University of Ghana http://ugspace.ug.edu.gh GENETIC STUDIES OF PEARL MILLET (Pennisetum glaucum (L.) R. Brown) DOWNY MILDEW RESISTANCE IN SENEGAL By GHISLAIN KANFANY (10496573) THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE AWARD OF DOCTOR OF PHILOSOPHY DEGREE IN PLANT BREEDING WEST AFRICA CENTRE FOR CROP IMPROVEMENT COLLEGE OF BASIC AND APPLIED SCIENCES (CBAS) UNIVERSITY OF GHANA LEGON December, 2017 University of Ghana http://ugspace.ug.edu.gh DECLARATION I hereby declare that, except for references to works of other researchers, which have been duly cited, this work is my original research and that neither part nor whole has been presented elsewhere for the award of a degree. .................................................. Ghislain Kanfany (Student) .................................................. Prof. Pangirayi B. Tongoona (Supervisor) .................................................. Prof. Eric Y. Danquah (Supervisor) .................................................. Prof. Samuel K. Offei (Supervisor) .................................................. Dr Agyemang Danquah (Supervisor) .................................................. Dr Ndiaga Cissé (Supervisor) i University of Ghana http://ugspace.ug.edu.gh ABSTRACT Pearl millet is an important cereal crop for food security in Senegal. However, the production of this crop has stagnated due to several factors, among them, downy mildew. A participatory rural appraisal and a survey conducted in the groundnut basin of Senegal involving 150 farmers revealed the occurrence of the disease across the main pearl millet growing areas. Almost all the farmers interviewed (99%) were able to recognize the disease and 94% of them observed it in their fields during the just ended season of 2014. However, they neglected the damages caused by the disease and did not rank it as the main constraint to pearl millet production. They rather ranked Striga and low soil fertility as the main pearl millet production constraints. They unanimously considered grain yield as the most important trait to be incorporated in a new pearl millet variety. The spatio- temporal pathogenic variation of the downy mildew disease was also studied using pearl millet downy mildew resistance differential lines. Significant differences were observed among the genotypes, location, year and their interactions for downy mildew incidence providing evidence of spatio-temporal pathogenic variation. Except for ICMR356, the downy mildew resistant differential lines differed in their resistance to the pathogen from the four agro-ecological zones across the years. The highest average downy mildew incidence over the two years was recorded in Nioro (18.7%), Kolda (17.6%) and Bambey (12.1%) research stations while the mean downy mildew incidence was the lowest in Sinthiou (6.9%). Therefore, Kolda, Nioro and Bambey research stations should be used for screening pearl millet materials. Thus, a set of 99 inbred lines, derived from landraces collected in several West and Central African countries, were evaluated under field conditions in Bambey and Nioro research stations during the rainy season of 2016. A highly significant differences were observed among lines for the downy mildew disease parameters and other agronomic traits. The lines were classified into 3 clusters with disease parameters and plant height as the most discriminant factors. One of the groups contained 38 lines ii University of Ghana http://ugspace.ug.edu.gh that were characterized as resistant, early flowering, and the tallest plants with longest panicles. Out of these 38 lines, 17 were crossed with SOSAT C88 and Souna3 following a line x tester mating design in order to study their combining ability. General combining ability (GCA) and specific combining ability (SCA) means squares were significant for most of the traits indicating that both additive and non-additive gene effects were involved in the control of the inheritance of these traits. However, the contribution of GCA to the total mean squares was higher than that of SCA for all the traits, meaning that additive gene action was more important in the inheritance of these traits. Lines IBL003-B-1, IBL091-1-1, IBL095-4-1, IBL110-B-1 and IBL206-1-1 had positive GCA effects for grain yield and negative GCA effects for downy mildew, flowering time and plant height. These lines can be used as parents to create synthetic varieties or F1 hybrids adapted to the groundnut basin agro-ecological zone. Positive mid-parent, best parent and standard heterosis were observed for hybrids IBL206-1-1 x Souna3; IBL091-1-1 x Sosat C88; IBL206-1-1 x Sosat C88; IBL001-4-1 x Souna3 and IBL003-B-1 x Sosat C88. An association analysis was conducted with 77 genotypes to identify single nucleotide polymorphisms (SNPs) markers associated with downy mildew resistance. Three sub-groups based on panicle length were identified and a rapid linkage disequilibrium decay was noted. Sixteen significant markers, located on four linkage groups (LG) were identified across locations. Among these SNPs, the ones located on LG6 explained 15 to 20% of the downy mildew phenotypic variation and were consistent across the two locations. These SNPs were located on the same LG that a Quantitative Trait Loci detected in a previous study against population of pathogens that originated from Senegal while the ones located on LG1, LG3 and LG4 were discovered for the first time under Senegalese environments and therefore could be considered as new markers. With the pearl millet genome sequenced, candidate genes surrounding these identified genomic regions can be identified and validated. iii University of Ghana http://ugspace.ug.edu.gh DEDICATION This thesis is dedicated to the Glory of The Almighty GOD; My beloved wife (Francine), kids (Greguy and Pascal) and family. Your love, prayers and support throughout this study made it easier for me to persevere; Thanks a lot for your support. iv University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENTS I would like to express my infinite gratitude to The Almighty GOD for making my dream a reality. Without your guidance and sustenance this modest work would not have reached its efficacious end. I could not have done this thesis without the funding I received from the West Africa Agricultural Productivity Programme (WAAPP) and the Dryland Cereal Programme in the form of a PhD fellowship through the West Africa Centre for Crop Improvement (WACCI), University of Ghana. I would like also to Thank IRD staff, especially Drs Laurent Laplaze and Maryline Debieu for support I received for the GBS data analysis used for the Genome Wide Association Studies under the New-Pearl millet Project. I am grateful to WACCI staff, lecturers and English proficiency professors. I deeply appreciate the strong moral support and knowledge I received during the one-year course in Ghana. It was not evident, but your full assistance made this academic journey an appreciable experience. I wish to express my sincere thanks to my academic supervisor committee composed of Professors Pangirayi Tongoona, Samuel Offei, Eric Danquah and Agyemang Danquah for their guidance, valuable inputs and constructive criticism. I am indeed grateful to my In-country supervisors’ Drs Ndiaga Cisse, Amadou Fofana and Ndjido Kane for their immense contributions to the completion of this study. You were more than supervisors and served as references. You believed in my capacities and gave me this opportunity. Your assistance during the fieldwork and preparation of this manuscript are fully acknowledged. Thanks to my mentor Dr Tom Hash. You gave me the opportunity to be trained on downy mildew screening techniques at ICRISAT Sadore and provided me with the germplasm used for this thesis. v University of Ghana http://ugspace.ug.edu.gh My sincere thanks and gratitude go to Professor Vernon Gracen, Drs Beatrice, Moctar Kante, Charles Nwankwo, Nana Abaka and Cyril Diatta for reading and advising me through this work. Special thanks to the Senegalese Agricultural Research Institute (ISRA), especially CNRA and CERAAS, for providing me all the setting for my field research. Thanks to my colleague and WACCI fellows, ISRA/CNRA and ISRA/CERAAS students for your encouragements and prayers. Wish you all the best, by the Will of GOD. I am also indebted to the technicians, BSc and MSc students especially Benjamin Badji, Ngor Sene, Ange Zoclanclounon, Georgina Ehemba, Phalone Meli, Diye Sow, Junior Bruno Ndiaye, Modou Ngom, Fatmata Ly, Bamba Sokhna, Barnabe Diatta and Ndiouga Samb for helping me in the field data collection. vi University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS DECLARATION ............................................................................................................................. i ABSTRACT .................................................................................................................................... ii DEDICATION ............................................................................................................................... iv ACKNOWLEDGEMENTS ............................................................................................................ v LIST OF TABLES ......................................................................................................................... xi LIST OF FIGURES ..................................................................................................................... xiii LIST OF ABBREVIATIONS ...................................................................................................... xiv CHAPTER ONE ............................................................................................................................. 1 1.0. GENERAL INTRODUCTION ............................................................................................................ 1 CHAPTER TWO ............................................................................................................................ 5 2.0. LITERATURE REVIEW .................................................................................................................... 5 2.1. Origin and distribution of pearl millet ...................................................................................... 5 2.2. Taxonomy and biology of pearl millet ...................................................................................... 5 2.3. Economic importance of pearl millet ....................................................................................... 6 2.4. Farmers’ participation in plant breeding .................................................................................. 7 2.5. Production constraints ............................................................................................................. 8 2.5.1. Abiotic constraints. ................................................................................................ 8 2.5.2. Biotic constraints ................................................................................................... 9 2.6. Downy mildew of pearl millet ................................................................................................ 10 2.6.1. Pathogen diversity ............................................................................................... 10 2.6.2. Downy mildew life cycle, epidemiology and control methods ........................... 10 2.6.3. Economic importance of downy mildew ............................................................ 12 2.6.4. Sources of downy mildew resistance .................................................................. 12 2.6.5. Mechanism of downy mildew resistance ............................................................ 13 2.6.6. Breeding for downy mildew resistance ............................................................... 14 2.6.7. Quantitative Trait Loci (QTL) mapping for downy mildew resistance .............. 15 2.6.8. Marker assisted deployment of downy mildew resistance .................................. 16 2.7. Combining ability study in pearl millet ................................................................................... 17 CHAPTER THREE ...................................................................................................................... 19 3.0. FARMERS’ PRODUCTION CONSTRAINTS, PERCEPTIONS ON DOWNY MILDEW AND VARIETAL PREFERENCES OF PEARL MILLET IN THE GROUNDNUT BASIN OF SENEGAL .......................................... 19 3.1. Introduction ............................................................................................................................ 19 vii University of Ghana http://ugspace.ug.edu.gh 3.2. Materials and Methods .......................................................................................................... 20 3.2.1. Study area ............................................................................................................ 20 3.2.2. Sampling method. ................................................................................................ 21 3.2.3. Data collection ..................................................................................................... 21 3.2.4. Data Analysis ...................................................................................................... 22 3.3. Results .................................................................................................................................... 23 3.3.1. Demographic characterization of farmers interviewed ....................................... 23 3.3.2. Importance of pearl millet ................................................................................... 25 3.3.3. Farmers’ perceptions on the downy mildew disease ........................................... 25 3.3.4. Production constraints ......................................................................................... 26 3.3.5. Farmers’ trait preferences .................................................................................... 27 3.3.6. Production level and fertilizer ............................................................................. 27 3.3.7. Cultivars grown and their provenance ................................................................ 28 3.3.8. Farming practices ................................................................................................ 29 3.4. Discussion ............................................................................................................................... 32 3.5. Conclusion .............................................................................................................................. 36 CHAPTER FOUR ......................................................................................................................... 37 4.0. ASSESSMENT OF PEARL MILLET DOWNY MILDEW RESISTANT DIFFERENTIAL LINES IN SENEGAL 37 4.1. Introduction ............................................................................................................................ 37 4.2. Materials and Methods .......................................................................................................... 38 4.2.1. Pearl millet differential lines ............................................................................... 38 4.2.2. Experimental design ............................................................................................ 39 4.2.3. Data collection ..................................................................................................... 40 4.2.4. Data analysis ....................................................................................................... 40 4.3. Results .................................................................................................................................... 41 4.3.1. Seasonal conditions ............................................................................................. 41 4.3.2. Variation in resistance for DM of tested lines ..................................................... 41 4.4. Discussion ............................................................................................................................... 43 4.5. Conclusion .............................................................................................................................. 45 CHAPTER FIVE .......................................................................................................................... 