I GENETIC ANALYSIS OF EARLINESS AND DROUGHT TOLERANCE IN GROUNDNUT (Arachis hypogaea L.) IN NIGER By Coulibaly Adama Mamadou (10293994) THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE AWARD OF PHD PLANT BREEDING DEGREE WEST AFRICA CENTRE FOR CROP IMPROVEMENT SCHOOL OF AGRICULTURE COLLEGE OF AGRICULTURE AND CONSUMER SCIENCES UNIVERSITY OF GHANA LEGON DECEMBER, 2013 University of Ghana http://ugspace.ug.edu.gh i 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. .................................................. Coulibaly Adama Mamadou Student .................................................. Prof. Eric Y. Danquah Supervisor .................................................. Prof. Kwadwo Ofori Supervisor .................................................. Prof. Vernon Gracen Supervisor .................................................. Dr. Ntare R. Bonny Supervisor University of Ghana http://ugspace.ug.edu.gh ii ABSTRACT End-of-season drought, characterized by low and erratic rainfall, is the most important factor limiting groundnut (Arachis hypogaea L.) production in Niger. Information about farmers' varietal preferences and production constraints are limited. Genotypes have not yet been screened for earliness and tolerance to end-of-season drought. Data on combining ability and heritability estimates are not available. Development of varieties that have better ability to use limited available water or that mature early and avoid drought is needed. A participatory rural appraisal (PRA) through Focus Group Discussion session was conducted in 2011 to assess farmers' perceptions on groundnut production constraints and to gather information on preferences for elite groundnut varieties. One hundred and fifty genotypes were screened in 2010 under well watered conditions in an alpha lattice design in an unreplicated trial to select five early maturing parental lines for genetic studies. One hundred intermediate maturing genotypes were selected and, evaluated in 2011 off- season under both well watered and end-of-season drought conditions in an alpha lattice design with two replications to select five drought tolerant parental lines for genetic studies. The selected parental lines were crossed with the three farmers preferred varieties in a North Carolina II mating design and 7 F4 populations for earliness and 7 F4 populations for drought tolerance were developed by selfing. Seven F3 populations were evaluated for earliness and agronomic traits in RCBD replicated three times under well watered conditions. Seven F3 and F4 populations were evaluated for drought related traits in RCBD replicated three times. Data recorded included: % emergence, 50% plants flowering, pod yield, seed weight, seed number, shelling %, pod length, maturity index, and biomass. Among the physiological traits measured University of Ghana http://ugspace.ug.edu.gh iii were SPAD chlorophyll meter reading and harvest index. The data were analyzed using appropriate statistical procedures and experimental designs. The PRA survey revealed that drought was a major production constraint followed by low soil fertility. In the study regions, the varieties 55-437, RRB and JL24, were the most grown. There was variation among the genotypes screened for the various characteristics. The PCA analysis revealed that 79.62% of the total variation among the genotypes was explained by maturity, pod yield, biomass, harvest index and 50% plants flowering. Out of the 150 genotypes, two extra early maturing genotypes (75 days) Chico and ICIAR19BT, three early maturing genotypes (80 days) ICG3584, 796, and ICGV02022 and five drought tolerant genotypes ICGV- SM99511, Tainan-9, ICG11249, ICG6703 and ICGIS01820 were identified. In the early maturity experiment, the analysis of variance for the eight traits revealed significant differences (P ≤ 0.01 and P ≤ 0.05) among parents for pod yield, seed weight, pod length and maturity index. Crosses showed highly significant differences for pod yield (P ≤ 0.001), seed number (P ≤ 0.001), shelling % (P ≤ 0.01) and pod length (P ≤ 0.05). Female and male General combining ability (GCA) mean squares were highly significant (P ≤ 0.01) only for pod yield. GCA mean squares (119.26) for females were greater than males GCA mean square (32.11) for pod weight indicating that the major contribution to additive variance for this trait was due to the female parents. Specific combining ability (SCA) mean squares differed significantly (P ≤ 0.01 and P ≤ 0.05) for seed number and shelling percentage indicating the importance of both additive and non-additive variance for these characters. Chico, ICIAR19BT, 55-437 and RRB were the best general combiners. The estimates of phenotypic coefficient of variation (PCV) were greater University of Ghana http://ugspace.ug.edu.gh iv than genotypic coefficients of variation (GCV) for all the traits studied. None of the traits recorded high PCV and GCV. PCV ranged from 4.36 to 19.34% and GCV from 2.02 to 11.99%. Narrow sense heritability estimates ranged from 35.4 to 95.5%; for days to emergence (95.5%), pod yield (85%), shelling percentage (84.3%), seed number (80%) and 50% plants flowering (72%). Moderate to low heritability estimates were obtained for maturity index (66.1%), pod length (54.4%) and seed weight (35.4%). Highly significant positive correlations were found between pod yield and seed weight (r = 0.97, P ≤ 0.001) and pod length (r = 0.72, P ≤ 0.01); between pod length and seed yield (r = 0.66, P ≤ 0.01); and between maturity index and days to emergence (r = 0.72, P ≤ 0.001). Correlation between shelling percentage and seed weight was positively significant at P ≤ 0.05 (r = 0.61). In the drought tolerant experiment, combined analysis of variance showed large and significant differences (P ≤ 0.01) between all 14 entries for all traits except for pod yield. Highly significant differences (P ≤ 0.01 and P ≤ 0.05) in chlorophyll content among the crosses at 60 and 80 DAS were found. The overall means of chlorophyll content under end-of-season drought stress conditions were 41.55 and 39.65, at 60 and 80 DAS respectively compared to 35.27 and 38.04 under irrigated conditions. The analysis of variance showed highly significant differences (P ≤ 0.01) between the crosses in water regimes for pod yield. The drought tolerance index for pod yield ranged from 0.44 to 1.37 for seven crosses. GCA mean squares for males and females were significant (P ≤ 0.05) for pod yield and biomass. GCA mean squares for females were greater than males GCA mean square for pod yield and total biomass indicating that the major contribution to additive variance for these traits was due the female parents. SCA mean squares differed significantly (P ≤ 0.05) for pod yield and harvest University of Ghana http://ugspace.ug.edu.gh v index indicating the importance of both additive and non-additive variance for these traits. Tainan-9, ICGV-SM99511, 55-437 and RRB were the best general combiners. The estimates of phenotypic coefficient of variation (PCV) were greater than genotypic coefficients of variation (GCV) for all the physiological traits. PCV ranged from 7.27 to 40.20% and GCV from 2.71 to 15.91%. Narrow sense heritability estimates for physiological traits were higher than for agronomic traits, and varied under both well-watered and end-of-season drought conditions. The heritability estimates for pod yield (0.26) and biomass (0.18) were low, but they were high for harvest index (0.78) and SPAD chlorophyll meter reading (SCMR), estimate of 0.71 and 0.68 respectively, at 60 and 80 days after sowing. Fifty percent (50%) to flowering showed moderate heritability estimate of 0.56. End-of-season drought decreased narrow sense heritability estimates for all the physiological traits. Heritability estimates decreased from 71.6% to 36.8% for SCMR 60 DAS; from 68.2% to 10.6% for SCMR 80 DAS and from 78.9% to 37.8% for harvest index. Highly significant positive association between pod yield and harvest index was found in both water regimes. The correlation coefficient was higher (r = 0.77, P ≤ 0.001) under end-of-season drought stress conditions than under well watered conditions (r = 0.74, P ≤ 0.01). Strong negative significant (P ≤ 0.05) correlation was found between biomass and harvest index under well watered and end-of-season drought conditions r = -0.66, -0.58, respectively. Weak positive and non-significant correlation was found between SCMR and pod yield, under well watered conditions (r = 0.38). The results show that development of early maturing and drought tolerant varieties may ensure better yields of groundnut in the Sahel. University of Ghana http://ugspace.ug.edu.gh vi ACKNOWLEDGEMENTS First and foremost, I would like to thank ALMIGHTY ALLAH without Whose blessing it would not have been possible for all my expectations to become reality. I would like to gratefully acknowledge the Alliance for Green Revolution in Africa (AGRA) that provided all financial support for my study at the West Africa Center for Crop Improvement (WACCI), University of Ghana, Legon. I am deeply indebted to my supervisors Dr. Bonny R. Ntare, Prof. Eric Y. Danquah, Prof. Vernon Gracen and Prof. Kwadwo Ofori, for their earnest and continuous guidance, critical comments, encouragement and timely suggestions that made this research a success. I would like to express my gratitude to my employer the National Agronomic Research Institute of Niger (INRAN) for giving me the opportunity to follow this programme and for all the administrative and finance supports. I am indebted to Dr Falalou Hamidou at ICRISAT, Niger for supervising my training on phenotyping, and Issa Karimou the technical director of INRAN, Maradi for his support. I am also thankful to all the WACCI staff, the AGRA officers, my colleagues at INRAN and my dear colleagues at WACCI who contributed significantly to the completion of this work. My sincere thanks go to Abdelkader Mahaman Souley, Jiga Mahaman Sani, Zabeirou Maman, Harouna Bakoye and Mahaman Sani Issaka for their support during field work and data collection. Indeed, I am very grateful. Finally, I am very grateful to my beloved wife and children for all their patience, encouragement and for sharing all difficulties. University of Ghana http://ugspace.ug.edu.gh vii DEDICATION Dedicated to my Family. University of Ghana http://ugspace.ug.edu.gh viii TABLE OF CONTENT DECLARATION I ABSTRACT II ACKNOWLEDGEMENTS VI DEDICATION VII TABLE OF CONTENT VIII LIST OF TABLES XIV LIST OF FIGURES XVI LIST OF ABBREVIATIONS XVIII CHAPTER ONE 1 1.0 GENERAL INTRODUCTION 1 CHAPTER TWO 6 2.0. LITERATURE REVIEW 6 2.1 Effect of drought on plant performance 6 2.2. Effect of drought during flowering and pod formation 7 2.2.1. Flowering 7 2.2.2. Pod formation 7 2.3. Environmental resource utilization 8 2.3.1. Groundnut water use 8 2.3. 2. Rainfall or soil moisture 9 2.3.3. Photoperiod or day length 9 2.4. Consequences of water stress 10 University of Ghana http://ugspace.ug.edu.gh ix 2.5. Mechanisms of drought stress avoidance 11 2.6. Approaches for combating drought stress 12 2.6.1. Physiological breeding approach 13 2.6.2. DNA marker-assisted selection 14 2.7. Farmer participation in variety development in Niger 16 2.8. Estimation of genetic parameters in groundnut 17 2.8.1.Combining ability estimates 17 2.8.2. Heritability estimate for agronomic and yield traits 19 CHAPTER THREE 22 3.0. IDENTIFICATION OF THE PRODUCTION CONSTRAINTS AND FARMERS’ PREFERRED VARIETIES OF GROUNDNUT IN NIGER 22 3.1. Introduction 22 3.2. Methodology 23 3.2.1. Study area 23 3.2.2. Study villages 25 3.2.3. Introduction to villages 25 3.2.4. Sampling method 26 3.2.5. Data collection and analysis 26 3.3. Results 28 3.3.1. Farmers preferred varieties in Magaria 28 3.3.2. Farmers preferred varieties in Madarounfa 29 3.3.3. Sources of Seed 30 3.3.4. Production constraints in Magaria 30 3.3.5. Production constraints in Madarounfa 31 3.4. Discussion 32 University of Ghana http://ugspace.ug.edu.gh x 3.5. Conclusions 35 CHAPTER FOUR 36 4.0. PHENOTYPING GROUNDNUT GERMPLASM FOR EARLINESS AND DROUGHT TOLERANCE 36 4.1. Introduction 36 4.2. Experimental site 37 4.3. Weather conditions 37 4.3.1. Temperature 37 4.3.2. Relative Humidity 38 4.3.3. Precipitation 39 4.4. Selection of early maturing varieties 39 4.4.1. Genetic material 39 4.4.2. Experimental conditions 39 4.4.3. Data collection 40 4.4.4. Data analysis 41 4.4.5. Results for earliness screening 41 4.5. Selection of drought tolerant varieties 45 4.5.1. Genetic material 45 4.5.2. Experimental conditions 45 4.5.3. Management of irrigation for treatment application 45 4.5.4. Field day 47 4.5.5. Data collection 47 4.5.6. Statistical Analysis 48 4.5.7. Results of drought tolerance screening 49 University of Ghana http://ugspace.ug.edu.gh xi 4.5.7.1. Farmers selection of the best genotypes under both water regimes 49 4.5.7.2. Pod yield and SCMRs of the drought tolerant genotypes 49 4.5.7.2.1. ANOVA 49 4.5.7.2.2. Drought effect on pod yield 50 4.5.7.2.3. Drought effect on SPAD chlorophyll meter readings 53 4.6. Discussion 57 4.6.1. Early maturity 57 4.6.2. Drought tolerance 58 4.7. Conclusions 61 CHAPTER FIVE 62 5.0. GENETIC ANALYSIS OF EARLINESS AND DROUGHT TOLERANCE 62 5.1. Introduction 62 5.2. Material and methods 64 5.2.1 Experimental site 64 5.2.2. Genetic material and hybridization techniques 65 5.2.3. Evaluation of populations for earliness and drought tolerance 67 5.2.3.1. Field experiments 67 5.2.3.2. Data collection 69 5.2.3.2.1. Early maturity 69 5.2.3.2.2. Drought tolerance 70 5.2.3.3. Data analysis 71 5.3. Results 73 5.3.1. Earliness 73 5.3.1.1. ANOVA 73 5.3.1.2. Mean performance 74 University of Ghana http://ugspace.ug.edu.gh xii 5.3.1.3. Phenotypic and genotypic coefficient of variation estimates 78 5.3.1.4. General and specific combining ability estimates 78 5.3.1.5. Narrow sense heritability estimates 80 5.3.1.6. Correlations 82 5.3.2. Drought tolerance 83 5.3.2.1. ANOVA 83 5.3.2.2. Mean performance 84 5.3.2.3. Effect of drought on SPAD Chlorophyll meter reading 86 5.3.2.4. Effect of drought on Pod yield 88 5.3.2.5. Phenotypic and genotypic coefficient of variation estimates 90 5.3.2.6. General and specific combining ability estimates 91 5.3.2.7. Narrow sense heritability estimates 93 5.3.2.8. Correlations 95 5.4. Discussion 97 5.4.1. Maturity 97 5.4.2. Drought Tolerance 101 5.5. Conclusions 106 CHAPTER SIX 108 6.0. GENERAL DISCUSSION 108 6.1. Identification of the production constraints and farmers’ preferred varieties of groundnut in Niger 108 6.2. Phenotyping groundnut germplasm for earliness and drought tolerance 109 6.3. Genetic analysis of earliness and drought tolerance 111 CHAPTER SEVEN 119 University of Ghana http://ugspace.ug.edu.gh xiii 7.0. GENERAL CONCLUSIONS AND RECOMMENDATIONS 119 7.1. General conclusions 119 7.2. Recommendations 121 BIBLIOGRAPHY AND APPENDICES 122 BIBLIOGRAPHY 122 APPENDICES 139 APPENDIX 1: CHECK LIST OF FOCUS GROUP DISCUSSION 139 APPENDIX 2: PERFORMANCE OF 150 GENOTYPES 140 APPENDIX 3: MEAN PERFORMANCE OF 100 GENOTYPES UNDER BOTH WATER REGIMES 144 University of Ghana http://ugspace.ug.edu.gh xiv LIST OF TABLES Table 3.1: The number of farmers by gender that participated in Focus group discussion sessions 26 Table 3.2: Groundnut seeds sources 30 Table 4.1: Eigenvectors from the three PC axes used to classified groundnut genotypes 42 Table 4. 2: The top 20 genotypes based on their maturity and their performance for other traits 43 Table 4. 3: Mean square from the combined analysis of variance for pod yield and SCMRs under both water regimes 50 Table 4.4: Pod yield of entries under drought stress and their performance under well water conditions and their respective drought tolerance indices 52 Table 4.5: Drought tolerance indices of entries at 45, 60 and 90 DAS 55 Table 5.1: Early maturity and drought tolerant parental lines 67 Table 5. 2: Early maturity and drought tolerant parental lines and F3 populations 68 Table 5.3: Entries used to estimate Narrow sense heritability 69 Table 5. 4: ANOVA in NC Design II 72 Table 5. 5: Mean squares from ANOVA for parental lines and F3 populations 73 Table 5.6: Mean squares from ANOVA for F3 and F4 populations for eight traits 74 Table 5.7: Mean performance of parental lines and F3 populations 75 Table 5.8: Range and mean of F3 and F4 groundnut populations for eight traits 76 Figure 5.1: Pod yield of F3 populations and their corresponding performance in F4 77 Figure 5.2: Maturity index of F3 populations and their corresponding values in F4 77 Table 5. 9: Components of variation for different groundnut traits 78 Table 5.10: Mean square of combining ability analysis for eight traits 79 University of Ghana http://ugspace.ug.edu.gh xv Table 5.11: General combining effect of parental lines for eight traits 79 Table 5.12: Specific combining ability effects of the crosses for eight traits 80 Table 5.13: Narrow sense heritability estimate from genetic component of traits 81 Table 5.14: Narrow sense heritability by parent-offspring regression and standard errors 82 Table 5.15: Correlation between earliness and agronomic traits 82 Table 5.16: Mean squares from ANOVA for parents and F3 populations for physiological traits 83 Table 5.17: Mean squares from the combined analysis of variance for the traits under well watered and end-of-season drought conditions 84 Table 5.18: Mean performance of parental lines and F3 populations for physiological traits 85 Table 5.19: Range, mean and drought tolerance indices of F3 and F4 populations under well- watered and end-of-season drought conditions for the six traits 86 Table 5.20: Pod yield (g) of F3 and F4 populations under well watered and their corresponding performance under end-of-season drought stress and DTI 89 Table 5.21: Components of variation for physiological traits 91 Table 5. 22: Mean square of Combining Ability analysis for the physiological traits 92 Table 5.23: General combining ability effect of parental lines for six traits 92 Table 5.24: Specific combining ability effects of the seven crosses for six traits 93 Table 5.25: Narrow sense heritability estimate from genetic component of physiological traits 94 Table 5.26: Narrow sense heritability by parent-offspring regression and standard errors of the seven groundnut crosses for physiological traits 95 Table 5.27: Correlations between physiological and agronomic traits under both water regimes 96 University of Ghana http://ugspace.ug.edu.gh xvi LIST OF FIGURES Figure 3.1: Niger map with study areas 24 Figure 3.2: Farmers preferred varieties ranked in Magaria 28 Figure 3.3: Farmers preferred varieties ranked at Madarounfa 29 Figure 3.4: Main production constraints ranked by gender in the 3 villages of Magaria 31 Figure 3.5: Main production constraints ranked by gender in the 3 villages of Madarounfa 32 Figure 4.1: Temperature variation during 2011 38 Figure 4.2: Humidity variation during 2011 38 Figure 4.3: Rainfall pattern from 1980 to 2009 at Tarna INRAN station 39 Figure 4.4: Dendrogram of selected early maturing genotypes 44 Figure 4.5: Drought stress imposition and irrigation frequencies 46 Figure 4. 6: Pod yield under both water regimes 51 Figure 4.7a: Chlorophyll content 45 DAS under WS and WW conditions 53 Figure 4.7b: Chlorophyll content pattern 60 DAS under WS and WW conditions 54 Figure 4.7c: Chlorophyll content pattern 90 DAS under WS and WW conditions 54 Figure 4.8: Dendrogram showing the selected drought tolerant parents 56 Figure 5.1: Pod yield of F3 populations and their corresponding performance in F4 77 Figure 5.2: Maturity index of F3 populations and their corresponding values in F4 77 Figure 5.3: Chlorophyll content at 60 DAS under well watered and end-of-season drought conditions 87 Figure 5.4: Chlorophyll content at 80 DAS under well watered and end-of-season drought conditions 88 Figure 5.5: Mean performance of crosses for pod yield under WS and WW conditions 90 University of Ghana http://ugspace.ug.edu.gh xvii LIST OF PLATES Plate 3.2: Focus group discussion with men at Wadata 27 Plate 3.1: Focus group discussion with women at N’cha Roua 27 Plate 4.1: Stressed plants versus non stressed plants (a, b); Plants after 2 weeks drought effect and resume irrigation (c, d) 46 Plate 4.2: Farmers selecting genotypes based on their preference criteria 47 University of Ghana http://ugspace.ug.edu.gh xviii LIST OF ABBREVIATIONS ABA Abscisic Acid AGRA Alliance for a Green Revolution in Africa ANOVA Analysis of Variance ATP Adenosine triphosphate BC Backcross CV Coefficient of Variation DAS Days after sowing DF Degree of freedom DNA Deoxyribonucleic acid DTI Drought Tolerance Index ESTs Expressed Sequence Tags FAO Food and Agriculture Organization FGD Focus Group Discussion GCA General Combining Ability GCV Genotypic Coefficient of Variance h2ns: heritability narrow sense ICRISAT International Crops Research Institute for Semi-Arid Tropics INRAN Institut National de Recherche Agronomique du Niger Kg/ha Kilogram per hectare LSD Least significant difference NGO Non-Governmental Organization PC Principal Component University of Ghana http://ugspace.ug.edu.gh xix PCV Phenotypic Coefficient of Variance PIC Polymorphic Information Content PN net photosynthesis PRA Participatory Rural Appraisal QTL Quantitative Trait Loci RAPD Random Amplified Polymorphic DNA RFLP Restriction Fragment Length Polymorphism SAT Semi-Arid Tropic SCA Specific Combining Ability SCMR SPAD Chlorophyll Meter Reading SLA Specific Leaf Area SPAD Soil Plant Analytical Development SSA sub-Saharan Africa SSR Simple sequence repeat TE Transpiration Efficiency T ha-1 Tons per hectare USDA United States Department of Agriculture WACCI West Africa Centre for Crop Improvement WS Water stress WUE Water-use efficiency WW Well water University of Ghana http://ugspace.ug.edu.gh 1 CHAPTER ONE 1.0 GENERAL INTRODUCTION Groundnut (Arachis hypogaea L., Fabaceae), is the 13th most important food crop and 4th most important oilseed crop of the world (Jharma et al., 2001; Nigam et al., 2005; Mace et al., 2006; Cuc et al., 2008). The crop is grown in all tropical and sub-tropical regions of the world as a result of its adaptability to a wide range of soil and climatic conditions. It is grown from latitude 40oN to 40oS (Frimpong, 2004; Arunyanark et al., 2008, Furlan et al., 2012). Groundnut is grown in 108 countries on nearly 26.4 million ha worldwide with the total production of 36.45 million tons and an average yield of 1520 kg/ha in 2009 (FAOSTAT, 2011). The global output comes from seasonally rainfed areas of tropical, subtropical and warm regions of the world (Upadhyaya, 2005, Upadhyaya et al., 2011). Developing countries account for 96.9% of the world groundnut area with 93.8% of total production (Nageswara and Nigam, 2003). Asia accounts for two-thirds of the world groundnut production while Africa produces only one-fifth (Mathur et al., 2004). According to USDA (2010), the world average yield of groundnut is about 1.3 t ha-1, to 3.36 t ha-1 in China, 0.98 t ha-1, to 3.36 t ha-1 in USA, to 3.84 t ha-1 in Nigeria and only 0.42 t ha-1 in Niger. Groundnuts are very important legumes and are nutritious and complement the major staples cereal in Africa, where cases of malnutrition are rampant. Groundnut seeds are a rich source of high quality edible oil (45 -56%), easily digestible protein of 12-36% and 10-20 % carbohydrates (Stalker, 1997; Nageswara and Nigam, 2003; Reddy et al., 2003a; Reddy et al., 2003b). Groundnut grains are a nutritional source of vitamin E, niacin, folacin, calcium, phosphorus, magnesium, zinc, iron, riboflavin, thiamine and potassium (Nwokolo, 1996). Groundnut haulms University of Ghana http://ugspace.ug.edu.gh 2 provide excellent fodder for animal feeding and add an extra income to smallholder farmers (Arslan, 2005). Groundnut also provides nutrients to the soil and its cultivation improves soil fertility by nitrogen fixation that makes it an important component for crop rotation. It is grown as a sole crop, intercrop, or mixed crop (Nageswara et al., 1990; Ghosh et al., 2005). These multiple uses of groundnut make it an excellent cash crop for domestic markets as well as for foreign trade in several developing and developed countries. Over two thirds of the global groundnut production occurs in rainfed regions that experience erratic and insufficient rainfall resulting in unpredictable droughts (Arunyanark et al., 2009; Songsri et al., 2009). Above 80% of the smallholder farmers are located in areas where rainfall is low and erratic, and soils tend to be infertile (Muzari et al., 2012). The natural capacity of groundnut to moderately tolerate drought renders it suitable to be grown largely under rain- dependent conditions, especially by the resource-poor farmers. Groundnut is the second most widely grown legume in Niger after cowpea. It is cultivated mainly during the rainy season from June to September. In Niger, groundnut farming experienced a remarkable development between 1950 and 1970, with a period of increased production in the 1960s due to a rapid increase in land areas and yield. This was followed in the seventies and eighties by a real slump in production and exports due to drought spells and lack of improved varieties to such a point that groundnut became a crop mostly processed traditionally into vegetable oil and oil-cakes for local consumption (Toukoua, 1986). Groundnut yields have dropped from 0.894 t ha-1 in the 1960s to 0.42 t ha-1 in 2008 (FAOSTAT, 2008). Since the 1970s, Niger has suffered from severe droughts which have triggered important University of Ghana http://ugspace.ug.edu.gh 3 food crises in particular in 1973, 1984 and 1991, and more recently in 2005, 2009 and 2010 (Leblois et al., 2011). Since groundnut’s water requirements are related to the variety, short- cycle varieties (90 days) are cultivated in the north of Niger (<500 mm), medium-cycle varieties (105 days) in the center (rainfall between 500 and 600 mm) and long-cycle varieties (120 days) in the southern part where annual rainfall exceeds 600 mm (Oumarou et al., 1990). The southwards shifting of isohyets according to Ben Mohamed et al., (2002) led to a reduced groundnut farming zone due to the unadaptability of the varieties to water deficits. Groundnut production in Niger fluctuates greatly. Yields per hectare are typically low, because of a combination of production factors such as mostly: unreliable and/or poorly distributed rainfall, lack of small scale irrigation facilities, traditional farming with little mechanization, and cultivation on marginal land. In addition there was also insufficient selection of suitable crop varieties, outbreaks of pests and diseases, inefficiencies in distribution of seeds and fertilizers coupled with, use of low-yielding seed varieties. Poor research-extension linkages, infertile soils, failure of the smallholder farmer to adapt to changing environments and adopt new technologies and political instability contributed significantly to reduce groundnut production in Niger (Dulvenbooden et al., 2002). Groundnut production is severely constrained by both biotic and abiotic stress factors. The most important constraint among the abiotic stress is drought. Drought patterns in groundnut have been classified into four groups: early season drought (immediately after germination to flowering), mid-season drought (flowering to pod formation), end-of-season drought (pod filling to pod maturity) and intermittent drought (Rachid et al., 2003; Hamidou et al., 2012). University of Ghana http://ugspace.ug.edu.gh 4 The most prevalent drought pattern in Niger is the end-of-season. One of the key challenges of the smallholder farmers in Niger is how to harness limited available resources to mitigate harsh environment conditions for the improvement of groundnut production to enhance incomes and reduce malnutrition and poverty. Consequently varieties that can escape end-of-season drought should be developed. In the context of climatic variation with rainfall reduction in terms of duration and amount, recurrent drought affects groundnut performance in Niger. Therefore, priorities in groundnut improvement in Niger should include tolerance to drought and earliness as drought escape mechanism. There is a lack of information on genetic parameters required for breeding groundnut for earliness and drought tolerance related traits in Niger. Because of the low groundnut productivity, smallholder farmers have reduced the amount of land allocated to groundnut and resulting in increased poverty. This has led Niger becoming a major importer of vegetable oil (Dulvenbooden et al., 2002). In seeking to solve the problem of drought stress in Niger for groundnut production, there is the need for breeders to involve farmers. One of the crucial aspects to consider in plant breeding is the way that breeders involve farmers in the improvement of existing cultivars and selection of new varieties. When farmers are not involved in breeding activities, breeder’s products may not be easily adopted. This is because the varieties developed may not meet farmers' and consumers' preferences. This research aimed to exploit existing germplasm pool and farmer knowledge and expectations to develop groundnut lines with early maturity and improved drought tolerance. University of Ghana http://ugspace.ug.edu.gh 5 The specific objectives of the study were to: i. Identify the production constraints and farmers preferred varieties of groundnut in Niger; ii. Identify and select early maturing and drought tolerant groundnut genotypes; iii. Determine the combining ability for earliness and drought tolerance traits for hybrid development; and iv. Estimate the heritability of earliness and drought tolerance traits to assess potential for selection of these traits. University of Ghana http://ugspace.ug.edu.gh 6 CHAPTER TWO 2.0. LITERATURE REVIEW 2.1 Effect of drought on plant performance Drought is the most damaging of all the abiotic stresses affecting groundnut production in Niger. The composite nature of drought stress limits its management through conventional breeding methods (Gosal et al., 2009). Drought stress is a complex syndrome, involving several climatic, edaphic and agronomic factors that severely limit groundnut productivity depending on: timing of its occurrence, duration and intensity (Nigam et al., 2005). The general complexity of drought is often aggravated under the Semi-Arid Tropic (SAT) conditions, by unpredictable rainfall and by the occurrence of high temperatures, high levels of solar radiation and poor soil characteristics (Nageswara and Nigam, 2003). Drought can have an effect from seedling establishment to seed filling and thereby severely reduce seed yield and biomass production (Abdalla et al., 2008). Both yield and quality of groundnuts are compromised by drought. Nigam et al., (2003) reported that drought affects yield and groundnut quality under rainfed conditions. Drought, occurring during the late reproductive phases; from pegging to pod filling, can drastically reduce pod yield (Plaut and Ben, 2005; Clavel et al., 2006; Painawadee et al., 2009; Wuma et al., 2009). Yield reduction is estimated at 56% to 85% (Nageswara et al., 1989). Nageswara and Nigam (2003) reported that annual groundnut production losses due to drought were estimated at over 250 million USD (at the market price of 1994). Global yield losses due to drought have been estimated as 6.7 million tons in groundnut (Subbarao et al., 1995). University of Ghana http://ugspace.ug.edu.gh 7 A survey of the literature suggests that sustainable strategies that enable farmers to manage drought at the most critical growth stages are required to increase yield of groundnuts and to enhance quality. Drought stress varies spatially and temporally at several different levels. Drought affects membrane lipids and photosynthetic responses and yield in groundnut (Lauriano et al., 2000). Water deficit affects thylakoid electron transport, phosphorylation, carboxylation and photosynthesis. Changes in the lipid content and composition are common in water-stressed plants and this increases membrane permeability. This causes damage and membrane disruption as well as reduction in photosynthesis (Lauriano et al., 2000). 2.2. Effect of drought during flowering and pod formation 2.2.1. Flowering The start of flowering is not delayed by drought stress (Boote and Ketring, 1990). The rate of flower production is reduced by drought stress during flowering but the total number of flowers per plant is not affected due to an increase in the duration of flowering (Meisner and Karnok, 1992). A significant burst in flowering on alleviation of stress occurs, particularly when drought is imposed just prior to reproductive development (Janamatti et al., 1986). When stress is imposed during 30-45 days after sowing, the first flush of flowers, produced up to 45 days, do not form pegs during that time. However, flowers produced after re-watering compensate for this loss (Gowda and Hegde, 1986). 