niger. MSc thesis. Faculty of the Social Sciences. University of Ibadan. Ibadan. Nigeria. 75p. 182 University of Ghana http://ugspace.ug.edu.gh APPENDIX Appendix 3.1. Focus group discussion interview guide REPUBLIQUE DU NIGER MINISTERE DE L’AGRICULTURE Institut National de Recherche Agronomique du Niger (INRAN) EVALUATION OF FARMERS’ PERCEPTION OF SOIL FERTILITY IMPACT ON CROP PRODUCTION FOCUS DISCUSSION INTERVIEW GUIDE Region Department Commune Village 1. List the crop grown in the area 2. Do you grow cowpea? Yes No 3. Type of varieties : Local______ Improved_____ Both ____ 4. Do you treat the seeds before planting? Yes No If yes how do you it? 5. How do you grow cowpea? 1= Sole cropping 2= In association Reasons : 183 University of Ghana http://ugspace.ug.edu.gh Which crop was mixed with cowpea? _____________________________ 6. Which cropping system is the best? Sole cropping____ Mixed cropping ____ 7. Do you use fertilizers ? Yes No 8. Type of fertilizer used : __________________ 9. What are the cowpea production constraints in your area? a. Drought b. Insects c. Diseases d. Striga e. Other witch weeds f. Low soil fertility g. Birds h. Lack of quality seeds i. Others 10. Rank five main production constraints based on yield loss caused 1_________________ 2 _______________ 3_____________ 4______________5______ 11. How do you solve the problems? 12. Is low soil fertility a major constraint to cowpea production ?____Yes _____No 13. What factors contribute to soil fertility deterioration? 14. What impact does low soil fertility have on cowpea production? 15. How do you distinguish the symptoms of low soil fertility on cowpea? 16. How do you improve the soil fertility of your farms? a. Fertilizer application b. Organic manure c. Compost d. Fallow e. Mixed cropping 184 University of Ghana http://ugspace.ug.edu.gh f. Leaving crop residues on the field g. None h. Others If you do not do any of the above, what are the reasons? a. Lack of money b. Not available in local markets c. Transportation problem d. High price e. Lack of livestock f. Low quality of inputs in local markets g. Others 17. What impact does fertilizer application have on cowpea production? 1. Grain and fodder yield Improvement 2. Improve grain yield only 3. Improve fodder yield only 4. Others 5. General comments 185 University of Ghana http://ugspace.ug.edu.gh Appendix 3.2. Household questionnaire 1. "Questionnaire Number" 2. "Enumerator code " 1 Abdou Souleymane" 2 "Issaka Hamani Boureima" 3. Region" 1 "Dosso" 2 "Maradi" 3 "Zinder" 4. Departement" 1 "Doutchi" 2 "Boboye" 3 "Loga" 4 "Dakoro" 5 "Mayahi" 6 "Guidan Roumdji" 7 "Tanout" 8 "Goure" 9 "Matameye" 5. Commune" . 6. Village" 7. Date" 8. Household head Name" 9. Age " Sex" 1 "M" 2 "F" 10. Marital status "1 "Maried" 2 "Single" 3 "Divorced" 4 "Widow" 11. Formal education level" 1 "None" 2 "Primary" 3 "Secondary" 4 "Tertiary" 12. Primary occupation" 1 "None" 2 "Student" 3 "Agriculture" 4 "Livestock rearing" 5 "Transportation" 6 "Small business" 7 "Retired worker 8 "Civil servant" 9 "Others" ; 13. Secondary occupation" 1 "None" 2 "Student" 3 "Agriculture" 4 " Livestock rearing " 5 "Transportation" 6 " Small business " 7 " Retired worker " 8 " Civil servant " 9 " Others " 14. Social status" 1 "Village chief" 2 "Opinion leader" 3 "Elected councilor" 6 " Just Farmer" 15. Member of farmer organization" Yes No Data on farms: information was collected on a maximum of five farms for each respondent. 