46 5.0. EVALUATION OF PEARL MILLET INBRED LINES IN DOWNY MILDEW PREVALENCE AREAS IN SENEGAL ................................................................................................................................................. 46 5.1. Introduction ............................................................................................................................ 46 viii University of Ghana http://ugspace.ug.edu.gh 5.2. Materials and Methods .......................................................................................................... 47 5.2.1. Plant material and geographical location of the study regions ............................ 47 5.2.2. Experimental design ............................................................................................ 48 5.2.3. Field evaluation for downy mildew resistance .................................................... 49 5.2.4. Data collection ..................................................................................................... 50 5.2.5. Data analysis ....................................................................................................... 51 5.3. Results .................................................................................................................................... 52 DMI = Downy mildew incidence; DMS = Downy mildew severity ....................................................... 57 5.4. Discussion ............................................................................................................................... 57 5.5. Conclusion .............................................................................................................................. 59 CHAPTER SIX ............................................................................................................................. 60 6.0. HETEROSIS AND COMBINING ABILITY FOR DOWNY MILDEW RESISTANCE AND GRAIN YIELD OF PEARL MILLET IN SENEGAL ..................................................................................................................... 60 6.1. Introduction ............................................................................................................................ 60 6.2. Materials and methods .......................................................................................................... 62 6.2.1. Plant material and mating design ........................................................................ 62 6.2.2. Study sites, experimental design, and field management .................................... 63 6.2.3. Data collection ..................................................................................................... 64 6.2.4. Data analysis ....................................................................................................... 65 6.3. Results .................................................................................................................................... 68 6.3.1. Performance of hybrids and parents across locations ......................................... 68 6.3.2. Combining ability analysis across locations ....................................................... 73 6.3.3. Relative contributions of mean squares for additive and non-additive effects ... 73 6.3.4. General combining ability effects ....................................................................... 74 6.3.5. Specific combining ability effects ....................................................................... 76 6.3.6. Heterosis for grain yield across locations ........................................................... 77 6.4. Discussion ............................................................................................................................... 79 6.5. Conclusion .............................................................................................................................. 82 CHAPTER SEVEN ...................................................................................................................... 83 7.0. ASSOCIATION ANALYSIS OF DOWNY MILDEW RESISTANCE IN PEARL MILLET UNDER SENEGALESE ENVIRONMENTS ................................................................................................................ 83 7.1. Introduction ............................................................................................................................ 83 7.2. Material and methods ............................................................................................................ 85 7.2.1. Plant materials ..................................................................................................... 85 ix University of Ghana http://ugspace.ug.edu.gh 7.2.2. Phenotypic data ................................................................................................... 85 7.2.3. Genotypic data ..................................................................................................... 85 7.2.4. Analysis of data ................................................................................................... 87 7.3. Results .................................................................................................................................... 88 7.3.1. Phenotypic variation ............................................................................................ 88 7.3.2. Marker distribution on the linkage group, linkage disequilibrium and population structure ............................................................................................................................. 89 7.3.3. Association mapping ........................................................................................... 90 7.4. Discussion ............................................................................................................................... 93 7.5. Conclusion .............................................................................................................................. 96 CHAPTER EIGHT ....................................................................................................................... 97 8.0. GENERAL CONCLUSION AND RECOMMENDATIONS .................................................................. 97 8.1. General conclusion ................................................................................................................. 97 8.2. Recommendations ................................................................................................................. 98 REFERENCES ........................................................................................................................... 100 APPENDICES ............................................................................................................................ 121 x University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES Table 3.1: Characteristics of selected pearl millet production areas in Senegal ........................... 21 Table 3.2: Characteristics of farmers interviewed ........................................................................ 24 Table 3.3: Land area planted per farmer (ha) and crops grown across the locations ................... 25 Table 3.4: Farmers’ perceptions of the downy mildew disease .................................................... 26 Table 3.5: Mean rank of pearl millet constraints across the locations .......................................... 27 Table 3.6: Mean rank of the most important preferred traits across the locations ........................ 27 Table 3.7: Estimated yield and NPK fertilizer applied ................................................................. 28 Table 3.8: Landraces and bred-cultivars grown and their provenance ......................................... 29 Table 3.9: Pearl millet production practices in the groundnut basin of Senegal .......................... 31 Table 4.1: List of plant material used in the study ........................................................................ 39 Table 4.2: Temperature and relative humidity of the four sites from sowing to 30 DAS ............ 41 Table 4.3: Effects of location, year, line, and their interactions for DMI ..................................... 42 Table 4.4: Average downy mildew incidence (%) and relative variation on tested pearl millet lines across locations during 2015 and 2016 rainy seasons .................................................................. 42 Table 5.1: List of genotypes and their respective origin ............................................................... 48 Table 5.2: F value for measured traits under artificial downy mildew infestation fields ............. 52 Table 5.3: Performance of the evaluated genotypes at Bambey and Nioro research stations ...... 53 Table 5.4: Performance of the downy mildew disease-free and highly susceptible pearl millet genotypes across Bambey and Nioro research stations ................................................................ 54 xi University of Ghana http://ugspace.ug.edu.gh Table 5.5: Correlation matrix between pairs of studied variables ................................................ 55 Table 5.6: Characteristics of the identified clusters based on the HAC analysis ......................... 55 Table 5.7: Lambda Wilk test from the six variables ..................................................................... 57 Table 6.1: List of parental lines and check used in the study ....................................................... 63 Table 6.2: Mean squares for studied traits across locations .......................................................... 68 Table 6.3: Performance of tested genotypes for studied traits across sites ................................... 70 Table 6.4: Mean flowering time, yield and related traits of genotypes per site ............................ 72 Table 6.5: Mean squares for combining ability for studied traits across locations ....................... 73 Table 6.6: Estimates of GCA effects of lines and testers evaluated across the two sites ............. 76 Table 6.7: Estimates of SCA effects for hybrids evaluated across the two sites .......................... 77 Table 6.8: Mean grain yield and heterosis of pearl millet hybrid across locations ...................... 78 Table 7.1: Analysis of variance for augmented design ................................................................. 87 Table 7.2: Mean square and broad sense heritability from ANOVA for DMI traits under artificial downy mildew infestation at Bambey and Nioro during the rainy season 2016 .......................... 88 Table 7.3: Pearson correlation coefficients between observed variables and membership probability ..................................................................................................................................... 90 Table 7.4: Markers significantly associated with downy mildew resistance at Bambey and Nioro research stations ............................................................................................................................ 92 xii University of Ghana http://ugspace.ug.edu.gh LIST OF FIGURES Figure 2-1: Downy mildew disease cycle (Singh et al., 1993) ..................................................... 11 Figure 3-1: Groundnut basin areas in Senegal included in the participatory rural appraisal and baseline survey .............................................................................................................................. 20 Figure 5-1: Field experimental at Bambey research station ......................................................... 49 Figure 5-2: Factorial discriminant analysis based on the study entries ........................................ 56 Figure 6-1: Proportion of total mean squares of studied traits attributable to GCAm, GCAf and SCA across locations ............................................................................................................................. 74 Figure 7-1: Linkage disequilibrium (LD) decay plot of the pearl millet genotypes ..................... 89 Figure 7-2: Estimated population structure of 77 pearl millet genotypes ..................................... 90 Figure 7-3: Manhattan plot, phenotypic distribution and quantile-quantile plots for downy mildew at Bambey (a) and Nioro (b) research stations .............................................................................. 91 xiii University of Ghana http://ugspace.ug.edu.gh LIST OF ABBREVIATIONS ANOVA: Analysis Of Variance ANSD: Agence Nationale de la Statistique et de la Démographie CERAAS: Centre d’Etude et de Recherches pour l’Amélioration de l’Adaptation à la Sècheresse DMI: Downy Mildew Incidence IRD: Institut de Recherche et de Développement ISRA: Institut Sénégalais de Recherches Agricoles FAO: Food and Agricultural Organization GBS: Genotype-By-Sequencing GCA: General Combining Ability GWAS: Genome Wide Association Study ICRISAT: International Crops Research Institute for Semi-Arid Tropics IPMDMVN: International Pearl Millet Downy Mildew Virulence Nursery LD: Linkage Disequilibrium LG: Linkage Group L x T: Line by Tester mating design MLM: Mixed Linear Model OPV: Open Pollinated Variety PRA: Participatory Rural Appraisal xiv University of Ghana http://ugspace.ug.edu.gh QTL: Quantitative Trait Loci RCBD: Randomized Complete Block Design RESOPP: Réseau des Organisations Paysannes et Pastorales du Sénégal SCA: Specific Combining Ability SD: Standard Deviation SE: Standard Error SNP: Single Nucleotide Polymorphism WCA: West and Central Africa WACCI: West Africa Centre for Crop Improvement xv University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE 1.0. GENERAL INTRODUCTION Pearl millet (Pennisetum glaucum (L.) R. Brown) is the 4th most important tropical cereal after rice, maize and sorghum (FAO, 2015). In 2014, global grain pearl millet production was estimated at 28 million tons, harvested from 32 million ha in Asia, Africa and the Americas. The average yield is about 900 kg ha-1. India and Africa are the most important producers with more than 85% of the total production in 2012 (FAO, 2015). Pearl millet, a staple food and source of fodder and feed for livestock for smallholders farmers in Africa and Asia, is considered as an important dual purpose cereal crop (Rajaram et al., 2013). The grain of pearl millet is a very rich source of protein, vitamins and minerals in comparison with other cereals and is used for human consumption in diverse ways (Nambiar et al., 2011). The flour is processed into paste, fried snacks and beverages (Angarawai et al., 2008). The grain is also used in processed foods and feed for animals (Wilson et al., 2008). The pearl millet stover is used as fuel, material for building, fencing and also as a soil additive to enhance soil fertility (Camara et al., 2006). It is also being experimented as a grain and forage crop in the USA, Canada, Mexico, India and West and North Africa (Yadav & Rai, 2013). In Senegal, pearl millet is mainly grown during the rainy season from June to October for early maturing varieties and from June to November-December for late maturing varieties. It is the most important cereal crop in terms of area harvested and production. In 2015, the production was estimated at 749 874 tons and represented more than 40% of Senegalese total cereal production (DAPSA, 2016). However, despite the importance of this cereal crop the production is still low, compared to other West African countries (FAO, 2015). In West Africa, Senegal is ranked fifth behind Niger, Nigeria, Mali and Burkina Faso in pearl millet production. 1 University of Ghana http://ugspace.ug.edu.gh Despite its importance in Senegal, pearl millet production is limited mainly by the use of unproductive cultivars as seed source, unreliable rainfall, poor soils, pests, inappropriate cultural practices, diseases and parasitic weeds (Mbaye, 1982; ISRA et al., 2005). Among these limiting factors, pearl millet downy mildew disease constitutes one of the major biotic problems that affects grain production and quality of pearl millet (Mbaye, 1994). The annual estimated worldwide losses due to downy mildew represents 20-40% of grain yield losses and could be higher when a susceptible cultivar is repeatedly grown in the same field (Hash et al. 1999). Downy mildew is observed in most pearl millet growing areas of Asia and Africa (Andrews et al., 1985). In Senegal, there is no recent available data on economic losses due to downy mildew. However, several reports from the Senegalese National Research Institute mentioned it as one of the main constraints to pearl millet production (Girard & Delassus, 1978; Mbaye, 1982, 1986; Badiane, 1999). In Senegal, yield loss due to downy mildew ranged from 0.2 to 21% (Mbaye, 1986) with an incidence of over 32% observed in farmers’ fields at Bambey (Badiane, 1999). Even though downy mildew occurs almost everywhere pearl millet is grown, it has been reported to be considerably less severe in the Northern and Southern parts of Senegal than in Sine-Saloum region, located in the Central part of the country (Girard & Delassus, 1978). The pathogen that causes the disease in pearl millet is Sclerospora graminicola (Sacc.) Schroet (Thakur et al., 2011). Several virulent populations have been reported around pearl millet growing areas. This diversity is due to its capacity to go through recombination causing a breakdown of resistance in pearl millet accessions (Thakur et al., 2009). Symptoms appear as chlorosis on the leaves starting from the base to the top of the leaves and a transformation of floral parts into leafy structures (Thakur et al., 2011). 