2.2.2. Pod formation Drought stress during pegging and pod development may lead to a drastic yield reduction and the magnitude of reduction would depend on the cultivars (Jogloy et al., 1996). Peg elongation, University of Ghana http://ugspace.ug.edu.gh 8 which is turgor dependent, is delayed due to drought stress (Boote and Ketring, 1990). Pegs failed to penetrate effectively into air-dry soil, especially in crusted soils. Often, within 4 days of withholding water, the soil surface becomes too dry for peg penetration. Rucker et al. (1995) mentioned that the quality of groundnut products decreases under drought stress. Adequate pod zone moisture is critical for development of pegs into pods and adequate soil water in the root zone cannot compensate for lack of pod zone water for the first 30 days of peg development (Gowda and Hegde, 1986). Dry pegging zone soil delays pod and seed development. Soil water deficits in the pegging and root zone decrease pod and seed growth rates by approximately 30% and decrease weight per seed from 563 to 428 mg (Janamatti et al., 1986). Sexton et al. (1997) reported that peg initiation and growth during drought stress demonstrated ability to suspend development during the period of soil water deficit and to re-initiate pod development after the drought stress was relieved. Nageswara et al. (1989) reported that under water stress, pegging and seed set responses of different groundnut cultivars varied substantially. This leads to a large reduction in pod yield, and the reduction percentage also varies among groundnut cultivars. 2.3. Environmental resource utilization 2.3.1. Groundnut water use Boote (1982) and Ghosh et al. (2005) described and defined the growth stages of groundnut as a series of vegetative and reproductive stages. The total water use by a groundnut crop is controlled by climatic conditions in addition to agronomic and varietal factors that include growth stages. Sivakumar and Sharma (1986) reported that water use of groundnut varies from 250 mm in the rainfed conditions to 830 mm under irrigated conditions (with irrigation at weekly intervals). Thus, groundnut needs water throughout its entire life cycle, but the reproductive stages are the most critical in terms of water use. University of Ghana http://ugspace.ug.edu.gh 9 2.3. 2. Rainfall or soil moisture Rainfall is the most significant climatic factor affecting groundnut production, as 70% of the crop area is under Semi-Arid Tropical conditions characterized by low and erratic rainfall (Ong, 1986). Low rainfall and prolonged drought spells during the crop growth period have been reported to be the main reasons for low average yields in most of the regions of Asia and Africa; for example in India (Reddy et al., 2003a), and several parts of Africa, (Camberlin and Diop, 1999). Camberlin and Diop (1999) reported that almost half of the variance in groundnut production is explained by rainfall variability, especially during the early part of the rainy season (July - August). Dulvenboooden et al. (2002) reported that groundnut production in Niger is significantly affected by rainfall during July to September. Therefore, to increase groundnut production, water availability must be considered as a critical factor in any production system. 2.3.3. Photoperiod or day length It is well-known that long days promote vegetative growth at the expense of reproductive growth and increased crop growth rate, decreased partitioning of photosynthesis to pods and decreased duration of effective pod filling phase (Ketring 1979, Nigam et al., 1998). Groundnut cultivars may show differences in response to photoperiod. Bagnall and King (1991) mentioned that flower, peg and pod numbers were consistently enhanced by short day treatments for a range of groundnut varieties. Flower and peg number at 60-70 days after sowing were almost doubled by 12 hour days exposure compared with plants in 16 hour days. Pod number and therefore yield was more influenced by photoperiod than was flower or peg formation. Bell et al. (1991) observed reduction on number of pegs, pods and total University of Ghana http://ugspace.ug.edu.gh 10 pod weight per plant in long (16 or 17 h) photoperiods, but no effect of photoperiod was evident on time to first flower in a cool subtropical environment. Nigam et al. (1994) studied the effect of temperature and photoperiod and their interaction on plant growth as well as partitioning of dry matter to pods. They observed that photoperiod did not significantly affect partitioning of dry matter to pods under low temperature regime (18/22ºC) but at higher temperatures (26/30ºC) partitioning to pods was significantly greater under short days (9 h) and this study provided evidence of genotypic variability for photoperiod and temperature interactions. 2.4. Consequences of water stress Reddy et al. (2003b) reported that canopy photosyntesis is reduced by water stress due to reduced stomatal conductance, and reductions in leaf area that lead to transpiration reduction. Periodic water stress leads to anatomic changes such a decrease in size of cells and intercellular spaces, thicker cell walls and greater development of epidermal tissue (Reddy et al., 2003b). Drought stress has been shown to cause alterations in the chemical composition and physical properties of the cell wall (Ingram and Bartels, 1996). The cellular water deficits results in the concentration of solutes, loss of turgor, change in cell volume, disruption of water potential gradients, change in membrane integrity, denaturation of proteins and several physiological and molecular components (Griffiths and Parry, 2002). The stress effects depend on: the timing, the intensity, the duration, the developmental stage of the plant, the genotype and the environmental interactions (Sharma, 2006). Rates of net photosynthesis (Pn) decrease when plants are subjected to moisture stress (Nautiyal et al., 2002; Vorasoot et al., 2004). Decreases in net photosynthesis due to moisture stress are accompanied by increased resistance to CO2 diffusion into the leaf (Bhagsari et al., 1976). Adenosine triphosphate (ATP) synthesis was the most University of Ghana http://ugspace.ug.edu.gh 11 affected process susceptible to drought stress. Collino et al. (2001) observed that the fraction of photosynthetically active radiation intercepted and harvest index were reduced under water stress. Abscisic acid (ABA) concentration increases under water stress as well as under some other abiotic and biotic stresses (Christmann et al., 2005). Lauriano et al. (2000) reported that drought affects membrane lipids and photosynthetic responses. Water deficit affects thylakoid electron transport, phosphorylation, carboxylation, and decreases photosynthesis. Clavel et al. (2004) reported that water deficit decreased leaf area index, relative water content and transpiration at about 3 weeks after the occurrence of water deficit at the soil level. Vorasoot et al. (2003) observed a drastic reduction in yield, as well as yield components like total dry weight and shelling percentage. Naveen et al. (1992) showed that water deficit reduced seed size, and increase the number of shriveled seed, if water stress occurs during the flowering stage. Water stress during reproductive development resulted in embryo abortion, reduced seed development and aggravated aflatoxin contamination, and finally, reduced yield (Stalker, 1997; Holbrook and Stalker, 2003; Vorasoot et al., 2003). In addition, drought conditions influence the growth of weeds, agronomic management, and the nature and intensity of insects, pests and diseases (Reddy et al., 2003b). Nageswara and Nigam (2003) reported that under drought conditions, the beneficial effects of improved crop management practices in terms of increased production are not fully realized. Calcium uptake by pods and nitrogen fixation processes are adversely affected by drought. Consequently water stress has several negative effects on the plant at different levels such as anatomic, physiological, molecular and agronomical. 2.5. Mechanisms of drought stress avoidance In groundnut, genetic differences in water use efficiency appear to be due to photosynthetic capacity. Nevertheless, variability in water use efficiency can be due to reduced stomatal University of Ghana http://ugspace.ug.edu.gh 12 diffusion or intrinsic photosynthetic capacity (Wright et al., 1994). A waxy layer on the top of the leaf cuticle is present to protect against water stress and help survival in harsh conditions (Tanka et al., 2004). The cuticle of flowering plants is a complex matrix that consists of lipids, cutins and waxes (Kolattukudy, 1980). Cutins are composed of hydroxy and epoxy fatty acids and serve as the barrier between the aerial parts of higher plants and their environment. Serraj and Sinclair, (2002) proposed osmotic adjustment as a mechanism for drought tolerance. The accumulation of osmolytes in the cells will result in a decrease in osmotic potential and in the maintenance of water absorption and turgor pressure. During water stress, abscisic acid is generated in the root and transported to the shoot to mediate stomatal closure as well as more general adaptive shoot responses (Christmann et al., 2005). Root structure, accumulation of osmolytes, leaf folding, and reduction in leaf area and regulation of transpiration are known mechanisms that increase drought tolerance (Samdur et al., 2003). 2.6. Approaches for combating drought stress According to Nageswara and Nigam (2003), three genetic enhancement approaches were developed and implemented simultaneously to enhance the adaptation to drought-prone environments. These are: (i) development of short-duration genotypes that can escape the end- season drought; (ii) development of genotypes with superior yield performance in drought prone regions following conventional breeding approaches and (iii) development of drought tolerant genotypes following physiological breeding approaches. In addition to these approaches, Sharma (2006) reported that molecular markers and Marker Assisted Breeding and Transgenic Technology have been used. University of Ghana http://ugspace.ug.edu.gh 13 2.6.1. Physiological breeding approach There has been significant progress in understanding the physiological basis of genotypic variability of drought response in groundnut, suggesting that selecting genotypes with traits contributing to superior performance under water-limited conditions is possible (Nageswara and Nigam, 2003). Substantial genetic variation has been observed in partitioning of dry matter to pods (Nageswara et al., 1993). Nageswara (1994) reported a significant genotypic variation in root system for the capacity to penetrate deeper soil layers. Significant genotypic variation in total amount of water transpired and transpiration efficiency has been shown under field conditions (Mathews et al., 1998). Physiological measurements such as water transpired, water-use efficiency (WUE) and harvest index (HI) have been identified and can be highly correlated with pod yield (Basu et al., 2003). Transpiration efficiency (TE) varied significantly among groundnut cultivars irrespective of whether plants were drought-stressed or well-watered (Hubick et al., 1988). Wright et al. (1994) confirmed that there are large cultivar differences in transpiration efficiency in groundnut grown in field conditions. Although a large variation has been found for each of these physiological traits in groundnut, there are significant difficulties in accurately measuring them because large number of plants are needed for selection programmes. However, there has been significant progress in understanding the mechanisms involved, and in developing novel and indirect selection tools for the model parameters (Nageswara and Nigam, 2003). For example, transpiration efficiency and carbon isotope discrimination in leaf are well correlated in groundnut (Hubick et al., 1986), suggesting the possibility of using carbon isotope discrimination in leaves as a rapid, nondestructive tool for selection of transpiration efficiency. Carbon isotope discrimination and specific leaf area (SLA) have been shown to be correlated with water-use University of Ghana http://ugspace.ug.edu.gh 14 efficiency of groundnut. SLA has been proposed as an indirect selection criterion for carbon isotope discrimination and water-use efficiency (Craufurd et al., 1999). Nageswara and Wright (1994) showed that the stability of relationship between carbon isotope discrimination in leaf and SLA over a wide range of cultivars and environments has raised the possibility of using SLA as an even more rapid and inexpensive technique for selection of transpiration efficiency. Madhava et al. (2003) suggested that the soil plant analytical development (SPAD) chlorophyll meter reading (SCMR) can be used as a rapid preliminary screening tool to select groundnut genotypes with high transpiration efficiency. Thus, through physiological approaches there are several techniques for combating water stress via indirect selection by targeting specific traits such as SLA, TE, HI, SCMR. 2.6.2. DNA marker-assisted selection The use of DNA markers for indirect selection offers the greatest potential gain for quantitative traits with low heritability, as these are the most difficult characters to work with in the field through phenotypic selection (Sharma et al., 2002). A large number of genes with a potential role in water stress tolerance have been identified and characterized (Ingram and Bartels, 1996). Genetic improvement for drought tolerance is crucial in environments where agriculture depends on limited water resources. Therefore, finding genes involved in the tolerance to drought and their insertion into the genetic background of agronomically preferred varieties could enhance and/or stabilize yields under drought-prone conditions. Once genomic regions contributing to the trait of interest have been assigned and the alleles at each locus designated, they can be inserted into locally adapted high-yielding cultivars by making requisite crosses (Mathur et al., 2004). The progenies that carry desired combinations of alleles will be selected for further evaluation using marker-assisted selection. University of Ghana http://ugspace.ug.edu.gh 15 Halward et al. (1993) developed the first genetic RFLP linkage map in groundnut from an F2 population of a cross between A-genome diploids Arachis stenosperma and Arachis cardenasii for use in genetic studies and breeding programmes to improve the cultivated species (Arachis hypogaea L.). A second map based on the previous parents used a BC1 population and added 167 RAPD markers to a skeleton of 39 RFLP markers from which the first map was constructed (Garcia et al., 2005). A third A-genome map was developed from the cross A. duranensis x A. stenosperma consisting of 170 SSR markers on 11 linkage groups, with a total map length of 1,231 cM ( Gomez et al., 2009). A total of 370 RFLP loci were mapped onto 23 linkage groups, for a map distance of 2210 cM (Burow et al., 2001). Proite et al. (2007) analyzed the transcriptome of wild species, Arachis stenosperma accession V10309. Almost nine thousand (8,785) ESTs were generated from four cDNA libraries. ESTs were classified into 23 different categories according to putative protein functions. Numerous sequences related to disease resistance and drought tolerance were identified. Cuc et al. (2008) have developed a set of 104 SSR markers for cultivated groundnut. The polymorphic SSR markers detected 2 to 5 alleles with an average of 2.44 per locus. These markers showed a good level of polymorphic information content (PIC) value in cultivated germplasm and, therefore, would be very useful for germplasm analysis, linkage mapping, diversity studies and phylogenetic relationships in cultivated groundnut as well as related Arachis species. Recently Varshney et al. (2009) constructed the first SSR-based genetic linkage map for cultivated groundnut (Arachis hypogaea L.). A total of 1,145 microsatellite (SSR) markers available in the public domain as well as unpublished markers from several sources were screened on two genotypes, TAG 24 and ICGV 86031, that are parents of a recombinant inbred line (RIL) mapping population. As a result, 144 (12.6%) polymorphic markers were identified and these University of Ghana http://ugspace.ug.edu.gh 16 amplified a total of 150 loci. As the mapping population used for developing the genetic map segregates for drought tolerance traits, phenotyping data obtained for transpiration, transpiration efficiency, specific leaf area and SPAD chlorophyll meter reading (SCMR) for two years were analyzed together with genotyping data. Two to five QTLs for each trait mentioned above were identified. Because of the complexity of drought, some molecular tools such as molecular markers are used to select genotypes with desirable genes playing an important role in water stress tolerance. Molecular tools can be used later on to introgress drought tolerance and earliness genes into farmers' preferred varieties. 2.7. Farmer participation in variety development in Niger Participatory rural appraisal (PRA) has been defined as a: ‘family of participatory approaches and methods which emphasize local knowledge and enable local people to do their own appraisal, analysis and planning (Brown et al., 2002). PRA uses group animation and exercises to facilitate information sharing, analysis and action among stakeholders. There has been lack of information about farmers' groundnut varieties preferences and production constraints in Niger. In addition, farmers had never involved in participatory groundnut varieties development in Niger. Farmers 'perceptions on sorghum production constraints and their traits preferences were identified through PRA in Niger (Aissata, 2013). Ahmadou (2013) used PRA tools such as focus group discussions and questionnaire to determine constraints and opportunities of pearl millet production and to get general information on the production systems from farmers in Niger. Participatory rural appraisal methods were used to establish farmer management strategies and perceptions on major constraints to pigeonpea productivity in Zimbabwe (Mapfumo et al., 2001). Participatory rural appraisals were conducted to identify the major characteristics of sorghum University of Ghana http://ugspace.ug.edu.gh 17 landraces. This was done in village meetings by Focused Group Discussions (FGDs), matrix ranking, and individual interviews in Malawi. The classical PRA tools (maps, sketches, diagrams, ranking) were used to investigate the indigenous soils and agroecological knowledge and identify determinants of land degradation and to identify specific land management queries in Lebanon (Zurayk et al., 2001). 2.8. Estimation of genetic parameters in groundnut In Niger, there is a lack of information on genetic parameters required for breeding groundnut for earliness and drought tolerance related traits. Therefore, combining ability and heritability estimates studies are needed. 2.8.1. Combining ability estimates Earliness is important for groundnut adaptation to a wide range of cropping systems in Semi- Arid Tropics. An early maturing variety may escape damage from drought. In any breeding programme knowledge of the genetic composition of the breeding stocks in hand and the nature of gene action involved in the expression of the trait to be improved is the basic requirement of the breeder. The combining abilities for physiological and earliness traits in groundnut are examined to understand the type of gene action governing these traits, and to identify groundnut genotypes suitable for use as parents in breeding for improvement in physiological and earliness traits. John, (2011) reported that the information on the precise nature of genetic control of physiological traits in groundnut is still lacking. General combining ability (GCA) of a line is defined as the mean performance in all its crosses, when expressed as deviation from the mean of all crosses. It is the average value of all crosses having this line as one parent, the value being expressed as a deviation from the overall mean of University of Ghana http://ugspace.ug.edu.gh 18 crosses. Any particular cross, then, has an expected value, which is sum of the general combining abilities of its two parental lines. The cross may deviate from this expected value to a greater or lesser extent. This deviation is called specific combining ability (SCA) of the two lines in combination (Griffing, 1956, Falconer and Mackey, 1996). This information provides guidelines for plant breeders to select parental lines to be used in breeding programmes and to use promising cross combinations for further selection. Jogloy et al. (1987) found that general combining ability was highly significant for pod yield, and shelling percentage and specific combining ability was significant for pod length and seed size. Swe and Branch (1986) found that estimates of general and specific combining abilities were significant for biomass, harvest index, total pod weight. In general, estimates of specific combining ability are more pronounced in the crosses of more diverse cultivars than in the closely related cultivars (Isleib et al., 1980). Mekonthehou (1987) studied the combining ability of early maturity for a selected group in Spanish and Virginia type of groundnut and found that GCA was always larger than SCA for most of the maturity parameters in F1 hybrid of Virginia x Spanish parents. Rachmeler (1988) suggested that selection for yield and earliness if practiced in later generation would be more effective. Therefore, he found greater and significant GCA than SCA estimated for early maturity in F2 and F3 generation. Bansal et al. (1991) showed that non additive gene effects were predominant for yield components although the magnitude of additive effects was considerable. They also reported that intra-group crosses were better than inter-group crosses. John et al. (2011) indicated that combining ability analysis was primarily non-additive gene action in the inheritance of SPAD (soil plant analytical development). Intermating of the F2 segregants followed by recurrent selection and pedigree breeding can harness the different kinds of gene - University of Ghana http://ugspace.ug.edu.gh 19 effects (John et al., 2011). They stated that repeated selection and inter-mating of segregating materials for two or three cycles, makes it possible to achieve simultaneous improvement in physiological attributes. Venkateswarlu et al. (2007) mentioned the involvement of both additive and non-additive gene action in the inheritance of physiological and pod yield traits. In general high magnitude of general combining ability variance showed the greater importance than additive gene action in the inheritance of physiological and pod yield traits. Through combining ability study information on the best general combiners' parental lines will be available. 2.8.2. Heritability estimate for agronomic and yield traits All broad and narrow‐sense heritabilities are defined on individual and family basis for prediction of response to selection for a target population of environments in space and time. The heritability of a trait within a population is the proportion of observable differences in a trait between individuals within a population that is due to genetic differences (Nyquist and Baker, 1991; Falconer and Mackay, 1996; Holland et al., 2003). Factors including genetics, environment and random chance can all contribute to the variation between individuals in their observable characteristics. Heritability is measured by estimating the relative contributions of genetic and non-genetic differences to the total phenotypic variation in a population. A consequence of the definition of heritability is that it depends on the population, because both the variations in additive and non-additive genetic factors, and the environmental variance, are population specific (Visscher et al., 2008). Genetic variance depends on segregation in a population of the alleles that influence the trait, the allele frequencies, and the mode of gene actions. All these variables can differ across populations. Similarly, environmental variance can vary across populations. Consequently, the heritability in one population does not, in theory, University of Ghana http://ugspace.ug.edu.gh 20 predict the heritability of the same trait in another population (Visscher et al., 2008). For example, heritabilities are higher for morphometric traits than for fitness traits; the former are often higher in more favorable environments. Heritabilities and the variances that contribute to them are parameters of a population. Heritability can be estimated from empirical data of the observed and expected resemblance between relatives. The expected resemblance between relatives depends on assumptions regarding its underlying environmental and genetic causes. Sometimes these assumptions are straightforward. Traditionally, heritability was estimated from simple and often balanced designs, such as simple functions of the regression of offspring on parental phenotypes, the correlation of full or half sibs, and the difference in the correlation of monozygotic and dizygotic identical pairs. Selection for drought tolerance using pod yield under drought conditions as the selection criterion has produced only slow progress due to large genotype x environment interactions (Wright et al., 1994; Nageswara and Wright, 1994; Sheshshayee et al., 2006). The integration of physiological traits or their surrogates in the selection scheme would be advantageous in selecting genotypes that are more efficient water utilizers or partitioners of photosynthates into economic yield (Nigam et al., 2005; Pimratch et al., 2009, Upadhyaya and Sharma, 2011). Improving water access and management are practically difficult since water is a rare resource. Therefore, breeding for drought tolerance is an important strategy in alleviating the problem and offers the best long-term solution. Selection of segregating populations under stress conditions has been a standard approach for developing cultivars with improved stress tolerance. More rapid progress may be achieved by using physiological traits (Nigam et al., 2005) such as harvest index or water use efficiency, and SPAD (Songsri et al., 2008). Both specific leaf area and University of Ghana http://ugspace.ug.edu.gh 21 SPAD have been used as surrogate traits for water use efficiency (Wright et al., 1994; Nigam et al., 2005; Sheshshayee et al., 2006).A strong and positive relationship between SPAD chlorophyll meter reading and WUE was found in groundnut (Sheshshayee et al., 2006). Upadhyaya, (2005) reported genetic variation for SCMR in groundnut. SPAD that is easy to measure and suitable as a selection criterion trait for drought tolerance because of high heritability, simplicity in gathering and positive correlation with pod yield and agronomic traits (Songsri et al., 2008; Girdthai et al., 2010). Upadhyaya and Sharma, (2011) found that additive effects were important for SPAD chlorophyll meter reading (SCMR) at 80 days after sowing (DAS) while predominance of dominance effects with duplicate epistasis was observed for SCMR at 60 DAS and specific leaf area (SLA) in groundnut. So they concluded that predominance of additive effect for SCMR at 80 DAS suggested effective selection could be practiced even in early generations, whereas SCMR at 60 DAS and SLA at both stages, it will be better to start selection in later generations. From these results, recording SCMR and SLA should be done between 60 and 80 DAS for screening germplasm for drought tolerance. University of Ghana http://ugspace.ug.edu.gh 22 CHAPTER THREE 3.0. IDENTIFICATION OF THE PRODUCTION CONSTRAINTS AND FARMERS’ PREFERRED VARIETIES OF GROUNDNUT IN NIGER 3.1. Introduction One of the crucial aspects to consider in plant breeding is to involve farmers in the improvement of existing cultivars and the selection of new varieties. There is limited information in Niger on groundnut production constraints and farmers' preferred groundnut varieties traits. This makes a study in this area is necessary. When farmers are not involved in breeding activities, breeder’s products may not be easily adopted; because they may not meet farmers and consumer expectations. Chambers (1994) described PRA as a growing family of approaches and methods to enable local (rural or urban) people to express, enhance, share and analyze their knowledge of life and conditions, to plan and to act. When PRA is well done, local people, and especially the poorer, enjoy the creative learning that comes from presenting their knowledge and their reality. Nkongolo et al. (2008) conducted a PRA to identify the major characteristics of sorghum landraces in Malawi by using focus group discussion (FGDs), matrix ranking, and individual interviews. Joshi and Witcombe, (1996) used Farmer participatory varietal selection to identify farmer-acceptable cultivars of rice and chickpea in India. Scientists’ selection criteria will not always be linked with that of farmers; therefore, integration of farmers’ opinion in breeding various crops as a way of enhancing variety adoption is becoming increasingly popular and effective in terms of cost and transferring the right varieties and technology to farmers (Efisue et al., 2008). The objectives of this study were to (i) determine production constraints in two groundnut production regions in Niger; (ii) identify the farmers preferred groundnut varieties. University of Ghana http://ugspace.ug.edu.gh 23 3.2. Methodology 3.2.1. Study area The PRA was carried out in two main regions of the Niger “groundnut basin”; Maradi and Zinder. Focus group discussions (FGDs) were conducted in six villages of Madarounfa (Maradi) and Magaria (Zinder) (Figure 3.1). The six villages were Angoual gamji, N’cha Roua, Wadata in Magaria (12°59’N and 8°56' E); and Angoual Roumdji, N’Yelwa and Yardaji in Madarounfa (13o18'27''N and 7o9'21'' E). The two districts are located in Sudano-Sahelian agro ecological zone with two main seasons: (i) a short rainy season from June to September, with an average annual rainfall estimated at 500-600 mm, (ii) a dry season spanning from October to May. This includes a cold period (December to March) and a hot period (April-May). The soil type is tropical ferruginous soils of clay loam in Madarounfa and mainly sandy and infertile in Magaria. University of Ghana http://ugspace.ug.edu.gh 24 Figure 3.1: Niger map with study areas N MADAROUNFA MAGARIA University of Ghana http://ugspace.ug.edu.gh 25 3.2.2. Study villages Selection of sites was done jointly by National Institute of Agricultural Research (INRAN) researcher and the agricultural extension agents. The six villages were selected by using the following criteria: - Having a Farmer’s organizations that would enable the study to build on existing trust of local farmers; - Groundnut was important in the local farming systems and farm household incomes; - They were representative of the regions in terms of agro-environmental and socio- economic conditions, - Closeness to INRAN substations. Once a village was selected, the investigator and the agricultural extension agent jointly identified one or two villagers (male and female) as potential local representatives for preparing the FGD session. 