186 University of Ghana http://ugspace.ug.edu.gh 16. Farm1 "Code " 17. Name of the farm" 18. List of crop grown" 19. "Area in hectare" 20. "Cowpea production (grain) in last cropping season" 21. Type of variety used" 1 "Improved" 2 "Local" 3 "NA" 22. Do you leave crop residues on the farm" Yes No 23. How do you judge your cowpea production on that farm?" 1 "Low" 2 "Medium" 3 "Good" 9 "NA" 24. If you think the production is low, what could be the reasons?" 1 "Low rainfall" 2 "Uneven rain distribution" 3 "Flood" 4"Insects" 5 "Disease" 6 "Late planting" 7 "Late weeding" 8 "Inadequate management of the field" 9 "Soil not adapted to cowpea production" 10 "Eroded soil 11 "Low soil fertility" 12 "Low yielding variety" 13 "Poor quality seeds" 15 "Farm far from home" 16 "Others" 99 "NA" 25. Type of soil" 1 "Clay" 2 "Sand" 3 "Silt" 4 "Clay-sand" 5 "Clay-silt" 6 "Sand-silt" 7 "Others" 9 "NA" 26. Mode of acquisition" 1 "Inheritance" 2 "Purchase" 3 "Gift from government" 4 "Borrow" 5 "Hire" 6 "Others" 9 "NA" ; 27. Fertility level" 1 "Very fertile" 2 "Fertile" 3 "Poor" 4 "very poor" 9 "NA" 28. Farm soil erosion level" 1 "Soil not eroded" 2 "Eroded soil" 3 "very eroded soil " 9 "NA" 29. Have you applied mineral fertilizer to your farm?" 1 "Yes" 2 "No" 30. Have you applied organic fertilizer to your farm?" 1 "Yes" 2 "No" 187 University of Ghana http://ugspace.ug.edu.gh 31. Have you applied insecticide to your farm?" 1 "Yes" 2 "No" 32. Sex of the farm manager" 1 "Masculin" 2 "Féminin" Varietal preference 33. Seed size" 1 "Large" 2 "Medium" 3 "Small" 34. Seed coat color" 1 "White" 2 "Dark Red" 3 "Red" 4 "Black" 5 " Speckled " 6 "Other" 35. Seed weight" 1 "Heavy" 2 "Medium" 3 "Light 36. Grain yield" 1 "High" 2 "Medium" 3 "Low" 37. Fodder yield" 1 "High" 2 "Medium" 3 "Low" 38. Drought tolerance" 1 "Early maturity" 2 "Tolerant" 3 "Susceptible" 39. Insects tolerance" 1 "Resistant" 2 "Tolerant 3 "Susceptible" 40. Resistance to Striga" 1 "Resistant" 2 "Tolerant" 3 "Susceptible" 41. Tolerance to low fertile soils" 1"Tolerant" 3 "Susceptible" 42. Maturity" 1 "Early" 2 "Medium" 3 "Late" 43. Dual purpose variety" 1 "Yes" 2 "No" 44. Taste" 1 "Good" 2 "Sweet" 3 "NA" 45. Cooking time" 1 "Long" 2 "Moyen" 3 "Short" 46. Market value" 1 "High" 2 "Medium" 3 "Low" 188 University of Ghana http://ugspace.ug.edu.gh 47. Choose the first five preferred traits and ranked it from 1 to 5 1 "Most preferred" 5 "Less preferred" 189 University of Ghana http://ugspace.ug.edu.gh Appendix 4.1. List of cowpea genotypes used in the main season trial ENTRY_No VARIETY NAME TYPE ORIGIN 1 Gorom Local Local Burkina faso 2 IT98K-131-4 Improved IITA 3 IT98K-205-8 Improved IITA 4 IT04K-332-2 Improved IITA 5 IT87D-1083 Improved IITA 6 IT98K-630 Improved IITA 7 IT91K186 Improved IITA 8 Sadore local Local Niger 9 Oubritenga Beng Raogol1 Local Burkina faso 10 LOCAL GOROUAL Local Niger 11 Local chical Local Niger 12 Halidou Djibo Local Niger 13 GEC Local USA 14 IT89KD-245 Improved IITA 15 IT81D-994 Improved IITA 16 Dan Illa INRAN Improved Niger 17 IT98D-1399 Improved IITA 18 IT 90K-277-2 Improved IITA 20 IT84S22-46-6 Improved IITA 21 IT90K-76 Improved IITA 22 Aloka Local IITA 23 Danila Local IITA 24 Dan Gorda Local Niger 25 Kwaama Shalla Local NIGER 26 IT07K-181-51 Improved IITA 27 Hannoun Mareni Shalla Local NIGER 28 Mai Diya da Harawa lido Local NIGER 29 Dan Moukaddass Zanne Local NIGER 30 Dan Tchaana Local NIGER 31 Yawa Wild NIGER 32 Matarka Nawa Issufuri Local NIGER 33 Dan Batsatsa Kornaka Local NIGER 34 Bougou Bougou Local NIGER 35 Dan Dam Ali neino Local NIGER 36 Tcheto Dan kassari Local NIGER 37 Tessa Abdoulaye mai goshi Local NIGER 38 Bagagi 21 ans Local NIGER 39 Sababba Sata Gamram Local NIGER 190 University of Ghana