2 University of Ghana http://ugspace.ug.edu.gh To assess the effect of downy mildew, both field and greenhouse screening techniques have been developed by the International Crops Research Institute for Semi-Arid Tropics (ICRISAT) (Williams et al., 1981; Singh et al., 1993; Singh et al., 1997; Thakur et al., 2011). During the past years, major downy mildew resistance breeding efforts have been made by ICRISAT and its partners, leading to the discovery of several sources of resistance through evaluation of a large set of germplasm accessions, across many environments in India and Africa (Thakur et al., 2006; Yadav & Rai, 2013). Some of these sources of resistance have been used in breeding programmes to develop downy mildew resistant hybrids and open pollinated varieties (Hash & Bramel-Cox, 1999). In recent years, with the advances in molecular biology, molecular breeding procedures are widely used for the development of pearl millet open pollinated and hybrid varieties. Hence, several mapping populations were developed and several downy mildew resistance quantitative trait loci (QTL) identified (Jones et al., 1995; Jones et al., 2002; Breese et al., 2002; Gulia et al., 2004; Yadav et al., 2004; Supriya et al., 2011). Some of these identified QTL were successfully transferred using marker assisted backcrossing, to the background of the parental lines of the hybrid “HHB 67 improved”. This hybrid was the first public-bred product of DNA marker based selection and is widely grown in India (Hash et al., 2006; Yadav et al., 2013). In Senegal, breeding for resistance to downy mildew was initiated in 1981 in collaboration with millet pathologist from India (Gupta, 1986). Different types of materials were screened in disease nursery field established in 1983 at Bambey station. Several sources of resistance were identified and used by the national pearl millet breeding programme. However, during the last decades, most of the breeding activities focused on the genetic improvement of grain yield and yield components (Gupta, 1986). Genetic improvement of pearl millet for downy mildew resistance has received 3 University of Ghana http://ugspace.ug.edu.gh little attention and few studies were done on the identification of new sources of resistance and genomic regions associated with downy mildew resistance under Senegalese environments. Presently, there is no ongoing breeding research activity related to the downy mildew problem, even though it has been stated as one of the most important disease limiting the national pearl millet production. Therefore, the identification of new sources of resistance to downy mildew, their combining ability and heterosis for downy mildew, yield and yield-related traits are needed to enhance pearl millet production. Before the identification of such sources of resistance to pearl millet downy mildew disease under Senegalese environments, information about variability of pearl millet downy mildew populations from the main pearl millet growing areas must be obtained. In fact, the pathogen is known to be highly diverse due to its capacity to undergo both asexual and sexual recombination. In addition, the identification of potential genomic regions associated with pearl millet downy mildew resistance that is efficient under Senegalese environments will help to address this constraint. There is also an urgent need to identify pearl millet production constraints and farmers’ preferred traits, information which will be essential to guide and assist in reorienting the pearl millet breeding programme in Senegal to meet farmers’ specific needs. The general aim of this research was to contribute to the improvement of pearl millet resistance to downy mildew disease under different Senegalese environments. The specific objectives were to: • identify constraints to pearl millet production and farmers’ preferred varieties and their perceptions on effects of pearl millet downy mildew disease; • assess variability of Sclerospora graminicola populations from Senegal; • identify sources of resistance to pearl millet downy mildew; • estimate combining ability and heterosis for downy mildew, grain yield and other agronomic traits; and • map genomic regions associated with downy mildew resistance. 4 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO 2.0. LITERATURE REVIEW 2.1. Origin and distribution of pearl millet Different centers of domestication of pearl millet have been reported across Africa and Asia. However, most of the authors agree that the pearl millet domestication took place in Africa with different geographical origins along the Sahelian zone from Mauritania to Sudan (Tostain & Marchais, 1989). Oumar et al. (2008) in their phylogenetic study using microsatelites, analyzed the relationship among accessions not showing introgression and concluded to a monophyletic origin of cultivated pearl millet in West Africa. This theory was recently confirmed by Manning et al. (2011) who reported the first evidence of pearl millet cultivation in Mali at around 4 500 before present. Then, from West Africa its cultivation spread and moved to India. Pearl millet is mainly cultivated in Africa and Asia (FAO, 2015) where it is used as a staple grain for human consumption (Sehgal et al., 2012). The greatest range of variability for downy mildew disease resistance and several other agronomic traits has been found in landraces collected across several countries in West and Central Africa, (Haussmann et al., 2006). In addition, the genetic distance observed betweeen wild and cultivated forms of pearl millet originating from West and Central Africa is lower than a large number of strains of pearl millet from India, providing further evidence that pearl millet is from West and Central Africa (Tostain & Marchais, 1989). 2.2.Taxonomy and biology of pearl millet Pearl millet belongs to the monocotyledonous group and is one of the most important species of the Poaceae family. It belongs to the Pennisetum genus and the Paniceae tribe of this family. The genus Pennisetum is divided into five sections and the pearl millet belongs to the Penicillaria section (Brunken, 1977). Pearl millet is a highly cross-pollinated (75-80% outcrossing) diploid 5 University of Ghana http://ugspace.ug.edu.gh crop (2n = 2x = 14) with a large genome size (2450 Mbp), which has been recently sequenced (Varshney et al., 2017). The pearl millet plant is hermaphroditic with a protogynous status, which facilitates the crossing process. The open pollination is done through wind or insects and occurs mainly between 7 to 12 a.m. During this period maximum pollen shedding is observed and the stigmata are well receptive. The crop is a C4 grass which shows an excellent photosynthetic efficency leading to high biomasse production. 2.3. Economic importance of pearl millet Pearl millet is one of the most important food crops cultivated in the arid and semi-arid areas. It ranks fourth among the cultivated tropical cereal crops (FAO, 2015) and is mostly grown as a subsistence crop in the arid and semi-arid tropical regions of Asia and Africa, in the hottest and driest areas, where no other cereals can grow (Thakur et al., 2011; Yadav et al., 2013). In 2013, global millet grain production was estimated 30 million tons, harvested from 33 million ha in Asia, Africa and the Americas (FAO, 2015). The pearl millet grain is a rich source of vitamins, micronutrient (iron and zinc) and proteins (8-19%) in comparison to other cereals and is used for human consumption in diverse ways (Nambiar et al., 2011; Tacko et al., 2015). It is also cultivated for its stover, used as forage, fuel and materiel for building (Yadav & Rai, 2013). In Senegal, pearl millet is the predominant crop among the cereals both in terms of surface area covered and production. The country is one of the top 10 pearl millet producers in the world (FAO, 2015). During the rainy season 2015, the national production reached 749 874 tons and represented approximatively 40% of Senegalese total cereal production (DAPSA, 2016). It is mainly produced under rainfall conditions in the north and central regions of the country where rainfall is low and erratic. The crop is a major staple food for many farmers and the grain is mainly used for making couscous while the stover constitutes a source of feed for cattle and materiel for building. 6 University of Ghana http://ugspace.ug.edu.gh 2.4. Farmers’ participation in plant breeding Plant breeders, to improve level of adoption of newly bred pearl millet cultivars, should work closely with farmers through participatory plant breeding where farmers are solicited and involved earlier and their technical knowledge is integrated during the cultivar development (Bidinger 1998). Nowadays, this technique is widely conducted by plant breeders to ensure adoption of newly improved varieties by farmers, particularly in marginal agro-ecological zones (Ceccarelli et al., 2007). In fact, this approach has been demonstrated to enhance adoption of newly bred- cultivars and may help plant breeders to identify farmers’ production constraints and preferred traits (Mulatu & Belete, 2001). This approach is also more effective in terms of cost and transferring the most suitable cultivars to farmers (Mangione et al., 2006). Participatory plant breeding approaches in pearl millet have been conducted in some pearl millet growing areas elsewhere in Africa. For example, the early maturing pearl millet cultivar Okashana 1, which is also responsive to both soil moisture and inorganic fertilizer, ensued from a participatory plant breeding approach used in Namibia (Bidinger, 1998). Farmers participated in early selection as per their preferences, while the Government of Namibia took pro-actively the rapid multiplication and dissemination of high quality seed, thus reducing costs and speeding adoption of Okashana 1. The adoption of this cultivar led to increasing pearl millet yield by 33% with available management practices, which contributed to household food security in the northern communal areas of Namibia (Matanyaire, 1998). Furthermore, an impact assessment study estimated the rate of returns (ROR) research-for-development (R4D) investment, including dissemination and adoption for Okashana 1, as 13.28% (Anandajayasekeram et al., 2007). This estimated ROR was 4.25% when the cost of investment increased by 10%, which shows the sensitivity of ROR to the R4D costs. Interestingly, a comparison of improved and traditional cultivars of pearl millets in Namibia 7 University of Ghana http://ugspace.ug.edu.gh revealed that farmers were able to preserve the characteristics of the improved cultivars after growing them along with their landraces because they grew them separately in the field (Uno, 2005). As noted by this research, the rapid adoption of a new pearl millet cultivar such as Okashana 1 in Namibia did not trigger the abandonment of old landraces because farmers kept them considering both the climate requirements and their lifestyle. 2.5. Production constraints Despite its importance, pearl millet production is limited by several biotic and abiotic constraints. In Asia and Africa, its production is characterized by unreliable rainfall, poor soils, pests, inappropriate cultural practices, diseases and parasitic weeds such as Striga. 2.5.1. Abiotic constraints Low soil fertility and water availability have been cited as the major abiotic constraints to pearl millet production, especially in West Africa (Issaka, 2012; Drabo, 2016). In this part of the world, pearl millet is mainly grown in soils of low inherent fertility and farmers in general do not apply the recommended amount of fertilizer. Moreover, this low soil fertility issue is amplified by the continuous pearl millet growing associated with demographic pressure on the arable lands. This long term pearl millet cultivation without soil improvement strategies contributes to the deficiency of the major mineral elements (NPK) which play an important role in pearl millet development. Drought is the most important abiotic stress that limits considerably plant production, especially the pearl millet production in West Africa (Issaka, 2012; Drabo, 2016). The Sahel region where pearl millet is largely grown is one part of the world that has experienced the most important severe droughts in the 1970s. Drabo (2016) reported drought as the main production constraint to pearl millet in Burkina Faso. The drought affects considerably the photosynthetic activity of pearl millet 8 University of Ghana http://ugspace.ug.edu.gh (Radhouane, 2009) and can destroy the entire plant while significantly reducing grain yield (Winkel et al., 1997) 2.5.2. Biotic constraints A large number of biotic stresses including insects, diseases and parasitic weeds such as Striga are known to attack pearl millet. Pearl millet is attacked by many insects from sowing to storage. A grain yield loss up to 95% has been reported in the Sahelian zone (Krall et al., 1995). Major pearl millet insect pests include pearl millet grasshoppers, leaf miner (Heliocheilus albipunctella de Joannis), stem borer (Coniesta ignesfusalis), and armyworms (Spodoptera exempta). Among these insects, head miner is the most important and can causes considerable grain yield loss (30-40%) in the Sudano Sahelian zone (Ndoye & Gahukar, 1986). The most serious weed pest problem for maize and pearl millet production in West Africa is the parasitic weed Striga hermonthica (Jamil et al., 2012; Issaka, 2012; Kountche et al., 2013; Drabo, 2016). It is widespread on pearl millet growing areas in Africa and Asia and can disturb the pearl millet photosynthetic activity (Haussmann et al., 2000). Its infestation is known to increase under N and P deficiency (Jamil et al., 2012). Many disease attack pearl millet, however, most of them are not economically important. Pearl millet is mainly affected by five pathogens which are Sclerospora graminicola causing downy mildew, Claviceps fusiformis causing ergot, Moesziomyces penicillariae causing smut, Puccinia substriata causing rust and Pyricularia grisea causing blast. Among these cited pathogens, Sclerospora graminicola constitutes the major problem of pearl millet that affects grain production and quality and is observed in most pearl millet growing areas in Africa and Asia (Andrews et al., 1985; Thakur et al., 2011). 9 University of Ghana http://ugspace.ug.edu.gh 2.6. Downy mildew of pearl millet 2.6.1. Pathogen diversity Downy mildew, an oomycetic disease caused by the pathogen Sclerospora graminicola (Sacc.) Schoret; was first reported by Butler (1907) on pearl millet in India. The disease was observed wherever pearl millet was grown in Asia and Africa (Andrews et al., 1985). The pathogen is an obligate parasite and highly variable across locations and years. This considerable variability was first reported by Bhat (1973) and was confirmed by several studies in India and several African countries through the pearl millet downy mildew virulence nursery using host differential lines and DNA markers (Thakur et al., 2006). It was demonstrated clearly that pathogens from West African countries were more virulent compared to those originated from India. Likewise, a field screening conducted by Thakur et al. (2009) showed that, out of 123 pearl millet germplasm accessions identified as resistant from 15 countries, only 21 remained so after around 13 years of cultivation. Indeed, the pathogen is able to reproduce by both sexual and asexual pathways, through oospores and sporangia, respectively (Thakur et al., 2011). 