3.2.3. Introduction to villages The extension agent introduced the principal investigator to the chief of each village and to the presidents of farmers’ organizations. The objectives were clearly explained: (i) identify the production constraints; (ii) determine farmers preferred varieties of groundnuts. Several discussions with farmers were held to explain FGDs procedures. Later an introductory meeting with farmers was scheduled in each village to explain the purpose of the research that pave the way for future contacts with them. University of Ghana http://ugspace.ug.edu.gh 26 3.2.4. Sampling method This study was conducted in January 2011 in collaboration with the agricultural extension unit, and INRAN. Because of common heterogeneity among agricultural households, stratified sampling was used. In each village, the village chief assisted in selecting the participants from the famer’s organizations. Thirty farmers were selected per village based on their experience in groundnut production. At least 15 men and 15 women participated in FGDs sessions, thus totaling 187 farmers (98 male and 89 female) for the six villages (Table 3.1). Table 3.1: The number of farmers by gender that participated in Focus group discussion sessions Department/Villages Male Female Total Magaria Angoual Gamji 17 14 31 N’cha Roua 15 21 36 Wadata 15 14 29 Madarounfa Angoual Roumdji 12 18 30 N’Yelwa 19 13 32 Yardaji 14 15 29 Total 98 89 187 3.2.5. Data collection and analysis The PRA technique used in this study was focus group discussion. The FGD supervisors included a moderator (the investigator), agronomist (recorder) from INRAN and an agricultural extension agent of the district. Before the focus group discussions sessions, a visit was made to the six villages. The objective was to identify the farmers’ organisations and establish selection criteria of the participants to the FGDs sessions. FGDs sessions were set jointly and a reminder was sent one week before the date. Two meetings were held in the village between 8:30 and 10:30 a.m. in the morning and 2:00 to 4:00 p.m. in the afternoon. FGD began with the moderator welcoming participants and briefing them on the process (e.g., that there are no right or wrong answers, that it’s important to speak one at a time and that the session will be recorded). Farmers University of Ghana http://ugspace.ug.edu.gh 27 sat in a semi-circle (Plate 3.1 and 3.2) to facilitate identification of key points raised by each farmer and also ensure full participation by all the participants. Participants were informed about the one to take records of the entire session. The discussions were held in Hausa (local language). Refreshments were served to participants at break in each session. Following a check list (Appendix1), the moderator moderates the session and ensured that all topics were covered. Participants were encouraged to express their views and even disagree with one another about the topics. The order in which topics were covered was flexible but generally the sessions started with more general issues and slowly moved into more specific ones. Towards the end, a few probing questions were asked to reveal more in-depth information or to clarify earlier responses. The topics discussed included main groundnut production constraints (biotic and abiotic stresses), varietal preferences, and marketing issues. A chart was used to rank the varieties and constraints. Each record was labeled with village name and date of the session. Plate 3.1: Focus group discussion with women at N’cha Roua Plate 3.2: Focus group discussion with men at Wadata University of Ghana http://ugspace.ug.edu.gh 28 3.3. Results 3.3.1. Farmers preferred varieties in Magaria RRB was the most preferred variety in the two villages followed by 55- 437. The varieties RRB and 55-437 were used by the farmers in all the three locations in Magaria. The variety RRB was preferred by 52.08% of the respondents followed by 55-437 which was preferred by 42.70% of the respondents (Figure 3.2). Only 5.22% of the respondents used local varieties. The respondents preferred the variety RRB to the variety 55-437 because of its oil content, tolerance to rosette disease and haulm production. All (100%) of the FGDs respondents required new groundnut varieties that are early maturity, drought tolerant and highly productive in terms of yield, oil content and haulm. Most of the respondents used the variety 55-437 because it was the first improved variety that was introduced since 1974 and well known to be drought tolerant initially. Figure 3.2: Farmers preferred varieties ranked in Magaria 0 5 10 15 20 25 Angoual Gamji N’cha Roua Wadata RRB 55- 437 Locale University of Ghana http://ugspace.ug.edu.gh 29 3.3.2. Farmers preferred varieties in Madarounfa The varieties 55-437 and JL24 were the best preferred and most cultivated by farmers, in all the three villages in Madarounfa followed by the variety RRB. These varieties were used by the farmers in all the three locations; RRB was less used in Yardaji. The variety 55-437 was preferred by 41.75% of FGD respondents, followed by the variety JL24, 32.96% and the variety RRB, 23.07% (Figure 3.3). The respondents preferred the variety 55- 437 to the varieties JL24 and RRB because of its yield potential, drought tolerance and earliness (90 days). Only 2.2% of the FGDs respondents in all the three localities used local groundnut varieties. Most of the respondents mentioned the need of new groundnut varieties that are high yielding, early maturing and drought tolerant. Figure 3.3: Farmers preferred varieties ranked at Madarounfa 0 2 4 6 8 10 12 14 16 Angoual Roumdji N’Yelwa Yardaji 55- 437 JL24 RRB Locale University of Ghana http://ugspace.ug.edu.gh 30 3.3.3. Sources of Seed Farmers in all the six villages bought groundnut seeds from different sources (Table 3.2). The majority of the respondents (52.4 %) purchased their groundnut seeds from the village markets, and 25.66% from their neighbors. Only 21.91% of the farmers used improved groundnut seeds from the National Institute for Agricultural Research. Other sources were, NGOs/Projects and seed shops. Table 3.2: Groundnut seeds sources Seeds sources % of respondents Village Market 52.40 Agricultural services 8.05 Farmers 25.66 INRAN 2.67 NGOs/Project 6.95 Seed shops 4.27 3.3.4. Production constraints in Magaria Drought, low soil fertility, insect/disease, weeds and lack of equipments were the most frequently mentioned constraints to groundnut production in Magaria. Drought was the greatest production constraint in all the three villages, followed by low soil fertility. Sixty-two percent (62.70%) of the respondents, mentioned drought, as the greatest production constraint; while 21.80% ranked low soil fertility as the second greatest constraint (Figure 3.4). The year to year variation in rainfall with frequent drought spells was the justification of ranking drought as the main production constraint. Low soil fertility was caused by a lack of manure and mineral fertilizer utilization, because most of the respondents are poor farmers and lack resources to purchase fertilizers. Insect/disease, weeds and lack of equipment were ranked 3, 4 and 5th with 9.30%, 5.2 and 1%, respectively. University of Ghana http://ugspace.ug.edu.gh 31 Figure 3.4: Main production constraints ranked by gender in the 3 villages of Magaria 3.3.5. Production constraints in Madarounfa Five production constraints were listed by the farmers in all the three villages. Out of the five production constraints, drought was the major constraint followed by low soil fertility. Drought was ranked first by 73.62% of the respondents as the major production constraint followed by low soil fertility (21.97%) (Figure 3.5). The reasons for ranking these constraints are the same as described in Section 3.4.4. Weeds and lack of equipment were mentioned by 3.29% and 1.09% respectively. 0 5 10 15 20 25 Angoual Gamji N’cha Roua Wadata Drought Low soil fertility Insect/disease weed Equipment University of Ghana http://ugspace.ug.edu.gh 32 Figure 3.5: Main production constraints ranked by gender in the 3 villages of Madarounfa 3.4. Discussion The varieties 55-437 and RRB were preferred in Magaria while in Madarounfa, JL24 and 55-437 were preferred by the farmers. There is a difference in groundnut varietal preferences in the two locations. None of the respondents in the three villages in Magaria mentioned the variety JL24, whereas 32.96% of the respondents in all the villages of Madarounfa mentioned the variety JL24. Among the varieties identified in this study, 55-437 was introduced in 1974 while JL24 and RRB were introduced in the last two decades by the International Crops Research Institute for the Semi-Arid-Tropics (ICRISAT) and INRAN. Overall, the three varieties 55-437, RRB and JL24 are the most farmers preferred varieties across the six villages according to FGD respondents. These varieties were preferred to local varieties because of their oil content (around 50%), yield potential and haulm production. Although, these varieties are the farmers' preferred varieties; they are still low yielding and susceptible to drought and rosette disease. The main production constraints are similar in all the six villages. From the ranking, drought is the major challenging 0 5 10 15 20 25 Angoual Roumdji N’Yelwa Yardaji Drought Low soil fertility Insect/disease weed Equipment University of Ghana http://ugspace.ug.edu.gh 33 constraint to groundnut production, followed by low soil fertility, insect/disease, then weeds and finally lack of equipment. The majority of FGD respondents believed that there was a reduction in rainfall during the last 10 years in time and space. They also believed that drought spells are more in recent times than in the past. Chiteka (1985), cited drought as the single most important constraint to groundnut production in Zimbabwe. The most prevalent drought in the study areas is the end-of season drought. Ndunguru et al. (1995) found that late-season drought has the largest impact; while Nageswara et al. (1985) have found that end-of-season droughts (pod-filling stage) result in poor yields. Unpredictable and unreliable rainfall distributions, and the recent change in weather conditions, have shortened the growing season in western Africa, rendering the existing long- duration cultivars unsuitable (Ntare and Waliyar 1994). To counteract late season drought, short- duration cultivars (80-100 days to maturity) have been introduced and are showing promise in short-season and drought-prone environments. Most FGD participants mentioned low soil fertility as one of the main obstacles to agriculture development particularly groundnut production. Low soil fertility is due predominantly sandy soils type. The respondents lack resources to purchase fertilizer or manure in order to improve soil fertility, consequently even though improved groundnut seeds will be available, they need also capacity building on how to enhance soil fertility. Soil fertility constraints to crop production are recognized widely as a major obstacle to food security and agro-ecosystem sustainability in sub-Saharan West Africa (Buerkert, 2002). Soil fertility depletion on smallholder farms has been cited as the fundamental biophysical root cause responsible for the declining per capita food production in Africa (Enyong, et al., 1999). Ajayi (2007) reported that University of Ghana http://ugspace.ug.edu.gh 34 low soil fertility is one of the greatest biophysical constraints to agricultural production in sub- Saharan Africa. The respondents in this study also mentioned the lack of government support to poor farmers to control pest damage as a major constraint. Most of the respondents cited rosette disease as one of the most damaging diseases. It is transmitted by the vector Aphis crassivora. The farmers faced serious yield losses (25-100%) due to this disease during the last seven years. Initially the farmers ignored the vector of the disease, but currently they understand how the disease is spread from farm to farm and best methods to control the disease. ICRISAT estimates that groundnut rosette disease causes greater yield loss than any other virus disease affecting groundnut in the semiarid tropics. Duivenbooden et al. (2002) cited rosette as the most destructive disease in Niger in 1975. During the same period, an epidemic in northern Nigeria that destroyed approximately 0.7 million ha of groundnuts, with an estimated loss of $250 million (Naidu et al., 1999). Similarly, the epidemic that occurred in 1995 in eastern Zambia affected approximately 43,000 ha, causing an estimated loss of $4.89 million. In the following year, in the central region of neighboring Malawi, groundnut production was reduced 23% by groundnut rosette disease. Yield losses due to groundnut rosette disease depend on the growth stage of the plant when infection occurs. A 100% loss in pod yield may result if infection occurs before flowering. Yield loss is variable if infection occurs between flowering and the pod maturing stage, whereas subsequent infections cause negligible effects. FGDs respondents in all the villages cited weeds among the main groundnut production constraints. Competition from weeds reduces yield and farmers complained about the lack of government support to control weeds. The large distribution of poor sandy soil in all the villages contributes to quick development of weeds that compete with the crop. The lack of equipment University of Ghana http://ugspace.ug.edu.gh 35 has been also cited by Angoual Roumdji farmers as one of the production constraints that reduces production. Most of FDG respondents mentioned the following local methods to alleviate drought in the six villagers: prayers, sowing before the onset of the rainy season, utilization of early maturing genotypes, early sowing just after the first rainfall, land ploughing before sowing. Although, utilization of early maturing varieties as an escape mechanism of drought, particularly the end- of-season drought; these early maturing genotypes are not available to the farmers. 3.5. Conclusions Drought and low soil fertility were the most important constraints contributing to low groundnut production in the study areas. The most popular farmers preferred varieties were 55-347, RRB, and JL24. These will be used in the crossing programme to develop new early maturing genotypes that are drought tolerant. It was also clear that drought is the major production constraint in both regions. The information gathered is useful to groundnut breeders to develop good quality groundnut varieties that take into account farmers’ preferences. University of Ghana http://ugspace.ug.edu.gh 36 CHAPTER FOUR 4.0. PHENOTYPING GROUNDNUT GERMPLASM FOR EARLINESS AND DROUGHT TOLERANCE 4.1. Introduction Groundnut is grown widely under rainfed conditions in the semi-arid tropics, where drought, particularly during the pod and seed forming stages, is a major constraint to productivity that significantly reduces pod yield. The yield of groundnut in Niger is frequently severely limited by drought arising from unpredictable rainfall, high evaporative demands and production on low water holding capacity soils. The relatively shorter growing seasons in semi-arid regions of West Africa as compared to the southern savannah ecologies prevent groundnut from maturing properly. Therefore, early maturing groundnut cultivars with improved yield are required for several agro-ecologies of the semi-arid regions of West Africa. However, information on the genetic variability for earliness among the early maturing candidate parental lines is lacking. Breeding early maturing cultivars is an important objective in most groundnut improvement programmes. Bailey and Bear (1973) demonstrated that the early onset of flowering and early accumulation of a given number of flowers (10 to 30) are important components of early maturity in groundnut. Tolerance to drought in late maturity varieties would also be useful. Drought is known to affect chlorophyll content and inhibit the photosynthetic capacity. The ability to maintain chlorophyll density under water deficit conditions has been suggested as a drought tolerance mechanism in groundnut (Jongrungklang et al., 2008). The SPAD chlorophyll meter readings (SCMR) could be University of Ghana http://ugspace.ug.edu.gh 37 used for indirect selection of drought tolerance as a rapid and cost effective tool for assessment of relative chlorophyll status in groundnut leaves (Arunyanark et al., 2008; Wunna et al., 2009). This study was conducted to screen 150 genotypes and identify drought tolerant and early maturity varieties. 4.2. Experimental site The experiment was conducted at Tarna station in Maradi's region (13o 28’N latitude and 7 o 10’E longitude) in May 2010. The annual rainfall ranges from 230 to 630 mm, a typical Sahelian climate. The soils at Maradi are sandy soils with low water holding capacity, low inherent soil fertility and organic matter content. The experimental site soil has a pH of 6.5 and contains 90% sand, 8% clay, and 2% organic matter. 4.3. Weather conditions The weather was recorded at the Maradi Airport station that is located 4 km from Tarna during 2011. 4.3.1. Temperature The hottest day of 2011 was April 19, with a maximum of 43°C (Figure 4.1). The hottest month of 2011 was May with an average daily temperature of 40°C. The bar at the top of the graph is red where both the daily high and low temperatures are above average, blue where they are both below average and white otherwise (Figure 4.1). The coldest day of 2011 was December 17, with a low temperature of 10°C. The coolest month of 2011 was January with an average daily minimum temperature of 14°C. University of Ghana http://ugspace.ug.edu.gh 38 Source: http://weatherspark.com/history/28574/2011/Maradi-Niger Figure 4.1: Temperature variation during 2011 4.3.2. Relative Humidity Hot and humid days feel even hotter than hot and dry days because the high level of water content in humid air discourages the evaporation. The least humid month of 2011 was March with an average daily low humidity of 9%, and the most humid month was August with an average daily low humidity of 60% (Figure 4.2). Source: http://weatherspark.com/history/28574/2011/Maradi-Niger Figure 4.2: Humidity variation during 2011 University of Ghana http://ugspace.ug.edu.gh 39 4.3.3. Precipitation The average annual precipitation and total rainfall days was collected at Tarna station (Figure 4.3). Figure 4.3: Rainfall pattern from 1980 to 2009 at Tarna INRAN station 4.4. Selection of early maturing varieties 4.4.1. Genetic material The plant material in this study included 150 accessions of which 43 were the best performing accessions from a reference set of 300 accessions evaluated under water stress condition during 2008 and 2009 by ICRISAT (Sadore) in Niger, 42 were promising drought tolerant and early maturing accessions from the ICRISAT Niger gene bank, and 65 were genotypes from the National Agricultural Research Institutes in West Africa. 4.4.2. Experimental conditions The fertilizers DAP [Diammonium phosphate (NH4)2HPO4] 150kg/ha (Jogloy et al., 2011) and farm yard manure 2000 kg/ha (Hamidou et al., 2012) were incorporated; after that, the land was 0 100 200 300 400 500 600 700 Rainfall (mm) Rainfall days University of Ghana http://ugspace.ug.edu.gh 40 disk-ploughed two times, leveled and harrowed. The field was irrigated with once before sowing. To prevent seed borne diseases, seeds were treated before planting with a combination of fungicide/insecticide: Thiram 10% (C6H12N2S4) and Imidacloprid 25% (C9H10ClN5O2) 25g for 10kg of seeds. The sowing depth was 2.5 cm to 3 cm as recommended by Ghosh et al, (2005). Seeds were hand planted in three rows per plot in a ridge-furrow system, 3 m long, with spacing of 0.5 m between rows and of 0.20 m between plants (Arunyanark et al., 2010). The plot size was 1.5 m2 (0.5 m x 3 m). All the trials were surrounded by two border rows. The design of experiment was an unbalanced α-lattice design unreplicated (10x15) with 10 blocks and each block contained 15 entries. The trial was irrigated twice a week during the first month, after that, once a week up to the harvesting by surface irrigation. The insecticide dimethoate (C5H12NO3PS2) was sprayed at 1l/ha two times to control aphids (Aphis craccivora) attack. 4.4.3. Data collection The following data were recorded: - % Emergence (% EMERG): The percentage of plants emerged was recorded from 3 to 7 days after sowing. - 50% Plants flowering (50% PF): Plots were regularly monitored to record the date at which 50% of the plants by plot flowered. - Plant height (PH): Five plants selected at random were used to record plant height 75 days after sowing. - Maturity date (MD): Seventy five days after sowing, one plant was harvested from each plot to determine the percentage of developed pods. The plots were harvested when at University of Ghana http://ugspace.ug.edu.gh 41 least 75% of the developed pods were mature as determined by the blackening of the internal shell wall (Williams and Drexler, 1981). - Biomass (BIO) weight: Above ground biomass was calculated from 15 plants harvested from each plot after 3 weeks air drying. - Pod yield (PY): Pod yield was determined from pod harvested from 15 plants in the middle of the plot after air drying to constant weight for two weeks. - Harvest index (HI): HI was calculated by using the following formula: HI= Total pod weight (g) / Total Biomass weight (Girdthai et al., 2010) 4.4.4. Data analysis Principal component analysis was used to classify the genotypes with GenStat 12th Edition software. The principal components were computed from the n eigenvalues and their corresponding eigenvectors are calculated from the correlation matrix. Each eigenvector defines a principal component. 4.4.5. Results for earliness screening Principal components measure the importance of each variable in accounting for the total variability. From the principal component analysis, the first three eigenvalues which were greater than 1 (2.63, 1.63 and 1.31), were identified and accounted for 79.62% of the total variation among the genotypes (Table 4.1). The first PC accounted for 37.59 % whereas the second and the third PC axes accounted for 23.30% and 18.73%, respectively. The first PC (Table 4.1) was positively associated with characters such as maturity and pod yield. The second PC was positively associated with days to emergence and negatively associated with harvest index and 50% plant flowering. The third PC correlated positively with pod yield and biomass. University of Ghana http://ugspace.ug.edu.gh 42 Table 4.1: Eigenvectors from the three PC axes used to classified groundnut genotypes Component Variable PC1 PC2 PC3 1. % emergence 7 DAS 0.34 0.43 -0.2 2. Harvest index 0.26 -0.63 -0.02 3. Maturity 0.52 0.09 0.17 4. 50% Plant flowering -0.3 -0.46 0.37 5. Plant height -0.32 0.27 0.41 6. Pod yield 0.49 -0.14 0.45 7. Biomass 0.3 0.29 0.63 Eigenvalues 2.63 1.63 1.31 Percentage of variance explained 37.59% 23.30% 18.73% PC: Principal Component, DAS: Days after sowing The performance of the 150 genotypes is presented in Appendix 2. The genotypes evaluated were classified into three groups based on the maturity: - First group: 5 extra early maturing genotypes (75 to 80 days), - Second group: 29 early maturing genotypes (85 days), - Third group: 105 intermediate genotypes (90 ≤ x ≤ 110 days). Eleven remaining genotypes were not adapted to Tarna conditions due to poor seed quality. Most of these accessions were collected from Mali National Agricultural Institute. The top five genotypes selected as early maturing varieties based on maturity were Chico, ICGV02022, 796, ICG3585 and ICIAR19BT. The genotype ICIAR19BT had the highest pod yield 1540 g followed by ICGV02022 and ICG 3584 with 700 and 620 g, respectively. Chico had the lowest pod yield (430 g). All the selected genotypes reached 50% flowering in 24 days after sowing. Seven days after sowing, all the selected genotypes had over 70% of emergence. The harvest indices ranged from 0.26 to 0.51. The best 20 genotypes (Table 4.2) based on the maturity were used to draw a dendrogram using DARWIN5 software version 5.0.158 (Figure University of Ghana http://ugspace.ug.edu.gh 43 4.4). The dendrogram showed three clusters of genotypes depending on the maturity, pod yield, harvest index, biomass weight, percentage of emergence, plant height and 50% plant flowering (Figure 4.4). Out of the best 30%, the top 5 extra early maturing genotypes were selected. Table 4. 2: The top 20 genotypes based on their maturity and their performance for other traits Ranking Genotypes % EMERG 7DAS ND 50% PF DM (days) PY (g) BIO (g) HI PH (cm) 1 Chico 75 24 75 430 900 0.47 39.33 2 ICG 02022 83 24 80 700 1800 0.38 33.33 3 ICG 3584 96 24 80 620 2380 0.26 49.66 4 796 100 24 80 520 1700 0.3 47.33 5 ICIAR 19BT GH 100 24 75 1540 3010 0.51 43.66 6 T127 - 83 58 26 85 480 1340 0.35 51.33 7 ICG-3736 67 26 85 630 1250 0.5 48.66 8 ICG 1415 75 26 85 720 1760 0.4 48.66 9 ICG 15236 67 28 85 540 2570 0.21 50 10 T44 - 88 83 28 85 530 2100 0.34 44.66 11 T49-87 46 28 85 600 1440 0.41 43.33 12 ICG 3312 92 26 85 420 1090 0.38 48.66 13 ICGV-IS 01827 92 26 85 240 830 0.28 47.33 14 T131 - 83 79 26 85 480 985 0.48 49.33 15 ICG 1519 79 24 85 100 520 0.19 47 16 ICG 9315 75 24 85 530 1830 0.28 49.33 17 TX 903652 62.5 26 85 490 1660 0.29 45.33 18 J 11 Niger 100 24 85 580 1930 0.3 45.33 19 ICG 15380 87.5 22 85 550 2320 0.23 42.66 20 ICG 9315 75 24 85 470 2520 0.18 55.33 % EMERG: Percentage of emergence, 50% PF: 50% plant flowering, DM: Date to maturity, PY: Pod yield, BIO: Biomass, HI: Harvest index, PH: Plant height University of Ghana http://ugspace.ug.edu.gh 44 Figure 4.4: Dendrogram of selected early maturing genotypes Cluster 3 Cluster 2 Cluster 1 University of Ghana http://ugspace.ug.edu.gh 45 4.5. Selection of drought tolerant varieties 4.5.1. Genetic material The top 100 intermediate maturing genotypes were selected from the activity, conducted previously in Section 4.4.5., based on their pod yield. 4.5.2. Experimental conditions Field preparation was done as described in section 4.4.2. The planting day was 21st February 2011. The crop was grown in two-row plots on ridges, 2 m long, with spacing of 0.5 m between rows and of 0.20 m between plants (Arunyanark et al., 2010). Seeds were hand planted in two environments (well watered and water stressed conditions) replicated two times in a balanced α- lattice design (10 x 10). The plot size was 1 m2 (0.5 m x 2 m), each block contains 10 genotypes. The distance between water regimes (well water and water stress) was 5 m and 5 m between replications. The trial was surrounded by two border rows. Dimethoate (C5H12NO3PS2) was applied at 1/ha five times to control aphids and caterpillar attack. Surface irrigation was used to apply water during the experiment. 4.5.3. Management of irrigation for treatment application The experiment was conducted during April, 2011 that is the most critical month with high temperature during the day (average of 40oC) described in Section 4.3.1. After sowing, the crop was irrigated twice a week up to when 50% plants flowed (30 DAS). After that, the plants were maintained fully irrigated until pod filling time by irrigating up to saturation weekly. The plants were exposed gradually to end-of-season drought from the pod filling (50DAS) until maturity. At 50 DAS, which corresponded with peg penetration and pod filling, drought stress was imposed for 14 days and irrigation was resumed at the 15th day to bring the soil up to saturation (Plate University of Ghana http://ugspace.ug.edu.gh 46 4.1a, b, c, d). Then, drought stress was imposed for 10 days, followed by irrigation up to saturation. After that, drought stress was imposed for 7 days followed by irrigation up to harvest (Figure 4.5). The well-watered plots were irrigated fully weekly until harvest stage. Figure 4.5: Drought stress imposition and irrigation frequencies Plate 4.1: Stressed plants versus non stressed plants (a, b); Plants after 2 weeks drought effect and resume irrigation (c, d) University of Ghana http://ugspace.ug.edu.gh 47 4.5.4. Field day During harvesting of drought screening experiment, a field day was organized to allow farmers to select drought tolerant genotypes based on their own criteria. A team was composed of six farmers; each was a member of a farmer organization in the 6 villages where the PRA was conducted. The farmers were divided in two groups of three each. Each group selected the best genotypes by using their own criteria in both well watered and water stressed plots (Plate 4.2). Plate 4.2: Farmers selecting genotypes based on their preference criteria 4.5.5. Data collection The following data were collected: - The SPAD chlorophyll meter reading [SCMR] Minolta (SPAD-502 meter, Tokyo, Japan) measures the greenness or relative chlorophyll content of leaves. SPAD is defined as soil plant analysis development. However, the leaf transmittance depends not only on the content of University of Ghana http://ugspace.ug.edu.gh 48 chlorophylls but also on the distribution in leaves. The chlorophyll content was recorded at 45 DAS (before water stress was imposed) and at 60 and 90 DAS to monitor the chlorophyll content. Five plants from each plot were sampled at random, and the second fully expanded leaf from the top of the main stem was used for SCMRs assessment during the morning period (0900±1200h) (Nageswara et al., 2001). The chlorophyll content was recorded on each of the four leaflets of the tetrafoliate leaf. An average SCMR for each plot was derived from 20 single observations (four leaflets x 5 plants per plot) (Arunyanark et al., 2008). While recording the SCMR, care was taken to ensure that the SPAD meter sensor fully covered the leaf lamina in order to avoid interference from veins and midribs (Nageswara et al., 2001). - Pod Yield (PY): Pod yield was determined from 10 plants selected at random from all the treatments. - Drought tolerance index (DTI): DTI was calculated for each trait as the ratio of the trait under water stress (WS) treatment to that under well-watered (WW) condition as suggested by Nautiyal et al. (2002). DTI for SCMR is as follows: DTI (SCMR) = SCMR under WS/ SCMR under WW 4.5.6. Statistical Analysis Combined analysis of variance was computed for the 100 entries across water regimes (Gomez and Gomez 1984) for pod yield and SCMR (45, 60 and 90 DAS) data using GENSTAT software version 12.0. In the combined analysis, water regimes, replications and blocks were treated as random effects while entries were considered as fixed effects. University of Ghana http://ugspace.ug.edu.gh 49 4.5.7. Results of drought tolerance screening The mean performance of one hundred genotypes for the traits measured under both water regimes is shown in Appendix 3. The best five drought tolerant genotypes were selected based on: - The highest yielding genotype under water stress condition, - The highest yielding genotype under well water condition, - The least yield difference between stressed and non-stressed conditions. 4.5.7.1. Farmers selection of the best genotypes under both water regimes The farmers were able to select the best genotypes in terms of haulm production and pod yield. The best genotypes selected were ICG6703 under stressed condition and ICGV-SM99511 under well watered condition. In addition, the genotypes Tainan-9 and ICG11249 were selected based on their good performance under both water regimes. The farmers focused on biomass production as one criterion because this trait adds economic value to the variety. 4.5.7.2. Pod yield and SCMRs of the drought tolerant genotypes 4.5.7.2.1. ANOVA Combined analysis of variance for pod yield and SCMRs are presented in Table 4.3. All the genotypes performed differently in both water regimes due to the significant interaction effect observed between water regimes and genotypes. This implied that the difference between genotypes performances was significantly affected by the water regimes and the water stress effect differed significantly among genotypes. Consequently, the genotypes showed different ranking from one water regime to another for all tested traits. University of Ghana http://ugspace.ug.edu.gh 50 Nevertheless, highly significant differences were seen among the genotypes for all the traits. The effect of the two water regimes was significant for all traits measured. Table 4.3: Mean square from the combined analysis of variance for pod yield and SCMRs under both water regimes Mean Square Source of variation DF PY SCMR 45DAS SCMR 60DAS SCMR 90DAS Replications 1 129278 188.067 52.95 381.535 Genotypes 98 17134** 17.744** 26.14** 17.669** Water regimes (Env 1 & 2) 1 328815** 220.207** 892.2** 1760.247** Genotypes x water regimes 98 7386* 8.77* 12. 91* 8.35* Residual 197 7423 8.071 15.78 7.067 Total 395 10885 11.426 19.7 15.116 Means 165a/222.6b 41.16a/39.66b 47.78a/47.77b 45.97a/41.75b CV % 46.11 7.03 8.59 6.06 LSD at 5% 169.9 5.603 7.835 5.242 **Significant at P ≤ 0.01, *Significant at P ≤ 0.05, a: under drought stress, b: under well water condition, Env: Environment 4.5.7.2.2. Drought effect on pod yield The genotype ICGV-SM 99511 had the highest yield (524 g) under well watered condition (Table 4.4). However, under end-of-season drought condition, the genotype ICG6307 was the best yielding (342 g) genotype. Interestingly, certain genotypes such as ICGV-IS01836, ICG4728, ICG5475, T13-89, ICGV-SM 99505, and ICGV-SM99511 performed well under both water regimes. Genotypes with least yield difference included Tainan-9, ICGV-IS01820 and ICG11249. Based on the selection criteria defined in section 4.5.7, the five best drought tolerant genotypes were ICG6703, ICGV-SM99551, Tainan-9, ICG11249 and ICGV-IS01820. The drought tolerance index ranged from 0.59 to 1.76 for five selected genotypes. ICG6703 had the highest DTI 1.76 and ICGV- SM 99511 the lowest DTI 0.59. Therefore, ICG6703 was the most University of Ghana http://ugspace.ug.edu.gh 51 drought tolerant genotype followed by ICG11249, then Tainan-9 and ICGV-IS01820 and finally ICGV-SM 99511. Genotypes pod yield performances are shown in Figure 4.6. Figure 4. 