http://ugspace.ug.edu.gh 40 Kafi Saley Kimba Local NIGER 41 Kwaama Maiki Local NIGER 42 Ouban Zaoura Dan saga Local NIGER 43 Farin wake Kantari Local NIGER 44 Lakade Karangiya Jimma Local NIGER 45 Gouma Sami Godje Local NIGER 46 Dan Bawada Daji Local NIGER 47 El Danke Maitsakoni Local NIGER 48 Dan Gawouna Gourgouzou Local NIGER 49 Dan Mallam Idi Dan saga Local NIGER 50 TAZAM Local ICRISAT 51 Farin wake Matankari Local NIGER 52 IT88D-867-11 Improved IITA 53 IT97K-499-37 Improved IITA 54 Salga Sabongari Local NIGER 55 Dan Baidari Nguiguimi Local NIGER 56 Dan Mangaza Wangarawa Local NIGER 57 Takatchame Jaja Local NIGER 58 Dan Bekori Bande Local NIGER 59 Darey kiota est Local NIGER 60 LOCALTAMOU Local NIGER 61 Baba Gamus Babul Local NIGER 62 Dan Baouchi Dan Saga Local NIGER 63 Arnen wake Matankari Local NIGER 64 Doungouri Bero Babiadey Local NIGER 65 SOUMBOUM BENG RAOGO 9 Local Burkina faso 66 Nakoussou Jaja Local NIGER 67 Dan Tassaoua Local NIGER 68 TN 2780 Improved NIGER 69 Gnagaou Lido Local NIGER 70 Dan Tonono Dan Saga Local NIGER 71 Hwarin wake kornaka Local NIGER 72 Dan Tahoua mayrere Local NIGER 73 MAMOUDA Local ICRISAT 74 Baadare Karanguiya Local NIGER 75 Dan galgagia Guidan Tagno Local NIGER 76 DAKIARA Local ICRISAT 77 Maitsakoni IT Local NIGER 78 Lakade Maiki Local NIGER 79 Dan Biri Faofao Local NIGER 80 Dan Baourota kornaka Local NIGER 191 University of Ghana http://ugspace.ug.edu.gh 81 Karadua Goure marche Local NIGER 82 Gnagaou Lido Local NIGER 83 Tanout marche Local NIGER 84 Yawa 04 Wild NIGER 85 Juin 2015 Local NIGER 86 Yawa20 Wild NIGER 87 Kolozongo 01 Wild NIGER 88 Tera 02 Wild NIGER 89 Isv 368/it81D-994 Improved NIGER 90 Tera Wild NIGER 91 Koukoki Local NIGER 92 Mallam Ibbo Local NIGER 93 Gada Dogo Local NIGER 94 Sabon Kafi sababba sata Local NIGER 95 Dan Biri Goure Marche Local NIGER 96 Dan Raha Local NIGER 97 KAOKIN LOCAL Local ICRISAT 98 NDIAMBOUR Local ICRISAT 99 SAKOURA LOCAL Local NIGER 101 Karadua Garam Tsofua Local NIGER 102 Sa babba sata Icrisat Local NIGER 103 Kelle kelle Mai Mundu Local NIGER 104 Dan Illa botsotso Local NIGER 105 Dan Sabo Jaja Local NIGER 106 Lakade karanguiya saidou souley Local NIGER 107 Dan Danko Karanguiya Local NIGER 108 Tamalalo Araouraye Local NIGER 109 Bagagi 12ans Local NIGER 110 MOUGNE Local NIGER 111 BF 18 Improved ICRISAT 112 ECI185-158 Improved ICRISAT 113 Maigondala Lido Local NIGER 114 ISV 128 Improved ICRISAT 115 Bahaouche Mai Tsakoni Local NIGER 116 Yagaou Dan saga Local NIGER 117 Darey Doungouri Local NIGER 118 Botsotsuwa Ishirnawa Local NIGER 119 Dan Malam Boura Gueza Local NIGER 120 Lakade karanguiya Local NIGER 121 Lakade Kornaka Local NIGER 122 Koubakodey Salamatou Local NIGER 192 University of Ghana http://ugspace.ug.edu.gh 123 Dan Missira Halbawa Local NIGER 124 Dan Dam Mai 2015 Local NIGER 125 Bonbohi Fabirdji Local NIGER 126 Garbey Doungouri Kalapate Local NIGER 127 Doungouri Bero Babiadey Local NIGER 128 KaraKAra Moussa Yagi Local NIGER 129 Ganda bery Local NIGER 130 Lakade Maiki Local NIGER 131 Dan Saley Kornaka Local NIGER 132 Sarkin Haoussa Tamalalo Local NIGER 133 Oloka Kantari Maigambo Local NIGER 134 Bata tatahi Rabi Ado Local NIGER 135 Lakade Dandaji Makaou Local NIGER 136 Lakade shalla Local NIGER 137 Kolalla Sa babba sata Local NIGER 138 Ajekoriya Local NIGER 139 Kollala Tanout Local NIGER 140 Kofadey Doungouri Yena Local NIGER 141 Kolalla Dan arbain Local NIGER 142 Dan Dam Guidan Tagno Local NIGER 143 Roumboukawa Local NIGER 144 Karadua Halbawa Tsaouni Local NIGER 145 Dan Kada Kouloumbouteye Local NIGER 146 Jan Pass Ismihou Local NIGER 147 Dan Watsatsa Kornaka