2.6.2. Downy mildew life cycle, epidemiology and control methods Oospores in the soil and contaminated seed lots are the primary sources of infection (Figure 2.1). The downy mildew disease generally appears during the vegetative stage in the form of chlorosis on the second leaf and then progresses to the subsequent leaves by production of spores (Thakur et al., 2011). Under high relative humidity (>95%) and moderate temperature (20-22oC), whitish growth of the pathogen in the form of sporangia and sporangiophores is observed on the abaxial leaf surface. These sporangia are disseminated by wind and rain splash causing secondary spread. Later on, the germinated sporangia release zoospores, which cause secondary infection. At panicle 10 University of Ghana http://ugspace.ug.edu.gh emergence, green ear symptoms appear in the infected plants and become visible. The florets are converted into leafy structures. Figure 2-1: Downy mildew disease cycle (Singh et al., 1993) Several methods including cultural practices and chemical treatments have been used for the control of downy mildew in pearl millet (Spencer-Phillips & Jeger, 2004). The disease can also be controlled by growing resistant varieties. The cultural practices such as roguing of infected plants, crop rotation, crop sanitation, manipulation of planting date, deep ploughing and soil solarisation have been used to control downy mildew in pearl millet (Jeger et al., 1998). For the chemical control, the systemic fungicide metalaxyl (Apron star) is used as a seed treatment for the control of the disease. It is reported as highly effective for about 30 days after sowing (Zarafi et al., 2004; Zarafi, 2005; Anaso & Anaso, 2010; Aliyu et al., 2011). A study conducted in Nigeria has shown significant differences in the disease incidence and severity in plots sown with treated seeds and 11 University of Ghana http://ugspace.ug.edu.gh those sown with non-treated seeds (Zarafi et al. 2004). The most cost-effective and eco-friendly control method for downy mildew disease is the use of resistant varieties. 2.6.3. Economic importance of downy mildew Downy mildew is the most important disease limiting pearl millet production in the world (Williams & Andrews, 1983; Thakur et al., 2011). Hash et al. (1999) reported that the estimated annual yield losses represents 20-40%, and the loss in grain production could be higher when a single cross hybrid F1 is grown in successive years. For instance, in 1971, an epidemic of downy mildew on pearl millet was observed on the first popular hybrid HHB 67 in India and the resulting loss in grain production was estimated at around 3.3 million metric tons (Singh et al., 1993). In Senegal, there is no recent data available on economic losses due to downy mildew. However, it constitutes the major problem that affects pearl millet grain production with a yield loss ranging between 0.2 and 21% (Mbaye, 1986). The losses in grain yield are correlated to the disease incidence with a significant correlation between pearl millet downy mildew incidence and grain yield loss (r = 0.9) (Mayee & Siraskar; 1982). 2.6.4. Sources of downy mildew resistance During the past decades, both field and greenhouse screening techniques have been developed and improved in ICRISAT (Williams et al., 1981; Thakur et al., 2011). These techniques were widely used in assessing the effect of downy mildew on a large number of germplasm accessions and breeding lines across Asia and Africa. Several sources of resistance to downy mildew have been identified (Singh et al., 1990; Singh et al., 1993; Zarafi, 2007; Thakur et al., 2009; Sharma et al., 2014). Some of these sources of resistance such as ICML 12, 13, 14, 15 and 16 showed a stable resistance to downy mildew and were used by the breeding programme of ICRISAT India to develop open pollinated varieties as well as hybrid varieties (Hash et al., 1996). For instance, 12 University of Ghana http://ugspace.ug.edu.gh ICML 16 was used as a parent to develop two varieties, ICMV 2 and ICMV 3. Among these identified sources of resistance, IP 18292 to IP 18298, are highly resistant to three major pathotypes of India and five of them remained free from disease about 30 days after sowing in preliminary tests conducted in Niger and Mali (Singh et al., 1993). Recovery resistance, a phenomenon in which the host and the pathogen co-exist without affecting neither the normal development of the plants nor their grain yield, was also observed in pearl millet lines. Singh and King (1988) have detected this trait in 28 out of 33 tested genotypes. Three of these 28 genotypes (SDN 503-2, P 1449, 841 A and 81 A) showed considerable percentage of recovery and may be used as sources of recovery resistance in pearl millet breeding programmes. 2.6.5. Mechanism of downy mildew resistance Despite the production of large amounts of inoculum by plant pathogens, majority of plants remain healthy. To withstand the hostile environment of pathogenic microorganisms, plants have developed several effective mechanisms of resistance. Indeed, the plants cope with pathogens by the combination of structural characteristics and biochemical reactions (Agrios, 2005). In pearl millet, the contribution to the defense against the pathogen of hydroxyproline-rich glycoproteins and polygalacturonase inhibitor proteins, plant defense proteins found in plant cell walls, have been investigated (Deepak et al., 2007; Prabhu & Kini, 2012; Prabhu et al., 2015). The amount of these proteins was higher in the resistant genotype than in the susceptible genotype. Melvin et al. (2015) demonstrated also that the pearl millet mitogen active protein kinase contributes to pearl millet defense against the downy mildew pathogen. Thus, the pearl millet varieties respond to downy mildew disease by producing a high amount of plant defense proteins. Phenolic compounds and enzyme activities are implicated in downy mildew resistance of pearl millet. The total phenol content was higher in susceptible genotypes than in resistant genotypes at 13 University of Ghana http://ugspace.ug.edu.gh post infection stage (Mahatma et al., 2011). Geetha et al. (2005) studied the phenylalanine ammonia lyase in pearl millet cultivars having different levels of downy mildew resistance. They observed that the enzyme activation varied between cultivars and was positively correlated with the degree of resistance to downy mildew disease. In another study, Shivakumar et al. (2003) measured the lytic factors in the coleoptile region of resistant and susceptible cultivars to downy mildew disease. They found that, the level of lytic factors were higher in the resistant cultivars than in the susceptible ones. Thus, they concluded that lytic factors are responsible for the lysis of the pathogen in the resistant cultivars and may provide resistance to downy mildew disease. 2.6.6. Breeding for downy mildew resistance The improvement of pearl millet for downy mildew resistance has been done by both conventional and molecular breeding (Weltzien & King, 1995; Hash et al., 2006). For conventional breeding, the pedigree method was most widely used (Yadav et al; 2013). It has been used for the development of parental lines for hybrid production. Recurrent selection has also been used but in fewer programme as compared to the pedigree method. The molecular breeding approach through Marker Assisted Backcrossing (MABC) for transferring downy mildew resistance QTL into elite hybrid parental lines has been widely used (Hash & Witcombe, 2001). Several QTL effective against diverse pathotypes were mapped and transferred into the hybrid parental lines (Yadav et al., 2013). This technique of MABC was successfully used in the development of the hybrid line “HHB 67 improved” which is resistant to downy mildew and widely grown in India (Howarth & Yadav, 2002). Mutation breeding has been rarely used in pearl millet breeding programme. In 1971, after the devastating damage caused by the downy mildew, a mutation breeding project was initiated (Murty et al., 1983). The objective of this mutation-breeding programme was to make Tift 23A, a 14 University of Ghana http://ugspace.ug.edu.gh commonly used male sterile parent, resistant to downy mildew. This breeding programme led to the development of a downy mildew resistant maintainer line MS 5071 B derived from Tift 23 B. 2.6.7. Quantitative Trait Loci (QTL) mapping for downy mildew resistance Molecular breeding procedures have been widely used in pearl millet breeding programme. Liu et al. (1994) developed the first linkage map of pearl millet using Restriction Fragment Length Polymorphism (RFLP) markers. This genetic linkage map contained 181 loci covering 303 cM. It was widely extended with a number of mapped pearl millet Simple Sequence Repeats (SSR) markers, Single Nucleotide Polymorphisms (SNP), Diversity Arrays Technology (DArT) and Sequenced Characterized Amplified Regions (SCAR) markers (Allouis et al., 2001; Budak et al., 2003; Qi et al., 2004; Rajaram et al., 2013; Jogaiah et al., 2014; Moumouni et al., 2015; Punnuri et al., 2016). This genetic map extension will facilitate QTL mapping and applied marker assisted selection for downy mildew disease resistance in pearl millet. Several QTL for downy mildew disease resistance have been mapped (Jones et al., 1995; Hash & Witcombe, 2001; Breese et al., 2002; Jones et al., 2002; Gulia et al., 2004; Supriya et al., 2011). Jones et al. (1995) investigated QTL responsible for resistance to pathogen populations from India, Nigeria, Niger and Senegal and showed that the disease resistance was quantitatively inherited. They detected QTL for resistance with large effect against the pathogen population from India on linkage groups (LG) 1, 6 and 7; against the pathogen populations from Niger and Nigeria on LG4; 1 and 6; and against the pathogen population from Senegal on LG2, 6 and 7. These QTL explained between 3.6% and 48.9% of the observed phenotypic variation. Some of these QTL were consistently detected in repeated screening but none of these QTL were effective against all four pathogen populations. However, Gulia et al. (2004) in their study detected nine putative downy mildew resistance QTL, and reported that one of them on LG4 was common for the eight pathogen 15 University of Ghana http://ugspace.ug.edu.gh populations from Africa and Asia except the one from Mali. Another QTL on LG2 had a major effect against most of the pathogen populations. In a different study where 114 F4:2 progenies were screened under field conditions in India and glasshouse in India and UK, Jones et al. (2002) identified the same two QTL. The common QTL on LG1 had a major effect and explained up to 60% of the phenotypic variation, while the second one on LG2 had a minor effect and explained up to 16% of the phenotypic variation. With recent advances observed in next generation sequencing technologies, millions of data points that are distributed throughout a genome, can be generated through Genotyping by Sequencing (GBS) (Elshire et al., 2011). These data points can be used for detecting high resolution QTL underlying complex traits through Genome Wide Association Study (GWAS). This approach is based on the principle of linkage disequilibrium (LD) and, compared to bi-parental mapping population, provides a higher resolution. In fact, this method makes use of past recombination events which occurred during the variety development. GWAS has been applied in several crops such as maize, wheat, rice, sorghum, cassava and foxtail millet (Shinada et al., 2015; Upadhyaya et al., 2015; Esuma et al., 2016). In pearl millet, there are fewer reports on the use of this approach because the genome sequence has just been published. However, this approach was used in pearl millet to identify markers associated with phenology, grain yield, stover, drought tolerance, zinc, iron and low phosphorus related-traits (Kannan et al., 2014; Gemenet et al., 2015; Sehgal et al., 2015; Satyavathi & Srivastava, 2017). 2.6.8. Marker assisted deployment of downy mildew resistance Marker Assisted Selection (MAS), a component of the molecular breeding, is the use of DNA markers in plant breeding. In fact, once a QTL is identified and validated, it can be transferred to elite lines using marker-assisted selection. This breeding approach presents several advantages 16 University of Ghana http://ugspace.ug.edu.gh over the conventional breeding. Indeed, a single plant can be selected earlier at the seedling stage and may be simpler than phenotypic selection (Collard & Mackill, 2008). However, before its application, tightly linked markers that predict the phenotype of the trait of interest should be identified and validated. In the case of downy mildew resistance in pearl millet, this technique has been deployed successfully for the improvement of the parental lines of the hybrid HHB 67 for downy mildew resistance (Yadav et al., 2013). Indeed, the MABC was used for the introgression of LG1 and LG4 downy mildew resistance QTL from the donor parent ICMP 451-P6 into the background of the elite line H 77 / 833-2 used as restorer line of the popular hybrid HHB 67. At the same time, the gene for downy mildew resistance from the line ICML 22 was added into the background of the line 843A used as female parent of the hybrid HHB 67 through a conventional backcrossing transfer. The MABC took only three years while the conventional backcrossing took up to nine years (Hash et al., 2006). In 2005, the hybrid was proposed for release in the state of Haryana as a downy mildew resistant and high yielding variety and was the first public-bred product of DNA marker selection (Hash et al., 2006). The efforts made on the improvement of the male parent of the hybrid HHB 67 have shown that the use of DNA markers can be a good strategy to increase the efficiency of conventional breeding method. 2.7.Combining ability study in pearl millet The understanding of the combining ability enables the plant breeder to design efficient breeding strategies for the development of improved varieties. When an inbred line combines adequately well with other lines, it gives an indication that particular line has a good general combining ability (GCA) and the additive gene action is involved in the inheritance of the trait. In contrast, when an inbred line combines well only for some crosses, that means that this line has a good specific 17 University of Ghana http://ugspace.ug.edu.gh combining ability (SCA) and the gene action is associated with non-additive genetic effects (Falconer, 1989). In pearl millet, combining ability for different traits such as grain quality, disease resistance, grain yield and dry fodder yield have been studied using different mating designs (Chaudhary et al., 2012; Parmar et al., 2013; Drabo, 2016; Lubadde et al., 2016). For downy mildew disease, most of the studies have concluded that dominance gene effects played a larger role than additive gene effects (Deswal & Govila, 1994; Jones et al., 1995; Singh & Talukdar, 1998; Gulia et al., 2004; Angarawai et al., 2008). However, Shetty et al. (2001) reported in their study that the inheritance of downy mildew disease resistance in pearl millet was not only controlled by additive and dominance genes but also by epistasic effects. These results were consistent with Issaka (2012) who also reported epistatic effects in the control of resistance to downy mildew. 18 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE 3.0. FARMERS’ PRODUCTION CONSTRAINTS, PERCEPTIONS ON DOWNY MILDEW AND VARIETAL PREFERENCES OF PEARL MILLET IN THE GROUNDNUT BASIN OF SENEGAL 3.1. Introduction Despite the availability of improved pearl millet cultivars having a grain yield potential of about 2 to 3 t ha-1, Senegalese farmers persistently grow local landraces that have a grain yield of about 60% of improved varieties (Kumar & Anand 1993). This is because of inadequate levels of awareness and minimal farmers’ involvement during cultivar development (Omanya et al., 2007). Most of the time, failure to involve farmers during pearl millet breeding often leads to cultivars that are unattractive to them. In Senegal, there is little formal documentation on the identification of main constraints to pearl millet production and farmers’ preferred traits. Assessment of constraints and farmers’ preferred traits for a specific crop does not have a global application. Plant breeders should develop cultivars adapted to a specific location rather than aiming at a wide range of environments (Omanya et al., 2007). Knowledge of production constraints and farmers’ preferred traits for a specific location are therefore essential to guide any pearl millet breeding programme. This approach will help the plant breeders to prioritize the traits for improvement in order to meet farmers’ specific needs. The specific objectives of this research were to: (i) assess farmers’ perceptions on pearl millet downy mildew disease, (ii) identify main production constraints, and (iii) identify farmers’ varietal preferences. 