6: Pod yield under well watered (WW) and water stressed (WS) conditions 0 100 200 300 400 500 600 IC G 6703 IC G 2738 IC G V 02266 IC G V - S M 99511 IC G V -87123 IC G 6022 IC G V -IS 01852 IC G 297 T 13-89 IC G V -IS01836 IC G 5475 IC G 3421 IG C 3386 IC G -3301 IC G 3027 IC G 4728 IC G 1534 IC G 1834 IC G V -IS 01820 IC G 3343 IC G -6222 IC G -12991 IC G V -87281 IC G 9666 IC G 3746 IC G V -IS01859 D ayo IC G 397 IC G 11249 IC G 8751 IC G V -SM 99507 IC G V - S M 99505 IC G 1823 T 119-83 T aim an - 9 Po d yi el d (g ) Genotypes Yield under water stress conditions (g) Yield under well water conditions (g) University of Ghana http://ugspace.ug.edu.gh 52 Table 4.4: Pod yield of entries under drought stress and their performance under well water conditions and their respective drought tolerance indices Rank Entry Pod yield (g) under WS Pod yield (g) under WW Least Pod yield (g) difference DTI 1 ICG 6703a 342 194 148 1.76 2 ICG 2738 326.5 191 135.5 1.71 3 ICGV 02266 310 203.5 106.5 1.52 4 ICGV - SM 99511b 309 524 215 0.59 5 ICGV-87123 302.5 192.5 110 1.57 6 ICG 6022 300 218.5 81.5 1.37 7 ICGV-IS 01852 282.5 221 61.5 1.28 8 ICG 297 279 286.5 7.5 0.97 9 T13-89 274.5 315.5 41 0.87 10 ICGV-IS01836 270.5 343 72.5 0.79 11 ICG 5475 261 335.5 74.5 0.78 12 ICG 3421 253.5 309.5 56 0.82 13 IGC 3386 240.5 208 32.5 1.16 14 ICG-3301 236.5 188.5 48 1.25 15 ICG 3027 234 146.5 87.5 1.6 16 ICG 4728 224.5 337.5 113 0.67 17 ICG 1534 224 264.5 40.5 0.85 18 ICG 1834 223 229.5 6.5 0.97 19 ICGV-IS 01820c 220 222.5 2.5 0.99 20 ICG 3343 213 280 67 0.76 21 ICG-6222 210 410 200 0.51 22 ICG-12991 207.5 196 11.5 1.06 23 ICGV-87281 207 213.5 6.5 0.97 24 ICG 9666 206 94.5 111.5 2.18 25 ICG 3746 206 265.5 59.5 0.78 26 ICGV-IS01859 205.5 299 93.5 0.69 27 Dayo 205.5 191 14.5 1.08 28 ICG 397 204.5 291 86.5 0.7 29 ICG 11249c 204 204 0 1 30 ICG 8751 197.5 240.3 42.8 0.82 31 ICGV-SM 99507 162 319 157 0.51 32 ICGV - SM 99505 161.5 302 140.5 0.53 33 Tainan - 9c 157 159 2 0.99 LSD 5% 155.17 195.71 - - CV % 47.4 44.3 - - a: best high yielding genotype under WS, b: best high yielding genotype under WW, c: Entries selected based the least yield difference between WW & WS, WW: Well Water, WS: Water stress University of Ghana http://ugspace.ug.edu.gh 53 4.5.7.2.3. Drought effect on SPAD chlorophyll meter readings Entries showed significant (P <0.01) differences in chlorophyll content at 45, 60 and 90 DAS, under both water regimes (Table 4.3). The overall means of chlorophyll content under drought stress condition of 41.16, 47.78 and 45.9, respectively, at 45, 60 and 90 DAS were higher than the overall means under irrigated condition, 39.66, 47.77 and 41.75, respectively. There was no genotype by water regimes interaction observed. Under drought condition, ICGV-IS01820 showed the highest SCMR reading (48.4) at 45DAS, followed by ICG11249 (44.3). The genotypes ICGVIS01852 (54.35) and T13- 89 (51.15) showed the highest SCMR reading at 60 and 90 DAS under drought stress condition, respectively. Tainan - 9 had a better SCMR, 48.6, with a rank of 2nd at 90D AS. The magnitude of SCMRs was higher at 90 DAS compared to 45 and 60 DAS (Figure 4.7.a., b., c.) Tainan-9 displayed the highest DTI of 1.13, followed by ICGV-SM99511 1.12, and ICG11249 that showed 1.11 at 60 DAS (Table 4.5). Figure 4.7a: Chlorophyll content 45 DAS under WS and WW conditions 0 10 20 30 40 50 60 IC G 67 03 IC G 2 73 8 IC G V 0 22 66 IC G V SM 9 95 11 IC G V 8 71 23 IC G -6 02 2 IC G V IS 0 18 52 IC G 2 97 T1 3- 8 9 IC G V IS 0 18 36 IC G 5 47 5 IC G 3 42 1 IC G 3 38 6 IC G 3 30 1 IC G 3 02 7 IC G 4 72 8 IC G 15 34 IC G 18 34 IC G V -I S 01 82 0 IC G 3 34 3 IC G -6 22 2 IC G 12 99 1 IC G V -8 72 81 IC G 9 66 6 IC G 3 74 6 IC G IS 0 18 59 D ay o IC G 3 97 IC G 1 12 49 IC G 8 75 1 IC G V -S M 9 95 07 IC G V - SM 9 95 05 IC G 1 82 3 T1 19 -8 3 Ta im an - 9 S C M R s Genotypes 45DAS WS 45DAS WW University of Ghana http://ugspace.ug.edu.gh 54 Figure 4.7b: Chlorophyll content pattern 60 DAS under WS and WW conditions Figure 4.7c: Chlorophyll content pattern 90 DAS under WS and WW conditions 0 10 20 30 40 50 60 IC G 67 03 IC G 2 73 8 IC G V 0 22 66 IC G V SM 9 95 11 IC G V 8 71 23 IC G -6 02 2 IC G V IS 0 18 52 IC G 2 97 T1 3- 8 9 IC G V IS 0 18 36 IC G 5 47 5 IC G 3 42 1 IC G 3 38 6 IC G 3 30 1 IC G 3 02 7 IC G 4 72 8 IC G 15 34 IC G 18 34 IC G V -I S 01 82 0 IC G 3 34 3 IC G -6 22 2 IC G 12 99 1 IC G V -8 72 81 IC G 9 66 6 IC G 3 74 6 IC G IS 0 18 59 D ay o IC G 3 97 IC G 1 12 49 IC G 8 75 1 IC G V -S M 9 95 07 IC G V - SM 9 95 05 IC G 1 82 3 T1 19 -8 3 Ta im an - 9 S C M R s Genotypes 60DAS WS 60DAS WW 0 10 20 30 40 50 60 IC G 67 03 IC G 2 73 8 IC G V 0 22 66 IC G V SM 9 95 11 IC G V 8 71 23 IC G -6 02 2 IC G V IS 0 18 52 IC G 2 97 T1 3- 8 9 IC G V IS 0 18 36 IC G 5 47 5 IC G 3 42 1 IC G 3 38 6 IC G 3 30 1 IC G 3 02 7 IC G 4 72 8 IC G 15 34 IC G 18 34 IC G V -I S 01 82 0 IC G 3 34 3 IC G -6 22 2 IC G 12 99 1 IC G V -8 72 81 IC G 9 66 6 IC G 3 74 6 IC G IS 0 18 59 D ay o IC G 3 97 IC G 1 12 49 IC G 8 75 1 IC G V -S M 9 95 07 IC G V - SM 9 95 05 IC G 1 82 3 T1 19 -8 3 Ta im an - 9 SC M R s Genotypes 90DAS WS 90DAS WW University of Ghana http://ugspace.ug.edu.gh 55 The genotypes that had pod yield above 150 g were used to draw a dendrogram using DARWIN5 software version 5.0.158. The five selected drought tolerant genotypes are in three different clusters (Figure 4.8). Table 4.5: Drought tolerance indices of entries at 45, 60 and 90 DAS Genotypes DTI SCMR 45DAS DTI SCMR 60DAS DTI SCMR 90DAS ICG6703 1.14 0.93 1.15 ICG 2738 0.91 1.06 1.09 ICGV 02266 0.96 0.96 1.02 ICG VSM 99511 1.2 1.12 0.96 ICGV 87123 1.1 1.06 0.97 ICG-6022 1.07 1.05 1.06 ICGV IS 01852 1 1.03 1.05 ICG 297 1.03 1.07 1.1 T13- 89 1.05 1.05 1.11 ICG V IS 01836 1.02 1.11 1.02 ICG 5475 0.94 0.99 1.11 ICG 3421 1.13 1.05 1.07 ICG 3386 1.02 0.91 1.06 ICG 3301 0.98 1.03 1.06 ICG 3027 0.97 1 1.08 ICG 4728 1.07 0.98 1.12 ICG1534 1.08 1.06 1.02 ICG1834 0.94 1.05 1.17 ICGV-IS 01820 1.1 1.05 1.04 ICG 3343 1.04 1.08 1.05 ICG-6222 1.07 1.03 1.13 ICG12991 1.05 1.08 1.04 ICGV-87281 1.05 1.03 1.09 ICG 9666 1 1.05 1.17 ICG 3746 1.11 1.08 1.08 ICGIS 01859 0.99 0.93 1.13 Dayo 0.92 1.06 1.13 ICG 397 1.03 1.03 1.04 ICG 11249 1.22 1.11 1.11 ICG 8751 0.97 0.92 1.14 ICGV-SM 99507 1.07 1.07 1.08 ICGV - SM 99505 1.01 0.93 1.11 ICG 1823 0.99 1.02 1.08 T119-83 0.94 1.03 1.1 Tainan - 9 0.98 1.13 1.21 DTI: Drought tolerance index, SCMR: SPAD chlorophyll meter reading University of Ghana http://ugspace.ug.edu.gh 56 Figure 4.8: Dendrogram showing the selected drought tolerant parents based on pod yield University of Ghana http://ugspace.ug.edu.gh 57 4.6. Discussion 4.6.1. Early maturity In this study, two extra early maturing varieties Chico and ICIAR19BT which matured in about 75 days after sowing and three early maturing ICG3584, ICGV02022 and 796 that matured in 80 days after sowing were identified. All selected early maturing varieties reached 50% flowering within 24 days after sowing. Chico is the most widely used source of earliness in several breeding programmes. It was used 1,180 times as a parent in developing early maturity lines at ICRISAT between 1976 to 2002 (Upadhyaya et al., 2006). Several studies reported Chico as the early maturing variety used to develop new early maturing varieties (N'Doye and Smith, 1993; Ali and Wynne, 1994; Upadhyaya and Nigam, 1994). The variety ICIAR19BT has been released under the name SAMNUT 24 in 2011 as an extra early, high yielding and rosette resistant variety in Nigeria. Most of popular varieties (55-437, RRB and JL24) identified through participatory rural appraisal are intermediate maturing varieties that matured between 90 to 95 days after sowing. Therefore, genes of earliness from the extra early maturing varieties identified in this study can be introgressed into farmers' preferred varieties to improve those varieties for earliness through hybridization. Days to maturity also determines the differences in pod yield. In general, late maturing genotypes yield better than early maturing genotypes (Culbreath et al., 1999; Padi, 2008). Kotzamanidis et al. (2006) found that early maturity in groundnut is associated with narrow pod distance from the main stem, leading to more synchronous maturity of the pods. Jogloy et al. (2011) reported that extremely early maturity is not desirable because it is generally associated with yield reduction. However, in this study among the selected early maturing varieties, some varieties exhibited good performance for pod yield. These varieties included ICIAR19BT (1540 University of Ghana http://ugspace.ug.edu.gh 58 g), followed by ICGV02022 (700 g). Chico is the lowest pod yielding varieity (430g) which is consistent with all the studies that used Chico as a parental in developing new early maturing varieties. The one hundred and fifty genotypes screened for earliness were classified into three maturity group that are extra early (5), early (29) and intermediate (105) maturing genotypes. The eleven remaining genotypes were not adapted to experimental conditions. The extra early maturing varieties can be used as parental lines to develop new early maturing varieties in a breeding programme by hybridazation. Information generated from this screening is useful for creating a mini core collection for further studies. 4.6.2. Drought tolerance In this study, some promising groundnut varieties with good performance for pod yield under drought stress (ICG6703) and well watered conditions (ICGV-SM99511) were identified. The varieties Tainan-9, ICG11249 and ICGV-IS 01820 performed well under both conditions. These varieties also showed good performance for SCMRs and showed higher drought tolerance indices for pod yield and chlorophyll content. Farmers' selected three genotypes ICGV- SM99511, ICG6703 and Tainan-9 based on pod yield and biomass production performance under both water regimes. The selected genotypes by the farmers are among the five drought tolerant varieties identified in this study. Therefore, farmers' involvement in this study was very useful because it enabled breeders to take farmers' preferred traits into consideration. Farmers were very happy to be involved in parental lines selection. Higher DTI values indicate the selected genotypes are drought tolerant (Nautiyal et al., 2002). In this study, the responses to drought among the groundnut genotypes for SCMR at 45, 60 and 90 DAS were rather similar in patterns but somewhat different in extents at 90 DAS. There was no University of Ghana http://ugspace.ug.edu.gh 59 significant difference among the genotypes for chlorophyll content (Figure 4.7a, b., c). An increase in drought stress period may probably show chlorophyll content difference among the genotypes. However, an increase in chlorophyll content under drought stress condition based on the increase of SCMR values was observed. Genotypes that showed high values of SCMRs and maintain high biomass production under drought stress may be considered as drought tolerant. Drought is known to affect cholorophyll content in many crops including wheat (Sarker et al., 1999), and cattail (Thypha latifolia) where it inhibits photosynthetic capacity (Epron and Dreyer,1993). The ability to maintain chlorophyll density under drought stress has been suggested as a drought tolerance mechanism in groundnut (Arunyanark et al., 2008; Sheshshayee et al.,2006). The results are in agreement with Moreshat et al. (1996) who found increased the chlorophyll content under mild drought stress. Jongrungklang et al.(2008) also found the increase of SCMR of groundnut under water limiting conditions and as this trait is related to photosynthetic capacity, they concluded that the increase of SCMR could be attributed to drought tolerance. However Reddy and Rao (1968) reported that severe drought stress decreased the levels of chlorophyll a, b and total chlorophyll. It could be hypothesized that groundnut genotypes with high SCMR have more photosynthetic machinery per unit leaf area and hence potential for greater assimilation under drought stress (Songsri et al., 2009). Boontang et al., (2010) found that groundnut genotypes with high SCMR under drought stress could maintain higher biomass production. Groundnut genotypes showing consistently high SCMRs values across water regimes and sampling dates were ICG-VSM 99511, ICGV87123, ICG6703, ICG 11249, ICGV-IS01820, University of Ghana http://ugspace.ug.edu.gh 60 Tainan-9, ICGV-IS01852, ICG12991, ICG3386, ICGV-IS01859, ICG9666, T13- 89, ICG3746 and ICG297 (Appendix 3). In this study, end-of-season drought caused pod yield reduction that varies from genotype to genotype. However, certain genotypes showed least pod yield difference in both water regimes. The varieties Tainan-9, ICG11249 and ICGV-IS01820 that showed least pod yield difference can be used by farmers in the short time as drought tolerant varieties prior to improvement of their popular varieties identified through the PRA study. In a participatory varietal selection conducted by INRAN in three locations in Niger in 2012, farmers in all the locations selected the variety Tainan-9 as the best drought tolerant variety and also high yielding with good biomass production among the varieties evaluated. University of Ghana http://ugspace.ug.edu.gh 61 4.7. Conclusions - Two extra early maturing varieties (75DAS) Chico and ICIAR19BT were identified. - Three early maturing varieties (80DAS) ICG3584, ICGV02022 and 796 were selected. These materials can be used as parental lines in a breeding programme to develop early maturing varieties. - ICG6703, ICGV-SM 99511, Tainan-9, ICG11249 and ICGV-IS01820 were selected as drought tolerant material based on their performance for yield and chlorophyll content under well water and end-of-season drought stress conditions. Therefore, these materials can be used as parental lines to develop drought tolerant groundnut varieties. - ICG6703, ICGV-SM 99551, Tainan-9 are the varieties selected by the farmers during the field day. - The varieties ICIAR19BT and ICGV-SM 99511 were the best high yielding among the parental lines. University of Ghana http://ugspace.ug.edu.gh 62 CHAPTER FIVE 5.0. GENETIC ANALYSIS OF EARLINESS AND DROUGHT TOLERANCE 5.1. Introduction Early maturing and high yielding cultivars are needed in most all groundnut growing areas to fit into a cropping system involving two crops a year, and to escape natural hazards such as drought and diseases. Improvement of drought tolerance in groundnut would be beneficial in rainfed regions where drought is a major constraint limiting productivity and quality. End-of- season drought is the most prevalent type in the major groundnut growing agroclimatic semi-arid environments. It occurs during the pod filling phase of groundnut (Nageswara et al., 1985; Wright et al., 1994; Ndunguru et al., 1995). Therefore, breeding for end-of-season drought tolerance can increase productivity in drought-prone environments and reduce aflatoxin contamination. With new knowledge of easily measurable surrogates of transpiration efficiency (TE), a trait associated with drought tolerance, specific leaf area (SLA) and soil plant analytical development (SPAD) chlorophyll meter readings (SCMR), it is now possible to integrate TE through the surrogates in breeding and selection schemes in groundnut (Nigam and Aruna, 2008). SCMR is an indication of the light-transmittance characteristics of the leaf which is dependent on the leaf chlorophyll content (Richardson et al. 2002). As a noninvasive surrogate of TE, SCMR is easy to operate, reliable, fairly stable and low cost (Sheshshayee et al., 2006). Recent studies indicated that specific leaf area (SLA) and soil plant analytical development (SPAD) chlorophyll meter reading (SCMR), that are easy to measure, are highly correlated with TE (Sheshshayee et al., 2006). Both traits (SLA, SCMR) have considerable genetic variation in groundnut (Upadhyaya 2005; Lal et al., 2006; Sheshshayee et al., 2006). Positive relationship between TE and SCMR University of Ghana http://ugspace.ug.edu.gh 63 has been reported (Bindu Madhava et al., 2003; Sheshshayee et al., 2006). Nageswara et al. (2001) and Upadyaya (2005) found a significant negative correlation between SCMR and SLA and suggested that this chlorophyll meter could be used as a rapid and reliable measure to identify genotypes with high water use efficiency in groundnut. SCMR is reported to be more stable than SLA (Upadhyaya, 2005). It is also correlated with pod yield in groundnut (Reddy et al., 2004; Upadhyaya 2005). Harvest index, has also been found to be correlated with SLA and SCMR under both well-watered and long-term drought conditions (Songsri et al., 2008). The concept of combining ability is important in designing plant breeding programmes. It is especially useful in testing procedures used to study and compare the performance of lines in hybrid combinations. Combining ability in crosses is defined as the ability of parents or cultivars to combine amongst each other during the process of hybridization so that favorable genes/characters are transmitted to their progenies (Panhwar et al., 2008). Specific combining ability is defined as the deviation in the performance of hybrids from the expected productivity based upon the average performance of lines involved in the hybrid combinations, whereas general combining ability is defined as average performance of a line in a series of crosses (Griffing, 1956; Falconer and Mackey, 1996). According to Sprague and Tatum (1942), general combining ability is due to genes which are largely additive in their effects and specific combining ability is due to the genes with dominance or epistatic effects. Rawlings and Thompson (1962) used line x tester analysis to estimate GCA and SCA of inbred parents. Since the development of new cultivars through hybridization is a continuous process, information on combining ability of new cultivars remains important. University of Ghana http://ugspace.ug.edu.gh 64 Estimates of heritability from segregating populations are useful in understanding the genetics of hybridization and inbreeding (Ali and Wynne, 1994). They can help breeders in selecting and utilizing superior individuals from a population. Heritability in the narrow sense is important, because the effectiveness of selection depends on the additive portion of genetic variance in relation to total variance (Fernandez and Miller, 1985). The parent-offspring regression method is commonly used to compute heritability estimates of quantitative characters in both self- and cross-fertilizing crops (Fernandez and Miller, 1985). Some of the routinely used parent-offspring combinations in self-fertilizing crops are: F1/F2, F2/F3, and F3/F4, (Smith and Kinman 1965). To develop the desired population for earliness and drought tolerance, knowledge of the combining abilities of lines is very important in selecting parents. Efficient utilization of the physiological traits for improving drought tolerance requires an understanding of the inheritance of the target traits. The present studies were carried out to estimate the combining ability and heritability of selected parents for earliness and drought tolerance traits for effective selection and use in a groundnut breeding programme. 5.2. Material and methods 5.2.1 Experimental site The experimental site was INRAN Maradi that is located in the south-central part of the country (Chapter IV: Section 4.2). Soil has a pH of 6.5 and contains 12% clay, 5% loam, 4% coarse silt, 77% sand and 2% organic matter (Raynaut et al., 1984). University of Ghana http://ugspace.ug.edu.gh 65 5.2.2. Genetic material and hybridization techniques The genetic material included three farmers' preferred varieties 55-437, RRB and JL24 from the PRA and five early maturing (Chico, ICIAR19BT, 796, ICG3584, ICGV02022) and five drought tolerant (ICG6307, ICGV-SM99511, Tainan-9, ICGV-IS01820, ICG11249) varieties selected from germplasm screening in Chapter IV (Table 5.1). The three farmers preferred varieties were crossed with each set of material (earliness and drought tolerance) in a 3 x 5 North Carolina II mating design to generate 15F1s in the case of earliness and 15F1s of drought tolerance combination given a total of 30F1s. Ten lines of groundnut viz., ICGV02022, ICG3584, Chico, ICIAR19BT, 796, ICGV-SM99511, Tainan-9, ICG11249, ICGV-IS01820 and ICG6703 (female parents) and three testers 55-437, JL24 and RRB (male parents) were crossed in July, 2010 at Tarna INRAN station. Before planting, the field was prepared as described previously in Chapter IV (section 4.4.2). Three rows 5 m long with 1 m between rows and 1 m spacing between hills were hand planted with female parents from the earliness and drought tolerant groups. One seed was planted per hill; each plot contained 15 plants. Each male parent was planted in one row of 25 plants along each two of the female parent. Seven days after planting, missing stands were replanted. Standard cultural practices were applied to keep the field free of weeds and pests. The crossing blocks were replicated two times to increase the F1 seeds. Necessary precautions were taken to avoid the contamination of genetic material at the time of crossing. Hand emasculation of the flower buds and pollination were made by skilled workers under the supervison of researcher. Pollination commenced on July 27, 2010 and lasted for about one month. Five to seven crosses were made on each individual female to increase hybrids seeds. The crosses were harvested during the first week of October 2010. F1Seed set ranged from 3 to 40 for the different combinations of parents. University of Ghana http://ugspace.ug.edu.gh 66 Ten F1 seeds of each cross combination along with their parents were hand planted in an unreplicated trial during 2011 off-season. Inter-plant and inter-row spacing were maintained at 0.25 and 0.75 m, respectively. One seed per hole were sown. The experimental population was kept under normal agronomic care from sowing to maturity. At harvest, All F1 plants were carefully examined for several morphological traits such us number of stem, plant height, leaf color, pod and seed characters and compared with both parents to confirm their hybridity. Out of the 30 F1s generated, based on the results of the hybridity test, seven F1 hybrids of early maturing combinations and seven F1 hybrids of drought tolerance were selected. The names of the crosses are indicated below: - Early maturity F1s - ICG 3584 x 55 - 437 - ICG 3584 x JL24 - ICIAR 19BT x 55 - 437 - ICIAR19BT x RRB - 796 x JL24 - 796 x RBB - Chico x RRB - Drought tolerance F1s - ICGV-SM 99511 x 55- 437 - Tainan-9 x 55- 437 - Tainan-9 x JL24 - Tainan-9 x RRB - ICG 11249 x 55- 437 - ICG 11249 x JL 24 - ICG 11249 x RBB Individual plants were selected from the segregating F2 populations. Seeds of selected F2 were planted as bulks to produce F3 seeds in an unreplicated trial. Similar procedure was followed to produce F4 seeds following bulk method of selection. University of Ghana http://ugspace.ug.edu.gh 67 Table 5.1: Early maturity and drought tolerant parental lines Parental lines Days to maturity Growth habit Botanical variety Origin Male JL 24 (ICG 7827) 90-95 Erect Vulgaris India RRB 90-95 Erect Spanish Nigeria 55-437 90 Erect Spanish Hungary Female Early maturity Chico (ICG 476) 75 Erect Vulgaris USSR ICIAR19BT 75 Erect Spanish Nigeria ICG 3584 80 Erect vulgaris India ICGV 02022 80 Erect fastigiata ICRISAT 796 80 Erect - Russia Drought tolerant Tainan-9 110 Erect Spanish Thailand ICGV-SM99511 95 Erect - ICRISAT ICG11249 90 Erect Spanish Tanzania ICG6703 95 Erect hypogaea Paraguay ICGV-IS01820 110 Erect - ICRISAT 5.2.3. Evaluation of populations for earliness and drought tolerance 5.2.3.1. Field experiments Field experiment was fully described in Chapter IV, Section 4.4.2. The seven crosses in the F3 populations and their parents were evaluated at Tarna INRAN station. The treatments of 27 entries (Table 5.2) were arranged in a randomized complete block design (RCBD) in three replications. University of Ghana http://ugspace.ug.edu.gh 68 Table 5. 2: Early maturity and drought tolerant parental lines and F3 populations Entry Early maturity Entry Drought tolerant F3 populations F3 populations 1 ICG3584 x 55 - 437 15 ICGV -SM 99511 x 55 - 437 2 ICG3584 x JL24 16 ICG11249 x 55 - 437 3 ICIAR19BT x 55 - 437 17 ICG11249 x JL24 4 ICIAR19BT x RRB 18 ICG11249 x RRB 5 796 x JL24 19 TAINAN -9 x 55 - 437 6 796 x RBB 20 TAINAN -9 x JL24 7 Chico x RRB 21 TAINAN -9 x RRB Parents Parents 8 ICG3584 22 55 - 437 9 ICIAR19BT 23 JL24 10 796 24 RRB 11 Chico 25 ICGV -SM 99511 12 55 - 437 26 ICG11249 13 JL24 27 TAINAN -9 14 RRB F3 and F4 populations were also evaluated at Tarna INRAN station. The treatments of 14 entries (Table 5.3) were arranged in a randomized complete block design in three replications. Drought tolerant entries (14) were planted in in α lattice design replicated three times in two environments (well watered and water stressed conditions). Water regimes treatments were described in Chapter IV, Section 4.5.3. Plant and rows spacing were 25 and 100cm respectively. Two row plots of 10 seeds each were hand planted. The plot size was 2.5m2 (2.5m x 1m). All recommended cultural practices were done during the experiment. The experiment was harvested 90 days after sowing. University of Ghana http://ugspace.ug.edu.gh 69 Table 5.3: Entries used to estimate Narrow sense heritability Entry Early maturity Entry Drought tolerant F3 populations F3 populations 1 ICG3584 x 55 - 437 15 ICGV -SM 99511 x 55 - 437 2 ICG3584 x JL24 16 ICG11249 x 55 - 437 3 ICIAR19BT x 55 - 437 17 ICG11249 x JL24 4 ICIAR19BT x RRB 18 ICG11249 x RRB 5 796 x JL24 19 TAINAN -9 x 55 - 437 6 796 x RBB 20 TAINAN -9 x JL24 7 Chico x RRB 21 TAINAN -9 x RRB F4 populations F4 populations 8 ICG3584 x 55 - 437 22 ICGV -SM 99511 x 55 - 437 9 ICG3584 x JL24 23 ICG11249 x 55 - 437 10 ICIAR19BT x 55 - 437 24 ICG11249 x JL24 11 ICIAR19BT x RRB 25 ICG11249 x RRB 12 796 x JL24 26 TAINAN -9 x 55 - 437 13 796 x RBB 27 TAINAN -9 x JL24 14 Chico x RRB 28 TAINAN -9 x RRB 5.2.3.2. Data collection 5.2.3.2.1. Early maturity A sample of 50 pods was collected from each plot in each replication. In order to reduce sampling error, two to three pods were taken at random from each plant rather than taking pods from the bulk plot. Data for the eight traits recorded were: - Emergence % (7DAS), - 50% plants flowering, - Length of 20 pods (cm), - Weight of 50 pods (g), - Number of seeds in 50 pods, - Weight of seeds in 50 pods (g), University of Ghana http://ugspace.ug.edu.gh 70 - Number of pods in each of 5 maturity classes (MC) based upon inner hull color ranging from white to back (white = MC1, Yellow = MC2, light brown = MC3, dark brown = MC4 and black = MC5). 1, 2, 3, 4 and 5 are Maturity classes. A sample of 50 pods were shelled and rated for maturity. From the recorded data, the following variables were computed: Maturity index = (MC1 x 1) + (MC2 x 2) + (MC3 x 3) + (MC4 x 4) + (MC5 x 5) For the maturity index, pods in each class were multiplied by its class number and summed. Higher values of maturity index reflect earlier maturity (Ali et al., 1995). 5.2.3.2.2. Drought tolerance Observations were recorded on fifteen plants selected at random among parents and F3 population. Recommended agronomic and plant protection measures were adopted during the experiment. Data measured were: - Days to 50% plants flowering, - SPAD chlorophyll meter reading at 60 and 80DAS (SCMR), - Pod yield, - Biomass, - Harvest index (HI), - Drought tolerance index defined as the ratio of trait value measured under water stressed conditions over value recorded under well watered conditions was computed for HI, biomass, pod yield and SCMR. High DTI values indicate drought tolerance and vice versa. Shelling % = Weight of seeds in 50 pods Weight 50 pods x 100 University of Ghana http://ugspace.ug.edu.gh 71 5.2.3.3. Data analysis Analysis of variance (ANOVA) was computed for the entries according to randomized complete block design using PROC GLM in SAS (version 9.2) considering entries as fixed effects, and replications and blocks within replications as random effects. Analysis of variance for traits measured was done separately (Gomez and Gomez, 1984). Combined analysis of variance of two water regimes data was done. Least square difference (LSD) at P =0.05 was used to compare means. Mean squares caused by difference among crosses were partitioned into difference due to male parents and female parents, which was attributed to general combining ability (GCA), and difference due to male x female interaction, which was attributed to specific combining ability (SCA). In case of the presence of significant GCA and SCA mean squares, GCA and SCA effects were calculated using the method described by Simmonds (1979). Crosses evaluation data were analyzed by regression analysis to estimate the narrow sense heritability by using GenStat 12th Edition software. The genetic model is as follows: X = μ + GCAP + GCAQ + SCAPQ - Where; X is the true mean of a cross between lines P and Q; - μ is the mean of all crosses and GCA and SCA are general and specific combining abilities, respectively. - The error term was assumed to be zero. If it is present, it should be associated with GCA and SCA effects. The mathematical model for NCDII is as follows: Yijkl = µ + mi + fj + (mf)ij + eijkl - Yijk is the observed value of the progeny of the ith male crossed with jth female in the kth replication; University of Ghana http://ugspace.ug.edu.gh 72 - µ is the overall population mean; - mi is the effect of the ith mother (effect of the general combining ability of the ith mother); - fj is the effect of the jth father mated to the ith mother (effect of the general combining ability of the jth father); - (mf)ij is the interaction effect between the ith mother and the jth father (effect of the specific combining ability of the ith mother and the jth father); - eijk is the experimental error. Table 5. 4: ANOVA in NC Design II Source of variation Degree of freedom Mean square Expected MS Replication r-1 GCAMale m-1 MSm σe 2 + rσfm 2 + rfσm 2 GCAFemale f-1 MSf σe 2 + rσfm 2 + rmσf 2 SCA =Male x Female (m-1)(f-1) MSmf σe 2 + rσfm 2 Error mf(r-1) MSe σe 2 Source: (Kearsey and Pooni, 1996) Estimating variance components in NCDII - σ2e = variation within full sibs = environmental variance - σ2f = variation between females = GCAf variance - σm2 = variation between males = GCAm variance - σfm2 = variation due to interaction between females and males = SCA variance From ANOVA table: - σm2 = (MSm – MSmf)/fr = covHS =1/4 σA2 - σF2 = (MSf – MSmf)/mr = covHS =1/4 σA2 From the foregoing genetic variances can be estimated as follows: - 4σm2 = 4covHS = σA2 - 4σF2 = 4covHS = σA2 - 4σmf2 = 4(covFS-2covHS) = σ2D University of Ghana http://ugspace.ug.edu.gh 73 - where σA2 and σ2D = VA and VD = additive and dominance variances, respectively Estimating narrow sense heritability Using variance ratios as follows: - Based on female additive variance - hf2 = 4σf2/(σ2e/r + 4σ2mf + 4σ2f) = VAf / VP Estimate of GCV and PCV GCV= ටఙಸ మ ௑ത X 100 , PCV= ටఙು మ ௑ത X 100 Correlations among traits were determined by Pearson correlations by using GenStat version 12.0. 5.3. Results 5.3.1. Earliness 5.3.1.1. ANOVA Results showed that the parental lines and F3 populations differed significantly for all the traits except for percentage of emergence and number of days 50% plants flowering (Table 5.5). Table 5. 5: Mean squares from ANOVA for parental lines and F3 populations Source of variation DF EMERSD NDFTPF WFTP WSFTP NSFTP SH PL MI Replication 2 100.59 0.74 25.5 27.46 27.59 18.73 0.009 3633.16 Entry 13 164.33 2.33 90.50** 44.24** 32.44* 27.90* 0.09** 1527.25* Error 26 187.13 1.27 23.46 15.23 14.67 11.62 0.005 631.16 *Significant at P ≤ 0.05, **Significant at P ≤ 0.01, %EMERSD: % of emergence 7DAS, NDFTPF: Number of days 50% plant flowering, WFTP: Pod weight, WSFTP: Seed Weight, NSFTP: Number of seed, SH: shelling %, PL: pod length, MI: Maturity index University of Ghana http://ugspace.ug.edu.gh 74 Mean squares from the ANOVA for F3 and F4 populations are shown in Table 5.6. The crosses responded similarly for days to emerge, 50% plants flowering, seed number and maturity index. Whereas the two populations differed significantly for pod yield, shelling percentage, seed weight and pod length. Table 5.6: Mean squares from ANOVA for F3 and F4 populations for eight traits Source of variation DF EMERSD NDFTPF WFTP WSFTP NSFTP SH PL MI Replication 2 100.59 0.74 25.5 27.46 27.59 18.73 0.009 3633.16 Entry 13 46.4 4.33 57.18*** 36.28*** 5.6 12.31** 0.029* 430.6 Residual 28 103.6 3.92 10.97 7.36 10.64 4.194 0.012 421.8 *Significant at P ≤ 0.05, **Significant at P ≤ 0.01, ***significant at P ≤ 0.001, %EMERSD: % of emergence 7DAS, NDFTPF: Number of days 50% plant flowering, WFTP: Pod weight, WSFTP: Seed Weight, NSFTP: Number of seed, SH: shelling %, PL: pod length, MI: Maturity index 5.3.1.2. Mean performance The crosses Chico x RRB, ICIAR19BT x RRB and ICIAR19BT x 55-437 showed the highest values of maturity indices of 223.33, 218 and 214.67, respectively indicating early maturity (Table 5.7). The cross ICG3584 x JL24 showed the lowest maturity index value (178.33). Fifty percent plants flowering was reached in 28 DAS for the cross 796 x RRB, followed by the crosses ICIAR19BT x RRB and ICIAR19BT x 55-437 in 28.33 DAS. The cross ICG3584 reached 50% plants flowering in 30DAS. In terms of pod weight, the cross ICIAR19BT x 55-437 showed the highest value 45.03g while Chico x RRB showed the lowest value 30.6g. The highest value of shelling percentage was obtained with the cross ICG3584 x JL24 (74.82%) followed by ICIAR19BT x 55-437 (70.82). The hybrid ICIAR19BT x 55-437 showed good performance for both agronomic and earliness traits. University of Ghana http://ugspace.ug.edu.gh 75 Table 5.7: Mean performance of parental lines and F3 populations Entries %EMERSD NDFTPF WFTP WSFTP NSFTP SH PL MI JL24 83.33 29.33 47.8 31.63 101.33 66.14 2.85 177 RRB 71.67 30.67 44.07 27.53 99.67 62.13 2.45 189.67 55-437 58.33 30.67 32.27 21.07 100.33 65.24 2.15 146 796 81.67 29.67 37.17 25.17 94.67 67.87 2.19 177.33 Chico 70 29.67 31.33 20.13 99.67 64.16 2.28 218 ICG3584 68.33 30 37.07 24.93 100 66.75 2.2 175.67 ICIAR 19BT 78.33 28.33 43.8 30.47 100 69.49 2.41 213.33 F3: 796 X JL24 61.67 29.67 38.9 26.67 107.67 68.37 2.33 210.33 F3: 796 X RRB 70 28 42.83 29.2 99.33 68.2 2.34 193.67 F3: Chico X RRB 70 28.67 30.6 21.17 100.67 69.16 2.23 223.33 F3: ICG 3584 X JL24 76.67 29 36.07 26.93 93 74.82 2.21 178.33 F3: ICG 3584 X 55- 437 61.67 30 41.13 28.1 99.67 67.98 2.26 204.33 F3: ICIAR19BT X RRB 75 28.33 43.73 29.2 100 66.55 2.48 218 F3: ICIAR19BT X 55- 437 71.67 28.33 45.03 31.9 101.33 70.82 2.4 214.67 Mean 71.31 29.31 39.41 26.72 99.81 67.69 2.34 195.69 Range 58.33 - 83.33 28.00 - 30.67 30.60 - 47.80 20.13 - 31.90 93.00 - 107.67 62.13 - 74.82 2.15 - 2.85 146.00 - 223.33 LSD 22.96 1.9 8.13 6.55 6.43 5.72 0.12 42.17 CV (%) 19.18 3.85 12.28 14.6 3.83 5.03 3.16 12.83 %EMERSD: % of emergence 7DAS, NDFTPF: Number of days 50% plant flowering, WFTP: Pod weight, WSFTP: Seed Weight, NSFTP: Number of seed, SH: shelling %, PL: pod length, MI: Maturity index Results of F3 and F4 populations' performance for eight traits are presented in Table 5.8. The crosses F3: ICIAR19BT X RRB, F3: ICIAR19BT X 55-437, F3: 796 X RRB, F3: 796 X JL24 and their corresponding F4 populations exhibited the highest means of pod weight and maturity index. The responses of F3 and F4 populations for pod yield (Figure 5.1) and maturity index (Figure 5.2) were rather similar in patterns but somewhat different in extents. However, the crosses F3: ICIAR19BT X RRB, F3: ICIAR19BT X 55-437, F3: 796 X RRB, F3: 796 X JL24, and their corresponding F4 generation displayed similarity in patterns and extents for pod weight and maturity index. The cross Chico X RRB showed good mean performance for maturity index in both populations. University of Ghana http://ugspace.ug.edu.gh 76 Table 5.8: Range and mean of F3 and F4 groundnut populations for eight traits Populations %EMER7DAS NUDFTPF WFTP NSFTP WSFTP SH PL MI F3: 796 X JL24 80 32 43.67 98 31.97 73.21 2.35 237.67 F3: 796 X RRB 76.67 33 41.1 100 27.93 68.03 2.41 221 F3: Chico X RRB 78.33 32.67 34.4 99 24.07 69.83 2.34 220.67 F3: ICG3584 X 55-437 76.67 33 38.1 98.33 25.97 68.12 2.26 212.67 F3: ICG3584 X JL24 78.33 33.33 36.53 95.67 25.6 69.99 2.23 210.67 F3: ICIAR19BT X 55-437 76.67 31.67 45.47 100 33.07 72.69 2.4 208.33 F3: ICIAR 19BT X RRB 70 32 46.07 99 32.47 70.48 2.49 203.33 F4: 796 X JL24 85 34 43.1 101.33 30.37 70.52 2.41 240.33 F4: 796 X RRB 80 30.33 42.2 99 28.27 66.96 2.47 228.33 F4: Chico X RRB 83.33 32 34.73 99 24.7 71.01 2.34 226.67 F4: ICG3584 X 55-437 76.67 34.67 34.57 100 23.3 67.05 2.19 207 F4: ICG3584 X JL24 81.67 34 41.07 99.67 29 70.59 2.5 213.67 F4: ICIAR19BT X 55-437 83.33 31 45.83 98.67 33.33 72.72 2.41 235.67 F4: ICIAR19BT X RRB 75 32.33 44.7 97.33 30.83 68.91 2.49 216 Mean 78.7 32.57 40.82 98.93 28.63 70.01 2.379 220.1 Range 70-85 30.33-34.67 34.40-46.07 95.67-101.33 23.30-33.33 66.96-73.21 2.19-2.50 203.33-240.33 LSD 5% 17.02 3.315 5.539 5.456 4.538 3.425 0.1825 34.35 CV % 12.9 6.1 8.1 3.3 9.5 2.9 4.6 9.3 %EMERSD: % of Emergence 7DAS, NDFTPF: Number of days 50% plant flowering, WFTP: Pod weight, WSFTP: Seed Weight, NSFTP: Number of seed, SH: shelling %, PL: pod length, MI: Maturity index University of Ghana http://ugspace.ug.edu.gh 77 Figure 5.1: Pod yield of F3 populations and their corresponding performance in F4 Figure 5.2: Maturity index of F3 populations and their corresponding values in F4 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 P od y ie ld ( g) Populations Pod Yield 180.00 190.00 200.00 210.00 220.00 230.00 240.00 250.00 M at ur ity in de x Populations Maturity Index University of Ghana http://ugspace.ug.edu.gh 78 5.3.1.3. Phenotypic and genotypic coefficient of variation estimates The estimates of phenotypic coefficient of variation (PCV) were greater than genotypic coefficients of variation (GCV) for all the characters studied (Table 5.9). None of the traits recorded high PCV and GCV. PCV ranged from 4.36 to 19.34% and GCV from 2.02 to 11.99%. The traits pod and seed weight showed moderate GCV and PCV values. Low PCV and GCV were observed for days to 50% plants flowering, shelling percentage and pod length. Table 5. 