Local NIGER 148 Oloka Maiki Local NIGER 149 Waken chaibou Local NIGER 150 Jiraini Ouban zaoura Jaja Local NIGER 193 University of Ghana http://ugspace.ug.edu.gh Appendix 4.2. List of cowpea lines added in the minor season trial ENTRY_No VARIETY NAME TYPE ORIGIN 151 boutique Matankari Local NIGER 152 Massaw Local NIGER 153 NELOC-153 Local NIGER 154 NELOC-154 Local NIGER 155 NELOC-155 Local NIGER 156 NELOC-156 Local NIGER 157 NELOC-157 Local NIGER 158 NELOC-158 Local NIGER 159 NELOC-159 Local NIGER 160 NELOC-160 Local NIGER 161 NELOC-161 Local NIGER 162 NELOC-162 Local NIGER 163 NELOC-163 Local NIGER 164 NELOC-164 Local NIGER 165 NELOC-165 Local NIGER 166 NELOC-166 Local NIGER 167 NELOC-167 Local NIGER 168 NELOC-168 Local NIGER 169 NELOC-169 Local NIGER 170 NELOC-170 Local NIGER 172 NELOC-172 Local NIGER 173 NELOC-173 Local NIGER 174 NELOC-174 Local NIGER 175 NELOC-175 Local NIGER 176 NELOC-176 Local NIGER 177 NELOC-177 Local NIGER 178 NELOC-178 Local NIGER 179 NELOC-179 Local NIGER 180 NELOC-180 Local NIGER 181 NELOC-181 Local NIGER 182 NELOC-182 Local NIGER 183 NELOC-183 Local NIGER 184 NELOC-184 Local NIGER 185 NELOC-185 Local NIGER 186 NELOC-186 Local NIGER 187 NELOC-187 Local NIGER 188 NELOC- 188 Local NIGER 189 NELOC-189 Local NIGER 194 University of Ghana http://ugspace.ug.edu.gh 190 NELOC- 190 Local NIGER 191 NELOC-191 Local NIGER 192 NELOC-192 Local NIGER 193 NELOC-193 Local NIGER 194 NELOC-194 Local NIGER 195 NELOC-195 Local NIGER 196 NELOC-196 Local NIGER 197 NELOC-197 Local NIGER 198 NELOC-198 Local NIGER 199 NELOC-199 Local NIGER 200 NEWCO-41 Wild NIGER 201 TN 256-87 Improved NIGER 202 IT99K-573-1-5 Improved IITA 203 TN3-78 Improved NIGER 204 TN88-63 Improved NIGER 205 IT90K-372-1-2 Improved IITA 206 KVX30-309 Improved BURKINA 207 Sa babba sata Icrisat1 Local NIGER 208 IT97K-499-35 Improved IITA 209 TN5-78 Improved NIGER 210 Var 28 multi décor Local NIGER 211 Dan Biri goure Local NIGER 212 V15 Cowpea square NA NIGER 213 CS114 NA NIGER 214 NEWCO-036 Wild NIGER 215 Gnagaou Lido1 Local NIGER 216 Doungouri Bero Babiadey1 Local NIGER 195 University of Ghana http://ugspace.ug.edu.gh Appendix 5.1. Combined ANOVA tables for traits Days to flowering Source DF Type III SS Mean Square F Value Pr > F Env 3 2071.36 690.45 67.64 <.0001 Genotype 14 1584.54 113.18 11.09 <.0001 Rep(Env) 6 209.92 34.99 3.43 0.0039 Block(Rep) 6 18.10 3.02 0.30 0.9378 Env*Genotype 42 2947.15 70.17 6.87 <.0001 Error 105 1071.74 10.21 Plant height Source DF Type III SS Mean Square F Value Pr > F Env 3 339469.09 113156.36 152.50 <.0001 Genotype 14 25655.39 1832.53 2.47 0.0048 Rep(Env) 6 44149.30 7358.22 9.92 <.0001 Block(Rep) 6 17128.27 2854.71 3.85 0.0017 Env*Genotype 42 41333.39 984.13 1.33 0.1269 Error 102 75687.30 742.03 Shoot dry weight Source DF Type III SS Mean Square F Value Pr > F Env 3 106338.00 35446.00 249.38 <.0001 Genotype 14 1839.09 131.36 0.92 0.5359 Rep(Env) 6 4497.56 749.59 5.27 <.0001 Block(Rep) 6 1750.69 291.78 2.05 0.0655 Env*Genotype 42 6846.05 163.00 1.15 0.2861 Error 100 14213.72 142.14 196 University of Ghana http://ugspace.ug.edu.gh Days to 50 % maturity Source DF Type III SS Mean Square F Value Pr > F Env 3 4951.45 1650.48 109.74 <.0001 Genotype 14 795.39 56.81 3.78 <.0001 Rep(Env) 6 176.43 29.41 1.96 0.0795 Block(Rep) 9 154.85 17.21 1.14 0.3398 Env*Genotype 42 2051.94 48.86 3.25 <.0001 Error 98 1473.88 15.04 Grain yield per plant Source DF Type III SS Mean Square F Value Pr > F Env 3 17879.48 5959.83 24.36 <.0001 Genotype 14 10120.10 722.86 2.95 0.0009 Rep(Env) 6 4948.92 824.82 3.37 0.0047 Block(Rep) 9 3306.00 367.33 1.50 0.1590 Env*Genotype 