19 University of Ghana http://ugspace.ug.edu.gh 3.2. Materials and Methods 3.2.1. Study area The study was conducted in February and March 2015 in the groundnut basin agro-ecological zone located in the central part of Senegal where groundnut and pearl millet are the predominant crops (Figure 3-1). Figure 3-1: Map of Senegal showing groundnut basin areas included in participatory rural appraisal and baseline survey Five representative pearl-millet growing regions of Senegal namely Fatick, Kaolack, Diourbel, Thies and Kaffrine were selected based on the importance of pearl millet in terms of production and area grown to the crop (Table 3.1). During 2014 cropping season, these five regions accounted for 74% of the national harvested pearl millet area and over 76% of total pearl millet production (DAPSA, 2016). In each region, one rural community was selected. Within each rural community, 20 University of Ghana http://ugspace.ug.edu.gh three villages were selected providing 15 study sites. Among all the selected sites most of farming activities are done during the rainy season, which is uni-modal from June to October, with annual rainfall between 500 to 700 mm. Soil textures vary, but are mostly sandy with low fertility. Table 3.1: Characteristics of selected pearl millet production areas in Senegal Altitude Pearl millet Pearl millet Average Regions (m) production (t) cultivated Soil type precipitation Selected area (ha) (mm) community Thies 71 31 747 81 482 Sandy 526 Malicounda Diourbel 7 46 231 105 277 Sandy 549 Ngogom Fatick 6 80 000 130 000 Sandy 638 Patar Lia Kaolack 6 77 613 132 508 Sandy-clay 655 Paoskoto Kaffrine 11 68 768 89 157 Sandy-clay 655 Kahi 3.2.2. Sampling method In each region, one rural community was sampled with the assistance of regional extension officers. The selection was based on the relative importance of pearl millet production in the rural community. The following rural communities were selected: Malicounda, Ngogom, Kahi, Paoskoto and Patar Lia (Table 3.1). In each rural community, three villages were sub-sampled with the help of the local extension officers. Within each village, 10 pearl millet farmers were selected providing 150 pearl millet farmers for the semi-structured interviews. A purposive sampling procedure was used to select farmers. Indeed, farmers were selected based on their experiences in pearl millet farming practices. In addition, a focus group discussion was held in each of the three selected villages that constituted one rural community and represented a region. Hence, 15 focus group discussions were carried out. The farmers who participated to these focus group discussions have been cultivating pearl millet for more than 10 years ago. 3.2.3. Data collection Data were collected through focus group discussions and farmers’ interviews. In each of the focus group discussions, five pearl millet farmers were involved making a total of 15 farmers per rural 21 University of Ghana http://ugspace.ug.edu.gh community. The discussions were held in order to determine or otherwise confirm the farmers’ agronomic practices, major crops and pearl millet cultivars grown, preferences for new pearl millet cultivars, main constraints limiting production and their perceptions on the downy mildew. After the group discussions, 10 farmers per village with their prior informed consent were interviewed individually using semi-structured questionnaires. The semi-structured questionnaires were administered to the farmers by a multidisciplinary research team comprising of a breeder, an agronomist, a plant pathologist, and a socio-economist. During the survey, pictures showing downy mildew symptoms at both vegetative and reproductive stages were used to illustrate and allow farmers to recognize the disease. 3.2.4. Data Analysis Data collected were analyzed using the Statistical Package for Social Sciences (SPSS) version 17. Variables were subjected to descriptive statistics, cross tabulations, one-sample t (two-sided) and Chi-square tests. 22 University of Ghana http://ugspace.ug.edu.gh 3.3. Results 3.3.1. Demographic characterization of farmers interviewed Of the 150 farmers interviewed, 93.3% were male (Table 3.2). Most of the respondents belonged to the Serere (42%) or Wolof (30%) ethnic groups. Among the total respondents, 73.3% were illiterate while 18% had finished primary school, 8% were educated up to post-primary level and less than 1% had attended secondary school. About 51% of the farmers belonged to a farmers’ association and 52% of them were above 50 years old. 23 University of Ghana http://ugspace.ug.edu.gh Table 3.2: Characteristics of farmers interviewed Variable Rural communities Total Percentage P* Paoskoto Kahi Ngogom Malicounda Patar Lia Gender Male 29 27 25 29 30 140 93.3 Female 1 3 5 1 0 10 6.7 0.073 Ethnic Wolof 14 17 0 14 0 45 30 Serere 5 1 29 7 21 63 42 Al pular 8 4 1 4 9 26 17.3 < 0.001 Others 3 8 0 5 0 16 10.7 Belongs to association Yes 11 15 10 24 17 77 51.3 No 19 15 20 6 13 73 48.3 0.002 Level of education Illiterate 22 26 25 17 20 110 73.3 Primary 6 2 3 8 8 27 18 Post primary 2 1 2 5 2 12 8 0.180 Secondary 0 1 0 0 0 1 0.7 Age < 35 4 3 4 5 2 18 12 36 - 50 15 13 8 10 8 54 36 0.415 > 50 11 14 18 15 20 78 52 *Probability values based on Chi square test 24 University of Ghana http://ugspace.ug.edu.gh 3.3.2. Importance of pearl millet The land covered per farmer by the five major crops grown is presented in Table 3.3. Overall, average land area planted per farmer for specific crops was higher for pearl millet (2.9 ha) followed by groundnut (2.4 ha). Cowpea, maize and sorghum were planted in smaller areas with a total of only 1.1 ha. The average land holding per crop differed significantly between locations. The dominant crop was pearl millet without any land allocated for maize production in Ngogom, Malicounda and Patar Lia, while in Paoskoto and Kahi groundnut was the major crop. Average land area planted to pearl millet was highest in Patar Lia (3.9 ha) followed by Kahi (3.4 ha). Table 3.3: Land area planted per farmer (ha) and crops grown across the locations Crops Rural communities Paoskoto Kahi Ngogom Malicounda Patar Lia Mean P Pearl millet 2.83 3.41 2.22 2.22 3.90 2.90 < 0.001 Groundnut 3.35 3.73 0.71 1.31 2.79 2.38 < 0.001 Maize 0.88 1.25 0.00 0.00 0.00 0.43 < 0.001 Cowpea 0.19 0.04 0.68 0.72 0.65 0.46 < 0.001 Sorghum 0.21 0.37 0.11 0.53 0.13 0.27 0.049 3.3.3. Farmers’ perceptions on the downy mildew disease After a comprehensive description of downy mildew disease of pearl millet using pictures to show symptoms at both vegetative and reproductive stages as support, almost all farmers (99%) had a prior knowledge of the disease (Table 3.4). Farmers in this category were asked whether this disease was present in their fields during the just ended season and the majority of them (94%) answered affirmatively. Though incidence of downy mildew disease was high among farmers interviewed, severity was deemed rather low by the majority (80%). Approximately, 46% of farmers did not attempt to alleviate the effects of the downy mildew disease, while the rest discarded infected plants. In Patar Lia, 75% of farmers who were aware of the disease discarded 25 University of Ghana http://ugspace.ug.edu.gh infected plants in case of attacks in their fields, while in Kahi the majority of them (68%) did not adopt any strategy to mitigate the effects of the disease. Table 3.4: Farmers’ perceptions of the downy mildew disease Rural communities Paoskoto Kahi Ngogom Malicounda Patar Total Percentage P Knowledge about DM Yes 29 28 28 30 27 142 99 No 1 2 2 0 3 8 1 0.249 DM observed on the field during this ended crop season Yes 28 27 26 26 26 133 94 No 1 1 2 4 1 9 6 0.465 Severity Important 9 8 2 2 7 28 20 Not important 20 20 26 28 20 114 80 0.034 Management strategies Nothing 13 19 11 15 7 65 46 Discard infected plants 16 9 17 15 20 77 54 0.033 3.3.4. Production constraints Constraints to pearl millet production revealed by the survey are summarized in Table 3.5. The parasitic weed Striga, low soil fertility and insect pests were the three most important constraints reported by pearl millet farmers. Availability of seeds was identified as the least important constraint. Other constraints across the five rural communities were lack of machinery, downy mildew disease, drought, limited access to land and birds. In addition, the data showed differences in ranking constraints between rural communities. Farmers in Patar Lia ranked low soil fertility as the most important constraint, followed by Striga, lack of machinery, insect pests, downy mildew, lack of seed availability, drought and birds; whereas in Paoskoto, Kahi and Ngogom Striga ranked first then, low soil fertility, insect pests and lack of machinery. In Malicounda, the major constraints were insect pests, Striga and birds. 26 University of Ghana http://ugspace.ug.edu.gh Table 3.5: Mean rank of pearl millet constraints across the locations Constraints* Rural communities Paoskoto Kahi Ngogom Malicounda Patar Mean Rank P Striga 2.2 2.4 2.7 2.3 2.8 2.5 1 0.269 Low soil fertility 2.6 2.6 3.1 5.2 1.6 3.0 2 <0.001 Downy mildew 6.6 5.8 6.2 6.2 5.5 6.0 6 0.162 Drought 5.4 6.7 5.5 4.7 6.5 5.7 5 0.004 Insects 3.9 3.8 4.5 2.2 3.9 3.6 3 <0.001 Lack machinery 5.1 3.6 4.6 6.5 3.7 4.7 4 <0.001 Land tenure 6.2 6.7 4.7 6.9 6.3 6.1 7 0.010 Seeds Availability 6.3 7.2 7.0 8.1 5.9 6.9 9 0.014 Birds 7.2 6.9 6.4 4.6 7.2 6.5 8 <0.001 *Constraint with smallest mean rank within a column is perceived to be the most important 3.3.5. Farmers’ trait preferences Significant differences were observed in ranking pearl millet trait preferences by farmers between rural communities, except for high yield, grain color and resistance to biotic stress traits (Table 3.6). Overall, high grain yield was by far the most important trait. Other important traits mentioned were adaptation to abiotic stresses such as drought and low soil fertility, early maturity, white and grey grain color, and host plant resistance to insects, pathogens and Striga. In Ngogom and Malicounda, earliness was identified as the second most important trait, whereas in Paoskoto, Patar Lia and Kahi host plant resistance to Striga and adaptation to low soil fertility were ranked after high grain yield. In general, farmers from the five different rural communities were not much concerned about the white and grey grain color. Table 3.6: Mean rank of the most important preferred traits across the locations Characteristics* Rural communities Mean Rank P Paoskoto Kahi Ngogom Malicounda Patar High yield 1.2 1.6 1.4 1.2 1.2 1.3 1 0.058 Grain color 3.7 4.2 3.7 4.3 3.8 3.9 5 0.136 biotic stress 3.3 2.7 3.4 3.3 3.3 3.2 2 0.072 abiotic stress 2.9 2.8 3.5 3.8 3.1 3.6 4 0.010 Early maturity 3.9 3.5 3.1 2.3 3.2 3.3 3 <0.001 *Characteristic with smallest mean rank within a rural commune is perceived to be the most important 3.3.6. Production level and fertilizer Table 3.7 lists average grain yield of pearl millet and quantities of NPK fertilizers applied. Generally, all of them differed significantly between rural communities with Malicounda and 27 University of Ghana http://ugspace.ug.edu.gh Paoskoto having both the highest estimated grain yield and amount of NPK fertilizer applied per hectare. The lowest amount of NPK fertilizer applied per hectare and grain yield were noted in Patar Lia with 22.8 kg ha-1 of NPK fertilizer applied and an estimated grain yield of 495 kg ha-1. Table 3.7: Estimated yield and NPK fertilizer applied Rural communities Variable (kg ha-1) Mean P Paoskoto Kahi Ngogom Malicounda Patar Yield 660.9 632.8 462.6 817.0 494.9 615.9 < 0.001 NPK 90.1 69.8 81.4 93.0 22.8 71.4 0.039 3.3.7. Cultivars grown and their provenance Both formal and informal surveys showed that farmers from the five rural communities grew local and improved cultivars (Table 3.8). They predominantly use local landraces (70%) except for Malicounda. Most farmers in other rural communities were using landraces and the seeds were from their own fields or local markets. In Malicounda, the most important cultivar grown by farmers was Souna 3, which is an improved cultivar. Among the improved cultivars, Souna 3 was the most widely used by farmers followed by Thialack 2 and IBV 8004. The seeds from these improved cultivars were donated to farmers by NGOs or other public projects. 28 University of Ghana http://ugspace.ug.edu.gh Table 3.8: Landraces and bred-cultivars grown and their provenance Rural communities Paoskoto Kahi Ngogom Malicounda Patar Total Percentage P Varieties* Thialack 13 7 0 0 0 20 13 Souna 14 16 25 4 27 86 57 Thialack 2 0 1 0 4 0 5 3 < 0.001 Souna 3 3 6 4 20 3 36 24 IBV 8004 0 0 1 2 0 3 2 Seeds provenance Own field 27 22 27 4 30 110 73 NGO/Project 2 6 3 26 0 37 25 Local < 0.001 market 1 2 0 0 0 3 2 Landraces are in bold 3.3.8. Farming practices Plant population varied from one locality to another (Table 3.9). The highest plant population (33,333 plants ha-1 cm) was mainly observed in Paoskoto and Kahi whereas the lowest one (11,111 plants ha-1) was in Ngogom and Patar Lia. In Malicounda, farmers had either 17,857 plants ha-1 or 11,111 plants ha-1 as plant population. Seed treatment varied from one rural community to the other. Approximately, 53% of the farmers sowed without treating their seeds. In Malicounda, the majority of them (97%) treated their seeds against insect pests and other pathogens before sowing, while in Patar Lia most farmers did not treat the pearl millet seeds before sowing (80%). In Paoskoto, Kahi and Ngogom both practices are present. Pearl millet farmers frequently use production inputs such as organic and mineral fertilizers except in Patar Lia. In this rural community, eight farmers out of the 30 interviewed farmers used mineral fertilizer. Pearl millet was mainly cultivated in rotation with groundnut in all localities (96%). Except for Paoskoto and Ngogom where mechanical weeding was done mainly twice during the crop cycle, the majority of fields were weeded three times. The overall results from both focus group discussion and formal survey indicated that farmers from the groundnut basin of Senegal sow pearl millet in June, before 29 University of Ghana http://ugspace.ug.edu.gh the rain begins. However, few of them in Paoskoto, Malicounda and Kahi prefer to sow at onset of rains, under the moist soil conditions. 