9: Components of variation for different groundnut traits Traits Mean Range σ2G σ 2 Ph PCV (%) GCV (%) %Emergence 7DAS 71.31 58.33-83.33 -7.6 179.53 18.79 - 50% plant flowering 29.31 28.00 - 30.67 0.35 1.63 4.36 2.02 Pod weight 39.41 30.60 - 47.80 22.35 45.81 17.17 11.99 Seed weight 26.72 20.13 -31.90 9.67 24.9 18.67 11.64 Number of seed 99.81 93.00 - 107.67 -7.54 47.53 6.91 - Shelling % 67.69 62.13 - 74.82 5.43 17.05 6.1 3.44 Pod length 2.34 2.15 - 2.85 0.03 0.04 8.54 7.4 Maturity index 195.69 146.0 - 223.33 298.7 929.86 15.58 8.83 σ2G: Genotypic variance, σ 2 Ph: Phenotypic variance, PCV: Phenotypic coefficient of variation, GCV: Genotypic coefficient of variation 5.3.1.4. General and specific combining ability estimates Analysis of variance of combining ability for the eight traits revealed significant differences (P ≤ 0.01 and P ≤ 0.05) among parents for pod weight, seed weight, pod length and maturity index (Table 5.10). The crosses showed highly significant differences (P ≤ 0.01) for pod weight, number of seed and pod length. GCA mean squares for males were highly significant (P ≤ 0.01) for only pod weight. Female GCA mean squares were highly significant (P ≤ 0.01) for pod weight. GCA for the maturity index was significant at 5% level (Table 5.10). Estimates of general combining ability effect for traits are presented in Table 5.11. Among the males, the variety RRB showed highest GCA effect (5.57) for maturity index followed by the University of Ghana http://ugspace.ug.edu.gh 79 variety 55-437 (3.4). The male 55-437 exhibited the highest GCA effect for pod and seed weight 3.33 and 2.4 respectively. Chico and ICIAR19BT showed the highest GCA effect for maturity index 17.24 and 10.24 respectively. The female parent ICIAR19BT showed the highest GCA effect for pod and seed weight 4.63 and 2.95 respectively. Table 5.10: Mean square of combining ability analysis for eight traits Source of variation DF %EMERSD NDFTPF WFTP WSFTP NSFTP SH PL MI Replications 2 100.59 0.73 25.5 27.46 27.59 18.73 0.009 3633.16 Entries 13 164.33 2.33 90.50** 44.24** 32.44* 27.90* 0.0972** 1527.25* Parents 6 228.96 1.96 118.66** 57.23** 13.93 17.62 0.180** 1821.93* Crosses 6 104.96 1.65 76.60** 33.28 14.67** 22.04 0.02** 729.3 Males 2 129.16 1.055 32.11** 7.16 40.05 18.84 0.003 194.66 Females 3 112.5 0.64 119.26** 47.52 78.7 8.36 0.04 712.94 Males x Females 6 200 3.55 0.01 8.26 93.38** 59.86* 0.008 1058 Error 26 187.13 1.27 23.46 15.23 55.07 11.62 0.005 631.16 *Significant at P ≤ 0.05, **Significant at P ≤ 0.01, %EMERSD: % of emergence 7DAS, NDFTPF: Number of days 50% plant flowering, WFTP: Pod weight, WSFTP: Seed Weight, NSFTP: Number of seed, SH: shelling %, PL: pod length, MI: Maturity index Table 5.11: General combining effect of parental lines for eight traits Genotypes MI %EMERSD NDFTPF WFTP WSFTP NSFTP SH PL Male JL24 -11.76 -0.36 0.48 -2.27 -0.8 0.1 2.18 -0.05 RRB 5.57 2.14 -0.52 -0.7 -1.07 -0.24 -1.44 0.03 55-437 3.4 -2.86 0.31 3.33 2.4 0.26 -0.01 0.01 Female 796 -4.1 -3.69 -0.02 1.11 0.34 3.26 -1.13 0.01 Chico 17.24 0.48 -0.19 -9.16 -6.43 0.43 -0.26 -0.09 ICIAR19BT 10.24 3.81 -0.52 4.63 2.95 0.43 -0.73 0.12 ICG3584 -14.76 -0.36 0.64 -1.16 -0.08 -3.9 1.99 -0.08 %EMERSD: % of emergence 7DAS, NDFTPF: Number of days 50% plant flowering, WFTP: Pod weight, WSFTP: Seed Weight, NSFTP: Number of seed, SH: shelling %, PL: pod length, MI: Maturity index Estimates of SCA effects of the seven crosses for eight traits measured are shown in Table 5.12. University of Ghana http://ugspace.ug.edu.gh 80 SCA ranged from -13.9 to 20.09 for maturity index; - 4.64 to 7.85 for percentage of emergence; - 0.97 to 0.52 for 50% plants flowering; -0.79 to 2.66 for pod weight; - 1.82 to 2.33 for seed weight; -3.92 to 4.07 for number of seed; -3.4 to 2.14 for shelling percentage and -0.05 to 0.04 for pod length. The best specific crosses for maturity index were 796 x JL24 and ICG3584 x 55- 437. Chico x RRB and ICIAR19BT x RRB were the specific crosses for 50% plant flowering, while 796 x RRB is the specific cross for pod and seed weight. 796 x JL24 and ICG3584 x 55- 437 were the best specific combinations for seed number. ICIAR19BT x 55-437 was the best specific cross for shelling percentage. Table 5.12: Specific combining ability effects of the crosses for eight traits CROSSES MI NDFTPF %EMERSD WFTP WSFTP NSFTP SH PL 796 x JL24 20.09 0.35 -3.8 0.3 -0.47 4.07 -2.09 0.04 ICG3584 x JL24 -1.23 -0.97 7.85 -0.25 0.21 -3.42 1.24 0.02 796 x RRB -13.9 -0.3 2.02 2.66 2.33 -3.92 1.36 -0.02 Chico x RRB -5.57 0.52 -2.14 0.7 1.07 0.23 1.44‚ -0.03 ICIAR19BT x RRB -3.9 0.52 -0.47 0.05 -0.27 -0.42 -0.68 0.01 ICIAR19BT x 55-437 -5.07 -0.3 1.19 -2.67 -1.05 0.4 2.14 -0.05 ICG3584 x 55-437 9.59 0.19 -4.64 -0.79 -1.82 3.07 -3.4 0.02 MI: Maturity index, %EMERSD: % of emergence 7DAS, NDFTPF: Number of days 50% plant flowering, WFTP: Pod weight, WSFTP: Seed Weight, NSFTP: Number of seed, SH: shelling %, PL: pod length. 5.3.1.5. Narrow sense heritability estimates Estimates from variance components Narrow sense estimates from variance components for different traits under study are presented in Table 5.13. Heritability ranged from 13.72% to 79.14%. The highest heritability estimates were obtained for 50% plant flowering (79.14%); shelling percentage (75.54); pod weight (72.04%); seed weight (70.04) and moderate estimates were found for maturity index and pod University of Ghana http://ugspace.ug.edu.gh 81 length respectively 42.57% and 40%. Seeds number and maturity index exhibited the lowest heritability estimate 13.72%. Table 5.13: Narrow sense heritability estimate from genetic component of traits Traits σ2M σ 2 F σ 2 MxF σ 2 Ph σ 2 A(F) σ 2 A(M) σ 2 D h 2 ns (%) %Emergence 7DAS -23.61 -29.17 4.29 179.53 -38.88 -23.61 -38.44 - 50% plant flowering -0.83 -0.97 0.76 1.63 1.29 -0.83 1.12 79.14 Pod weight 10.7 39.75 -7.82 45.81 33 10.68 5.12 72.04 Seed weight -0.37 13.09 -2.32 24.9 17.44 -0.36 0.85 70.04 Number of seed -17.78 -4.89 12.77 47.53 6.52 -17.76 80 13.72 Shelling % -13.67 -17.17 16.08 17.05 12.88 -13.67 7.96 75.54 Pod length 0 0.01 0 0.04 0.016 -0.0016 0.0035 40.00 Maturity index -287.78 -115.02 142.28 829.86 353.32 -287.76 133. 12 42.57 σ2M: Male genotypic variance, σ 2 F: Female genotypic variance, σ 2 MxF: Male x female genotypic variance, σ 2 Ph: Phenotypic variance, σ2A (F): Additive variance of Female, σ 2 A (M): Additive variance of Male, σ 2 D: Dominance variance, h 2 ns: Narrow sense heritability Estimates by parent-offspring regression (F3:F4) The estimates of narrow sense heritability by parent-offspring regression are presented in Table 5.14. Heritability estimates of the eight traits studied ranged from 35.4 to 95.5%. The highest estimates were found for days to emergence (95.5%), pod yield (85%), shelling percentage (84.3%), number of seed (80%) and 50% plant flowering respectively (72%). Moderate to low heritability estimates were obtained for Maturity index (66.1%), pod length (54.4%) and seed weight (35.4%). University of Ghana http://ugspace.ug.edu.gh 82 Table 5.14: Narrow sense heritability by parent-offspring regression and standard errors Traits h2ns (%) SE % Emergence 7DAS 95.5 0.29 50% Plants flowering 72 1.12 Pod weight 85 0.23 Seed weight 35.4 0.33 Seed number 80 0.22 Shelling % 84.3 0.29 Pod length 54.1 0.48 Maturity index 66.1 0.38 h2ns: Narrow sense heritability, SE: Standard error 5.3.1.6. Correlations Highly significant positive correlations were found between pod weight and seed weight (r = 0.97, P ≤ 0.001) and pod length (r = 0.72, P ≤ 0.01); between pod length and seed weight (r = 0.66, P ≤ 0.01); between maturity index and days to emergence (r = 0.72, P ≤ 0.001) (Table 5.15). Correlation between shelling percentage and seed weight was positively significant at P ≤ 0.05 (r = 0.61). However, seed number was found not to be correlated with all the other traits. But, 50% plants flowering showed non-significant and weak negative correlations with seed weight and pod length. Table 5.15: Correlation between earliness and agronomic traits %EMER7DAS NSFTP WSFTP SH PL 50%PF WFTP NSFTP 0.237 WSFTP -0.060 0.098 SH 0.297 -0.045 0.612* PL -0.037 0.242 0.664** 0.184 50%PF 0.017 0.215 -0.466 -0.271 -0.426 WFTP -0.145 0.120 0.975*** 0.423 0.722** -0.464 MI 0.772*** 0.185 0.231 0.325 0.164 -0.284 0.183 *Significant at P ≤ 0.05, **Significant at P ≤ 0.01, ***significant at <0.001, %EMERSD: % of emergence 7DAS, 50%TPF: Number of days 50% plant flowering, WFTP: Pod weight, WSFTP: Seed Weight, NSFTP: Number of seed, SH: shelling %, PL: pod length, MI: Maturity index University of Ghana http://ugspace.ug.edu.gh 83 5.3.2. Drought tolerance 5.3.2.1. ANOVA Mean squares from the ANOVA for physiological traits and pod yield and biomass are presented in Table 5.16. Results showed that the parents and F3 populations differed significantly (P ≤ 0.01) for all the physiological traits except for SPAD chlorophyll meter reading at 60DAS and 50% plants flowering. Table 5.16: Mean squares from ANOVA for parents and F3 populations for physiological traits Mean square Source of variation DF 50%PF SMCRSD SMCRED PY BIO HI Replication 2 0.02 16.67 5.84 144.36 793.19 0.004 Entry 12 1.26 7.3 8.07** 353.05* 1115.78** 0.026** Error 24 1.6 4.92 4.45 268.67 521.57 0.016 *Significant at P ≤ 0.05, **Significant at P ≤ 0.01, 50% PF: 50% Plant Flowering, SMCRSD: SPAD chlorophyll meter reading at 60DAS, SMCRED: SPAD chlorophyll meter reading at 80DAS, PY: Pod Yield, BIO: Biomass, HI: Harvest index Combined analysis of variance showed large and significant differences (P ≤ 0.01) between all 14 genotypes for all traits (Table 5.17) except for pod yield. The progenies showed high variation. Significant difference in water regimes for pod yield, total biomass and SCMR were also found (P ≤ 0.05 to P ≤ 0.001), but no significant differences for 50% plant flowering and harvest index (Table 5.17). The hybrids responded differently to both moisture regimes. This was shown by the significant interaction effect obtained between water regimes and crosses for chlorophyll content, pod yield, biomass and harvest index. Consequently the crosses exhibited different ranking from one water regime to another, for chlorophyll content, pod yield, biomass and harvest index. University of Ghana http://ugspace.ug.edu.gh 84 Table 5.17: Mean squares from the combined analysis of variance for the traits under well watered and end-of-season drought conditions Mean Square Source of variation DF 50%PF SMCR 60DAS SMCR 80DAS PY BIO HI Replication 2 33.03 14 4.18 22825 231878 0.06 Crosses 13 7.58* 23.77** 9.74* 5615 50235*** 0.05*** Water regime 1 2.68 828.17*** 54.53** 37855** 60380* 0.04 Crosses x water regime 13 2.22 11.91* 7.33* 4997* 18103* 0.03* Residual 54 3.4 9.77 4.99 4003 14855 0.01 Total 83 4.58 22.08 6.51 5270 26668 0.02 *Significant at P ≤ 0.05, **Significant at P ≤ 0.01, ***significant at <0.001, 50% PF: 50% Plant Flowering, SMCRSD: SPAD chlorophyll meter reading at 60DAS, SMCRED: SPAD chlorophyll meter reading at 80DAS, PY: Pod Yield, BIO: Biomass, HI: Harvest index 5.3.2.2. Mean performance The mean performance of parental lines and F3 populations for physiological traits is presented in Table 5.18. Among the males, 55-437 showed the highest values for pod yield (178.23g) and total biomass (554.53); whereas the genotype ICGV-SM 99511exhibited the highest values for the same traits among the females 180.43g and 637.6 respectively. The cross ICG11249 x RRB showed the highest values for pod yield (238.97g) and biomass (574.2g) among the crosses. SPAD chlorophyll meter reading at 80DAS values ranged from 34.49 - 39.71 and the highest valued was recorded with the cross ICGV-SM 99511 x 55-437 (39.71). SPAD chlorophyll meter reading at 80DAS values were greater than SPAD chlorophyll meter reading at 60DAS for all the entries. The highest harvest index was shown by the cross ICGV-SM 99511 x 55-437 (0.54). University of Ghana http://ugspace.ug.edu.gh 85 Table 5.18: Mean performance of parental lines and F3 populations for physiological traits Mean performance ENTRIES NDFPF SMCRSD SMCRED PY BIO HI JL24 30.33 30.95 37.83 167.13 356.73 0.47 RRB 31 33.73 39 127.73 380.98 0.31 55- 437 30 33.01 37.12 178.23 554.53 0.33 ICG11249 30.33 30.64 34.57 133.83 443.07 0.31 ICGVSM 99511 32 33.62 38.52 180.43 637.6 0.3 TAIMAN-9 31.33 34.43 37.91 136.72 507.03 0.27 ICG11249 X JL24 31 31.39 36.02 113.73 504.13 0.21 ICG11249 X RRB 31.33 32.87 34.49 238.97 574.2 0.42 ICG11249 X 55- 437 31.33 34.04 37.24 148.2 404.47 0.4 ICGVSM 99511 X 55- 437 30 34.95 39.71 193.83 362.8 0.54 TAINAN-9 x JL24 30.33 30.24 35.39 192.77 442.97 0.44 TAINAN-9 x RRB 30.33 34.25 37.99 104.47 366.07 0.29 TAINAN-9 x 55- 437 30 33.12 37.95 171.8 590.7 0.3 Mean 30.72 32.87 37.21 160.6 471.18 0.35 Range 30.00 - 32.00 30.24 - 34.95 34.49 - 39.71 104.47 - 238.97 356.73 - 637.60 0.21 - 0.54 LSD 2.13 3.74 3.55 102.67 171.47 0.217 CV % 4.12 6.75 5.66 37.93 21.59 36.76 50% PF: 50% Plant Flowering, SMCRSD: SPAD chlorophyll meter reading at 60DAS, SMCRED: SPAD chlorophyll meter reading at 80DAS, PY: Pod Yield, BIO: Biomass, HI: Harvest index Results of F3 and F4 populations mean performance for six traits under both water regimes are shown in Table 5.19. At 60 DAS, the crosses F3: Tainan-9 X 55-437, F3: ICG11249 X JL24, F4: Tainan-9 X JL24 exhibited the highest values of drought tolerance indices for SAPD SCMR respectively 1.35, 130 and 1.27. The lowest value of DTI was obtained with the hybrid Tainan-9 X RRB. The crosses ICG11249 X JL24 and Tainan-9 X 55-437 showed the highest values of DTI for SPAD SCMR at 80 DAS. High values of DTI for total biomass were recorded for the crosses ICGVSM 99511 X 55-437, ICG11249 X 55-437 and Tainan-9 X JL24. The lowest value of DTI for total biomass was found for the hybrid ICG11249 X JL24. Drought tolerance indices values ranged from 1.72 to 0.6 for harvest index. The crosses ICG11249 X RRB and ICGV- SM99511 X 55-437 showed the highest values of DTI for harvest index 1.72 and 1.41, University of Ghana http://ugspace.ug.edu.gh 86 respectively. The lowest value of DTI for harvest index was shown by the hybrid Tainan-9 X RRB. Table 5.19: Range, mean and drought tolerance indices of F3 and F4 populations under well- watered and end-of-season drought conditions for the six traits Groundnut crosses SCMR 60DAS DTI SCMR 80DAS DTI Biomass DTI Harvest index DTI WW WS WW WS WW WS WW WS F3: ICGVSM 99511 X 55-437 38.92 42.07 1.08 40.25 40.93 1.02 739.03 516.03 0.84 0.41 0.45 1.41 F3: ICG 11249 X 55-437 37.81 43.67 1.17 38.65 39.77 1.03 779.93 760.07 0.97 0.29 0.22 0.98 F3: ICG 11249 X JL24 34.99 45.14 1.3 36.57 37.32 1.02 582.7 487.7 0.86 0.47 0.28 0.76 F3: ICG 11249 X RRB 34.16 38.32 1.12 36.6 37.22 1.02 667.03 681.97 1.03 0.18 0.24 1.72 F3: Tainan-9 X 55-437 38.15 44.03 1.1 8 38.97 41.92 1.08 688.03 510.87 0.79 0.38 0.21 0.83 F3: Tainan-9 X JL24 34.53 40.33 1.17 37.5 37.55 1 547.27 561.37 1.05 0.36 0.38 0.74 F3: Tainan-9 X RRB 35.69 37.73 1.06 40.44 40.98 1.01 604.77 556.03 0.93 0.31 0.24 0.6 F4: ICGVSM 99511 X 55-437 35.61 40.27 1.14 40.41 39.17 0.97 492.77 623.8 1.24 0.34 0.29 0.89 F4: ICG 11249 X 55-437 35.81 41.25 1.15 36.19 38.72 1.07 488.9 572.1 1.17 0.41 0.26 0.67 F4: ICG 11249 X JL24 33.12 40.01 1.21 34.99 41.26 1.18 519.13 379.37 0.73 0.51 0.46 0.81 F4: ICG 11249 X RRB 32.47 39.22 1.22 38.26 39.35 1.03 771.83 588 0.76 0.21 0.24 1.55 F4: Tainan-9 X 55-437 35.3 47.71 1.35 38.27 40.96 1.07 571.13 416.97 0.75 0.33 0.31 0.93 F4: Tainan-9 X JL24 32.45 40.77 1.27 38.86 39.11 1.01 415.4 456.9 1.14 0.54 0.52 0.81 F4: Tainan-9 X RRB 34.74 41.15 1.18 39 40.9 1.05 547.2 553.27 0.99 0.34 0.36 0.83 Mean 35.27 41.55 1.2 38.04 39.65 1.04 601 547 0.95 0.36 0.32 0.97 Maximum 38.92 47.71 1.35 40.44 41.92 1.18 779.93 760.07 1.24 0.54 0.52 1.72 Minimum 32.45 37.73 1.06 34.99 37.22 0.97 415.4 379.37 0.73 0.18 0.21 0.6 LSD 5% 4.68 5.79 0.27 3.22 4.16 0.14 250.4 252.5 0.51 0.188 0.212 0.99 CV % 7.9 8.3 11.9 5.1 6.3 7.4 24.9 27.6 28.6 30.9 39.6 53.9 DTI: Drought tolerance index, 50% PF: 50% Plant Flowering, SMCRSD: SPAD chlorophyll meter reading at 60DAS, SMCRED: SPAD chlorophyll meter reading at 80DAS, PY: Pod Yield, BIO: Biomass, HI: Harvest index 5.3.2.3. Effect drought on SPAD Chlorophyll meter reading Highly significant differences in chlorophyll content among the crosses at 60 and 80 DAS was found (P ≤ 0.05 to P ≤ 0.01) (Table 5.17). The overall means of chlorophyll content under end- of-season drought stress condition, 41.55 and 39.65 respectively, at 60 and 80DAS were higher than the overall means under irrigated conditions, 35.27 and 38.04 respectively. The hybrids University of Ghana http://ugspace.ug.edu.gh 87 Tainan-9 X 55-437, ICG 11249 X JL24 and ICG 11249 X 55-437 exhibited the highest values of SPAD chlorophyll meter reading at 60 and 80 days after sowing under end-of-season drought stress conditions (Table 5.19). The hybrid Tainan-9 X 55-437 ranked first in terms of chlorophyll content at 60 and 80 DAS with 47.71 and 41.92, respectively under end-of-season drought stress conditions. While, under the same conditions, the cross ICG11249 X JL24 ranked second followed by the hybrid Tainan-9 X RRB. The magnitude of SPAD chlorophyll meter reading was higher at 60 and 80 DAS compared between both water regimes (Figure 5.3 and 5.4). Figure 5.3: Chlorophyll content at 60 DAS under well watered (WW) and water stressed (WS) conditions 0 5 10 15 20 25 30 35 40 45 50 SC M R 6 0D A S Populations WW WS University of Ghana http://ugspace.ug.edu.gh 88 Figure 5.4: Chlorophyll content at 80 DAS under well watered (WW) and water stressed (WS) conditions 5.3.2.4. Effect of drought on Pod yield Analysis of variance revealed highly significant differences in water regime among the crosses for pod yield (Table 5.17). The F3 cross ICGV-SM99511 X 55-437 was the highest yielding (270.37 g) under well water conditions (Table 5.20). However, once end-of-season drought stress was imposed, the F4 cross Tainan-9 X JL24 was the highest yielding cross at 240.53 g. The crosses ICG11249 X RRB and Tainan-9 X 55-437 showed the lowest pod yield 119.63 g and 111.23 g, respectively, under well watered and end-of-season drought stress conditions. Certain hybrids such as ICGV-SM99511 X 55-437, Tainan-9 X JL24, ICG11249 X RRB and Tainan-9 X RRB performed well for pod yield under both water regimes. The cross Taiman-9 X RRB showed the least pod yield difference among all the crosses. All crosses pod yield performances are shown in Figure 5.5. 30 32 34 36 38 40 42 S C M R 8 0D A S Populations WW WS University of Ghana http://ugspace.ug.edu.gh 89 The drought tolerance index for pod yield ranged from 0.44 to 1.37 for seven crosses. The crosses ICG11249 X RRB and Tainan-9 X JL24 displayed the highest value of DTI for pod yield 1.37 and 1.27, respectively. While the hybrids Tainan-9 X 55-437 and ICG11249 X JL24 exhibited the lowest values of pod yield DTI 0.44 and 0.5, respectively. Table 5.20: Pod yield (g) of F3 and F4 populations under well watered and their corresponding performance under end-of-season drought stress and DTI Groundnut crosses Pod yield WW Pod yield WS Yield difference DTI F3: ICGVSM 99511 X 55-437 270.37 221.07 49.3 0.85 F3: ICG 11249 X 55-437 227.7 145.3 82.4 0.63 F3: ICG 11249 X JL24 267.1 133.03 134.07 0.5 F3: ICG 11249 X RRB 119.63 161.27 41.64 1.37 F3: Tainan-9 X 55-437 250.6 111.23 139.37 0.44 F3: Tainan-9 X JL24 213.07 199.57 13.5 1.27 F3: Tainan-9 X RRB 183.93 125.6 58.33 0.73 F4: ICGVSM 99511 X 55-437 167.3 200.6 33.33 1.11 F4: ICG 11249 X 55-437 194.13 148.2 45.93 0.79 F4: ICG 11249 X JL24 258.7 165.43 93.27 0.66 F4: ICG 11249 X RRB 161.9 149.17 12.73 1 F4: Tainan-9 X 55-437 184.43 121.63 62.8 0.65 F4: Tainan-9 X JL24 219.27 240.53 21.6 1.12 F4: Tainan-9 X RRB 194.17 195.27 1.1 1.11 Mean 208 166 0.87 Maximum 270.37 240.53 1.37 Minimum 119.63 111.23 0.44 LSD 5% 110.9 117.8 0.88 CV % 31.9 42.5 52.9 WW: well watered conditions, WS: water stressed conditions, DTI: Drought tolerance index University of Ghana http://ugspace.ug.edu.gh 90 Figure 5.5: Mean performance of crosses for pod yield under well watered (WW) and water stressed (WS) conditions 5.3.2.5. Phenotypic and genotypic coefficient of variation estimates The estimate of phenotypic coefficient of variation (PCV) was greater than genotypic coefficient of variation (GCV) for all the physiological traits studied (Table 5.21). The traits harvest index, pod yield and biomass showed high PCV estimates of 40.20, 35.81 and 26.91% respectively. Moderate GCV estimates were obtained for the same traits 15.9, 11.02 and 14.12 respectively. The trait SPAD chlorophyll meter reading at both dates (60 or 80 DAS) were low for GCV and PCV estimates. 0 50 100 150 200 250 300 P od y ie ld ( g) Populations Pod yield under WW Pod yield under WS University of Ghana http://ugspace.ug.edu.gh 91 Table 5.21: Components of variation for physiological traits Characters Mean Range σ2G σ 2 Ph PCV (%) GCV (%) 50% plant flowering 30.72 30.00 - 32.00 -0.113 1.5 3.98 - SPAD SCMR 60DAS 32.87 30.24 - 34.95 0.79 5.71 7.27 2.71 SPAD SCMR 80DAS 37.21 34.49 - 39.71 1.2 5.66 6.39 2.95 Pod yield 160.6 104.47 - 238.97 28.12 296.8 35.81 11.02 Biomass 471.18 356.73 - 637.60 198.06 719.65 26.91 14.12 Harvest index 0.35 0.21 - 0.54 0.003 0.02 40.2 15.91 σ2G: Genotypic variance, σ 2 Ph: Phenotypic variance, PCV: Phenotypic coefficient of variation, GCV: Genotypic coefficient of variation 5.3.2.6. General and specific combining ability estimates Analysis of variance of combining ability for the six traits revealed significant differences among parents for SPAD chlorophyll meter reading, pod yield, total biomass and harvest index (Table 5.22). All the crosses were highly significantly different for all the physiological traits, except for 50% plants flowering and chlorophyll content at 60DAS. GCA mean squares for males and females were highly significant for pod yield and biomass. GCA mean squares for females were greater than males GCA mean square for pod yield and total biomass indicating that the major contribution to additive variance for these traits was due the female parents. SCA mean squares differed significantly for pod yield and harvest index indicating the importance of both additive and non-additive variance for these traits. University of Ghana http://ugspace.ug.edu.gh 92 Table 5. 22: Mean square of Combining Ability analysis for six the physiological traits for WW Mean Square Source of variation DF 50%PF SCMRSD SCMRED PY BIO HI Replications 2 0.025 4.92 5.84 144.36 793.19 0.0043 Entries 12 1.26 7.3 8.07** 353.05* 1115.78** 0.026** Parents 5 1.7 7.38 7.36* 271.29** 1601* 0.0152* Crosses 6 1.04 8.34 9.57** 453.76* 820.76* 0.0367** Males (GCAm) 2 0.06 15.16 5.7 45.93* 675.96* 0.0012 Females (GCAf) 2 2.69 2.04 7.69 236.49* 1151.83* 0.0351 Males x Females (SCA) 2 0.17 2.96 6.64 1122.51* 1047.6 0.0603* Error 24 1.69 4.92 4.45 268.67 521.57 0.0167 *Significant at P ≤ 0.05, **Significant at P ≤ 0.01, 50% PF: 50% Plant Flowering, SCMRSD: SPAD chlorophyll meter reading at 60DAS, SCMRED: SPAD chlorophyll meter reading at 80DAS, PY: Pod Yield, BIO: Biomass, HI: Harvest index Estimates of combining ability effects of parents for traits measured are presented in Table 5.23. Among the males, RRB showed highest GCA effect for 50% plant flowering. The males RRB and 55-437 exhibited positive GCA effect for pod yield and total biomass respectively. The female ICGV-SM99511 displayed the highest positive GCA effect for SPAD chlorophyll meter reading and pod yield. While Tainan-9 showed the highest GCA effect for biomass and harvest index. Table 5.23: General combining ability effect of parental lines for six traits for WW General Combining Ability Parents 50%PF SCMRSD SCMRED PY BIO HI Male JL24 0.047 -2.16 -1.27 -0.67 -7.9 -0.04 RRB 0.21 0.58 -0.74 0.54 3.3 -0.01 55-437 -0.17 1.06 1.33 0.084 3.07 0.04 Female ICG11249 0.6 -0.21 -1.05 -3.36 -1.91 -0.04 ICGV-SM 99511 -0.62 1.97 2.74 9.73 -17.41 -0.02 TAINAN-9 -0.39 -0.44 0.14 0.12 7.72 0.042 50% PF: 50% Plant Flowering, SCMRSD: SPAD chlorophyll meter reading at 60DAS, SCMRED: SPAD chlorophyll meter reading at 80DAS, PY: Pod Yield, BIO: Total Biomass, HI: Harvest index University of Ghana http://ugspace.ug.edu.gh 93 Estimates of SCA effects of the seven crosses for six traits are shown in Table 5.24. SCA ranged from -0.27 to 0.28 for 50% plant flowering; from -0.47 to 1.13; -0.49 to 1.61 for SPAD chlorophyll meter readings at 60 and 80 DAS respectively; from -9.04 to 17.04 for pod yield; form -13.39 to 22.17 for total biomass and from -0.08 to 0.14 for harvest index. ICG11249 x 55- 437 was the best specific cross for 50% plant flowering; while Tainan-9 x RRB was the best specific combination for SPAD chlorophyll meter reading. Tainan-9 x JL24 and ICG11249 x RRB were the best specific crosses for pod yield, while the cross Tainan-9 x 55-437 was the best specific combination for biomass. Table 5.24: Specific combining ability effects of the seven crosses for six traits for WW Specific Combining Ability Crosses 50%PF SCMRSD SCMRED PY BIO HI ICG 11249 x JL24 -0.27 0.79 1.36 -9.04 1.44 -0.08 Tainan-9 x JL24 0.06 -0.13 -0.45 12.28 -7.24 0.14 ICG 11249 x RRB -0.1 -0.47 -0.69 17.04 7.59 0.09 Tainan-9 x RRB -0.1 1.13 1.61 -13.8 -13.39 -0.03 ICG 11249 x 55-437 0.28 0.21 -0.007 -7.96 -7.5 0.01 ICGV-SM99511 x 55-437 0.17 -1.05 -1.33 -0.08 -3.06 -0.04 Tainan-9 x 55-437 -0.04 -0.47 -0.48 1.56 22.17 -0.08 50% PF: 50% Plant Flowering, SCMRSD: SPAD chlorophyll meter reading at 60DAS, SCMRED: SPAD chlorophyll meter reading at 80DAS, PY: Pod Yield, BIO: Biomass, HI: Harvest index 5.3.2.7. Narrow sense heritability estimates Estimates from variance components Narrow sense estimate from variance components for physiological traits and pod yield and biomass ranged from 20.33 to 74.44% (Table 5.25). The highest value was obtained for 50% plant flowering (74.44%); while moderate heritability values were found for pod yield (65.28%), harvest index (50%) and SPAD chlorophyll meter reading (43.46%). Biomass exhibited the lowest heritability value (20.33%). University of Ghana http://ugspace.ug.edu.gh 94 Table 5.25: Narrow sense heritability estimate from genetic component of physiological traits for WW Characters σ2M σ 2 F σ 2 MxF σ 2 Ph σ 2 A(F) σ 2 A(M) σ 2 D h 2 ns (%) 50% plant flowering -0.04 0.84 -0.48 1.5 1.12 -0.049 -0.087 74.44 SPAD SCMR 60DAS 4.07 -0.32 -0.65 5.71 -0.42 5.42 0.28 - SPAD SCMR 80DAS -0.32 0.35 0.73 5.66 2.46 -0.42 0.74 43.46 Pod yield - 358.86 -295.34 284.61 296.8 193.76 -478.48 12.04 65.28 Biomass -123.88 34.74 175.34 719.65 146.32 -165.16 138.2 20.33 Harvest index -0.02 -0.01 0.01 0.02 0.01 -0.02 0.0076 50 σ2M: Male genotypic variance, σ 2 F: Female genotypic variance, σ 2 MxF: Male x female genotypic variance, σ 2 Ph: Phenotypic variance, σ2A (F): Additive variance of Female, σ 2 A (M): Additive variance of Male, σ 2 D: Dominance variance, h 2 ns: Narrow sense heritability Estimates by parent-offspring regression In this study, heritability estimates for physiological traits were higher than for agronomic traits, and varied under both well-watered and end-of-season conditions (Table 5.26). The heritability estimates for pod yield (0.26) and biomass (0.18) were low, but they were high for harvest index (0.78) and SPAD chlorophyll meter reading 0.71 and 0.68 respectively at 60 and 80 days after sowing. Fifty percent (50%) Plants flowering showed moderate heritability estimate of 0.56. End -of-season drought stress effect on heritability estimates decreased narrow sense heritability values for all the physiological traits. Interestingly increased heritability estimates for agronomic traits, pod yield and biomass, under end-of-season drought stress conditions. As a result of drought effect, the heritability estimates varied from low 0.26 to moderate heritability 0.68 for pod yield and 0.18 to 0.49 for biomass. University of Ghana http://ugspace.ug.edu.gh 95 Table 5.26: Narrow sense heritability by parent-offspring regression and standard errors of the seven groundnut crosses for physiological traits Traits Well water condition Drought stress condition h2ns (%) SE h 2 ns (%) SE 50% Pant flowering 56.3 0.2 28.2 0.32 SCMR 60 DAS 71.6 0.12 36.8 0.4 SCMR 80 DAS 68.2 0.41 10.6 0.234 Pod Yield 26.1 0.25 68.6 0.32 Biomass 18.4 0.58 49.5 0.35 Harvest index 78.9 0.4 37.8 0.48 h2ns: Narrow sense heritability, SE: Standard error 5.3.2.8. Correlations Pod yield displayed highly significant positive association with harvest index in both water regimes; the correlation coefficient was higher (r = 0.77, P ≤ 0.001) under end-of-season drought stress conditions than under well watered conditions (r = 0.74, P ≤ 0.01). Strong negative significant (P ≤ 0.05) correlation was found between biomass and harvest index under well watered and end-of-season drought conditions r = -0.66, -0.58, respectively (Table 5.27). Weak positive and non-significant correlation was found between SCMR and pod yield under well watered conditions (r = 0.38). University of Ghana http://ugspace.ug.edu.gh 96 Table 5.27: Correlations between physiological and agronomic traits under both water regimes Under well water conditions SCMR60DAS SCMR 80DAS Biomass Harvest Index SCMR 80DAS 0.434 Biomass 0.459 0.202 Harvest Index -0.064 -0.204 -0.606* Pod Yield 0.387 -0.079 0.029 0.745** Under end of-season drought conditions SCMR60DAS SCMR80DAS Biomass Harvest Index SCMR 80DAS 0.190 Biomass -0.287 -0.373 Harvest Index -0.077 0.103 -0.589* Pod Yield -0.340 -0.186 -0.036 0.777*** *Significant at P ≤ 0.05, **Significant at P ≤ 0.01, ***significant at <0.001, SCMR: SPAD chlorophyll meter reading, PY: Pod Yield, BIO: Biomass, HI: Harvest index University of Ghana http://ugspace.ug.edu.gh 97 5.4. Discussion 5.4.1. Maturity Four hybrids, out of the seven, 796 x RRB, Chico x RRB, ICIAR19BT x RRB and ICIAR19BT x 55-437, showed good performance for both earliness and agronomic traits. The hybrids ICIAR19BT x RRB, ICIAR 19BT x 55-437, 796 x RRB, 796 x JL24, Chico x RRB exhibited the highest means for pod yield, seed weight and maturity index in both populations. High values of maturity index indicated early maturity. Therefore these hybrids were promising early maturing materials. The estimates of phenotypic coefficient of variation were higher than genotypic coefficient of variation indicating the influence of environmental factors. None of the traits recorded high PCV and GCV. Pod yield, seed weight and maturity index exhibited moderate PCV estimates and low GCV estimates. 