42 12261.47 291.94 1.19 0.2390 Error 93 22754.85 244.68 Fodder yield per plant Source DF Type III SS Mean Square F Value Pr > F Env 2 12013.80 6006.90 10.40 <.0001 Genotype 15 99628.90 6641.93 11.50 <.0001 Rep(Env) 4 3893.35 973.34 1.68 0.1615 Block(Rep) 9 11354.69 1261.63 2.18 0.0315 Env*Genotype 30 57968.24 1932.27 3.35 <.0001 Error 81 46790.32 577.66 197 University of Ghana http://ugspace.ug.edu.gh Number of seeds per pod Source DF Type III SS Mean Square F Value Pr > F Env 1 2.76 2.76 1.21 0.2761 Genotype 14 40.00 2.86 1.26 0.2704 Rep(Env) 2 6.63 3.32 1.46 0.2433 Block(Rep) 9 31.75 3.53 1.55 0.1586 Env*Genotype 14 64.10 4.58 2.01 0.0376 Error 47 106.98 2.28 Pod length Source DF Type III SS Mean Square F Value Pr > F Env 1 0.30 0.30 0.12 0.7352 Genotype 14 107.07 7.65 2.96 0.0027 Rep(Env) 2 6.87 3.44 1.33 0.2743 Block(Rep) 9 19.11 2.12 0.82 0.5992 Env*Genotype 14 42.83 3.06 1.18 0.3184 Error 47 121.44 2.58 Pod yield per plant Source DF Type III SS Mean Square F Value Pr > F Env 1 607.84 607.84 3.75 0.0587 Genotype 14 10591.48 756.53 4.67 <.0001 Rep(Env) 2 3283.20 1641.60 10.14 0.0002 Block(Rep) 9 1098.01 122.00 0.75 0.6588 Env*Genotype 14 6618.17 472.73 2.92 0.0030 Error 47 7610.84 161.93 Harvest Index Source DF Type III SS Mean Square F Value Pr > F Env 1 0.04 0.04 1.51 0.2232 Genotype 20 1.72 0.09 3.64 <.0001 Rep(Env) 2 0.08 0.04 1.64 0.2011 Block(Rep) 6 0.08 0.01 0.57 0.7561 Env*Genotype 20 1.08 0.05 2.27 0.0059 Error 73 1.73 0.02 198 University of Ghana http://ugspace.ug.edu.gh Threshing percentage Source DF Type III SS Mean Square F Value Pr > F Env 1 40.82 40.82 0.58 0.4478 Genotype 20 2948.49 147.42 2.10 0.0119 Rep(Env) 2 181.57 90.78 1.30 0.2801 Block(Rep) 9 1475.73 163.97 2.34 0.0227 Env*Genotype 20 3279.94 164.00 2.34 0.0048 Error 70 4903.66 70.05 199 University of Ghana http://ugspace.ug.edu.gh Appendix 5.2. AMMI analysis of variance table for traits Days to flowering Source df SS MS F F_prob Total 179 8263 46.2 * * Treatments 59 6900 117.0 11.91 0.00000 Genotypes 14 1782 127.3 12.96 0.00000 Environments 3 2073 691.0 20.24 0.00000 Block 8 273 34.1 3.48 0.00129 Interactions 42 3045 72.5 7.39 0.00000 IPCA 16 2288 143.0 14.56 0.00000 IPCA 14 564 40.3 4.10 0.00001 Residuals 12 194 16.1 1.64 0.08936 Error 111 1090 9.8 * * Shoot dry weight per plant Source df SS MS F F_prob Total 179 140961 787 * * Treatments 59 120065 2035 13.51 0.00000 Genotypes 14 2120 151 1.01 0.45344 Environments 3 111290 37097 60.18 0.00000 Block 8 4931 616 4.09 0.00028 Interactions 42 6655 158 1.05 0.40739 IPCA 16 4226 264 1.75 0.04757 IPCA 14 1735 124 0.82 0.64295 Residuals 12 694 58 0.38 0.96671 Error 106 15964 151 * * 200 University of Ghana http://ugspace.ug.edu.gh Plant height Source df SS MS F F_prob Total 179 580948 3246 * * Treatments 59 442029 7492 8.72 0.00000 Genotypes 14 29348 2096 2.44 0.00509 Environments 3 372401 124134 21.54 0.00000 Block 8 46104 5763 6.71 0.00000 Interactions 42 40281 959 1.12 0.32053 IPCA 16 19718 1232 1.43 0.13956 IPCA 14 15824 1130 1.32 0.21031 Residuals 12 4739 395 0.46 0.93388 Error 108 92816 859 * * Days to 50 % maturity Source df SS MS F F_prob Total 179 11205 62.6 * * Treatments 59 9365 158.7 10.43 0.00000 Genotypes 14 910 65.0 4.27 0.00001 Environments 3 6364 2121.4 80.01 0.00000 Block 8 212 26.5 1.74 0.09698 Interactions 42 2091 49.8 3.27 0.00000 IPCA 16 1387 86.7 5.70 0.00000 IPCA 14 380 27.2 1.78 0.05026 Residuals 12 324 27.0 1.77 0.06213 Error 107 1629 15.2 * * 201 University of Ghana http://ugspace.ug.edu.gh Grain yield per plant Source df SS MS F F_prob Total 179 80086 447 * * Treatments 59 47250 801 