30 University of Ghana http://ugspace.ug.edu.gh Table 3.9: Pearl millet production practices in the groundnut basin of Senegal Agronomic practice Rural communities Paoskoto Kahi Ngogom Malicounda Patar Percentage P Number of plants ha-1 33,333 28 24 1 5 1 39 17,857 2 1 5 11 7 17 < 0.001 11,111 0 5 24 14 22 43 Seed treatment Yes 11 16 11 29 4 47 No 19 14 19 1 26 53 < 0.001 Organic Fertilizer Yes 24 28 28 25 29 89 No 6 2 2 5 1 11 0.226 Mineral fertilizer Yes 24 22 16 24 8 63 No 6 8 14 6 22 37 < 0.001 Number of weeding 2 17 0 22 10 2 34 3 12 26 7 19 28 61 < 0.001 4 1 4 1 1 0 5 Crop rotation Millet/millet 0 0 3 2 1 4 0.207 Millet/groundnut 30 30 27 28 29 96 Date of sowing Before rainy 19 17 30 27 30 82 Onset rainy 11 13 0 3 0 18 < 0.001 31 University of Ghana http://ugspace.ug.edu.gh 3.4. Discussion This survey conducted in the main pearl millet growing area in Senegal showed that almost all the farmers recognized the downy mildew disease symptom causing the transformation of the floral parts into leafy structures. A considerable proportion of them had observed this symptom in their fields during the just ended cropping year. In fact, downy mildew of pearl millet is stated to be present everywhere pearl millet is grown in Africa and Asia and is considered as among the most important diseases affecting this crop (Thakur et al. 2011). In Senegal, an incidence of over 60% and a grain yield loss up to 21% were reported in farmers’ fields (Girard & Delassus, 1978; Mbaye, 1982, 1986; Badiane, 1999). In India, a grain yield loss observed in the popular hybrid HHB 67 was estimated to 3.3 million metric tons during the pearl millet downy mildew epidemic observed in 1971 (Singh et al., 1993). However, despite the occurrence of the disease in their fields, most of the Senegalese farmers were not generally aware of the damages that it caused and did not rank it as among the main constraints. This farmers’ perception may be explained by the fact that they could not recognize properly the foliar chlorosis symptom on the leaves which affects considerably the photosynthesis mechanism and focused more on the panicle transformation symptom. Further, most of them grew open pollinated populations, which despite their low yielding potential compared to hybrid varieties, afford a certain level of resistance. Thus, despite the crop rotation strategy and destruction of transformed leafy structures plants in the fields initiated by some farmers, there is a need to strengthen the farmers in order to create awareness about the grain yield loss caused by the disease. In fact, before the transformation of panicles to leafy structures, the disease has already destroyed the plants. Damages due to Striga and low soil fertility were ranked as major constraints limiting pearl millet production in the groundnut basin of Senegal, prompting the need to include these two constraints 32 University of Ghana http://ugspace.ug.edu.gh in any pearl millet improvement programme for these targeted regions of the country. These major pearl millet production constraints reported by farmers have been also described as major problems in Asia and Africa pearl millet growing areas. Rai et al. (2012) showed that low soil fertility was one of the most important constraints leading to low productivity of 500-700 kg ha-1 in Africa. Issaka (2012) also reported that decreased soil fertility was a challenging constraint for pearl millet production in Niger. However, research conducted by Lubadde et al. (2015) in Uganda showed that ergot disease was the most important field production constraint reported by farmers followed by birds attack. In Burkina Faso, Drabo (2016) reported drought as the main constraint to pearl millet production. This finding confirms the deteriorating state of soil fertility and persistence of Striga infestation in major pearl millet production areas in Senegal and generally in West Africa. The most probable causes of this situation is overexploitation of farmland resulting from demographic pressure coupled with limited soil conservation practices. In fact, due to limited resources, farmers lack the cash to buy the optimum mineral fertilizer needed, even though they are aware that the nutrient deficiency could increase Striga infestation. This farmers’ perception agrees with earlier research in maize conducted by Jamil et al. (2012), who pointed out that the percentage of Striga seed germination increased under N and P deficiency. Available control methods to tackle Striga infestation under farmers’ fields are limited. The traditional measures taken by some of the farmers are hand pulling, crop rotation with groundnut and the use of fertilizer to contribute to soil fertility. Tackling these constraints need concerted research efforts. In recent years, some attempts have been made by the International Crop Research Institute for the Semi- Arid Tropics (ICRISAT) in Sadore (Niger) to develop Striga resistant cultivars. Hence, pearl millet Striga resistant experimental cultivars have been bred and can now be tested in Striga infested farmers’ fields in Senegal (Kountche et al., 2013). These experimental cultivars may be suitable 33 University of Ghana http://ugspace.ug.edu.gh for Senegal, and along with sound agronomic practices like the use of mineral and organic fertilizers, weeding or pulling Striga plants before flowering may contribute to integrated Striga control in pearl millet growing areas and thus increase productivity of this crop. Insect infestation is another constraint that pearl millet farmers were dealing with in the groundnut agro-ecological zone in Senegal. For instance, in Malicounda, insects were perceived as the main constraints limiting pearl millet production. Hence, to sustain pearl millet production, there is a need to consider these farmers’ identified constraints by the breeding programme. As noted by Van Ginkel et al. (2013), a holistic participatory approach to constraints analysis allows identifying possible interventions and their interactions across individual, household and national levels. It also facilitates cooperation and trust among stakeholders, and helps mobilizing resources for implementing interventions. Most of the farmers agreed that high grain yield was the most important trait in adopting pearl millet cultivars. Participatory breeding research conducted in some pearl millet growing areas in some West African countries and Namibia also found that high grain yield was an important criterion in selecting pearl millet cultivars by farmers (Monyo et al., 2001; Omanya et al., 2007; Issaka, 2012). Despite the need of having a high grain yield, most of the farmers (70%) still used seeds from local landraces, carefully selected in the previous season. The use of improved cultivars could be an effective way to increase grain yield and consequently meet farmers’ needs. Several projects have shown how the use of improved pearl millet bred-cultvars has considerably improved pearl millet production compared to local landraces in Africa using either participatory plant breeding (Bidinger, 1998) or a crossbreeding approach (Shiferaw et al., 2004). Recently, a project on dissemanation of new bred varieties run by the national extension service has shown that farmers who used improved cultivars doubled their pearl millet production compared to those who 34 University of Ghana http://ugspace.ug.edu.gh used local landraces in Senegal. In India, where more than 70% of pearl millet area is sown with single-cross hybrids, grain production considerably increased while areas sown with this crop have declined (Hash et al. 2006; Pray & Nagarajan 2009). Further, Kumar & Anand (1993) demonstrated that continuous cultivation of landraces leads to a grain yield decrease of about 40%. However, the superiority of the improved cultivars is often observed in optimal growing conditions, when the recommended cultural practices are used, as observed in Malicounda. In this part of the country, RESOPP, a farmer association provided appropriate cultural practices including the use of genetically enhanced seed-embedded technology and the estimated grain yield is this community was higher compared to the national average (MAER, 2012). Such initiatives have to be replicated in all of the pearl millet growing areas and may result in the greater use of bred-cultivars by farmers. Besides preference for high-yielding cultivars, farmers selected varieties based on their resistance to biotic stress (Striga and downy mildew) and earliness. Short growth cycle was identified as the second most important trait in Malicounda and Ngogom, located in the central part of the groundnut basin, probably because of the rainfall pattern with shorter rainy periods (< 3 months), compared to the other rural communities. Hence, the development of new improved pearl millet varieties should be based on these findings on farmers’ preferred traits and constraints. Their considerations will probably stimulate the pearl millet sector and improve livelihoods of growers in Senegal, specifically in the groundnut basin agro-ecological zone where farmers grew mainly pearl millet. Nevertheless, participatory plant breeding, where farmers are involved at different stages of the breeding process, will probably promote the ownership of these new developed cultivars and consequently enhance their adoption. 35 University of Ghana http://ugspace.ug.edu.gh 3.5. Conclusion A considerable proportion of the interviewed farmers were aware of the pearl millet downy mildew disease and were able to recognize the panicle transformation symptom. Indeed, almost all of them have experienced the disease in their pearl millet field during the just ended rainy season. However, despite the occurrence of the disease across the study sites, farmers ranked it as the sixth most important constraint. Based on their knowledge, the pearl millet production is affected by a myriad of constraints and they ranked Striga and low soil fertility as the main pearl millet production constraints. High grain yield was considered as the most important varietal preferred trait for farmers in all the study areas. 36 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR 4.0. ASSESSMENT OF PEARL MILLET DOWNY MILDEW RESISTANT DIFFERENTIAL LINES IN SENEGAL 4.1. Introduction Downy mildew caused by the obligate oomycete Sclerospora gramicola is known to be largely heterothallic. However, homothalism has been noted (Michelmore et al., 1982). This heterothallic characteristic make it highly variable (Sudisha et al., 2008). Pathogenic and genetic variability in S. graminicola have been largely studied during the past years using downy mildew resistance differential lines or DNA markers. A considerable variability of the pathogen according to zones, cultivars and periods of time have been observed in different pearl millet growing countries (Ball, 1983; Ball et al., 1986; Sivaramakrishnan et al., 2003; Thakur et al., 2004; Pushpavathi et al. 2006; Sharma et al., 2015). The first evidence of this pathogenic variation was reported by Bhat (1973) in a F1 hybrid cultivar which was found to be resistant at Mysore but susceptible on other locations in India. Likewise, a field screening conducted in India has shown that out of 123 pearl millet germplasm accessions from 15 countries across Africa and Asia that were identified as resistant, only 21 remained resistant when the screening was conducted again 14 years later (Thakur et al., 2009). Therefore, a well-known resistant cultivar in a given geographic region or time may not be resistant in a different zone or in another period owing to the pathogenic variability of the fungus. In Senegal, the pathogenic variability of this fungus following the different agro-ecological zones where pearl millet is mainly grown has not yet been investigated thoroughly. For effective resistance breeding programme, it would be necessary to study this pathogenic variability across different locations in Senegal in order to identify the suitable hotspots where field screening for 37 University of Ghana http://ugspace.ug.edu.gh downy mildew resistance could be conducted. The objective of the present study was to assess the reaction of differential pearl millet downy mildew resistant lines available, across the main pearl millet growing areas in Senegal. 4.2. Materials and Methods 4.2.1. Pearl millet differential lines The experimental material consisted of a set of 19 entries, including 11 downy mildew resistant (DMR) differential lines, one international susceptible check, five improved varieties and two local checks (Table 4.1). Seeds of the DMR differential lines, the susceptible check and the five improved varieties were obtained from the Genetic Resources Division of ICRISAT Niger while the seeds of the two checks were obtained from the national programme. For the local checks, Souna 3 is the most popular variety, while IBMV 8402 is known to have resistance to downy mildew under Senegal environment. The lines were evaluated for downy mildew resistance under field conditions during the rainy season 2015 and 2016 in four different agro-ecological zones in Senegal. 38 University of Ghana http://ugspace.ug.edu.gh Table 4.1: List of plant material used in the study Designation Description Seed source 700651 DMRDL ICRISAT, Niger 852B DMRDL ICRISAT, Niger ICMB 841B-P3 DMRDL ICRISAT, Niger 843B DMRDL ICRISAT, Niger ICMB 88004 DMRDL ICRISAT, Niger ICMB 90111-P2 DMRDL ICRISAT, Niger 81B-P6 DMRDL ICRISAT, Niger 7042 DMR DMRDL ICRISAT, Niger ICMP 451-P6 DMRDL ICRISAT, Niger ICMP 85410-P7 DMRDL ICRISAT, Niger ICMR 356 DMRDL ICRISAT, Niger 7042 S Susceptible check ICRISAT, Niger ICMV IS 90309 Improved variety ICRISAT, Niger ICMV IS 90311 Improved Variety ICRISAT, Niger ICMV IS 92326 Improved Variety ICRISAT, Niger SOSAT-C88 (Pat) Improved Variety ICRISAT, India SOSAT-C88 (ISC) Improved Variety ICRISAT, Niger Souna 3 Popular Senegalese variety (check) Bambey IBMV 8402 Popular Senegalese variety (check) Bambey DMRDL = Downy mildew resistant differential line 4.2.2. Experimental design Field evaluations were conducted for two consecutive rainy seasons (2015 and 2016) in four locations where pearl millet is mainly cultivated: Bambey (13° 49’ 12’’ North, 13° 55’ 11,99’’East), Nioro (13° 45’ 0’’ North, 15° 48’ 0’’ West), Kolda (12° 52’ 48’’ North, 14° 39’ 0’’ West) and Sinthiou Maléme (14° 43’ 12’’ North, 16° 0’ 36’’ East). Bambey and Nioro research stations are both located in the central part of the country, where pearl millet is mainly cultivated. However, these two sites are different in terms of rainfall pattern. Bambey is located in the Sahelian zone with an annual rainfall ranged from 200 to 500 mm while Nioro is in the Soudano-Sahelian zone with an annual rainfall ranged from 500 to 900 mm. Kolda and Sinthiou Maleme are located in the southern and eastern part of the country, respectively. The climate at these sites is soudanian with an annual rainfall between 900 and 1200mm. For each of the individual experiments, the land was ploughed, harrowed and then compound fertilizer (15N-15P-15K) was broadcasted over the fields at a rate of 150 kg ha-1. The 19 genotypes were evaluated in a randomized complete block design with three replications. Each entry was sown in a plots of 2 rows of 4.8 m long with spacing of 0.80 m between rows and 0.40 m between 39 University of Ghana http://ugspace.ug.edu.gh plants within a row. Urea was later applied as top dressing at a rate of 100 kg ha-1 and was done twice (50% at 15 days after sowing and 50% at 30 days after sowing). All recommended cultural practices were applied. Sprinkler irrigation was provided when needed to maintain high relative humidity for favourable disease development. 4.2.3. Data collection The disease incidence was recorded 30 days after sowing (DAS) by counting the total number of plants and the number of infected plants. Entries were classified as described by Sharma et al. (2015): (i) resistant (≤ 10% incidence), (ii) moderately resistant (10.1 to 20% incidence), (iii) susceptible (20.1 to 50% incidence), and (iv) highly susceptible (>50% incidence). Relative variation, an indicator for the stability of resistance in the cultivars, was calculated as described by Thakur et al. (2004) by dividing standard deviation over years across locations with the square root of [mean incidence x (1 - mean incidence)]. The air temperature and relative humidity from sowing to 30 days after sowing (DAS) were recorded. 4.2.4. Data analysis Analysis of variance (ANOVA) for DM incidence (%) was computed using SAS 9.4 to determine significant differences among years, locations, genotypes and their interactions. Test of homogeneity of variances was confirmed with the Bartlett test for homogeneity of group variances before combined analysis. Significant effects of treatments were determined by magnitude of F 40 University of Ghana http://ugspace.ug.edu.gh values (P ≤ 0.05). Relative downy mildew variation was compared among the pearl millet lines, locations and years. 4.3. Results 4.3.1. Seasonal conditions The weather conditions from sowing to 30 DAS are presented in Table 4.2. In general, they were the same over years and locations and were favourable for DM disease development. The minimum temperature ranged between 22.9 0C and 24.8 0C and the maximum between 32.1 0C and 33.1 0C. Relative humidity was also observed during the same period. Its minimum and maximum values ranged between 66.3% - 68.0% and 95.1 - 100%, respectively. Table 4.2: Temperature and relative humidity of the four sites from sowing to 30 DAS Relative Humidity (%) Temperature (oC) Site Year Minimum Maximum Mean Minimum Maximum Mean Bambey 65.1 99.7 82.4 24.8 33.1 28.9 Nioro 68.0 96.1 82.1 22.9 32.1 27.5 2015 Sinthiou 65.4 100.0 82.7 23.5 32.6 28.0 Kolda 66.3 95.1 80.7 23.6 32.7 28.1 Bambey 67.0 99.4 83.2 24.8 33.0 28.9 Nioro 67.5 97.1 82.3 24.0 33.0 28.5 2016 Sinthiou 66.6 97.3 82.0 23.8 32.4 27.2 Kolda 67.0 97.5 82.3 24.1 32.9 28.5 4.3.2. Variation in resistance for DM of tested lines A highly significant differences (P<0.01) between location, genotypes, years and their interactions for DMI (Table 4.3). 