50% plants flowering, shelling percentage and pod length, low magnitude of difference observed indicated that these characters were less influenced by the environments (Narasimhulu et al., 2012). Similar results were reported for low PCV and GCV for 50% flowering by Vetriventhan and Nirmalakumari (2007) and Sumathi et al., (2010). Genotypic coefficient of variation measures the amount of variation present in a particular character (Narasimhulu et al., 2012) while PCV measures total relative variation (Roychowdhury and Tah, 2011). Several authors reported comparable results where PCV values were higher than GCV values for all the traits which reflect the influence of environment on the expression of traits (Yousaf et al. 2008; Sumathi et al. 2010; Roychowdhury et al. 2011 and Narasimhulu et al. 2012). Analysis of variance of combining ability for the eight traits revealed significant differences (P ≤ 0.05 to P ≤ 0.01) among parents for all the traits except for percentage of emergence and 50% University of Ghana http://ugspace.ug.edu.gh 98 plants flowering. This indicated the presence of genotypic variability among cultivars used as parents. The significance of GCA mean squares indicated the importance of additive gene effect governing the inheritance of pod yield, seed yield, pod length, and maturity index. Selection of superior genotypes in segregating generations should be possible for these traits. These results are in agreement with those of Jogloy et al. (2005) who found significant GCA mean squares which indicated the importance of additive gene effects governing the inheritance of pod yield, seed yield and 100-seed weight. Jogloy et al. (1999) found that GCA mean squares were important for pod length and seed size. Green et al. (1983) and Swe and Branch (1986) found that GCA and SCA mean squares were significant for yield and yield component traits. Similar results were also reported. Holbrook (1990) found that GCA and SCA were also significant for yield and yield components. The analysis revealed differences among the crosses for almost all the traits. This indicated the presence of genotypic variability among F3 populations. F3 and F4 populations differed significantly for pod yield, shelling percentage, seed weight and pod length. GCA mean squares for females were greater than for the males for pod weight indicating that the major contribution to additive effect for this trait was due the female parent. SCA mean squares were significant for number of seed and shelling percentage at P ≤ 0.01and P ≤ 0.05, respectively, indicating the importance of both additive and non-additive effects for these traits. These results support the findings of Mekonthehou, (1987) who studied the combining ability of early maturity for a selected group of groundnut and found that GCA was always larger than SCA for most of the maturity parameters. Jogloy et al. (1987) found that general combining ability was highly significant for pod yield, and shelling percentage and specific combining ability was significant for pod length and seed size. Rachmeler, (1988) found greater and significant GCA than SCA University of Ghana http://ugspace.ug.edu.gh 99 estimated for early maturity in F2 and F3 generations. Bansal et al., (1991) showed that non- additive gene effects were predominant for yield components although the magnitude of additive effects was considerable. GCA effects clearly identified the variety 55-437 as the best male parent with good combining ability for most traits under study. RRB was the best combiner for maturity index and days to emerge. While the variety JL24 was the best general combiner for shelling percentage. GCA effects revealed that the variety ICIAR19BT as the best female parent with general good combining ability for most traits under study. Chico was the best general combiner for maturity index with the highest positive GCA effect. Ali et al. (2001) reported Chico to be the best combiner for maturity index. Several authors reported the use of Chico as parental line in breeding programme for developing early maturing varieties such as N'Doye and Smith (1993), Ali and Wynne (1994); Upadhyaya and Nigam (1994) and Upadhyaya et al. (2006). SCA effects for individual hybrids identified at least one specific hybrid for each trait. The hybrids 796 x JL24 and ICG 3584 x 55-437 were the good specific combiners for maturity index. The hybrid 796 x RRB was the good for pod and seed weight. ICIAR19BT x 55-437 was the best specific combination for shelling percentage. Hybrids 796 x JL24 and ICG 3584 x 55-437 were the good specific combinations for number of seeds. Specific combinations are expected to give transgresive segregation in later generations. Estimates of SCA effects were larger in magnitude than GCA effects for 50% plant flowering, number of seed and shelling percentage indicating that non-additive genetic variance was more important than additive genetic variance for these traits. Redona and Lantican, (1985) reported that the non-additive genetic variation due to dominance interactions should reduce in comparison with the additive genetic variation as inbreeding progresses. University of Ghana http://ugspace.ug.edu.gh 100 High to moderate heritability was obtained for days to emergence, pod yield, shelling %, number of seed, 50% plants flowering and maturity index from the estimates from variance components and by parent-offspring regression. The lowest heritability value was obtained with the traits seed weight and seed number. Because of higher heritabilities, selection for yield and maturity should be possible among the populations. The results agree with the findings of Coffelt and Hammons (1974), Wynne et al., (1975), Kale and Dhoble (1988), Tsaur et al. (1989) who reported high heritability for pod yield, pod length and 100 seed weight. Results also agree with the studies of Rachmeler, (1988) and Mohammed et al., (1978) who found relatively high narrow sense heritability for early maturity in groundnut. Holbrook et al., (1989) reported also high heritabilities for maturity while studying F1 and F2 plants of reciprocal crosses involving Chico (extra early maturing) and PI383421(late maturing) groundnut lines. Ali and Wynne, (1994) similarly found high narrow sense heritability estimates for the cross NC7/Chico for 100 seed weight (71%) and pod length (52%), intermediate for maturity index (44%). John et al., (2011), in F6 generation found moderate broad sense heritability estimates for pod yield high heritability for pod yield (76%) and 100 seed weight (71%) and intermediate estimates for maturity index (44%) and seed number (31%) depending on the crosses. This study results are in agreement with the findings of Khan et al., (2000) and Jogloy et al., (2011). Burton (1951) suggested that the genetic coefficient of variation together with heritability estimates gave the better picture of the extent of heritable variation. Positive associations between pod yield with seed weight and pod length; between maturity index and days to emergence; and between pod length with seed weight and shelling percentage indicated the possibility for simultaneous improvement of these traits. Waranyuwat and Tongsri (1990) reported highly significant associations between pod and seed yield, pod yield and University of Ghana http://ugspace.ug.edu.gh 101 number of mature seeds per plant, and seed yield and number of mature seeds per plant, but shelling percentage had variable correlation with pod and seed traits in different generations. They also suggested selection based on high number of pods per plant for high yield. These results are in good agreement with those found by Khan et al., (2000) and Saleh and Masiron (1994) on groundnut lines evaluations in Malaysia. 5.4.2. Drought Tolerance The hybrid ICG 11249 x RRB showed the highest values for pod yield and total biomass under drought followed by ICGV-SM 99511 x 55-437. SPAD chlorophyll meter readings values at 80DAS were greater than SPAD chlorophyll meter reading values at 60DAS for all crosses. The highest harvest index was found with the hybrid ICGV-SM 99511 x 55-437 (0.54), while the highest value of leaf chlorophyll content was recorded with ICGV-SM 99511 x 55-437 (39.71). Combined analysis of variance under well water and end of-season drought conditions showed large and significant differences between all 14 genotypes for all the physiological traits except for pod yield. Therefore, the tested progenies displayed high variation. Hybrids ICG11249 x JL24, ICG11249 x RRB, Tainan-9 X 55-437 and ICGVSM 99511 x 55-437 showed the highest values of drought tolerance index for SCMR, biomass and harvest index. Therefore, these hybrids are promising drought tolerant progenies. Nageswara et al. (1994) reported that the productivity of drought resistant groundnut lines under drought-stress conditions, as measured by total biomass, were higher than those of drought-sensitive genotypes. Thus, with differential responses to drought stress, high biomass production under drought stress of a tolerant genotype could be due to its ability to produce high biomass under well-watered conditions. University of Ghana http://ugspace.ug.edu.gh 102 In the present study, pod yield, biomass, and harvest index decreased under end-of-season drought stress, whereas SCMR increased as reported in previous studies under various environments (Nageswara and Wright 1994; Craufurd et al., 1999; Nigam et al., 2008; Songsri et al., 2009; Girdthai et al., 2010;). However, DTI is useful in explaining how some genotypes had higher pod yield under drought. Previous reports on inheritance of drought resistance traits suggested a predominant role of additive gene effects in SLA and HI inheritance (Nigam et al., 2001; Surihan et al., 2005). In early generations (F3 and F4), Cruickshank et al. (2004) reported that broad sense heritability of transpiration efficiency (TE) and harvest index (HI) were varied among groundnut crosses and traits depending on levels of genetic variation in parents. The results are also in agreement with Moreshat et al. (1996) who found increasing chlorophyll content under mild drought stress. Jongrungklang et al. (2008) also found that the more severe the drought stresses the more was the increase in SCMR. The same authors, reported relatedness between increase of SCMR and photosynthetic capacity. They concluded that the increase of SCMR could be attributed to drought tolerance. Painawadee et al. (2009) reported that the water loss from cells might affect the concentration of chlorophyll content. Duy Nang, (2004) also found SCMR increased in rice during stress periods but declined rapidly within three days after re-irrigation. It could be hypothesized that groundnut genotypes with high SCMR have more photosynthetic machinery per unit leaf area and hence potential for greater assimilation under drought stress (Songsri et al., 2009). Boontang et al. (2010) reported that groundnut genotypes with high SCMR under drought stress could maintain higher biomass production. The estimates of phenotypic coefficient of variation were higher than genotypic coefficient of variation indicating the influence of environmental factors. Pod yield, biomass and harvest index showed highest PCV estimates respectively 35.81, 26.91 and 40.20 and moderate GCV values University of Ghana http://ugspace.ug.edu.gh 103 for the same characters. The low magnitude of difference between PCV and GCV observed with SCMR indicated that this character was less influenced by the environment (Narasimhulu et al., 2012). High values of GCV suggest better improvement scope for these traits by selection. A number of authors reported comparable results where PCV values were higher than GCV values for all the traits which reflect the influence of environment on the expression of characters Roychowdhury and Tah (2011); Thirumala et al. (2012); Azharudheen et al. (2013) and Reddy et al. (2013). Analysis of variance of combining ability for the physiological characters revealed significant differences (P ≤ 0.05 to P ≤ 0.01) among parents for all the traits except for percentage of emergence and 50% plant flowering. This indicated the presence of genotypic variability among cultivars used as parents. The significance of GCA mean squares indicated the importance of additive gene effect governing the inheritance of pod yield, biomass, harvest index and SCMR. The analysis also revealed differences among the crosses for almost all the characters. This indicated the presence of genotypic variability among F3. GCA mean squares for males and females were highly significant for pod yield and biomass. GCA mean squares for females were greater than males GCA mean square for pod yield and biomass indicating that the major contribution to additive variance for these traits was due to the female parents. SCA mean squares differed significantly for pod yield and harvest index indicating the importance of both additive and non-additive variance for these characters. The findings are comparable to Swe and Branch (1986) who reported estimates of general and specific combining abilities to be significant for biomass, harvest index, total pod weight. John et al., (2011) indicated primarily non-additive gene action was important in the inheritance SCMR. Venkateswarlu et al. (2007) University of Ghana http://ugspace.ug.edu.gh 104 reported that high magnitude of general combining ability variance showed greater importance than additive gene action in the inheritance of physiological and pod yield traits. GCA effects identified the variety 55-437 as the best male parent with good combining ability for all physiological characters under study. The variety RRB was the best combiner for biomass and pod yield. The female ICGV-SM 99511 was the best general combiner for SPAD chlorophyll meter reading and pod yield, while Tainan-9 was the best combiner for biomass and harvest index. The hybrid Tainan-9 x JL24 was the best specific combination for pod yield and harvest index. The hybrid Tainan-9 x RRB was the best specific combination for SCMR. ICG 11249 x RRB and Tainan-9 x 55-437 were the best specific combination for total biomass production. GCA Mean squares for females were greater than males for pod yield and biomass indicating the major contribution to additive variance for this trait was by the female parents. Similar results were found by Ali et al. (1995) and Ali et al. (2001). The study showed that estimates of SCA effects were larger in magnitude than GCA effects for pod yield and biomass indicating that non-additive genetic variance was more important than additive genetic variance. These results are in agreement with John et al., 2011 who found that the variance due to specific combining ability was greater than the variance due to general combining ability for all the physiological traits and concluded the predominance of non- additive gene action for these traits. Mechanisms of drought resistance are diverse under different timing and periods of drought conditions (Clavel et al., 2004; Subbarao et al., 1995). Therefore, the inheritance of end-of- season drought tolerance traits might be different from other types of drought; high and moderate narrow sense heritability estimates were found for the physiological traits under study through University of Ghana http://ugspace.ug.edu.gh 105 both methods of estimation. Nonetheless, parent-offspring regression gave better estimates of heritability compare to variance components estimates. Heritability estimates were higher by parent-offspring regression for physiological traits (SCMR, HI) than for agronomic traits (pod yield, Biomass) under well watered conditions. However, heritability estimates decreased for all physiological traits when compared between different water levels; and increased for agronomic traits. Therefore, selection for higher yield or higher biomass production under drought stress conditions should be possible. More rapid breeding progress may be achieved by using physiological or surrogate traits, such as HI, SLA, and SCMR, which may be used to select for drought tolerance (Wright et al., 1994; Sheshshayee et al., 2006; Arunyanark et al., 2009). The results agree with Songsri et al., (2008) who found high heritability estimates for physiological traits. End-of-season drought stress decreased heritability estimates for all the physiological traits except for pod yield and biomass. These findings are similar with Girdthai et al. (2012) who reported increase of broad sense heritability for pod yield and biomass under drought stress conditions. Boontang et al. (2010) also reported the maintenance of high chlorophyll under drought stress would benefit groundnuts. Significant and positive correlation between pod yield and harvest index under both water regimes indicated it will be possible to identify relative performance of groundnut lines for drought tolerance. Harvest index had highly positive direct effects on pod yield. Hence, improvement of this character would also improve pod yield and direct selection of pod yield using this trait will be effective. Negative significant association was also found between biomass and harvest index. Wunna et al. (2009) reported moderate association between HI and biomass under early drought and irrigated conditions. These results support the findings of Ravi University of Ghana http://ugspace.ug.edu.gh 106 Kumar et al. (2012) for SCMR, harvest index. Ahamed (1995), Abdalla et al. (2008) and Jayalakshmi et al. (2000) have reported similar results for harvest index. 5.5. Conclusions Additive gene effects were found to govern the inheritance of pod yield, seed weight, pod length, maturity index. Moderate narrow sense heritability of seed weight and pod length coupled with their positive association with pod yield, indicated that these traits would be good criteria for selection programmes to improve yield of groundnut. Even though maturity index, giving a high heritability, it was not associated with the yield traits. However maturity index could be a good criterion for earliness selection because it showed a strong positive correlation with days to emergence. The variety 55-437 was identified as the best male parent for pod yield, seed weight, and number of seed. RRB was the best male parent for maturity index and days to emergence. ICIAR19BT was the best female parent for days to emergence, pod yield, seed weight, pod length and Chico was the best female parent for maturity index. Most physiological traits measured in these seven groundnut crosses had high heritability, indicating that breeding progress should be possible. The results of the present study indicated that harvest index, and SPAD chlorophyll meter reading observations can be recorded at both stressed and nonstressed conditions. This gives groundnut breeders flexibility to record these observations in a large number of segregating populations and breeding lines, thus making it easy to incorporate these physiological traits associated with drought tolerance in breeding and University of Ghana http://ugspace.ug.edu.gh 107 selection schemes in groundnut. SPAD chlorophyll meter reading should be particularly useful as a selection criterion for drought tolerance in groundnut because of high heritability and the simplicity in gathering data. Groundnut lines with the ability to maintain high chlorophyll content and high biomass under drought conditions may also show better tolerance to drought. High heritability estimate for harvest index coupled with strong positive association with pod yield under both water regimes indicated the harvest index should also be considered as a selection criterion to ensure progress for pod yield. The variety 55-437 was identified as the best male parent with good general combining ability for all physiological characters under study. The female ICGV-SM 99511 was the best female parent with good general combining ability for SPAD chlorophyll meter reading and pod yield, while Tainan-9 was the best combiner for biomass and harvest index. University of Ghana http://ugspace.ug.edu.gh 108 CHAPTER SIX 6.0. GENERAL DISCUSSION 6.1. Identification of the production constraints and farmers’ preferred varieties of groundnut in Niger Farmers in Niger recognized the importance of drought as an important constraint to groundnut productivity and yields. Farmers were growing three varieties, namely 55-437, RRB and JL24, because of their adaptation to production systems. Some of the farmers were aware that they could get other types of groundnut varieties from research institutions. The farms were all family owned and the fields were small (≤ 2ha), resulting in low production levels. Groundnut is allocated a very small portion of the farm or generally intercropped with other crops, which is the most common cropping system in the Sahel. The farmers felt that producing groundnut on a larger scale would be “risky” due the frequent drought spells and rosette disease infestation. The study revealed that the farmers were ready to accept new technologies of groundnut that can mitigate the changing environment. PRA results showed that drought and lack of improved groundnut varieties caused many farmers to abandon groundnut production and shift to other crops such cowpea and sesame. The low productivity affected women farmers more than men, because they are highly engaged in groundnut production and are not able to contribute consistently to the welfare of the household without groundnut income. In the groundnut "basin" that includes Dosso, Maradi and Zinder regions, groundnut was the major source of income. Recurrent drought has instigated a significant decrease in farmers' incomes. Groundnut production decline affected farmers' livelihood and reduced social celebrations. University of Ghana http://ugspace.ug.edu.gh 109 Involving farmers in the developing breeding programme objectives is very important because it will (i) improve relationships between breeders and farmers, (ii) guide breeders to focus on priority constraints of farmers, and (iii) allow breeders to identify farmers' criteria for adopting new varieties. Farmers' participation in the selection process can change the classical breeder's approach. 6.2. Phenotyping groundnut germplasm for earliness and drought tolerance Early maturity: PCA analysis revealed that 79.62% of the total variation among the genotypes was explained by pod yield, biomass, harvest index and 50% plants flowering. The genotypes evaluated under well water condition were classified into three groups based on their maturity: (i) extra early maturing genotypes (75 to 80 days), (ii) early maturing genotypes (85 days) and (iii) intermediate genotypes (90 110 days). Out of the 150 genotypes, 2 extra early (75 days) and 3 early (80 days) genotypes were identified. ICIAR19BT had the highest pod yield 1540g/15plants. ICIAR19BT was released in Nigeria in 2011 as an extra early and rosette resistant variety under the name SAMNUT 24. Chico had the lowest pod yield 430 g/15plants. It is the most widely used source of earliness in several breeding programmes. It was used 1180 times as a parent in developing early maturity lines at ICRISAT (Upadhyaya et al., 2006). The selected early and extra early lines were crossed onto three farmers preferred varieties (55-437, JL24, RRB) to generate 7 F3 and 7 F4 populations for genetic studies. Drought tolerance: Genotypes were evaluated under both water regimes to select the best drought tolerant lines. Farmers were involved in selecting these drought tolerant lines. Highly significant differences were found between the genotypes for all the traits and water regimes.  x University of Ghana http://ugspace.ug.edu.gh 110 ICGV-SM99511 was the highest yielding (524 g/10plants) under well water conditions. ICG6307 was the best yielding (342 g) under end-of-season drought stress. Tainan-9, ICG11249 and ICGV-IS 01820 performed well under both well watered and end-of-season drought conditions. The five best drought tolerant genotypes were ICG6703, ICGV-SM99551, Tainan-9, ICG11249 and ICGV-IS01820. Their drought tolerance indices ranged from 0.59 to 1.76. ICG6703 had the highest DTI (1.76) and ICGV- SM 99511 the lowest DTI (0.59). Higher DTI values indicate drought tolerance (Nautiyal et al., 2002). In this study, the responses to drought among the groundnut genotypes for SCMR at 45, 60 and 90 DAS were rather similar in patterns but somewhat different in extents at 90DAS. Increase of chlorophyll contents under drought stress condition was observed. Genotypes that showed high values of SCMRs under drought stress may be considered as drought tolerant. The findings are in agreement with Jongrungklang et al.(2008) who found an increase of SCMR of groundnut under water limiting conditions and as this trait is related to photosynthetic capacity, they concluded that the increase of SCMR could be attributed to drought tolerance. It could be hypothesized that groundnut genotypes with high SCMR have more photosynthetic machinery per unit leaf area and hence more potential for greater assimilation under drought stress (Songsri et al., 2009). Boontang et al., (2010) found that groundnut genotypes with high SCMR under drought stress could maintain higher biomass production. The selected drought tolerant lines were crossed onto farmers preferred varieties to develop 7 F3 and 7 F4 populations for genetic studies. Some varieties have been developed at Tarna INRAN station. They are early maturing and performed well under end-of-season drought stress. Unfortunately they have not reached the farmers and seed companies yet. Six varieties out of the 10 lines selected in this study, were used in a farmer's participatory varietal selection in three locations: Angoual Doua, Dan Saga and University of Ghana http://ugspace.ug.edu.gh 111 Elguéza. Farmers were impressed by the performance of ICIAR19BT, Tainan-9 and ICGV- SM99511based on their earliness, drought tolerance and high yielding potential. Selection of a preferred variety is only part of the story. For these varieties to reach a large number of farmers, seed must be available to those who want to grow them. This must involve various actors to ensure that breeder seed is available to seed companies to produce subsequent classes of seed. 6.3. Genetic analysis of earliness and drought tolerance Early maturity: Four hybrids, out of the seven, showed good performance for both earliness and agronomic traits. Pod yield ranged from 34.40 to 46.07g/50pods and maturity indices ranged from 203.33 to 240.33 for the hybrids ICIAR19BT x RRB, ICIAR 19BT x 55-437, 796 x RRB, 796 x JL24, and Chico x RRB in both F3 and F4 populations. Significant GCA mean squares indicated the importance of additive gene effect governing the inheritance pod yield, seed yield, pod length and maturity index. Selection of superior genotypes in segregating generations should be possible for these traits. These results are in agreement with those of Jogloy et al., (2005) who found significant GCA mean squares for the inheritance of pod yield, seed yield and 100-seed weight. Jogloy et al. (1999) found that GCA mean squares were important for pod length and seed size. Green et al. (1983) and Swe and Branch (1986) found that GCA and SCA mean squares were significant for yield and yield component traits. Specific combining ability (SCA) mean squares differed significantly (at P ≤ 0.01and P ≤0.05) for number of seed and shelling percentage indicating the importance of both additive and non- additive variance for these traits. These results support the findings of Jogloy et al. (1987) who University of Ghana http://ugspace.ug.edu.gh 112 found that general combining ability was highly significant for pod yield, and shelling percentage and specific combining ability was significant for pod length and seed size. Rachmeler, (1988) found greater GCA than SCA estimates for early maturity in F2 and F3 generations. Bansal et al., (1991) showed that non additive gene effects were predominant for yield components although the magnitude of additive effects was considerable. The variety 55-437 had the best general combining ability for most traits under study. RRB was the best general combiner for maturity index and days to emergence. JL24 was the best general combiner for shelling percentage. ICIAR19BT was the best general combiner for days to emergence, pod yield, seed weight, pod length. Chico was the best general combiner for maturity index corroborating the findings of Ali et al. (2001). Several authors reported the use of Chico as parental line in breeding programmes for developing early maturing varieties such as N'Doye and Smith (1993), Upadhyaya and Nigam (1994) and Upadhyaya et al. (2006). Estimates of SCA were lager in magnitude than GCA for 50% plant flowering, number of seed and shelling percentage indicating that non-additive genetic variance was more important than additive genetic variance for these traits. The estimates of phenotypic coefficient of variation (PCV) were greater than genotypic coefficient of variation (GCV) for all the traits. PCV ranged from 4.36 to 19.34% and GCV from 2.02 to 11.99%. Several authors reported comparable results which reflect the influence of environment on the expression of characters (Yousaf et al. 2008; Sumathi et al. 2010; Roychowdhury et al. 2011 and Narasimhulu et al. 2012). University of Ghana http://ugspace.ug.edu.gh 113 Narrow sense heritability estimates ranged from 35.4 to 95.5% with days to emergence being 95.5%, pod yield (85%), shelling percentage (84.3%), number of seed in 50 pods (80%) and 50% plants flowering (72%). Moderate to low heritability estimates were obtained for maturity index (66.1%), pods length (54.4%) and seed weight (35.4%). Because of higher heritabilities, selection for higher yield and maturity should be possible among the populations. The results agree with the findings of Tsaur et al. (1989) who reported high heritability for pod yield, pod length and 100 seed weight. Results also support the findings of Holbrook et al. (1989) who reported high heritability for maturity while studying F1 and F2 plants of reciprocal crosses involving Chico and PI383421(late maturing) groundnut lines. Ali and Wynne, (1994) similarly found high narrow sense heritability estimates for the cross NC7/Chico for 100 seed weight (71%) and pod length (52%), intermediate for maturity index (44%). Highly significant positive correlations were found between pod yield and seed weight (r = 0.97, P ≤ 0.001) and pod length (r = 0.72, P ≤ 0.01); between pod length and seed weight (r = 0.66, P ≤ 0.01); and between maturity index and days to emergence (r = 0.72, P ≤ 0.001). Correlation between shelling percentage and seed weight was positively significant at P ≤ 0.05 (r = 0.61). Simultaneous improvement of these traits should be possible. Waranyuwat and Tongsri (1990) reported highly significant associations between pod and seed yield, pod yield and number of mature seeds per plant, and seed yield and number of mature seeds per plant, but shelling percentage had variable correlation with pod and seed traits in different generations. Drought tolerance: Pod yield, biomass, and Harvest index decreased under end of-season drought stress, whereas SCMR increased as reported in previous studies under various environments (Nigam et al., 2008; University of Ghana http://ugspace.ug.edu.gh 114 Songsri et al., 2009; Girdthai et al., 2010). Drought tolerance index (DTI) was useful in explaining how some genotypes had higher pod yield under drought. The crosses ICG11249 x JL24, ICG11249 x RRB, Tainan-9 X 55-437 and ICGVSM 99511 x 55-437 had the highest DTI values for SCMR, biomass and harvest index. Therefore, these crosses are promising drought tolerant progenies. Previous reports on inheritance of drought resistance traits suggested a predominant role of additive gene effects in SLA and HI inheritance (Nigam et al., 2001; Surihan et al., 2005). Painawadee et al., (2009) reported that the water loss from cells might affect the concentration of chlorophyll. It could be hypothesized that groundnut genotypes with high SCMR have more photosynthetic machinery per unit leaf area and hence potential for greater assimilation under drought stress (Songsri et al., 2009). The significance of GCA mean squares indicated the importance of additive gene effect governing the inheritance pod yield, biomass, harvest index and SCMR. GCA mean squares for males and females were highly significant for pod yield and biomass. GCA mean squares for females were greater than males for pod yield and biomass indicating that the major contribution to additive variance for these traits was due the female parents. SCA mean squares differed significantly for pod yield and harvest index indicating the importance of both additive and non-additive variance for these traits. The findings agree with Swe and Branch (1986) who found that estimates of general and specific combining abilities to be significant for biomass, harvest index, total pod weight. John et al., (2011) indicated primarily non-additive gene action was important in the inheritance SCMR. University of Ghana http://ugspace.ug.edu.gh 115 The variety 55-437 was the best general combiner for all physiological characters. RRB was the best general combiner for biomass and pod yield. The ICGV-SM99511 was the best general combiner for SPAD chlorophyll meter reading and pod yield while Tainan-9 was the best combiner for biomass and harvest index. The estimates of phenotypic coefficient of variation (PCV) were greater than genotypic coefficients of variation (GCV) for all the physiological traits. PCV ranged from 7.27 to 40.20% and GCV from 2.71 to 15.91%. The traits pod yield, biomass and harvest index showed highest PCV estimates, respectively, 35.81, 26.91 and 40.20% and moderate GCV estimates 11.02, 14.12 and 15.91%, respectively, for the same traits. The low magnitude of difference between PCV (6.39%) and GCV (2.95%) observed with SPAD chlorophyll meter readings indicated that this trait was less influenced by the environment (Narasimhulu et al., 2012). High values of GCV suggest that these traits can be easily improved by selection. Several authors reported comparable results where PCV values were higher than GCV values for all the traits which reflect the influence of environment on the expression of characters including Thirumala et al. (2012); Azharudheen et al. (2013) and Reddy et al. (2013). Narrow sense heritability estimates for physiological traits were higher than for agronomic traits, and varied under both well-watered and end-of-season