3.13 0.00000 Genotypes 14 12061 861 3.37 0.00018 Environments 3 22499 7500 8.86 0.00003 Block 8 6775 847 3.31 0.00208 Interactions 42 12690 302 1.18 0.24585 IPCA 16 6388 399 1.56 0.09288 IPCA 14 3384 242 0.95 0.51335 Residuals 12 2918 243 0.95 0.49957 Error 102 26061 255 * * 202 University of Ghana http://ugspace.ug.edu.gh Appendix 6.1: Combined analysis of variance table for traits across the planting dates of 2016 and 2017. Grain yield Source DF Type III SS Mean Square F Value Pr > F Env 5 56513.96 11302.79 92.56 <.0001 Genotype 18 28318.29 1573.24 12.88 <.0001 Rep(Env) 5 926.07 185.21 1.52 0.1925 Block(Rep) 8 1037.79 129.72 1.06 0.3966 Env*Genotype 90 31683.23 352.04 2.88 <.0001 Error 90 10990.50 122.12 Shoot dry weight Source DF Type III SS Mean Square F Value Pr > F Env 5 125839.69 25167.94 91.82 <.0001 Genotype 18 6765.77 375.88 1.37 0.1642 Rep(Env) 5 9018.04 1803.61 6.58 <.0001 Block(Rep) 8 5836.42 729.55 2.66 0.0111 Env*Genotype 90 24428.18 271.42 0.99 0.5179 Error 95 26038.48 274.09 Plant height Source DF Type III SS Mean Square F Value Pr > F Env 5 89514.87 17902.97 16.97 <.0001 Genotype 18 60687.89 3371.55 3.20 0.0001 Rep(Env) 5 101192.30 20238.46 19.19 <.0001 Block(Rep) 8 21131.84 2641.48 2.50 0.0160 Env*Genotype 90 119765.41 1330.73 1.26 0.1293 Error 99 104415.40 1054.70 203 University of Ghana http://ugspace.ug.edu.gh Days to 50 % maturity Source DF Type III SS Mean Square F Value Pr > F Env 4 9543.32 2385.83 500.76 <.0001 Genotype 17 3477.89 204.58 42.94 <.0001 Rep(Env) 4 29.52 7.38 1.55 0.1971 Block(Rep) 4 51.05 12.76 2.68 0.0383 Env*Genotype 67 1660.77 24.79 5.20 <.0001 Error 73 347.80 4.76 204 University of Ghana http://ugspace.ug.edu.gh Appendix 6.2: Combined analysis of variance tables for traits across the planting dates of 2016 Days to flowering Source DF Type III SS Mean Square F Value Pr > F Env 2 5195.45 2597.73 205.86 <.0001 Genotype 19 3145.61 165.56 13.12 <.0001 Rep(Env) 2 36.02 18.01 1.43 0.2498 Block(Rep) 8 131.63 16.45 1.30 0.2638 Env*Genotype 38 806.65 21.23 1.68 0.0433 Error 49 618.32 12.62 Days to 50 % maturity Source DF Type III SS Mean Square F Value Pr > F Env 2 9816.76 4908.38 1497.84 <.0001 Genotype 18 2308.99 128.28 39.15 <.0001 Rep(Env) 2 11.81 5.91 1.80 0.1764 Block(Rep) 8 31.89 3.99 1.22 0.3109 Env*Genotype 36 833.50 23.15 7.07 <.0001 Error 46 150.74 3.28 Shoot dry weight Source DF Type III SS Mean Square F Value Pr > F Env 2 4183.01 2091.51 17.17 <.0001 Genotype 19 2261.06 119.00 0.98 0.5028 Rep(Env) 2 2057.69 1028.84 8.45 0.0008 Block(Rep) 8 1569.20 196.15 1.61 0.1493 Env*Genotype 38 2898.69 76.28 0.63 0.9281 Error 44 5358.17 121.78 205 University of Ghana http://ugspace.ug.edu.gh Plant height Source DF Type III SS Mean Square F Value Pr > F Env 2 12599.67126 6299.83563 4.93 0.0113 Genotype 19 52723.96819 2774.94569 2.17 0.0156 Rep(Env) 2 83396.83784 41698.41892 32.61 <.0001 Block(Rep) 8 26172.15822 3271.51978 2.56 0.0208 Env*Genotype 38 42227.22863 1111.24286 0.87 0.6705 Error 48 61383.0986 1278.8146 Grain yield per plant Source DF Type III SS Mean Square F Value Pr > F Env 2 2697.34 1348.67 22.77 <.0001 Genotype 16 5603.42 350.21 5.91 <.0001 Rep(Env) 2 55.74 27.87 0.47 0.6293 Block(Rep) 8 446.13 55.77 0.94 0.4986 Env*Genotype 32 3550.17 110.94 1.87 0.0456 Error 29 1717.54 59.23 206 University of Ghana http://ugspace.ug.edu.gh Appendix 6.3. Combined analysis of variance table for traits across the planting dates of 2017 Plant height Source DF Type III SS Mean Square F Value Pr > F Env 2 33701.64 16850.82 14.42 <.0001 Genotype 19 49666.46 2614.02 2.24 0.0122 Rep(Env) 2 3634.78 1817.39 1.56 