41 University of Ghana http://ugspace.ug.edu.gh Table 4.3: Effects of location, year, line, and their interactions for DMI Source of variation df Type III SS Mean Square F Value Pr > F Replication (Location) 8 290.6 36.3 0.26 0.9769 Year 1 14290.8 14290.8 103.92 <.001 Location 3 12202.9 4067.6 29.58 <.001 Year x Location 3 6025.6 2008.5 14.61 <.001 Genotype 18 174653.8 9702.9 70.56 <.001 Genotype x Year 18 16608.5 922.6 6.71 <.001 Genotype x Location 54 23608.3 437.2 3.18 <.001 Genotype x Year x Location 54 23138.8 428.5 3.12 <.001 Error 289 39741.0 137.5 Df=degree of freedom; SS=Sum of squares; Pr=Probability, DMI = Dpwny mildew Incidence The tested lines showed differential responses to downy mildew disease. The 11 differential downy mildew resistance lines varied in their relative positions for DM and have confirmed their status as differential lines, except for ICMR 356 which was resistant across locations and years (Table 4.4). The international downy mildew susceptible check 7042 S was highly susceptible during the two years in all the four locations. The improved open pollinated varieties were consistently resistant to moderately resistant across location during the two years. Table 4.4: Average downy mildew incidence (%) and relative variation on tested pearl millet lines across locations during 2015 and 2016 rainy seasons Variety Bambey Kolda Nioro Sinthiou Mean SE Relative variation 700651 1.4 18.3 11.0 0.0 7.7 3.1 0.57 852B 1.7 19.3 20.0 4.0 10.9 5.9 0.87 ICMB 841B-P3 1.8 23.7 20.0 0.0 11.0 4.3 0.65 843B 5.1 23.5 6.0 1.7 9.1 2.5 0.43 ICMB 88004 4.7 24.9 21.6 1.3 13.1 4.4 0.64 ICMB 90111-P2 13.6 6.9 14.1 0.9 8.9 3.1 0.53 81B-P6 14.4 12.6 8.3 2.3 9.4 3.6 0.60 7042 DMR 12.6 21.7 55.2 9.7 24.8 5.0 0.57 ICMP 451-P6 39.4 3.3 40.0 0.0 21.6 8.4 0.89 ICMP 85410-P7 0.8 7.1 22.9 1.2 8.0 3.8 0.68 ICMR 356 1.0 1.5 7.9 1.0 2.8 1.1 0.33 7042 S 94.8 100.0 97.2 83.6 93.9 1.9 0.38 ICMV-IS 90309 10.8 17.9 6.9 1.5 9.2 3.0 0.48 ICMV-IS 90311 5.4 0.8 0.0 3.6 2.7 0.9 0.28 ICMV-IS 92326 1.4 11.1 8.1 2.9 5.9 2.3 0.48 SOSAT-C88-Pat 2.5 6.6 16.8 4.7 7.1 2.6 0.33 SOSAT-C88-ISC 6.0 13.3 7.3 1.9 7.6 1.7 0.47 Souna 3 10.5 18.2 2.7 7.3 9.7 2.8 0.46 IBMV 8402 2.0 4.8 3.6 2.6 3.3 1.3 0.34 Mean across entries 12.1 17.6 18.7 6.9 Standard error 2.3 1.8 2.6 2.9 Relative variation 0.76 0.72 0.78 0.75 42 University of Ghana http://ugspace.ug.edu.gh Among the tested pearl millet lines, the improved varieties from ICRISAT along with 700651, 843B, ICMB 88004, ICMB 90111-P2, ICMP 85410-P7 and ICMR 356 were the most resistant ones with downy mildew incidence ranging from 2.8% to 9.4% across location. The line 7042 S was highly susceptible and recorded an average disease incidence of 93.9%. The relative variation in the incidence of downy mildew of tested lines varied between 0.28 and 0.89. The two downy mildew resistance differential lines ICMP 451-P6 and 852 B exhibited higher relative variation. In contrast, lines ICMV-IS 90311 and ICMR 356 showed low relative downy mildew variation. The highest DMI over the two years was recorded at Nioro (18.7%), Kolda (17.6%) and Bambey (12.1%) while the mean DMI was lowest at Sinthiou Maléme (6.9%). The differences in DMI over the two years were significantly different. 4.4. Discussion The weather conditions, particularly the temperature and relative humidity, are known to play an essential role and can have an impact on the development and spread of the pearl millet downy mildew disease. The analysis of these two parameters at the respective trial sites during the rainy seasons of 2015 and 2016 showed that conditions for pearl millet downy mildew development were fully met according to Thakur et al. (2011) who stated that the downy mildew disease development is observed in high relative humidity (80-90% RH) and moderate temperature (20- 30°C) conditions. The average temperature and relative humidity recorded for the four environments and over years were 28.2 0C and 82.2 %, respectively. In the present study, downy mildew incidence in the sensitive downy mildew line 7042 S was high with an average of 94% across the four locations over the two years and served as an indicator of good disease pressure. This high downy mildew incidence on this international susceptible line 43 University of Ghana http://ugspace.ug.edu.gh has also been reported in Senegal during the past years by Gupta (1981) and confirmed the occurrence of the disease in the main pearl millet growing area in Senegal. He conducted an international pearl millet DM nursery at Bambey and Nioro during three consecutive years and recorded a downy mildew incidence varying between 74 and 95% on this line. All the downy mildew differential lines, except ICMR 356, differed in their reaction to downy mildew in the different locations and years and therefore indicated pathogenic variation across locations and years in Senegal. This evidence of pathogenic variation was also reported in pearl millet lines evaluated in different environments across Africa and Asia and other crops such as cucumber (Shetty et al., 2002; Gwary et al., 2007; Thakur et al., 2007; Sharma et al., 2013). Some of these differential lines were tested in different locations in India and West Africa by ICRISAT and their partners under the International Pearl Millet Downy Mildew Nursery collaborative research. Results from this work provided also evidence of pathogenic variability of the pearl millet downy mildew disease across and within Indian and African countries. The higher downy mildew incidence average was observed at Bagauda in Nigeria and Durgapura in India and lower downy mildew incidence average at Coimbatore and Aurangabad in India ( Singh et al., 1993; Singh, 1995; Thakur et al., 2004). This study showed also that the improved OPVs included in the field evaluations were resistant across the locations. This result implied that recurrent selection used in the improvement of these OPVs was effective and confirmed that inheritance of resistance to the pearl millet downy mildew disease is under additive gene action (Issaka, 2012). The pearl millet differential line ICMR 356 was identified as the most resistant line and it also exhibited lower relative variation among the tested lines. Therefore, this line could be used as a donor parent for stable downy mildew resistance by the Senegalese pearl millet improvement programme. In contrast, ICMP 451-P6 and 852 B lines showed the highest relative variation across 44 University of Ghana http://ugspace.ug.edu.gh the tested locations. This indicated that the resistance observed in these lines is specific and they could not be used as parents in developing pearl millet lines with stable downy mildew resistance. However, these two lines were used as downy mildew and rust resistance donors in the pearl millet hybrid programme at ICRISAT Patancheru (Kumar et al., 1995). The line ICMP 451-P6 has also shown resistance to downy mildew in Kolda and Sinthiou Maléme and therefore could be used by the pearl millet programme to develop pearl millet lines which have both downy mildew and rust resistances effective for the southern part of the country. The downy mildew incidence was higher in Nioro, Kolda and Bambey locations compared to Sinthiou Maléme. This finding is consistent with Girard & Delassus (1978) and Mbaye (1986) who reported that a high downy mildew incidence, up to 60%, was observed in the groundnut basin where Bambey and Nioro are located, while incidence as low as 3% was observed in Sinthiou Maléme. Therefore, the field screening process for identifying breeding materials for resistance to DM which is effective in Senegal should be conducted in Bambey, Nioro and Kolda research stations which seem to be hotspots. 4.5. Conclusion The international susceptible check 7042 S exhibited high downy incidence across the locations and years, providing evidence of presence of the disease across the locations. All the differential downy mildew resistance lines varied in their relative positions for downy mildew incidence across locations, except for ICMR 356 which showed stable resistance across locations. The pearl millet downy mildew incidence was higher in Kolda, Bambey and Nioro compared to Sinthiou Maléme. Thus, the field screening of the pearl millet breeding materials should be conducted in these locations. 45 University of Ghana http://ugspace.ug.edu.gh CHAPTER FIVE 5.0. EVALUATION OF PEARL MILLET INBRED LINES IN DOWNY MILDEW PREVALENCE AREAS IN SENEGAL 5.1. Introduction Pearl millet production is always constrained with many biotic challenges; the most important being downy mildew caused by Sclerospora graminicola (Thakur et al., 2011). This pathogen induces systemic infection causing two types of symptoms; foliar chlorosis and panicle transformation (Singh, 1995). Grain yield losses up to 60% have been reported in India and several African countries (Singh et al., 1993). In Senegal, a comprehensive report for grain yield loss due to this disease is not available. However, a report from the national research institute reported a downy mildew incidence up to 60% in the central part of the country where pearl millet is mainly cultivated (DAPSA, 2016; Mbaye, 1986). Unlike in India where farmers treat pearl millet grains with metalaxyl to reduce pearl millet downy mildew incidence, in Senegal pearl millet seed is often sown directly into the soil without seed treatments. Additionally, most resource-poor farmers in Senegal cannot afford the recommended insecticide treatment for downy mildew control due to lack of financial resources. Furthermore, the use of this insecticide may be harmful to the environment because of residual toxicity. Thus, the use of resistant varieties as a management strategy against Sclerospora graminicola is considered to be cost-effective and eco-friendly. Breeding for pearl millet downy mildew resistance has been an important goal for the Indian pearl millet breeding programme because of several downy mildew epidemics which have threatened hybrid technology, particularly the one observed in mid-1970 (Yadav & Rai, 2013). In Africa, such epidemics have not yet been observed and pearl millet downy mildew has not received the 46 University of Ghana http://ugspace.ug.edu.gh needed attention compared to India because of the use of open-pollinated varieties which despite their low yielding potential compared to hybrid varieties afford a certain level of resistance. However, with the climate change and reduction of pearl millet growing areas in several West African environments, breeders are putting more attention on pearl millet hybrid development. Therefore, identifying new sources of resistance will definitively facilitate the development of pearl millet hybrid varieties combining high yield and resistance to downy mildew disease. The objective of this study was to identify sources of resistance from pearl millet inbred lines developed from a collection of landraces originated from West and Central African countries under downy mildew prone areas in Senegal. 5.2. Materials and Methods 5.2.1. Plant material and geographical location of the study regions A total of 101 lines were used for this study (Table 5.1). Among these lines, SOSAT C 88 and 7042 S were used as tolerant and susceptible checks, respectively. The other entries were inbred lines developed by ICRISAT and Institut de Recherche et de Developpement (IRD) teams from a collection of landraces originating from West and Central Africa, which are known as centres of origin and diversity of pearl millet (Haussmann et al., 2006; Oumar et al., 2008). These landraces originated from seven countries: Burkina-Faso, Cameroon, Central African Republic, Mali, Mauritania, Niger and Senegal. The inbred lines derived from these landraces were also evaluated under low P conditions and some of them performed well (Gemenet et al., 2014). Field evaluation of these lines for resistance against downy mildew and other agronomic traits were conducted at two different locations: Bambey (13° 49’ 12’’ North, 13° 55’ 12’’ West) and Nioro (13° 45’ 0’’ North, 15° 48’ 0’’ West). These locations were identified as hotspots for pearl millet downy mildew from a previous study on characterization of downy mildew isolates from four locations. 47 University of Ghana http://ugspace.ug.edu.gh Furthermore, Bambey and Nioro are located in the groundnut basin (central part of the country) where pearl millet is mainly cultivated in Senegal. 5.2.2. Experimental design The experiment was laid out in an augmented design of eleven blocks during the rainy season of 2016. Pearl millet downy mildew susceptible and resistant lines 7042 S and SOSAT C-88, respectively were used as checks. In each block, nine tested genotypes were grown along with the two checks (Figure 5-1). Each of the lines was grown in a single-row plots and consisted of 14 hills. The distance between rows was 80 cm while between plants within row was 40 cm. Table 5.1: List of genotypes and their respective origin Entry Name Country of origin Entry Name Country of origin 1 SOSAT C-88 NIGER 52 IBL 100-5-1 NIGER 2 7042 S NIGER 53 IBL 101-3-1 NIGER 3 IBL 001-4-1 UNKNOW 54 IBL 102-3-1 NIGER 4 IBL 003-B-1 UNKNOW 55 IBL 105-3-1 NIGER 5 IBL 011-4-1 UNKNOW 56 IBL 105-B-1 NIGER 6 IBL 012-1-1 UNKNOW 57 IBL 106-B-1 NIGER 7 IBL 012-2-1 UNKNOW 58 IBL 107-B-1 NIGER 8 IBL 015-1-1 UNKNOW 59 IBL 110-B-1 NIGER 9 IBL 021-3-1 UNKNOW 60 IBL 111-3-1 NIGER 10 IBL 023-2-1 UNKNOW 61 IBL 114-6-1 NIGER 11 IBL 024-3-1 UNKNOW 62 IBL 117-2-1 NIGER 12 IBL 026-1-1 UNKNOW 63 IBL 119-B-1 SENEGAL 13 IBL 026-2-1 UNKNOW 64 IBL 121-2-1 SENEGAL 14 IBL 028-B-1 UNKNOW 65 IBL 125-B-1 BENIN 15 IBL 033-1-1 UNKNOW 66 IBL 131-6-1 MALI 16 IBL 037-4-1 UNKNOW 67 IBL 133-2-1 MALI 17 IBL 037-5-1 UNKNOW 68 IBL 138-B-1 MALI 18 IBL 040-1-1 UNKNOW 69 IBL 141-B-1 MALI 19 IBL 040-5-1 UNKNOW 70 IBL 143-1-1 MALI 20 IBL 047-1-1 CAMEROUN 71 IBL 143-2-1 MALI 21 IBL 050-1-1 CENTRAFRIQUE 72 IBL 149-1-1 MALI 22 IBL 053-2-1 CENTRAFRIQUE 73 IBL 150-B-1 MALI 23 IBL 053-3-1 CENTRAFRIQUE 74 IBL 151-2-1 MALI 24 IBL 055-4-1 MALI 75 IBL 155-2-1 MALI 25 IBL 056-2-1 MALI 76 IBL 160-1-1 BURKINA FASO 26 IBL 058-5-1 MALI 77 IBL 161-1-1 BURKINA FASO 27 IBL 061-1-1 MALI 78 IBL 165-1-1 BURKINA FASO 28 IBL 064-1-1 MALI 79 IBL 167-5-1 BURKINA FASO 29 IBL 065-B-1 MALI 80 IBL 170-1-1 BURKINA FASO 30 IBL 066-3-1 MALI 81 IBL 170-B-1 BURKINA FASO 31 IBL 066-4-1 MALI 82 IBL 173-1-1 BURKINA FASO 32 IBL 067-2-1 NIGER 83 IBL 173-3-1 BURKINA FASO 33 IBL 067-B-1 NIGER 84 IBL 174-3-1 BURKINA FASO 34 IBL 069-4-1 NIGER 85 IBL 179-2-1 BURKINA FASO 35 IBL 070-1-1 NIGER 86 IBL 179-3-1 BURKINA FASO 36 IBL 071-4-1 NIGER 87 IBL 180-2-1 MAURITANIE 37 IBL 073-B-1 NIGER 88 IBL 181-2-1 MAURITANIE 38 IBL 077-1-1 BURKINA FASO 89 IBL 183-4-1 MAURITANIE 39 IBL 079-B-1 BURKINA FASO 90 IBL 183-5-1 MAURITANIE 40 IBL 081-2-1 BURKINA FASO 91 IBL 185-3-1 MALI 48 University of Ghana http://ugspace.ug.edu.gh Entry Name Country of origin Entry Name Country of origin 41 IBL 082-B-1 SENEGAL 92 IBL 186-1-1 MALI 42 IBL 084-1-1 SENEGAL 93 IBL 188-1-1 MALI 43 IBL 091-1-1 NIGER 94 IBL 198-1-1 UNKNOW 44 IBL 092-3-1 NIGER 95 IBL 198-2-1 UNKNOW 45 IBL 093-1-1 NIGER 96 IBL 200-3-1 UNKNOW 46 IBL 094-2-1 NIGER 97 IBL 206-1-1 UNKNOW 47 IBL 095-1-1 NIGER 98 SL 2-B-1 SENEGAL 48 IBL 095-4-1 NIGER 99 SL 4-3-1 SENEGAL 49 IBL 098-1-1 NIGER 100 SL 5-1-1 SENEGAL 50 IBL 098-3-1 NIGER 101 SL 5-4-1 SENEGAL 51 IBL 099-3-1 NIGER 5.2.3. Field evaluation for downy mildew resistance Field evaluation for downy mildew resistance in the two locations was conducted according to method described by Thakur et al. (2011). An infector row was established as border throughout the entire length of the field in order to increase the disease pressure using a mixture of two lines: a local landrace (Souna 3) and a highly susceptible line (7042 S). After three weeks, when downy mildew incidence in the infector row was over 70%, the test lines were sown (Figure 5-1). Infector row Test row Figure 5-1: Field experimental at Bambey research station After one week from sowing of the test lines, plants were thinned down to one plant per hill. Mineral fertilizer, NPK (15-15-15), was applied at the recommended rates of 150 kg ha-1. Urea, at the rate of 100 kg ha-1, was applied later as topdressing after thinning (50%) and at booting stage 49 University of Ghana http://ugspace.ug.edu.gh (50%). Weeds were managed using hoes on the third and the sixth weeks after sowing of test lines. High average relative humidity of over 82% and average temperature of 28% were observed for about 30 days after sowing. 