drought conditions. The heritability estimates for pod yield (0.26) and biomass (0.18) were low, but they were high for harvest index (0.78) and SPAD chlorophyll meter reading 0.71 and 0.68 respectively at 60 and 80 days after sowing. Fifty percent (50%) Pants flowering showed moderate heritability estimate of 0.56. End of-season drought decreased narrow sense heritability estimates for all the physiological traits. Heritability estimates decrease from 71.6% to 36.8% for SCMR60DAS; from 68.2% to 10.6%for University of Ghana http://ugspace.ug.edu.gh 116 SCMR80DAS and from 78.9% to 37.8% for harvest index. Results agree with Songsri et al. (2008) who found high heritability estimates for physiological traits. End of-season drought stress decreased heritability estimates for all the physiological traits except for pod yield and biomass. These findings are similar with Girdthai et al., (2012) who found increase of broad sense heritability for pod yield and biomass under drought stress conditions. Highly significant positive association between pod yield and harvest index was found in both water regimes. The correlation coefficient was higher (r = 0.77, P ≤ 0.001) under end of-season drought stress condition than under well watered conditions (r = 0.74, P ≤ 0.01). Strong negative significant (P ≤ 0.05) correlation was found between biomass and harvest index under well watered and end of-season drought conditions r = -0.66, -0.58, respectively. Weak positive and non-significant correlation was found between SCMR and pod yield, under well watered conditions (r = 0.38). Wunna et al. (2009) reported moderate association between HI and biomass under early drought and irrigated condition. These results support the findings of Ravi Kumar et al. (2012) for SCMR, harvest index. Ahamed (1995), Abdalla et al. (2008) and Jayalakshmi et al. (2000) have reported similar results for harvest index. In the past, breeders focused on earliness as a drought escape mechanism, particularly for the end-of-season drought that was easily predictable in the past. But, now with climate variability, this is changing. Rainfall is very erratic, floods are frequent, intermitted drought spells are frequent. Therefore, the drought escape strategy is not adequate because it is difficult to predict the end-of-season drought. However, drought escape mechanisms are still useful; for example, during rainy season 2013, most of groundnut growing areas in Niger experienced end- of-season drought. Interestingly, FESA seed Company reported that they had already harvested University of Ghana http://ugspace.ug.edu.gh 117 the extra early variety identified in the study that matured in 75 days, before end-of-season drought. The manager of this seed company was very impressed with the potential of this variety and organized a field day to share his findings with farmers around his farm. On station the variety was planted 15th July and harvested September 30th 2013. The success of the variety ICIAR19BT to escape drought means that earliness as a drought avoidance mechanism remains supportable. Promising early maturity and drought tolerant crosses were identified in this study and could be exploited to develop new varieties following their evaluation across several environments. Through conventional breeding, genetic variability for drought tolerance among groundnut cultivars can be identified and the genetic variation identified can be introduced through different mating designs into cultivars with good agronomic characteristics. Relationships that have been established between farmers and seed companies and INRAN’s seed unit must be maintained to implement a viable groundnut breeding programme in Niger. Farmers' confidence in groundnut production should be restored by development of new improved early or drought tolerant groundnut varieties. Conventional plant breeding is a highly time-consuming as well as a cost- and labor-intensive approach. When transferring desired genes from one plant to other through the conventional plant breeding, a number of undesired genes are also transferred. To achieve gains through traditional breeding, a number of selection and breeding cycles are required. The limited success in improving crop drought tolerance is because drought tolerance is controlled by multiple genes having additive effects and a strong interaction exists between the genes for drought tolerance and those involved in yield potential. Thus, there is a need to seek more efficient approaches for University of Ghana http://ugspace.ug.edu.gh 118 genetically tailoring crops for enhanced drought tolerance. Through marker-assisted breeding (MAB), it is now possible to examine thousands of genomic regions of a crop under water limited regimes, which was, previously not possible (Ashraf, 2010). Drought tolerance QLTs have been reported, this information can be exploited to introgress drought tolerance related traits such as transpiration, TE, SLA, SCMR into elite early maturing variety. Ravi et al. (2010) reported a number of QTLs for drought tolerance related traits in groundnut, including transpiration (7 QTLs), TE (7), SLA (13), SCMR (29), Biomass (7), Pod yield (7), Seed weight (5), and Haulm (6). From farmers' perceptions about earliness and drought tolerance, breeding efforts can be focused on "ideal" variety and can combine earliness, drought tolerance and high yielding. Development of ideal variety can be achieved more efficiently by using Marker assisted backcrossing. An extra early variety (ICIAR19BT) with good combining ability for most traits studied and drought tolerant varieties also have been identified. Molecular screening of the selected drought tolerant varieties will be crucial to identify genotypes carrying drought tolerance QTLs to facilitate marker assisted selection. Information generated from this study can be used to develop new groundnut varieties combining both traits. In the context of climate change, breeding efforts should focus on drought and high temperatures in the near future. Therefore, prior to establishing a viable breeding programme, a better understanding of interactions between the abiotic stresses is required. In addition to this understanding, relationships between biotic and abiotic stress must be established. University of Ghana http://ugspace.ug.edu.gh 119 CHAPTER SEVEN 7.0. GENERAL CONCLUSIONS AND RECOMMENDATIONS 7.1. General conclusions Through this study, the main constraints to groundnut production and preferred varieties have been identified. Drought and low soil fertility were the most important constraints contributing to low groundnut production in the study areas. The most preferred varieties were 55-347, RRB, and JL24. These will be used in a breeding programme to develop new early maturing genotypes that are drought tolerant. It was also clear that drought is the major production constraint in both regions of the study. The information gathered is useful for groundnut breeders to develop good quality groundnut varieties that incorporate farmers’ preferences. The genetic variation found in the germplasm screened for early maturity and drought tolerance will help breeders to develop higher yielding early- maturing and drought- tolerant groundnut varieties adapted to the Sahelian environment characterized by a short crop period and intermittent drought spells. Two extra early varieties namely Chico and ICIAR19BT, and three early maturing varieties ICG3584, ICGV02022 and 796 were identified. Five drought tolerant varieties namely ICG6703, ICGV-SM 99551, Tainan-9, ICG11249 and ICGV-IS01820 were selected. These materials can be used as parental lines in a breeding programme to develop early maturing and drought tolerant varieties. The varieties ICIAR19BT and ICGV-SM 99511 were the best high yielding. University of Ghana http://ugspace.ug.edu.gh 120 The results from the genetic analysis show that it is possible to select for both earliness and drought tolerance in early generations. This process could be enhanced using marker assisted selection schemes if and when QTLs/genes for the traits are found and markers are developed. Additive gene effects primarily controlled the inheritance of pod yield, seed yield, pod length, and maturity index. Based on heritability of seed weight and pod length and their positive association with pod yield, these traits could be used to improve yield of groundnut. Maturity index, although giving a high heritability estimate, was not associated with the yield traits, but showed a strong correlation with days to emergence. It could be a good criterion for earliness selection. Most physiological traits evaluated had high heritability, indicating that breeding progress should be possible. SPAD chlorophyll meter readings should be particularly useful as a selection criterion for drought tolerance in groundnut because of high heritability and the simplicity in gathering data. Groundnut lines with the ability to maintain high chlorophyll content and high biomass under drought conditions may also show better tolerance to drought. High heritability estimates for harvest index coupled with strong positive association with pod yield under both water regimes indicated the harvest index should also be considered as a selection criterion for pod yield. The variety 55-437 contributed to superior pod yield, seed weight, number of seed, harvest index and SCMR. RRB and Chico contributed superior maturity index and days to emerge. ICIAR19BT was the best general combining ability for days to emergence, pod yield, seed weight, and pod length. ICGV-SM 99511 had the best general combining ability for SPAD chlorophyll meter reading and pod yield, while Tainan-9 was the best combiner for biomass and harvest index. University of Ghana http://ugspace.ug.edu.gh 121 7.2. Recommendations - Three farmers' varieties, 55-437, JL24 and RRB, should be used in further studies to combine earliness and drought tolerance to mitigate the effects of climate variability in Niger and similar environments in West Africa. - One extra early maturing variety, ICIAR19BT (75 Days), and drought tolerant variety, Tainan-9, identified as high yielding and promising varieties among the parental lines should be widely evaluated in farmer participatory variety trials in the Niger groundnut main regions for release to boost groundnut production through increased productivity and contribute to smallholder farmers' livelihoods. - QTLs for earliness and drought tolerance should be mapped in the population developed in this study. This would facilitate the application of marker assisted selection if large effects of QTL's will be found. - A study should be carried out to combine earliness and drought tolerance traits to develop varieties that are capable of producing high yields to allow them to grow in harsher environments similar to those of Niger. - The best specific crosses for earliness and drought tolerance should be advanced up to F5 before selection through bulk method to develop new pure lines varieties. University of Ghana http://ugspace.ug.edu.gh 122 BIBLIOGRAPHY AND APPENDICES BIBLIOGRAPHY Abdalla, E.A., Ibrahim, A.E.S., & Abuelgasim, E.H. (2008). Performance of nine groundnut mutants and their parents under terminal drought stress conditions in western Sudan, Gezira Journal of Agricultural Science, 6, 1-8. Ahmadou, I. (2013). Development of downy mildew resistant, F1 pearl millet (Pennisetum glaucum (L.) R. brown) hybrids in Niger. (Unpublished doctoral thesis). University of Ghana. Ahamed, S.N. (1995). Heterosis combining ability and interrelationships among yield and yield attributes in groundnut (Arachis hypogaea L.) (Unpublished Master thesis). Andhra Pradesh Agricultural University, Hyderabad. Aissata, M. I. (2013). Evaluation of Nigirien sorghum [Sorghum bicolor (L.) Moench] landraces and their testcross hybrids for combining ability. (Unpublished doctoral thesis). University of Ghana. Ajayi, O. C. (2007). User Acceptability of Sustainable Soil Fertility Technologies: Lessons from Farmers 'Knowledge, Attitude & Practice in Southern Africa, Journal of Sustainable Agriculture, 30, 21-40. Ali, N. & Wynne, J.C. (1994). Heritability estimates and correlations studies of early maturity and other agronomic traits in two crosses of peanut (Arachis hypogaea L.), Pakistan Journal of Botany, 26, 75-82. Ali, N., Wynne, J.C. & Murphy, J.P. (1995). Combining ability estimates for early maturity and agronomic traits in peanut (Arachi hypogaea L.), Pakistan Journal of Botany, 27, 111- 119. Ali, N., Nawaz, M.S., Bashir, K., & Mirza, M.Y. (2001). Combining ability estimates in F2 and F3 generations for early maturity and agronomic traits in peanut (Arachis hypogaea L.). Pakistan Journal of Botany, 33, 93-99. Arslan, M. (2005). Effect of haulm cutting time on haulm and yield of peanut. Journal of Agronomy, 4, 39-43. Arunyanark, A., Jogloy, S., Akkasaeng, C., Vorasoot, N., Kesmala, T., Nageswara, R.R.C., Wright, G.C., & Patanothai, A. (2008). Chlorophyll stability is an indicator of drought tolerance in peanut. Journal of Agronomy and Crop Science, 194, 113-125. University of Ghana http://ugspace.ug.edu.gh 123 Arunyanark, A., Jogloy, S., Wongkaew, S., Akkasaeng, C., Vorasoot, N., Wright, G.C., Rachaputi Rao, C.N., & Patanothai, A. (2009). Association between aflatoxin contamination and drought tolerance traits in peanut. Field Crops Research, 114, 14-22. Arunyanark, A., Jogloya, S., Wongkaewb, S., Akkasaenga, C., Vorasoota, N., Kesmalaa, T., & Patanothaia, A. (2010). Heritability of aflatoxin resistance traits and correlation with drought tolerance traits in peanut. Field Crops Research, 117, 258-264. Ashraf, M. (2010). Inducing drought tolerance in plants: Recent advances. Biotechnology Advances, 28,169-183. Azharudheen, T.P.M., Gowda, M.V.C., Lingaraju, S., Bhat, A.R.S., Bhat R. & Motagi, B.N. (2013). Evaluation of sister homozygous lines for oil content and oil quality in groundnut. Karnataka Journal Agricultural Sciences, 26, 1-5. Bailey, W.K. & Bear, J.E. (1973). Components of earliness of maturity in peanuts, Arachis hypogaea L. Journal American Peanut Research and Education Association. 5, 32-39. Bagnall, D.J. & King, R.W. (1991). Response of peanut (Arachis hypogaea L.) to temperature, photoperiod & irradiance. Effect on peg and pod development. Field Crops Research, 26, 279-293. Bansal, U.K., Satija, D.R., & Gupta, V.P. (1991). Combining ability for yield in inter and intra- growth habit crosses in groundnut. Tropical Agriculture, 68, 71-73. Basu, M.S., Mathur, R.K., & Manivel, P. (2003). Evaluation of Trait-based and Empirical Selections for Drought Resistance at the National Research Centre for Groundnut, Junagadh, Gujarat, India, in Breeding of Drought-resistant Peanuts-Edited by A.W. Cruickshank, N.C. Rachaputi, G.C. Wright & S.N. Nigam, ACIAR Proceedings. 112, 23-25. Bell, M.J., Bagnall, D.J. & Harch, G. (1991). The effects of photoperiod on reproductive development of Peanut (Arachis hypogaea L.) in a cool subtropical environment & temperature interactions, Australian Journal of Agricultural Research, 42, 1151-1161. Ben Mohamed, A., Duivenbooden, N. V., & Abdoussallam, S. (2002). ‘Impact of Climate on Agricultural Production in the Sahel - Part 1. Methodological Approach and Case Study for Millet in Niger’, Clim. Change, 54, 148-160. Bhagsari, A.S., Brown, R.H. & Schepers, J.S. (1976). Effect of moisture stress on photosynthesis and some related physiological characteristics in peanut. Crop Science, 16,712-715. Bindu Madhava, H., Sheshshayee, M.S., Shankar, A.G., Prasad T.G., & Udayakumar, M. (2003). Use of SPAD chlorophyll meter to assess transpiration efficiency of peanut. In: Cruickshank AW, Rachaputi NC, Wright GC, Nigam SN (eds) Breeding of drought University of Ghana http://ugspace.ug.edu.gh 124 resistant peanuts. ACIAR Proceedings No. 112. Australian Centre for International Agricultural Research, Canberra, 3-9. Boontang, S., Girdthai, T., Jogloy, S., Akkasaeng C., Vorasoot, N., Patanothai, A., & Tantisuwichwong, N. (2010). Responses of released cultivars of peanut to terminal drought for traits related to drought. Asian Journal of Plant Sciences, 9, 423-431. Boote, K.J. (1982). Growth stages of Peanut (Arachis hypogaea L.), Peanut Science, 9, 35-40 Boote, K.J. & Ketring, D.L. (1990). Peanut. In: Stewart B.A. And Nielson D.R. (Eds), Irrigation of Agricultural Crops. Asa-Groundnut. A Global Perspective. International Crops Research CSSA-SSSA, Madison. Brown, D., Howes, M., Hussein, K., Longley C. & Swindell, K. (2002). Participatory methodologies and participatory practices: assessing PRA use in the Gambia, Agricultural Research & Extension Network, Network Paper, 124, 20p. Buerkert, A., (2002). Multi-site time-trend analysis of soil fertility management effects on crop production in sub-Saharan West Africa, Experimental Agriculture, 38, 163-183. Burow, M.D., Simpson, C.E., Starr, J.L. & Paterson, A.H. (2001). Transmission genetics of chromatin from a synthetic amphidiploid to cultivated peanut (Arachis hypogaea L.): broadening the gene pool of a monophyletic polyploid species. Genetics, 159, 823- 837. Burton, G.W. (1951). Quantitative inheritance in Pearl millet (P. glaucum). Journal of Agronomy, 43, 409-417. Camberlin, P. & Diop, M. (1999). Inter-Relationships between groundnut yield in Senegal, inter annual rainfall variability and sea surface temperatures. Theoretical and Applied Climatology, 63,163-181. Chambers, R. (1994). Participatory Rural Appraisal (PRA): Analysis of Experience. World Development, 22, 1253-1268. Chiteka, Z.A. (1985). The present and future status of groundnut breeding and research in Zimbabwe, pp. 125-133 in Proceedings of the Regional Groundnut Workshop for Southern Africa, 26-29 Mar 1984, Lilongwe, Malawi. Patancheru, A.P. 502 324, India: International Crops Research Institute for the Semi-Arid Tropics. Christmann, A., Hoffmann, T., Teplova, I., Grill, E. & Müller, A. (2005). Generation of active pools of abscisic acid revealed by in vivo imaging of water stressed Arabidopsis. Plant Physiology, 137, 209-219. University of Ghana http://ugspace.ug.edu.gh 125 Clavel, D., Sarr, B., Marone, E. & Ortiz, R. (2004). Potential agronomic and physiological traits of Spanish groundnut varieties (Arachis hypogaea L.) selection criteria under end-of- cycle drought conditions. Agronomie, 24, 101-111. Clavel, D. Diouf, O., Khalfaoui, J.L. & Braconnier, S. (2006). Genotypes variations in fluorescence parameters among closely related groundnut (Arachis hypogaea L.) lines and their potential for drought screening programs. Field Crops Research, 96, 296-306. Coffelt, T.A. & Hammons, R.O. (1974). Correlation and heritability studies of nine characters in parental and interspecific cross populations of Arachis hypogaea L. Oleagineux, 29, 23- 26. Collino, D.J., Dandanelli, J.L., Sereno, R. & Racca, R.W. (2001). Physiological responses of argentine peanut varieties to water stress. Light interception, radiation use efficiency and partitioning assimilate. Field Crops Research, 70, 177-184. Craufurd, P. Q., Wheeler T. R., Ellis R.H., Summerfield, R.J., Williams & J.H. (1999). Effect of temperature and water deficit on water-use efficiency, carbon isotope discrimination, and specific leaf area in peanut. Crop Science, 39, 136-142. Cruickshank, A., Dowkiw, A., Wright, G.C., Nageswara, R.C.R. & Nigam, S.N. (2004). Heritability of drought–resistance traits in peanut. In T. Fischer, N. Turner, J. Angus, L. McIntyre, M. Robertson, A. Borrell, D. Lloyd (ed.) New directions for a diverse planet: Proceedings for the 4th Int. Crop Science Congress, Brisbane, Australia. 26 Sept.-1 Oct. 2004. Cuc, M .L., Mace E.S., Crouch J.H., Quang V.D., Long D.T. & Varshney R.K. (2008). Isolation and characterization of novel microsatellite markers and their application for diversity assessment in cultivated groundnut (Arachis hypogaea L). Plant Biology, 8, 55-59. Culbreath, A.K., Todd, J.W., Gorbet, D.W., Brown, S.L., Baldwin, J.A., Pappu, H.R., Holbrook, C.C. & Shokes, F.M. (1999). Response of early, medium, and late maturing peanut breeding lines to field epidemics of tomato spotted wilt. Peanut Science, 26,100- 106. Dulvenbooden, N.V., Abdoussalam, S. & Moamed, A.B. (2002). Impact of climate change on agricultural production in the Sahel-Part 2.case study for groundnut and Cowpea in Niger, Climatic Change, 24, 349-368. Duy-Nang, N. (2004). Nitrogen uptake, leaf nitrogen, chlorophyll content and leaf color of rice (Oriza sativa L.) as affected by drought stress. (unpublished Master thesis ) University of Philippine, Los Banos. Efisue, A., Tongoona, P., Derera, J., Langyintuo, A., Laing, M. & Ubi, B. (2008). Farmers' Perceptions on Rice Varieties in Sikasso Region of Mali and their Implications for Rice Breeding. Journal of Agronomy and Crop Science,194, 393-400. University of Ghana http://ugspace.ug.edu.gh 126 Enyong, L.A., Debrah, S.K. & Bationo A. (1999). Farmers’ perceptions and attitudes towards introduced soil-fertility enhancing technologies in western Africa, Nutrient Cycling in Agroecosystems, 53, 177-187. Epron, D. & Dreyer, E. (1993). Long-term effects of drought on photosynthesis of oak trees (Quercus petraea and Q. robur) in nature stand. New Phytologist, 125, 381-389. Falconer, D.S. & Mackey, T.F.C. (Eds.) (1996). Introduction to quantitative genetics. Longman Group Ltd., London FAO, (2008). FAOSATAT, Statistical Data Base. Food and Agricultural Organizations of the United Nations, Rome. FAO, (2011). FAOSATAT, Statistical Data Base. Food and Agricultural Organizations of the United Nations, Rome. Farshid, A. (2010). Barriers of Agricultural Development in Iran: A Case study of Fars Province. Journal of American Science, 6, 1034-1037. Fernandez, G.C., & Miller, J.C. (1985). Estimation of heritability by parent-offspring regression. Theoretical Applied Genetic, 70, 650-654. Frimpong A., 2004 - Characterization of groundnut (Arachis hypogaea L.) in Northern Ghana, Pakistan Journal of Biological Sciences, 7, 838-842. Furlan, A., Lianes, A., Luna, V. & Castro, S. (2012). Physiological and Biochemical Responses to Drought Stress and Subsequent Rehydration in the Symbiotic Association Peanut- Bradyrhizobium sp. International Scholarly Research Network Agronomy, 8p. doi:10.5402/2012/318083. Garcia, G.M., Stalker H.T., Schroeder, E., LYERLY, J.H. & Kochert, G. (2005). A RAPD- based linkage map of peanut based on a backcross population between the two diploid species Arachis stenosperma and A. cardenasii. Peanut Science, 32, 1-8. Girdthai, T., Jogloy S., Vorasoot N., Akkasaeng C., Wongkaew S., Holbrook C.C., & Patanothai A. (2010). Heritability of, and genotypic correlations between, aflatoxin traits and physiological traits for drought tolerance under end of season drought in peanut (Arachis hypogaea L.). Field Crops Research, 118, 169-176. Girdthai, T., Jogloy, S., Vorasoot, N., Akkasaeng, C., Wongkaew, S., Holbrook, C.C., &Patanothai, A. (2010). Associations between physiological traits for drought tolerance and aflatoxin contamination in peanut genotypes under terminal drought. Plant Breeding, 129, 693-699. University of Ghana http://ugspace.ug.edu.gh 127 Ghosh, P.K., Dey, R., Pal, K.K. & Mathur, R.K. (2005). Groundnut, In: Techniques and management of field crop production (Eds.Rathore, P.S.). Agrobios (India) Publishers, Jodhpur, India 153-177. Girdthai, T., Jogloy, S., Vorasoot, N., Akkasaeng, C., Wongkaew, S., Patanothai, A. & Holbrook, C.C. (2012). Inheritance of the physiological traits for drought resistance under terminal drought conditions and genotypic correlations with agronomic traits in peanut. Journal of Breeding and Genetic, 44, 240-262. Gomez, K.A. & Gomez, A.A. (1984). Statistical procedures for agricultural research. John Wiley & Sons, New York. Gomez, S.M., Narayana M., Schubert A.M., Ayers J.L., Baring M.R., & Burow M.D. (2009). Identification of QTLs for pod and kernel traits in cultivated peanut by bulked segregant analysis. Electronic Journal of Biotechnology, 12, 1-10. Gosal, S.S., Wani S.H. & Kang M.S. (2009). Biotechnology and drought tolerance. Journal of Crop Improvement, 23, 9-54. Gowda, A. & Hegde, B.R. (1986). Moisture Stress and Hormonal Influence on the Flowering Behavior and Yield of Groundnut. Plant Physiology, 66, 835-837. Green, C.C., Wynne, J.C. & Beute, M.K. (1983). Genetic variability and heritability estimates based on the F2 generation from crosses of large-seeded Virginia-type peanuts with lines resistant to Cylindrocladium black rot. Peanut Science, 10, 47-51. Griffing, B. (1956). Concept of general and specific combining ability in relation to diallel crossing systems, division of plant industry, C.S.I.R.O., Australian Journal of Biological Sciences, 9, 463-493. Griffiths, H. & Parry, M.A.J. (2002). Plant responses to water stress. Annals of Botany, 89, 801- 802. Halward, T., Stalker H.T., & Kochert, G. (1993). Development of an RFLP linkage map in diploid peanut species. Theoretical and Applied Genetics, 87, 379-384. Hamidou, F., Halilou, O. & Vadez, V. (2012). Assessment of Groundnut under Combined Heat and Drought Stress. Journal of Agronomy & Crop Science, 1-11. Holbrook, C.C., Kvien C.S. & Branch, W.D. (1989). Genetic control of peanut maturity as measured by hull-scape procedure. Oleagineux, 44, 359-362. Holbrook, C.C. (1990). Utility of early generation diallel analysis for predicting parental potential for yield and yield components in peanuts. Peanut Science, 17, 9-11. Holbrook, C. & Stalker H. (Eds.) (2003). Peanut Breeding and Genetic Resources, Plant Breeding Reviews, 22, John Wiley & Sons, Inc, 327-328. University of Ghana http://ugspace.ug.edu.gh 128 Holland, J.B., Nyquist W.E. & Cervantes-Martinez C.T. (Eds.) (2003). Estimating and interpreting heritability for plant breeding: An update, Plant breeding reviews, Wiley & Sons, 22:9-20. Hubick, K.T., Farquhar, G.D., & Shorter, R. (1986). Correlation between water-use efficiency and carbon isotope discrimination in diverse peanut (Arachis) germplasm. Australian Journal of Plant Physiology, 13, 803-816. Hubick, K.T., Farguhar G.D. & Shorter R. (1988). Heritability of genotypic x environmental interactions of carbon isotope discrimination and transpiration efficiency in peanut (Arachis hypogea L.). Australian Journal of Plant Physiology, 15, 779- 813. Ingram, J. & Bartels, D. (1996). The molecular basis of dehydration tolerance in plants. Annual Review of Plant Physiology and Plant Molecular Biology, 47: 377-403. Isleib, T.G., Wynne, J.C., Elkan, G.H. & Schneeweis, T.J. (1980). Quantitative genetic aspects of nitrogen fixation in peanuts (Arachis hypogaea L.). Peanut Science, 7, 101-105. Jalgaonkar, R., Jamadagni, B.M. & Salvi, M.J. (1990). Genetic variability and correlation studies in turmeric. Ind. Cocoa. Arecanut and Spices Journal, 15, 20-22. Janamatti, V.S., Sashidharv, R., Prasad, I.G. & Sastry, K.S.K. (1986). Effect of Cycles of Moisture Stress on Flowering Pattern, Flower Production, Gynophore Length and Their Relationship to Pod Yield In Bunch Types Of Groundnut. Journal of Agricultural Research, 1, 136-142. Jayalakshmi, V., Reddy, C.R., Reddy, P.V. & Reddy, G.L. (2000). Character association among morpho-physiological attributes in parental genotypes and groundnut hybrids. Legume Research, 23, 102-105. Jharma D.E., Chlowdhury B.L.D., Haque M.A., Bhuiyan M.R.H. & Husian M.M. (2001). Biochemical screening of some groundnut (Arachis hypogaea L.) genotypes for drought tolerance, Journal of Biological Sciences, 1, 1009-1011. Jogloy, S., Wynne, J.C. & Beute, M.K. (1987). Inheritance of late leaf spot resistance and agronomic traits in peanut. Peanut Science, 14, 86-90. Jogloy, S.; Patanothai, A., Toomsan, S. & Isleib, T.G. (1996). Breeding peanut to fit into Thai cropping systems. Proc. of the Peanut Collaborative Research Support Program- International Research Symposium and Workshop, Two Jima Quality Inn, Arlington, Virginia, USA, 25-31 March, 1996, 353-362. Jogloy, S., Trisuvanwat, W., Wongkaew, S. & Patanothai, A. (1999). Combining ability and heritability of rust resistance and agronomic traits in peanut (Arachis hypogaea L.). Thai Journal of Agriculture Science, 32, 487-494. University of Ghana http://ugspace.ug.edu.gh 129 Jogloy, S., Tula, W. & Kesmala, T. (2005). Combining ability analysis and phenotypic correlation of nodule parameters and agronomic traits in peanut (Arachis hypogaea L.) Songklanakarin Journal of Science and Technology, 27, 213-221. Jogloy, C., Jaisil, P., Akkasaeng, C., Kesmala T. & Jogloy, S. (2011). Heritability and Correlation for Maturity and Pod Yield in Peanut. Journal of Applied Sciences Research 7, 134-140. John K., Reddy P.R., Reddy P.H., Sudhakar P. & Reddy N.P.E. (2011). General and specific combining ability estimates of physiological traits for moisture stress tolerance in groundnut (Arachis hypogaea L.). International Journal of Applied Biology and Pharmaceutical Technology, 2, 470-481. Jongrungklang, N., Toomsan, B., Vorasoot, N., Jogloy S., Kesmala, T. & Patanothai, A. (2008). Identification of peanut genotypes with high water use efficiency under drought stress conditions from peanut germplasm of diverse origins. Asian Journal of Plant Sciences, 7, 628-638. Joshi, A. & Witcombe, J.R. (1996). Farmer Participatory Crop Improvement II. Participatory Varietal Selection, a Case Study in India, Experimental Agriculture, 32, 461-477. Kale, U.V. & Dhoble, M.V. (1988). Genetic variability and character association in groundnut varieties. Journal of Maharashtra Agricultural Universities, 13, 239-240. Kearsey, M.J, & Pooni H.S. (1996). The Genetical Analysis of Quantitative Traits. Chapman and Hall, London. Ketring, D.L. (1979). Light effects on development of an indeterminate plant. Plant Physiology, 64, 665-667. Khan, A., Rahim, M., Khan, M.I., & Tahir, M. (2000). Genetic variability and criterion for selection of high yielding peanut genotypes. Pakistan Journal of Agricultural Research, 16, 9-12. Kolattukudy, P.E. (1980). Biopolyester membranes of plants: cutin and suberin. Science, 208, 990-1000. Kotzamanidis, S.T., Stavropoulos N. & Ipsilandis, C.G. (2006). Correlation studies of 21 traits in F2 generation of groundnut (Arachis hypogaea L.). Pakistan Journal of Biological Sciences, 9, 929-934. Lal, C., Hariprasanna, K., Rathnakumar, A.L., & Samdur, M.Y. (2006). High yielding, water use efficient Spanish groundnut (Arachis hypogaea) genotypes for rainfed production system. Indian Journal of Agricultural Science, 76, 148-150. University of Ghana http://ugspace.ug.edu.gh 130 Lauriano, J.A., Lidon, F.C., Carvalho, C.A., Campos, P.S. & Céo Matos, M.D. (2000). Drought effects on membrane lipids and photosynthetic activity in different peanut cultivars. Photosynthetica, 38, 7-12. Leblois, A., Quirion, P., Alhassane, A. & Traore, S. (2011). Weather index drought insurance: an ex ante evaluation for millet growers in Niger in International Congress, August 30- September 2, 2011, Zurich, Switzerland 27p. Mace E. S., Yuejin W., Boshou L., Upadhyaya H., Chandra S. and Crouch J. H., (2006). Simple sequence repeat (SSR)-based diversity analysis of groundnut (Arachis hypogaea L.) germplasm resistant to bacterial wilt, Plant Genetic Resource,5, 27-36. Madhava, B.H., Sheshshayee M.S., Shankar A.G., Prasad T.G. & Udayakumar M. (2003). Use of SPAD Chlorophyll Meter to Assess Transpiration Efficiency of Peanut, in Breeding of Drought-resistant Peanuts-Edited by A.W. Cruickshank, N.C. Rachaputi, G.C. Wright and S.N. Nigam, ACIAR Proceedings,112, 3-9 Mapfumo, P., Campbell, B.M. & Mpepereki, S. (2001). Legumes in soil fertility management: The case of pigeonpea in smallholder farming systems of Zimbabwe. Crop Science and Production, 9, 629-644. Mathews, R.B., Harris D., Nageswara R.R.C., Williams J.M. & Wadia K.D.R. (1998). The physiological basis for yield differences between four groundnut genotypes in response to drought. I. Dry matter production and water use. Experimental Agriculture, 24, 191-202. Mathur, B.P., Devi M.J., Serraj R., Yamaguchi S.K., Vadez V., & Sharma K.K. (2004). Evaluation of transgenic groundnut lines under water limited conditions. International Arachis Newsletter, 24, 33-35. Meisner, C.A. & Karnok, K.J. (1992). Peanut root response to drought stress. Journal of Agronomy, 84, 159-165. Mekontchou, T. (1987). Inheritance and combining ability of early maturity and seed dormancy for a selected group of peanut lines (Arachis hypogaea L.) (Unpublished Master thesis), North Carolina State University, Raleigh Mohammed, J., Wynne J.C. & Rawlings J.O. (1978). Early generation variability and heritability estimates in crosses of virginia and spanish peanuts. Oleagineux, 35, 81-86. Moreshat, S., Bridges D.C., Ne Smith D.S. & Huang B. (1996). Effects of Water stress on competitive interaction of peanut and sickle pod. Agronomy Journal, 88, 636-644. Muza, L., Dhliwayo, H.H., & Twomlow, S.J. (Eds) (1996). Dryland Maize Response to Different Combinations of Tillage and Weeding Methods. Fifth Eastern and Southern Africa Regional Maize Conference 3rd-7th June, 110-114. University of Ghana http://ugspace.ug.edu.gh 131 Muzari, W., Gatsi, W. & Muvhunzi, S. (2012). The Impacts of Technology Adoption on Smallholder Agricultural Productivity in Sub-Saharan Africa: A Review. Journal of Sustainable Development, 5, 69-77. Nageswara, R.R.C., Sadar Singh, R.C., Sivakumar, M.V.K., Srivastava, K.L. & Williams, J. H. (1985). ‘Effect of Water Deficit at Different Growth Phases of Peanut: I. Yield Responses’. Journal of Agronomy, 77, 782-786. Nageswara, R.R.C. Williams J.H. & Singh M. (1989). Genotypic sensitivity to drought and yield potential of peanut. Journal of Agronomy, 81, 887-893. Nageswara R.R.C., Wadia K.D.R. and Williams J.H., (1990). Intercropping short and long duration groundnut (Arachis hypogaea) genotypes to increase productivity in environments prone to end-of-season droughts, Experimental Agriculture, 26, 63-72. Nageswara, R.R.C., Williams J.H., Wadia K.D.R., Hubick K. & Farquhar G.D. (1993). Crop growth, water use efficiency and carbon isotope discrimination in groundnut (Arachis hypogaea L.) genotypes under end-off-season water deficit. Annals of Applied Biology, 122, 357-67. Nageswara, R.R.C. (1994). Drought research on groundnut at ICRISAT center (Wright G.C. & Nageswara, R.R.C., Eds). Report of a workshop held at ICRISAT Center Andhra Pradesh, India. 