0.2214 Block(Rep) 8 12529.25 1566.16 1.34 0.2466 Env*Genotype 38 64498.06 1697.32 1.45 0.1088 Error 49 57261.07 1168.59 Striga infestation Source DF Type III SS Mean Square F Value Pr > F Env 2 0.42 0.21 1.00 0.3755 Genotype 19 82.79 4.36 20.90 <.0001 Rep(Env) 2 0.52 0.26 1.24 0.2985 Block(Rep) 8 4.13 0.52 2.48 0.0242 Env*Genotype 38 14.18 0.37 1.79 0.0276 Error 49 10.22 0.21 Grain yield Source DF Type III SS Mean Square F Value Pr > F Env 2 48417.46 24208.73 127.13 <.0001 Genotype 19 24405.91 1284.52 6.75 <.0001 Rep(Env) 2 1000.84 500.42 2.63 0.0828 Block(Rep) 8 1025.07 128.13 0.67 0.7126 Env*Genotype 38 19322.61 508.49 2.67 0.0008 Error 47 8950.26 190.43 207 University of Ghana http://ugspace.ug.edu.gh Shoot dry weight Source DF Type III SS Mean Square F Value Pr > F Env 2 12800.67 6400.33 15.30 <.0001 Genotype 18 10839.70 602.21 1.44 0.1588 Rep(Env) 2 3756.35 1878.17 4.49 0.0165 Block(Rep) 8 5746.85 718.36 1.72 0.1198 Env*Genotype 36 12144.17 337.34 0.81 0.7463 Error 46 19240.93 418.28 Days to 50 % maturity Source DF Type III SS Mean Square F Value Pr > F Env 1 129.81 129.81 21.59 <.0001 Genotype 19 2431.94 128.00 21.29 <.0001 Rep(Env) 1 11.16 11.16 1.86 0.1836 Block(Rep) 8 104.86 13.11 2.18 0.0596 Env*Genotype 19 293.48 15.45 2.57 0.0108 Error 29 174.38 6.01 Fodder yield per plant Source DF Type III SS Mean Square F Value Pr > F Env 2 174944.23 87472.12 84.23 <.0001 Genotype 19 126103.54 6637.03 6.39 <.0001 Rep(Env) 2 14226.48 7113.24 6.85 0.0024 Block(Rep) 8 15252.86 1906.61 1.84 0.0927 Env*Genotype 38 73586.35 1936.48 1.86 0.0202 Error 49 50883.92 1038.45 Seed number per pod Source DF Type III SS Mean Square F Value Pr > F Env 2 528.68 264.34 44.88 <.0001 Genotype 18 154.19 8.57 1.45 0.1675 Rep(Env) 2 40.15 20.07 3.41 0.0444 Block(Rep) 8 9.19 1.15 0.20 0.9898 Env*Genotype 36 129.19 3.59 0.61 0.9280 Error 35 206.14 5.89 208 University of Ghana http://ugspace.ug.edu.gh Pod length Source DF Type III SS Mean Square F Value Pr > F Env 2 130.30 65.15 17.96 <.0001 Genotype 18 169.65 9.42 2.60 0.0076 Rep(Env) 2 25.21 12.60 3.47 0.0420 Block(Rep) 8 23.12 2.89 0.80 0.6093 Env*Genotype 36 174.89 4.86 1.34 0.1947 Error 35 126.95 3.63 209 University of Ghana http://ugspace.ug.edu.gh Appendix 6.4 AMMI analysis of variance tables for traits across the planting dates of 2016 and 2017 Grain yield Source df SS MS F F_prob Total 227 140911 621 * * Treatments 113 127729 1130 9.21 0.00000 Genotypes 18 33661 1870 15.24 0.00000 Environments 5 58890 11778 61.25 0.00000 Block 6 1154 192 1.57 0.16496 Interactions 90 35178 391 3.18 0.00000 IPCA 22 24832 1129 9.20 0.00000 IPCA 20 5460 273 2.22 0.00516 Residuals 48 4886 102 0.83 0.76171 Error 98 12028 123 * * Plant height Source df SS MS F F_prob Total 227 541616 2386 * * Treatments 113 307616 2722 2.32 0.00001 Genotypes 18 71070 3948 3.37 0.00005 Environments 5 90295 18059 1.00 0.42199 Block 6 108453 18075 15.41 0.00000 Interactions 90 146251 1625 1.38 0.05307 IPCA 22 55558 2525 2.15 0.00515 IPCA 20 40565 2028 1.73 0.03942 Residuals 48 50129 1044 0.89 0.66916 Error 107 125547 1173 * * 210 University of Ghana http://ugspace.ug.edu.gh Shoot dry weight per plant Source df SS MS F F_prob Total 227 212127 934 * * Treatments 113 170443 1508 4.87 0.00000 Genotypes 18 8077 449 1.45 0.12469 Environments 5 136161 27232 16.66 0.00000 Block 6 9809 1635 5.28 0.00009 Interactions 90 26205 291 0.94 0.61529 IPCA 22 12320 560 1.81 0.02500 IPCA 20 6072 304 0.98 0.49087 Residuals 48 7813 163 0.53 0.99280 Error 103 31875 309 * * Fodder yield per plant Source df SS MS F F_prob Total 119 519361 4364 * * Treatments 59 430651 7299 6.29 0.00000 Genotypes 19 164648 8666 7.47 0.00000 Environments 2 174944 87472 11.63 0.00006 Block 3 22573 7524 6.48 0.00075 Interactions 38 91059 2396 2.07 0.00641 IPCA 20 52475 2624 2.26 0.00849 IPCA 18 38584 2144 1.85 0.04113 Residuals 0 0 * * * Error 57 66137 1160 * * 211 University of Ghana http://ugspace.ug.edu.gh Appendix 6.5. AMMI analysis of variance tables for traits across the planting dates of 2016 Days to 50 % maturity Source df SS MS F F_prob Total 113 13757 121.7 * * Treatments 56 13550 242.0 71.54 0.00000 Genotypes 18 2578 143.2 42.34 0.00000 Environments 2 10015 5007.7 604.10 0.00000 Block 3 25 8.3 2.45 0.07330 Interactions 36 957 26.6 7.86 0.00000 IPCA 19 751 39.5 11.69 0.00000 IPCA 17 206 12.1 3.58 0.00018 Residuals 0 0 * * * Error 54 183 3.4 * * Grain yield Source df SS MS F F_prob Total 101 18417 182.3 * * Treatments 50 16199 324.0 5.54 0.00000 Genotypes 16 7888 493.0 8.43 0.00000 Environments 2 3062 1530.9 85.27 0.00000 Block 3 54 18.0 0.31 0.82014 Interactions 32 5249 164.0 2.80 0.00144 IPCA 17 3776 222.1 3.80 0.00034 IPCA 15 1473 98.2 1.68 0.09937 Residuals 0 0 * * * Error 37 2164 58.5 * * 212 University of Ghana http://ugspace.ug.edu.gh Days to flowering Source df SS MS F F_prob Total 119 10462 87.9 * * Treatments 59 9664 163.8 12.45 0.00000 Genotypes 19 3576 188.2 14.31 0.00000 Environments 2 5195 2597.7 162.19 0.00000 Block 3 48 16.0 1.22 0.31174 Interactions 38 892 23.5 1.78 0.02340 IPCA 20 616 30.8 2.34 0.00637 IPCA 18 276 15.3 1.16 0.32017 Residuals 0 0 * * * Error 57 750 13.2 * * Height Source df SS MS F F_prob Total 119 333360 2801 * * Treatments 59 144794 2454 1.57 0.04563 Genotypes 19 69323 3649 2.33 0.00740 Environments 2 13702 6851 0.20 0.81649 Block 3 101010 33670 21.54 0.00000 Interactions 38 61769 1625 1.04 0.44064 IPCA 20 40165 2008 1.28 0.22789 IPCA 18 21604 1200 0.77 0.72676 Residuals 0 0 * * * Error 56 87555 1563 * * 213 University of Ghana http://ugspace.ug.edu.gh Shoot dry weight Source df SS MS F F_prob Total 119 23110 194.2 * * Treatments 59 13010 220.5 1.66 0.03289 Genotypes 19 3486 183.5 1.38 0.17967 Environments 2 4753 2376.4 2.25 0.11587 Block 3 3173 1057.6 7.94 0.00019 Interactions 38 4771 125.5 0.94 0.57126 IPCA 20 3681 184.0 1.38 0.17447 IPCA 18 1090 60.5 0.45 0.96591 Residuals 0 0 * * * Error 52 6927 133.2 * * 214 University of Ghana http://ugspace.ug.edu.gh Appendix6.6: AMMI analysis of variance tables for traits across the planting dates of 2017 Fodder yield per plant Source df SS MS F F_prob Total 143 246651 1725 * * Treatments 47 183403 3902 6.04 0.00000 Genotypes 15 115333 7689 11.90 0.00000 Environments 2 12014 6007 7.06 0.00142 Block 6 5103 851 1.32 0.25796 Interactions 30 56056 1869 2.89 0.00006 IPCA 16 38287 2393 3.70 0.00003 IPCA 14 17769 1269 1.96 0.02959 Residuals 0 0 * * * Error 90 58145 646 * * Pod lenght Source df SS MS F F_prob Total 113 769.2 6.81 * * Treatments 56 589.4 10.52 3.02 0.00013 Genotypes 18 248.6 13.81 3.96 0.00011 Environments 2 152.1 76.07 7.67 0.00141 Block 3 29.7 9.91 2.84 0.04895 Interactions 36 188.6 5.24 1.50 0.10091 IPCA 19 104.9 5.52 1.58 0.10581 IPCA 17 83.7 4.92 1.41 0.17891 Residuals 0 0.0 * * * Error 43 150.1 3.49 * * 215 University of Ghana http://ugspace.ug.edu.gh Number of seeds per pod Source df SS MS F F_prob Total 113 1247.8 11.04 * * Treatments 56 975.1 17.41 3.48 0.00002 Genotypes 18 225.4 12.52 2.50 0.00704 Environments 2 585.8 292.89 15.33 0.00001 Block 3 57.3 19.11 3.82 0.01644 Interactions 36 163.9 4.55 0.91 0.61292 IPCA 19 84.4 4.44 0.89 0.59948 IPCA 17 79.5 4.67 0.93 0.54332 Residuals 0 0.0 * * * Error 43 215.3 5.01 * * 216