5 -7 May 1993. ACIAR Technical report no 27: The Australian center for International Agricultural Research. 70p. Nageswara, R.R.C. & Wright G.C. (1994). Stability of the relationship between specific leaf area and carbon isotope discrimination across environments in peanut. Crop Science, 34, 98-103. Nageswara, R.R.C., Talwar, H.S., & Wright, G.C. (2001). Rapid assessment of specific leaf area and leaf N in peanut (Arachis hypogaea L.) using chlorophyll meter. Journal of Agronomy and Crop Science, 189, 175-182. Nageswara, R.R.C. & Nigam, S.N. (2003). Genetic Options for Drought Management in Groundnut. In: Management of Agricultural Drought - Agronomic and Genetic Options. Science Publishers, Inc, 123-141. Naidu, R.A., Kimmins F.M., Deom, C.M., Subrahmanyam, P., Chiyembekeza, A.J. & Van der Merwe, P.J.A. (1999). Groundnut rosette A virus disease affecting groundnut production in Sub-Saharan Africa. Plant Disease, 83, 700-709. Narasimhulu, R., Kenchanagoudar, P.V. & Gowda, M.V.C. (2012). Study of genetic variability and correlations in selected groundnut genotypes. International Journal of Applied Biology and Pharmaceutical Technology, 355-358. University of Ghana http://ugspace.ug.edu.gh 132 Nautiyal, P.C., Nageswara, R.R.C., & Joshi, Y.C. (2002). Moisture deficit–induced changes in leaf water content, leaf carbon exchange rate and biomass production in groundnut cultivars differing in specific leaf area. Field Crops Research, 74, 67-79. Naveen, P., Daniel, K.V., Subramanian, P. & Kumar, P.S. (1992). Response of irrigated groundnut (Arachis hypogaea L.) to moisture stress and its management. Indian Journal of Agronomy, 37, 82-85. N'Doye, O. & Smith, D. (1993). A Note on the Earliness of Offspring from Crosses among Five Short Growth-Duration Peanut Lines. Peanut Science, 20, 132-137. Ndunguru ,B.J., Ntare B.R., Williams J.H. & Greenberg D.C. (1995). Assessment of groundnut cultivars for end-of-season drought tolerance in a Sahelian environment. Journal of Agricultural Science, 125, 79-85. Nigam, S.N., Rao, R.C.N., Wynne, J.C., Williams, J.H., Fitzner, M., & Nagabhushanam, G.V.S. (1994). Effect and interaction of temperature and photoperiod on growth and partitioning in three groundnut (Arachis hypogaea L.) genotypes. Annals of Applied Biology, 125, 541-552. Nigam, S.N., Nageswara R.R.C. & Wynne J.C., (1998). Effects of temperature and photoperiod on vegetative and reproductive growth of groundnut (Arachis hypogaea L.). Journal of Agronomy and Crop Science, 181, 117-124. Nigam, S.N., Upadhyaya, H.D., Chandra, S., Nageswara, R.R.C., Wright, G.C. & Reddy, A.G.S. (2001). Gene effects for specific leaf area and harvest index in three crosses of groundnut (Arachis hypogaea). Annals of Applied Biology, 139, 301-306. Nigam, S.N., Chandra S., Manohar B., Talwar H.S., Reddy A.G.S. & Kanchi R., (2003). Evaluation of Trait-based and Empirical Selections for Drought Resistance at ICRISAT Centre, Patancheru, Andhra Pradesh, India, in Breeding of Drought-resistant Peanuts- Edited by A.W. Cruickshank, N.C. Rachaputi, G.C. Wright, & S.N. Nigam, ACIAR Proceedings, 112, 43-51. Nigam, S.N., Chandra S., Srievi K.R., Manohar B., Reddy A.G.S., Nageswara R.R., Wright G.C., Reddy P.V., Deshmukh M.P., Mathur R.K., Basu M.S., Vasundhara S., Varman P.V. & Nagda A.K. (2005). Efficiency of physiological trait-based and empirical selection approaches for drought tolerance in groundnut. Annals of Applied Biology, 146, 433-439. Nigam, S.N. & Aruna, R. (2008). Stability of soil plant analytical development (SPAD) chlorophyll meter reading (SCMR) and specific leaf area(SLA) and their association across varying soil moisture stress conditions in groundnut (Arachis hypogaea L.). Euphytica, 160, 111-117. University of Ghana http://ugspace.ug.edu.gh 133 Nkongolo, K.K., Chinthu, K.K., L., Malusi, M. & Vokhiwa, Z. (2008). Participatory variety selection and characterization of Sorghum (Sorghum bicolor (L.) Moench) elite accessions from Malawian gene pool using farmer and breeder knowledge, African Journal of Agricultural Research, 3, 273-283. Ntare, B.R. & Waliyar, F. (1994). The Role of Genetic Enhancement in Sustainable Groundnut Production in Western Africa, 14-19 in Sustainable Groundnut Production in Southern and Eastern Africa Proceedings of a Workshop 5-7 Jul 1994. Mbabane, Swaziland. Nwokolo, E. (1996). Peanut (Arachis Hypogaea L.). In Food and Fee from Legumes and Oilseeds. E. Nwokolo and J. Smartt, Eds. 49-63. New York: Chapman and Hall. Nyquist, W.E. & Baker, R.J. (1991). Estimation of heritability and prediction of selection response in plant populations Critical Reviews. Plant Sciences, 10, 235-322. Ong, C.K., (1986). Agroclimatological factors affecting phenology of groundnut, In: Agrometeorology of Groundnut. Proceedings Symposium. ICRISAT Sahelian Centre, Niamey, Niger, 115-125. Oumarou, H., Hamma, H., N’Diaye A. & Mounkaila. A. (1990). Development of groundnut production in Niger. In ICRISAT (International Crops Research Institute for the Semi- Arid Tropics), Summary Proceedings of the First ICRISAT Regional Groundnut Meeting for West Africa, 13-16 September 1988, Niamey, Niger, pp. 57-59. Padi, F.K. (2008). Genotype × environment interaction for yield and reaction to leaf spot infections in groundnut in semiarid West Africa: Genotype × environment interaction and leaf spot resistance in groundnut. Euphytica, 164, 143-161. Painawadee, M., Jogloy S., Kesmala T., Akkasaeng C. & Patanothai, A. (2009). Identification of traits related to drought resistance in peanut (Arachis hypogaea L.). Asian Journal of Plant Sciences, 8, 120-128. Panhwar, S.A., Baloch, M.J., Jatoi, W.A., Veesar, N.F. & Majeedano, M.S. (2008). Combining ability estimates from line x tester mating design in upland cotton. Proceedings of the Pakistan Academy of Sciences, 45, 69-74. Pimratch, S., Jogloy S., Vorasoot N., Toomsan B., Kesmala T., Patanothai A., & Holbrook C.C. (2009). Heritability of N2 Fixation Traits, and Phenotypic and Genotypic Correlations between N2 Fixation Traits with Drought Resistance Traits and Yield in Peanut. Crop Science, 49, 791-800. Plaut, Z. & Ben, H.M., (2005). Irrigation Management of Peanut with a Moving Sprinkler System: Runoff, Yield, and Water Use Efficiency. Journal of Agronomy, 97, 1202-1209. University of Ghana http://ugspace.ug.edu.gh 134 Proite K., Leal-Bertioli S.C.M., Bertioli D.J., Moretzsohn M.C., Silva F.R., Martins N. F. and Guimarães P.M., (2007). ESTs from a wild Arachis species for gene discovery and marker development, BMC Plant Biology, 7, 71-85. Rachid, S., Tom C.H., Hutokshi K.B., Francis R.B., Rolf T.F., Subhash C., Pooran M.G., Junichi K., Shyam N.N., Aruna R., & Jonathan H.C. (2003). Marker-assisted breeding for crop drought, Proceedings of the International Congress “In the Wake of the Double Helix: From the Green Revolution to the Gene Revolution”, 27-31 May 2003, Bologna, Italy, 217-238. Rachmeler, D.N. (1988). The inheritance of early maturity and flatty acid composition in peanut (Arachis hypogaea L.), (Unpublished doctoral thesis), North Carolina State University., Raleigh. Ravi Kumar, D., Sekhar, M.R., Reddy, K.R. & Ismail, S. (2012). Character association and path analysis in groundnut (Arachis hypogaea L.). International Journal of Applied Biology and Pharmaceutical Technology, 3, 385-389. Ravi, K., Vadez V., Isobe, S., Mirr S.S., Guo, Y., Nigam, S.N., Gowda, M.VC., Radhakrishnan, T., Bertioli, D.J., Knapp, S.J. & Varshney R.K. (2010). Identification of several small main-effect QTLs and a large number of epistatic QTLs for drought tolerance related traits in groundnut (Arachis hypogaea L.).International Journal of Plant Breeding Research. doi:10.1007/s00122-010-1517-0 Rawlings, J.O. & Thompson, D.L. (1962). Performance level as criterion for the choice of maize testers. Crop Science, 5, 217-220. • Raynaut, C., Koechlin, J., Basset, P., Cheung, C. & Stigliano, M. (1984). Le développement rural de la Région de Maradi : Analyser et comprendre la diversité. GRID, Université de Toulouse. Reddy, A.J. & Rao, I.M. (1968). Influence of induced water stress chlorophyll components of proximal and distal leaflets of groundnut plants. Current Science, 5, 118-121. Reddy, T.Y., Reddy V.R. & Anbumozhi, V. (2003a). Physiological responses of groundnut (Arachis hypogea L.) to drought stress and its amelioration: A critical review. Plant Growth Regulation, 41, 75-88. Reddy, T.Y., Reddy V.R. & Anbumozhi, V. (2003b). Physiological responses of groundnut (Arachis hypogaea L.) to drought stress and its amelioration: A review. Acta Agronomica Hungarica, 51, 205-227. Reddy, P.V., Sudhakar, P., Sujatha, D., Babitha, M. & Latha, P. (2004). Relationship of SPAD chlorophyll meter reading with specific leaf area, leaf total chlorophyll and pod yield in groundnut (Arachis hypogaea L.). In: Reddy PV, Rao JSP, Reddy KB, Sudhakar P, Rama Rao G, Babitha M, Latha P, Jyotsna MK (eds) Proceedings of the national University of Ghana http://ugspace.ug.edu.gh 135 seminar on physiological interventions for improved crop productivity and quality: opportunities and constraints. Indian Society for Plant Physiology, Division of Plant Physiology, Indian Agricultural Research Institute, New Delhi, India, 24-27. Reddy, B.P.K., Begum, H., Sunil, N. & Reddy, M.T. (2013). Variance component analysis of quantitative traits in muskmelon (Cucumis melo L.). Trakia Journal of Sciences, 2, 118- 124. Redona, E.D. & Lantican, R.M. (1985). Genetic analysis of some quantitative traits in peanut, (Arachis hypogaea L.) I. General and Specific combining ability estimates. Philippine Journal of Crop Science, 10, 81-86. Richardson, A.D, Duigan, S.P. & Berlyn, G.P. (2002). An evaluation of noninvasive methods to estimate foliar chlorophyll content. New Phytologist, 153, 185-194. Roychowdhury, R. & Tah, J. (2011). Genetic variability study for yield and associated quantitative characters in mutant genotypes of Dianthus caryophyllus L. International Journal of Biosciences, 1, 38-44. Roychowdhury R., Tah, J., Dalal T. & Bandyopadhyay, A. (2011). Selection response and correlation studies for metrical traits in mutant carnation (Dianthus caryophyllus L.) genotypes. Continental Journal of Agricultural Science, 5, 6-14. Rucker, K.S.; Kevin, C.K., Holbrook, C.C. & Hook, J.E. (1995). Identification of Peanut Genotypes with Improved Drought Avoid- Ance Traits. Peanut Science, 22, 14-18. Saleh G.B., & Masiron, N.A. (1994). Performance and heritability studies on some peanut lines and cultivars. Journal of Islamic Academy of Sciences, 7, 39-43. Samdur, M.Y., Manivel, P., Jain, V.K., Chikani B. M., Gor, H. K., Desai, S., & Misra, J.B. (2003). Genotypic Differences and Water-Deficit Induced Enhancement in Epicuticular Wax Load in Peanut. Crop Science, 43, 1294-1299. Sarker, A.M., Rahman M.S. & Paul, N.K. (1999). Effect of soil moisture on relative leaf water content, chlorophyll, proline and sugar accumulation in Wheat. Journal of Agronomy and Crop Science, 183, 225-229. SAS Institute. (2009). SAS Proprietary Software Release 9.2. SAS Institute, Inc., Cary, NC. Serraj, R. & Sinclair, T.R. (2002). Osmolyte accumulation: can it really help increase crop yield under drought conditions. Plant, Cell and Environment, 25, 333-341. Sexton, P.J., Benett, J.M. & Boote, K.J. (1997). The Effect of Dry Use Efficiency and Carbon Isotope Discrimination in Peanut under Pegging Zone Soil on Pod Formation. Peanut Science, 24, 19-24. University of Ghana http://ugspace.ug.edu.gh 136 Sharma, H.C., Crouch, J.H., Sharma, K.K., Seetharama, N. & Hash, C.T. (2002). Review: Applications of biotechnology for crop improvement: prospects and constraints. Plant Science, 163, 381-395. Sharma K.K. (2006). Drought Resistant Crops Drought Crops, third Asian Biotechnology conference November 9-10, 2006, Manila, Philippines. Sheshshayee, M., Bindumadhava H., Rachaputi N., Prasad T., Udayakumar M., Wright G. & Nigam S. (2006). Leaf chlorophyll concentration relates to transpiration efficiency in peanut. Annals of Applied Biology, 148, 7-15. Simmonds, N.W. (Eds.) (1979). Principles of crop improvement. John Wiley and Sons, Inc., New York 110-111. Sivakumar, M.V.K., & Sharma, P.S. (1986). Studies on water relations of groundnut. In: Agrometeorology of Groundnut. Proc. Int. Symp. ICRISAT Sahelian Centre, Niamey, Niger 83-98. Smith, J.D. & Kinman M.L., (1965). The use of parent-offspring regression as an estimator of heritability. Crop Science, 5, 595-596. Songsri, P., Jogloy S., Kesmala T., Vorasoot N., Akkasaeng C., Patanothai A. & Holbrook C. C.(2008). Heritability of Drought Resistance Traits and Correlation of Drought Resistance and Agronomic Traits in Peanut. Crop Science, 48, 2245-2253. Songsri, P., Jogloy, S., Holbrook, C.C., Kesmala T., Vorasoot, N., Akkasaeng, C. & Patanothai, A. (2009). Association of root, specific leaf area and SPAD chlorophyll meter reading to water use efficiency of peanut under different available soil water. Agricultural water management, 96, 790-798. Sprague, G.F. & Tatum, L.A. (1942). General versus specific combining ability in single crosses of corn. Journal of American Society of Agronomy, 34, 923- 932. Stalker, H.T. (1997). Peanut (Arachis hypogaea L.). Field Crop Research, 53, 205-217. Subbarao, G.V., Johansen, C., Slinkard, A.E., Nageswara, R.R.C., Saxena, N.P. & Chauhan, Y.S. (1995). Strategies for improving drought resistance in grain legumes. Plant Science, 14, 469-523. Sumathi, P., Madineni, S. & Veerabadhiran, P. (2010). Genetic Variability for Different Biometrical Traits in Pearl Millet Genotypes (Pennisetum glaucum L. R. BR.). Electronic Journal of Plant Breeding, 1, 437-440. Surihan, B., Patanothai, A. & Jogloy, S. (2005). Gene effect for specific leaf area and harvest index in peanut (Arachis hypogaea L.). Asian Journal of Plant Sciences, 4, 667-672. University of Ghana http://ugspace.ug.edu.gh 137 Swe, S.T. & Branch, W.D. (1986). Estimates of combining ability and heterosis among peanut cultivars. Peanut Science, 13, 70-74. Tanka, T., Tanaka H., Machida C., Watanabe M. & Machida Y. (2004). A new method for rapid visualization of defects in leaf cuticle reveals five intrinsic patterns of surface defects in Arabidopsis. The Plant Journal, 37, 139-146. Thirumala R.V., Bhadru, D., Murthy, K.G.K. & Bharathi, D. (2012). Genetic variability and association among the various characters in groundnut (Arachis hypogaea L). International Journal of Applied Biology and Pharmaceutical Technology, 337-341. Toukoua, D. (1986). ‘Institut National de Recherches Agronomiques du Niger (INRAN)’, in Agrometeorology of groundnut, Proceedings of an international Symposium, 21-26 August 1985, ICRISAT Sahelian Center, Niamey, Niger. ICRISAT, Patancheru, India, 11 p. Tsaur, W.L., Yang, K.H. & Lu, H.S. (1989). Heritability and correlation based on the F2 population from two crosses between large seeded Virginia bunch type and small seeded Spanish type peanuts. Journal Agricultural Research of China, 38, 191-200. Upadhyaya, H.D. & Nigam, S.N. (1994). Inheritance of two components of early maturity in groundnut (Arachis hypogaea L.). Euphytica, 78, 59-67. Upadhyaya, H.D. (2005). Variability for Drought Resistance Related Traits in the Mini Core Collection of Peanut. Crop Science, 45, 1432-1440. Upadhyaya, H.D., Reddy, L.J., Gowda C.L.L. & Singh, S. (2006) - Identification of diverse groundnut germplasm: Sources of early maturity in a core collection. Field Crops Research, 97, 261-271. Upadhyaya, H.D. & Sharma, S. (2011). Inheritance of drought resistance related traits in two crosses of groundnut (Arachis hypogaea L.). Euphytica, 177, 55-66. USDA, (2010). Top ten producers of peanut in 2008/2009. Foreign Agricultural Services. In: Wikipedia encyclopedia. en.wikipedia.org/wiki/peanut. Varshney, R.K., Bertioli, D.J., Moretzsohn, M.C., Vadez, V., Krishnamurthy, L., Aruna, R., Nigam, S. N., Moss, B. J., Seetha K., Ravi, K., He, G., Knapp, S. J., & Hoisington, D. A. (2009). The first SSR-based genetic linkage map for cultivated groundnut (Arachis hypogaea L.). Theoretical and Applied Genetics, 118, 729-739. Venkateswarlu, O., Reddy, R.K., Reddy P.V., Vasanthi R.P., Reddy Prasada H.K. Reddy, & Reddy Eswara, N.P. (2007). Identification of superior donor parents for drought tolerance and yield through combining ability analysis in groundnut. Legume Research 30, 128-132. University of Ghana http://ugspace.ug.edu.gh 138 Vetriventhan, M. & Nirmalakumari, A. (2007). Studies on variability parameters in pearl millet (Pennisetum glaucum (L.) R. Br.). Madras Agricultural Journal, 94, 118-120. Visscher, P.M., Hill, W.G., & Wray, N.R. (2008). Reviews: Fundamental concept s in genetics Heritability in the genomics era concepts and misconceptions. Nature Reviews Genetics, 9, 255-266. Vorasoot, N., Songsri, P., Akkasaeng, C., Jogloy, S., & Patanothal, A. (2003). Effect of water stress on yield and agronomic characters of peanut. Journal of Science and Technology, 25, 283-288. Vorasoot, N., Akkasaeng C., Songsri P., Jogloy S., & Patanothai A., (2004). Effect of available soil water on leaf development and dry matter partitioning in 4 cultivars of peanut (Arachis hypogaea L.). Journal of Science and Technology, 26, 787-794. Waliyar, F., Hamma, A., Bonkoungou, H. & Bose, J.P., (1994). Sources of resistance to Aspergillus flavus and aflatoxin contamination in groundnut genotypes in West Africa. Plant Disease, 78, 704-708. Warunyuwat, A., & Tongsri, S. (1990). Correlations of yield and yield component traits in advanced generations of peanut crosses. Thai Journal of Agriculture Science, 23, 19-22. Williams, E.J. & Drexler J.S. (1981). A non-destructive method for determining peanut pod maturity, pericarp, mesocarp, color, morphology, and classification. Peanut Science, 8, 134-141. Wright, G.C., Nageswara, R.R.C. & Farquar., G.D. (1994). Water-use efficiency and carbon isotope discrimination in peanut under water deficit conditions. Crop Science, 34, 92-97. Wuma, H., Jogloy S., Toomsan, B., & Sanitchon. J. (2009). Response to early drought for traits related to Nitrogen fixation and their correlation to yield and drought tolerance in Peanut (Arachis hypogaea L.). Asian Journal of Plant Sciences, 8, 138-145. Wynne, J.C., Emery, D.A. & Rice, P.W. (1970). Combining ability estimates in Arachis hypogaea L.II, Field performance of F1 hybrids. Crop Science. 10,713-715. Wynne, J.C., Rawlings, J.O. & Emery, D.A. (1975). Combining ability estimates in Arachis hypogaea L. III. F2 generation of intra and inter-subspecific crosses. Peanut Science, 2, 50-54. Yousaf A., Atta, B.M., Akhter, J., Monneveux, P. & Lateef, Z. (2008). Genetic variability, association and diversity studies in wheat (Triticum aestivum L) germplasm. Pakistan Journal of Botany, 40, 2087-2097. Zurayk, R., El-Awar, F., Hamadeha, S., Talhouk , S., Sayegh, C., Chehabb, A.G. & Al Shab K. (2001). Using indigenous knowledge in land use investigations: a participatory study in a semi-arid mountainous region of Lebanon. Agriculture, Ecosystems and Environment, 86, 247-262. University of Ghana http://ugspace.ug.edu.gh 139 APPENDICES APPENDIX 1: check list of focus group discussion Number Check list of the Focus Group Discussion Varieties used and seeds sources 1. Varieties used a. Identification and characterization of varieties - How many groundnut varieties do you use? - Can you describe each of those varieties? b. Assessment of varietal change over time - Varieties were cultivated for how many times before you change to a new variety? c. Varieties ranking - Can you rank all the varieties that are used in your village? 2. Seed management: a. Seed selection - How do you select groundnut seeds for next cropping season? b. Seed storage - Are you having any problem of storage? 3. Seed access - What are the sources of seeds used? - Are you having any problem to access seeds? 4. Socio-economic data a. Farmland size - Can you tell us the size of land roughly allocated to groundnut production? b. Equipment - Which types of equipment do you use for groundnut cultivation? 5. Farming system - Can you describe the farming system currently used in your village? 6. Planting & harvesting time - What are your groundnut planting times? - At which period is groundnut mostly harvested? Production constraints 7. Biotic stresses - Do you have any biotic stresses in groundnut production? - How do you manage insects and diseases? 8. Abiotic stresses - What are the major abiotic stresses that limit groundnut production? - Can you describe the methods used to manage drought locally? 9. Fertiliser application - Do you apply fertilizer? - Which types do you apply? 10. Weeding - How many times is weeding done in your groundnut filed? 11. Constraints ranking - Can you rank the main production constraints? Marketing 12. - Are you having any difficulties to sell your production? University of Ghana http://ugspace.ug.edu.gh 140 APPENDIX 2: Performance of 150 genotypes Number Entry % EMERG 7DAS ND 50% PF MD PY (g) BIO (g) HI PH (cm) 1 796 100 24 80 560 2000 0.28 52.33 2 ICIAR 19BT GH 100 24 75 930 1570 0.59 54.33 3 J 11 Niger 100 24 85 390 1860 0.2 54.33 4 ICG 3584 96 24 80 570 2050 0.27 54.33 5 ICGV-IS01835 96 24 85 690 1700 0.4 48 6 ICG 3312 92 26 85 450 1220 0.36 44 7 ICGV-IS 01827 92 26 85 570 1258 0.45 44.66 8 ICGV-IS 01852 92 24 85 670 1930 0.34 56.66 9 TS 32 -1 Niger 92 24 85 300 2100 0.14 58.66 10 ICG 12879 92 24 90 570 720 0.79 31.66 11 55-437 Niger 92 24 90 570 1450 0.39 48 12 ICG 15380 87.5 22 85 740 2340 0.31 56.66 13 SHITAOCHI 87.5 26 85 710 1260 0.56 49.33 14 ICIAR 12AT 87.5 28 85 210 700 0.3 46 15 ICG 1487 87.5 24 85 580 2760 0.21 45.33 16 ICG 4764 87.5 24 85 600 1380 0.43 54.66 17 ICG 81 87.5 24 90 480 985 0.48 49.33 18 ICGV-IS01859 87.5 24 90 830 1900 0.43 53.33 19 ICG 8106 86 24 85 300 2000 0.15 43 20 ICG 02022 83 24 80 580 1580 0.36 50.33 21 ICIAR 19 BT NG 83 24 80 580 1100 0.52 54 22 T44 - 88 83 28 85 530 1830 0.28 49.33 23 T169 - 83 83 26 85 790 1560 0.5 46.66 24 ICG 10384 83 26 90 440 1930 0.22 47.33 25 ICG 2019 83 24 90 630 1680 0.375 55.33 26 ICG 3421 83 26 90 1210 2760 0.43 58.66 27 Dayo early 83 26 90 870 1700 0.51 48.66 28 T131 - 83 79 26 85 720 2354 0.3 43.33 29 ICG 1519 79 24 85 550 1420 0.38 50.66 30 T9 - 89 79 26 85 630 1380 0.45 52.66 31 ICG 405 79 26 85 760 1180 0.64 46 32 ICG 11249 79 24 90 810 1570 0.51 46.66 33 ICGV-SM 99507 Niger 79 24 91 930 1910 0.48 44.33 34 Chico 75 24 75 540 1860 0.29 39.33 35 ICG 1415 75 26 85 730 2100 0.34 44.66 36 ICG 9315 75 24 85 470 2520 0.18 55.33 37 ICGV-IS01836 75 28 90 750 2250 0.33 50.66 38 T6 - 96 75 24 90 1070 1400 0.76 44 39 Fleur 11 Niger 71 26 85 590 1760 0.33 43 40 ICG-3736 67 26 85 630 1250 0.5 48.66 University of Ghana http://ugspace.ug.edu.gh 141 41 ICG 15236 67 28 85 420 1090 0.38 48.66 42 T 177 - 83 67 26 90 800 1960 0.4 46.66 43 ICG 9809 67 24 90 890 2040 0.43 47.33 44 TX 903652 62.5 26 85 500 2330 0.21 42.33 45 ICG 14106 62.5 24 85 450 1050 0.42 51.33 46 ICG-12991 62.5 28 90 1000 1880 0.53 53.33 47 T127 - 83 58 26 85 480 1340 0.35 51.33 48 T2 - 2007 58 26 85 1240 1780 0.69 52 49 ICG 6703 58 22 90 650 1540 0.42 44 50 T13-89 54 28 85 720 1760 0.4 48.66 51 O-20 54 26 90 390 1720 0.22 54.66 52 T45 - 87 54 26 90 530 1470 0.36 44 53 T181 - 83 50 26 90 700 1940 0.36 47.33 54 T49-87 46 28 85 1220 1987 0.61 42.66 55 ICG 12921 46 24 85 350 1470 0.23 45.33 56 T119-83 46 26 91 690 1587 0.43 51.33 57 T4 -96 41 28 90 990 3040 0.32 48.66 58 ICG 4670 37.5 28 85 700 1820 0.38 52 59 ICG 9346 37.5 28 85 740 1360 0.54 55.33 60 ICG 1823 33 28 91 800 2230 0.35 46.33 61 ICG 4728 29 28 90 600 1850 0.32 48.66 62 ICG 2738 25 26 90 570 1790 0.31 50 63 IGC 3386 21 28 90 870 2540 0.34 60.33 64 ICG 3775 4 28 90 490 1660 0.29 45.33 65 CMA 8705 0 0 0 0 0 0 0 66 CMA 87111 0 0 0 0 67 ICG-4747 0 0 0 0 0 68 CMA 87082 0 0 0 0 0 0 0 69 CMA 87063 0 0 0 0 0 0 0 70 CMA 95044 0 0 0 0 0 0 0 71 CMA 87007 0 0 0 0 0 0 0 72 CMA 95006 0 0 0 0 0 0 0 73 CMA 92072 0 0 0 0 0 0 0 74 CMA 95008 0 0 0 0 0 0 0 75 ICG 434 0 0 0 0 0 0 0 76 RRB Niger 87.5 26 95 680 1460 0.46 42.66 77 ICGV-IS 96 -895 75 26 95 710 1860 0.38 41.33 78 ICGV 02266 54 26 95 620 2380 0.26 49.66 79 ICGV 97182 67 26 95 480 1520 0.31 42.66 80 KPANIELLI-ICGV90084 33 26 95 540 2570 0.21 50 81 ICGV-87281 41 30 95 650 1380 0.47 50 82 ICG 397 87.5 24 95 870 1580 0.55 40.33 83 ICG 1415 4 28 95 360 1320 0.27 39.33 University of Ghana http://ugspace.ug.edu.gh 142 84 ICGV-92101 67 26 95 430 976 0.44 44.66 85 ICG 7906 12.5 30 95 1100 2550 0.43 50.33 86 SH470P 71 26 95 1090 2040 0.53 42.66 87 ICG 3657 29 30 95 830 1720 0.48 51.33 88 ICG 3343 75 26 95 840 1350 0.62 52 89 ICGV - SM 99511 50 24 95 890 1980 0.44 48.33 90 ICG 1914 8 38 95 440 830 0.53 33.66 91 ICGV-87003 83 26 95 1330 1720 0.77 48.66 92 ICG 9777 29 30 95 220 890 0.24 43.66 93 T3 - 96 62.5 26 95 760 1460 0.52 50.66 94 ICG-6222 83 24 95 930 2500 0.37 62 95 ICG-476 4 32 98 700 1800 0.38 33.33 96 ICG 9666 29 28 98 550 2320 0.23 42.66 97 JL24 79 24 100 1160 1850 0.62 46.33 98 ICGV-87123 0 24 100 1290 1830 0.7 47.33 99 Te3 83 28 100 900 1940 0.46 42.66 100 GUSIE-BALIN(ICGV92099) 79 28 100 910 2460 0.36 42 101 Edorpo-Munikpa (SARGV 88001) 71 28 100 680 3080 0.22 50 102 ICG 9346 92 26 100 390 2300 0.16 56 103 T47 - 87 92 26 100 1110 1300 0.85 46 104 ICG 5494 12.5 28 100 440 1070 0.41 42.33 105 ICG 3746 58 28 100 450 2380 0.18 52.66 106 ICGV-IS 01850 75 24 100 1140 1900 0.6 48.66 107 T1 - 95 0.75 28 100 660 1080 0.61 50 108 ICGV-SM 99504 Niger 0.92 24 100 1080 1860 0.58 55.33 109 TX 903620 0.83 30 100 470 910 0.51 38 110 T5 - 96 71 26 100 960 2250 0.42 49 111 ICIAR 7B 92 26 100 830 1540 0.53 46.66 112 ICGV-IS01851 79 24 100 640 880 0.72 43.66 113 ICG 296 46 26 110 730 1760 0.41 47.33 114 ICG 7756 83 28 110 510 3500 0.14 48 115 ICG 76 54 32 110 470 2680 0.17 29.66 116 NC 7 71 28 110 890 1760 0.5 40.66 117 ICG 15232 0 30 110 430 900 0.47 39.33 118 MANIPINTAR 58 28 110 220 3580 0.06 47.33 119 ICGV-IS 96 -814 33 28 110 1020 2230 0.45 43.66 120 ICGV - SM 99505 71 24 110 600 1440 0.41 43.33 121 ICG 1142 0 42 110 100 520 0.19 47 122 ICG 3240 87.5 24 110 580 1930 0.3 45.33 123 ICG 2716 17 26 110 260 2630 0.09 50 124 ICGV-91225 92 28 110 590 3130 0.18 53.33 125 ICGV-IS01835 58 30 110 780 2190 0.35 47.66 126 ICG 15233 0 38 110 340 870 0.39 42 University of Ghana http://ugspace.ug.edu.gh 143 127 NKATIESARI 71 28 110 890 2360 0.37 43 128 ICG-10142 TIFFON-8 4 39 110 280 1960 0.142 30.33 129 ICGV(FDRS)-20X F-MIX-39 83 26 110 840 1870 0.44 40 130 ICG 6813 33 30 110 140 1560 0.08 32.66 131 ICG 5475 79 26 110 840 1500 0.56 50 132 Tainan - 9 4 30 110 720 1270 0.56 42 133 ICG 3027 33 32 110 560 1770 0.31 44.66 134 ICGV-IS92093 62.5 28 110 860 1620 0.53 44 135 ICG 1534 92 26 110 540 2710 0.19 51 136 ICGV-92087 54 28 110 630 2580 0.24 52.66 137 ICG 8751 87.5 26 110 510 810 0.62 38 138 MIX SINK 24 67 26 110 960 1540 0.62 40 139 ICG 297 62.5 28 110 650 1280 0.5 48 140 ICG 1703 0 30 110 150 250 0.6 33.33 141 ICGV-IS 01820 87.5 26 110 1000 1550 0.64 51.66 142 F-MIX 75 28 110 720 2000 0.36 39 143 ICG-3301 12.5 30 110 480 2520 0.19 47.33 144 ICG 1834 100 26 110 1000 1580 0.63 56.66 145 ICG-1697 8 32 110 660 1260 0.52 44.66 146 CMA 87073 0 38 110 50 120 0.41 26 147 ICG 862 46 30 110 540 1270 0.42 22.33 148 ICG 11088 0 38 110 30 420 0.07 22 149 ICG 6022 17 28 110 590 900 0.65 48.66 150 ICG 2773 54 32 111 410 2110 0.19 35.33 % EMERG: Percentage of emergence, 50% PF: 50% plant flowering, DM: Date to maturity, PY: Pod yield, BIO: Biomass, HI: Harvest index, PH: Plant height University of Ghana http://ugspace.ug.edu.gh 144 APPENDIX 3: Mean performance of 100 genotypes under both water regimes Genotypes Pod Yield SCMR 45 DAS SCMR 60 DAS SCMR 90 DAS ws ww ws ww ws ww ws ww Dayo Early 205.5 191 37.9 41.4 51.4 44.5 44.15 39.15 Edorpo-Munikpa (SARGV 88001) 97.5 134.5 42.3 39.7 51.05 46.35 47.25 44.45 F-MIX 128.5 111 38.5 38.7 46.15 44.85 43.75 43.6 Fleur 11 Niger 182.5 318 42.85 40.8 46.15 43.95 49.15 41.4 GUSIE-BALIN(ICGV92099) 102 128 40.05 36.35 49.95 42.7 45.35 45.95 ICG 11249 204 204 44.3 36.2 48.75 44.05 46.25 41.7 ICG 12879 119 305 40.15 34.6 48.15 39.95 45.85 36.05 ICG 14106 102.5 186.5 39.9 39.9 44.9 44.75 45.15 41.1 ICG 1415 137.5 224.5 38.75 40.3 49 41.85 48.35 43.95 ICG 1519 172.5 266.5 40.75 39.4 52.8 44.8 40.75 38.9 ICG 15232 146.5 164 41.75 40.85 53.7 50 45.45 47.35 ICG 15236 56.5 173 44.65 41.65 51.6 44.95 47.1 41.1 ICG 1534 224 264.5 34.55 32.05 40.65 38.45 45.25 37.8 ICG 15380 133.5 221.5 38.25 38 51.9 43 44.55 40.75 ICG 1823 160.5 253 39.6 39.85 47.2 46.25 45 41.8 ICG 1834 223 229.5 36.2 38.4 44.5 42.45 46.4 39.6 ICG 1914 116 233.5 46.25 40.5 51.15 49.5 47.45 43 ICG 2716 186 345.5 40.3 42.6 45.4 43.9 45.3 40.45 ICG 2738 326.5 191 36.85 40.35 47.75 44.95 45.7 41.8 ICG 2773 98.5 142.5 43.35 40.2 49.8 46.95 47.45 45.85 ICG 296 105 120 40.8 36.55 44.35 44.1 44.9 40.5 ICG 297 279 286.5 40.65 39.3 40.85 38.1 41.5 37.7 ICG 3027 234 146.5 41.15 42.35 49 49.1 46.75 39.55 ICG 3240 121 269.5 38.8 41.55 43.95 45.5 43.4 41.15 ICG 3343 213 280 43 41.3 51.25 43.35 46.8 40.7 ICG 3421 253.5 309.5 44.05 38.9 47.4 45.15 43.45 40.55 ICG 3657 167.5 301 42.25 37 50.7 44.9 43 42.5 ICG 3746 206 265.5 42.8 38.4 47.6 40.35 45.2 41.7 ICG 3775 63.5 242 41 38.55 49.05 44 45.75 42.55 ICG 397 204.5 291 38.85 37.7 42.6 41.25 39.7 38.2 ICG 405 138.5 267.5 44.2 41.5 50.4 46.25 47.05 45.25 ICG 4728 224.5 337.5 40.15 37.65 45.45 46.2 44.85 39.9 ICG 5475 261 335.5 40.1 42.55 46.1 46.75 46 41.5 ICG 5494 195.5 235 41.8 42.55 47.8 40.85 48.3 37.65 ICG 6022 300 218.5 42.35 39.75 44.4 42.25 44.25 38.25 ICG 6703 342 194 39.95 35.05 40.5 43.75 44.5 38.85 ICG 76 128.5 92 38.1 40.6 48.65 44.6 49.35 40.85 ICG 7756 63.5 105 45.1 41.25 48.9 53.5 48.75 46.45 ICG 7906 191 172.5 41.25 39.25 50.45 43.25 45.35 41.4 University of Ghana http://ugspace.ug.edu.gh 145 ICG 862 62 98 41.2 36.55 51.15 41.05 50.7 41.95 ICG 8751 197.5 283 36.6 37.85 42.05 45.5 45.05 39.65 ICG 9346 162.5 315 38.75 40.05 48.65 45.75 46.2 41.65 ICG 9666 206 94.5 41.5 41.55 47.15 44.85 50.5 43.25 ICG 9777 53.5 78 40.9 40.8 44.4 44.4 49.85 41.15 ICG 9809 163 138.5 42.6 40.4 50.55 44.9 46.75 46.25 ICG-10142 TIFFON-8 54.5 64.5 43.05 41.5 53.75 46.45 51.95 45.9 ICG-12991 207.5 196 40.5 38.55 48.6 45.05 43.05 41.35 ICG-1697 118 95.5 43.15 43.35 41.7 41.5 40.4 41.25 ICG-3301 236.5 188.5 36.05 36.65 44.55 43.4 41.05 38.8 ICG-476 110.5 321.5 36.65 34.45 40.95 42.95 41.45 38.55 ICG-6222 210 410 41.55 38.9 45.6 44.1 44.3 39.15 ICGV - SM 99505 161.5 302 40.7 40.1 43.55 46.8 44.7 40.35 ICGV - SM 99511 309 524 42.3 35.2 48.25 43.05 44.6 46.35 ICGV 02266 310 203.5 42.15 43.8 45.75 47.45 46.55 45.65 ICGV 97182 149.5 314.5 42.65 40.15 44.8 44.3 45.45 43.2 ICGV(FDRS)-20X F-MIX-39 82.5 194.5 41.7 41.3 50.2 45.75 46.15 42.4 ICGV-87003 138 261.5 39.9 36.4 45.8 44.3 44.3 38.8 ICGV-87123 302.5 192.5 40.85 37.3 50.35 43.4 41.5 42.6 ICGV-87281 207 213.5 41.2 39.25 41.9 40.6 43.95 40.3 ICGV-91225 194.5 155 36.65 38.7 47.15 45.3 46.5 45.25 ICGV-92087 91 73.5 37.6 37.7 47.1 45.9 45.5 44.1 ICGV-92101 87 172 47.2 40.55 50.5 45.45 47.6 46.35 ICGV-IS 01820 220 222.5 48.4 43.85 50.25 48 47.45 45.7 ICGV-IS 01850 115 335 43 40.6 52.1 48.45 46.15 41.65 ICGV-IS 01852 282.5 221 42.4 42.45 54.35 52.8 46.75 44.65 ICGV-IS 96 -814 65 69.5 42.85 39.1 47.25 37.7 46.95 41.45 ICGV-IS 96 -895 82 257.5 40.25 38.7 48.1 44.75 48.35 43.6 ICGV-IS01835 183 117 41.9 44.3 49.05 44 43.95 40.1 ICGV-IS01836 270.5 343 41.25 40.6 48.7 43.75 48.2 39.5 ICGV-IS01851 34.5 104 43.95 42.3 56.2 50 53.9 46.35 ICGV-IS01859 205.5 299 44.05 44.4 48.9 52.75 48.7 43.05 ICGV-IS92093 122 105 38 40.7 48.65 47.85 48.1 44.95 ICGV-SM 99504 Niger 194.5 260.5 43.25 40.65 51.2 42.6 45.65 40.9 ICGV-SM 99507 Niger 162 319 43.45 40.5 47.9 44.7 45.75 42.35 ICIAR 7B 106.5 298 41.3 41.4 45.6 46.8 49.6 43.95 IGC 3386 240.5 208 39.4 38.75 42.8 46.8 42.4 40.1 KPANIELLI-ICGV90084 59.5 114.5 40.55 41.55 46.35 45.25 45.45 43.7 MANIPINTAR 97.5 84.5 45.7 43 50.55 45.3 45.7 45.65 MIX SINK 24 118.5 129 39.85 35.7 42.5 42.1 46.95 43.9 NC 7 123 171.5 40.9 41.35 46.7 46.35 47.45 44.4 NKATIESARI 110 139 38.25 38.05 45.65 46.05 49.4 45.45 SH470P 164.5 327.5 41.05 39.55 47 45.35 44.5 38.75 University of Ghana http://ugspace.ug.edu.gh 146 SHITAOCHI 142 246 42.15 31.3 45.9 47.25 45.6 38.2 T1 - 95 165 205 42.65 40.3 51.7 41.8 47.5 40.45 T119-83 158 190.5 39.55 42.05 48.4 43.45 43.55 39.45 T13-89 274.5 315.5 40.4 38.45 46.95 44.6 51.15 40.8 T169 - 83 170 295 40.05 37.7 51.65 49.6 43.65 43.15 T181 - 83 136.5 238.5 39 39.35 46.8 45.4 46.65 39.8 T2 - 2007 145 283 40.15 40.5 49.55 42.4 47.5 40.1 T4 -96 134.5 309.5 40.95 39.5 49 44.35 46.5 40 T45 - 87 133.5 271 43.3 39.75 53.05 46.2 46.95 40.75 T47 - 87 188 405.5 38.8 40.25 53.3 42.5 45.75 39.8 T49-87 179 265.5 39.45 38.15 44.9 39.45 45.6 40.65 T5 - 96 184 160 45.2 40.4 47.8 43.6 47.5 39.35 T6 - 96 119.5 233.5 48.7 38.7 48.2 48.35 45.3 41.8 T9 - 89 166 210 38.65 42.6 46.15 45.7 43.95 40.95 Tainan - 9 157 159 40.65 41.55 47.5 42.05 48.6 40.1 Te3 116 236.5 44.65 44.5 47.45 44.7 43.4 39.5 TX 903620 181.5 245.5 41.45 39.65 45.9 46.3 47.7 43.1 Mean 165 222.6 41.16 39.66 47.78 44.77 45.97 41.75 CV % 46.11 7.03 8.59 6.06 SE 89.37 2.841 3.973 2.658 University of Ghana http://ugspace.ug.edu.gh