5.2.4. Data collection The climatic data such as temperature, relative humidity and rainfall were recorded daily. The numbers of infected seedlings in each plot was counted and expressed as a percentage of infected seedlings at 30 days after sowing. Entries were classified as described by Sharma et al. (2015): (i) resistant (≤ 10% incidence), (ii) moderately resistant (10.1 to 20% incidence), (iii) susceptible (20.1 to 50% incidence), and (iv) highly susceptible (>50% incidence). The disease severity score was taken after flowering on each individual plants in each plot using a 1-5 scale as described by Williams et al. (1981), where: 1 = no downy mildew disease symptoms, 2= symptoms on aerial tillers only, 3 = symptoms on less than 50% basal tillers, 4 = symptoms on more than 50% basal tillers, and 5 = total destruction of stand or no production of normal head. Disease mildew severity (DMS) index (%) was calculated using formula described by Williams et al. (1981), as follows: n1(1 − 1) + n2(2 − 1) + n3(3 − 1) + n4(4 − 1) + n5(5 − 1) DMS = N(5 − 1) Where 50 University of Ghana http://ugspace.ug.edu.gh • n1...n5 = number of plants with different disease grades described in the 1–5 scale above; • N = total number of plants assessed. Other agronomic traits such as plant height, panicle length, number of productive tillers and flowering time were also recorded. 5.2.5. Data analysis Data analysis was carried out using SAS version 9.4 (SAS Institute Inc. 2013). The analysis of variance (ANOVA) was performed for each location using the SAS macro for analysis of data from augmented block designs. Data from the two locations were subjected to homogeneity chi square test of significance before pooling them. The analysis of variance of pooled data was then performed using restricted maximum likelihood estimation method (REML; PROC MIXED) to determine significant differences among genotypes, locations and their interactions with genotypes and locations as fixed while block was considered as random. For the combined analysis, the following model was used: 1234 = µ + 62 + 73 + 6723 + β4 + 9234 Where: Yijk= Plot value of genotype i at location j in the incomplete block k; µ = general mean; Gi= effect of genotype i; Lj =effect of location j; G*Lij= interaction effect of genotype i with location j; βk = the effect of the incomplete block k; eijk= residual error. 51 University of Ghana http://ugspace.ug.edu.gh The correlation matrix was computed for all the collected data using PROC CORR procedure of SAS 9.4. Hierarchical Ascendant Clustering (HAC) was carried out using Ward’s method based on Euclidean distances. A Factorial Discriminant Analysis (FDA) was performed in order to refine the clustering and classify the pearl millet lines into different groups based on observed data. The HAC and FDA were performed using the XLSTAT software (Addinsoft, 2012, Paris). 5.3. Results Highly significant differences (P<0.01) among tested lines were observed for all the traits measured (Table 5.2). The effects of Genotype x Location interaction (G x L) were also significant (P<0.05) for all the traits except for plant height. The downy mildew parameters and the productive tillers were not affected by location. Table 5.2: F value for measured traits under artificial downy mildew infestation fields F Value Effect DF DMI DMS Flowering Plant Height Panicle Length Productive tillers Genotype (G) 100 96.27*** 27.07*** 177.72*** 7.42*** 23.23*** 4.43** G x L 100 18.08*** 4.08** 28.48*** 1.05 5.42*** 1.98* Location (L) 1 3.04 1.89 50.57*** 38.67*** 11.85** 3.74 DF= degree of freedom; DMI = Downy mildew incidence; DMS = Downy mildew severity; *, **, ***, Significant at 0.05 and 0.01 and 0.001 probability levels, respectively At both research stations, an important variation was observed among the tested lines for downy mildew disease parameters and other agronomic traits (Table 5.3). DMI of the tested lines across the two locations varied from 0 to 100% whereas the productive tillers ranged from 0 to 6. DMS at Bambey ranged from 0 to 97% while at Nioro it varied from 0 to 100%. Flowering time of the tested lines ranged from 38 to 71 DAS at Bambey and from 44 to 83 DAS at Nioro. Mean panicle length of the tested lines at Bambey ranged from 10 to 55 cm with an overall mean of 30 cm while 52 University of Ghana http://ugspace.ug.edu.gh at Nioro it varied from 8 to 65 cm with an average of 27 cm. For plant height, it varied from 55 to 257 cm at Bambey research station while at Nioro research station it ranged from 60 to 305 cm. Table 5.3: Performance of the evaluated genotypes at Bambey and Nioro research stations Site Parameters Mean Minimum Maximum SE DMI (%) 23 0 100 2.9 DMS (%) 15 0 97 2.3 Flowering (das) 50 38 71 0.7 Bambey Plant height (cm) 178 55 257 3.7 Panicle length (cm) 30 10 55 0.9 Productive tillers 1.8 0 6 0.09 DMI (%) 21 0 100 2.7 DMS (%) 16 0 100 2.6 Flowering (das) 56 44 83 0.9 Nioro Plant height (cm) 198 60 305 3.8 Panicle length (cm) 27 8 65 0.9 Productive tillers 2 0 6 0.1 DMI = Downy mildew incidence; DMS = Downy mildew severity; SE = Standard Error Most of the genotypes showed less than 10% DMI and DMS at both locations. At Bambey research station, out of the 101 tested lines, 58 including SOSAT C 88 were resistant, 12 were moderately resistant, 17 were susceptible and 14 including 7042 S were highly susceptible (Appendix 5.1). At Nioro research station, 65 genotypes exhibited less than 10% DMI while only 14 including 7042 S had more than 50% DMI (Appendix 5.2). The resistant check SOSAT C 88 was moderately resistant with 11% DMI. Among the lines classified as resistant across the two locations, 20 of them were disease-free at both locations and performed better than the highly susceptible lines (Table 5.4). For instance, inbred line IBL 174-3-1 had short plants (58 cm) with small panicles (13 cm) and did not produce any productive tillers under downy mildew infested fields while inbred line IBL 098-1-1 had long panicles (30 cm) and produced up to four productive tillers plant-1. 53 University of Ghana http://ugspace.ug.edu.gh Table 5.4: Performance of the downy mildew disease-free and highly susceptible pearl millet genotypes across Bambey and Nioro research stations Entry DMI (%) DMS (%) Flowering Plant Height Panicle Productive (das) (cm) length (cm) tillers IBL 001-4-1 0 0 61 207 33 2 IBL 040-5-1 0 0 46 168 22 1 IBL 055-4-1 0 0 50 207 36 1 IBL 065-B-1 0 0 70 183 27 4 IBL 084-1-1 0 0 49 183 25 2 IBL 095-4-1 0 0 44 231 47 2 IBL 098-1-1 0 0 58 186 30 4 IBL 106-B-1 0 0 55 258 46 2 IBL 111-3-1 0 0 49 161 24 3 IBL 114-6-1 0 0 70 210 30 1 IBL 133-2-1 0 0 75 196 33 2 IBL 141-B-1 0 0 48 194 21 2 IBL 143-1-1 0 0 45 171 22 3 IBL 143-2-1 0 0 47 187 22 3 IBL 160-1-1 0 0 68 176 25 2 IBL 161-1-1 0 0 66 212 17 2 IBL 179-2-1 0 0 54 209 26 2 IBL 183-4-1 0 0 49 181 27 3 IBL 186-1-1 0 0 49 224 26 3 IBL 188-1-1 0 0 49 153 38 2 IBL 012-1-1 54 14 62 159 25 1 IBL 012-2-1 63 13 64 155 36 2 IBL 174-3-1 63 53 74 58 13 0 IBL 138-B-1 65 56 67 172 25 1 IBL 121-2-1 78 81 72 63 10 1 IBL 069-4-1 79 26 57 180 35 2 7042 S 86 85 45 159 14 1 IBL 064-1-1 86 59 68 161 22 1 IBL 024-3-1 92 39 59 193 34 2 SL 4-3-1 95 34 63 192 43 1 DMI = Downy mildew incidence; DMS = Downy mildew severity The correlation matrix computed using the combined data for all the observed variables over the two locations during the rainy season of 2016 is presented in Table 5.5. Results showed that DMI was strongly and positively correlated with the DMS. The downy mildew parameters (DMI and DMS) were significantly and negatively correlated with plant height and productive tillers at the < 0.05 levels of probability. Productive tillers were significantly and negatively correlated with 54 University of Ghana http://ugspace.ug.edu.gh flowering time, but significantly and positively correlated with plant height. Panicle length was strongly and positively correlated with the plant height but negatively correlated with the DMS. Table 5.5: Correlation matrix between pairs of studied variables DMS Flowering Plant Panicle Length Productive Height tillers DMI 0.64*** 0.31* -0.33** -0.09 -0.36** DMS 0.20 * -0.41*** -0.37** -0.38** Flowering -0.08 -0.04 -0.25* Plant Height 0.51*** 0.26** Panicle length 0.10 DMI = Downy mildew incidence; DMS = Downy mildew severity; *, **, ***, Significant at 0.05 and 0.01 and 0.001 probability levels, respectively Hierarchical ascendant cluster (HAC) analysis based on data recorded at Bambey and Nioro during the rainy season of 2016 sorted out the 101 lines into three clusters based on the seven variables (Table 5.6). Table 5.6: Characteristics of the identified clusters based on the HAC analysis DMI Flowering Plant height Panicle Productive Cluster Number DMS (%) (%) (das) (cm) length (cm) tillers 1 38 10 7 55 225 32 2 2 8 81 49 63 147 25 1 3 54 13 7 54 167 29 2 DMI = Downy mildew incidence; DMS = Downy mildew severity Cluster I contains 38 lines out of the 101 tested entries, including one line from Senegal. It is characterized by DM resistant lines with an average of 10% and 7% for DMI and DMS, respectively and includes the resistant check SOSAT C 88. It also had the tallest plants (225 cm) with the longest panicles (32 cm), and early flowering lines with an average of 55 DAS and a moderate number of productive tillers. 55 University of Ghana http://ugspace.ug.edu.gh Cluster II encompasses the lowest number of entries (8), including the susceptible check 7042 S and two lines originated from Senegal. Lines belonging to this cluster were susceptible to downy mildew disease (DMI = 81%) and produced a lower number of productive tillers. It showed also the late flowering (63 DAS) lines with short plants (147 cm) and small panicles (25 cm). Cluster III contains the highest number of lines (54). This cluster comprises moderately resistant lines with an average of 13% for DMI and produced on average 2 productive tillers plant-1. It showed, however, early flowering plants (54 DAS) with moderate panicle length (29 cm). It also shows plants with more than 1.5 m height but less than 2 m compared to plants in cluster I. Observations (axes F1 and F2 : 100,00 %) 8 7 6 5 22 4 16 963 46443 86 80 2357 27 2 43 782 5707 C3lu 52 8sterI37 62 10 68 28 21 18 58 4886551924 693 9459271475955 401 12 991167 0 87 ClusterII150309679 -3.5 -2.5 -1.51 4927778-316 418302-951.8115124 82 95320 619133200.5 382651 1.154 2.5 6 3.5 344.5 5.5 6.5 7.5 8.5 9.5 754471908686 Clu 55 3 st69e54r6III 3950 47 7 60-2 64 93 39-330 98 8472 31 -4 17 -5 -6 -7 F1 (59,58 %) ClusterI ClusterII ClusterIII Barycentres Figure 5-2: Factorial discriminant analysis based on the study entries To identify the most discriminating variables of the three identified clusters, a factorial discriminant analysis was performed. Based on the Lambda Wilk test from the six variables studied, the disease parameters (DMI and severity) and plant height were the most discriminant parameters (Table 5.7). 56 F2 (40,42 %) University of Ghana http://ugspace.ug.edu.gh The axis F1 shows 60% total variability and was mainly correlated with the disease parameters and the flowering time. The second axis F2 revealed 40% of the overall variability and was mostly correlated with plant height and panicle length. The three clusters identified based on the two axes have confirmed the clustering made by HAC method for most of the tested lines ( Figure 5-2). Table 5.7: Lambda Wilk test from the six variables Variable Lambda F p-value DMI 0.3260 101.2922 < 0.0001 DMS 0.5101 47.0538 < 0.0001 Flowering 0.9118 4.7394 0.0108 Plant Height 0.4221 67.0929 < 0.0001 Panicle length 0.9324 3.5506 0.0325 Productive tillers 0.9129 4.6733 0.0115 DMI = Downy mildew incidence; DMS = Downy mildew severity 5.4. Discussion The present study aimed at identifying sources of resistance for pearl millet downy mildew by exploring the phenotypic variability among pearl millet inbred lines derived from West and Central African pearl millet landraces for downy mildew and agronomic traits. The pearl millet lines displayed tremendous phenotypic variability for all the traits assessed. For instance, the downy mildew incidence ranged from 0 to 100% across the two locations. This huge variation for the different observed traits was also noted in the accessions where these inbred lines were derived. This is due to their origins known to be the centre of origin and diversity of pearl millet (Haussmann et al., 2006; Pucher et al., 2015). The phenotypic variability observed indicated that these lines can be useful resources in pearl millet breeding programmes to create genetic variation and improve elite pearl millet germplasm. Furthermore, these contrasting morphological phenotypes for the observed traits could be exploited by developing mapping population for any genetic analysis (Xu, 2010). The downy mildew incidence ranged from 0 to 100% across the two 57 University of Ghana http://ugspace.ug.edu.gh locations. Among the tested lines, twenty inbred lines including IBL 055-4-1, IBL 095-4-1, IBL 098-1-1, IBL 106-B-1, IBL 111-3-1, IBL 160-1-1, IBL 161-1-1 and IBL 186-1-1 were disease- free and exhibited some good agronomic and farmer preferred traits (Issaka, 2012). These inbred lines could be used as parents to form synthetic or open pollinated pearl millet varieties which are resistant to the downy mildew disease under Senegalese environments. A strong but negative correlation was noted between downy mildew parameters and plant height and productive tillers parameters. The presence of significant association among these parameters is consistent with previous findings and indicated that the disease reduced growth parameters of pearl millet (Wilson et al., 2008). Gupta & Singh (1996) reported a plant height reduction of 28% due to the disease. A similar result was obtained by Yadav et al. (1997) who reported up to 53% reduction in plant height in a downy mildew susceptible pearl millet variety. Productive tillers were significantly and negatively correlated with flowering time, and additionally significantly and positively correlated with plant height. It was also observed that the productive tillers were positively correlated with plant height. The considerable phenotypic variation observed in the present study allowed ranking the 101 entries into three clusters, which mainly contrasted in terms of downy mildew incidence, severity and plant height. Considering their resistance to the disease and plant height, cluster I contained the tallest and resistant entries, cluster II the shorter and highly susceptible entries and cluster III the moderately resistant entries with intermediate plant height. This finding is consistent with Kumari et al. (2017) who also classified 221 Indian pearl millet accessions using hierarchical clustering method into three clusters based on agronomic and downy mildew resistance traits. In this study, inbred lines belonging to cluster I, in addition of being resistant to the disease, have longer panicles and flowered earliest. Furthermore, among these 37 inbred lines which belonged 58 University of Ghana http://ugspace.ug.edu.gh to this cluster, six of them were ranked by Gemenet et al. (2014) as among the 15 best inbred lines under low P conditions, one of the main constraints limiting pearl millet production in the Sahel (Bationo & Mokwunye, 1991). These inbred lines included IBL 047-1-1; IBL 055-4-1; IBL 065- B-1; IBL 161-1-1; IBL 155-2-1; IBL 149-1-1 and IBL 003-B-1. They should thus be good candidates for developing hybrids and synthetic varieties which will perform well under low P conditions and combine early flowering time, long panicles and resistance to pearl millet downy mildew. However, the per se performance of a given inbred line does not always predict the performance of their hybrids (Falconer, 1989). This performance is determined by its combining ability which can be general or specific. Therefore, for the development of pearl millet hybrids and synthetic varieties using these inbred lines, it is essential to study their genetic potential in hybrid combinations through combining ability and heterosis analyses. 5.5. Conclusion A large phenotypic variation was observed among the lines for downy mildew disease and agronomic parameters. Out of the 101 genotypes, 20 of them such as IBL 055-4-1, IBL 095-4-1, IBL 098-1-1, IBL 106-B-1, IBL 111-3-1, IBL 160-1-1, IBL 161-1-1 and IBL 186-1-1 exhibited 0% disease incidence and severity and were more resistant than SOSAT C 88. The lines were grouped into 3 clusters mainly based on the disease parameters and plant height and lines which belonged to cluster I showed resistance to the disease, good agronomic performance and exhibited some farmer preferred traits such as earliness. Therefore, these lines would be good candidates for the development of downy mildew tolerant pearl millet varieties adapted to Senegalese environments or for the improvement of farmer preferred varieties that are susceptible to the disease. 59 University of Ghana http://ugspace.ug.edu.gh CHAPTER SIX 6.0. HETEROSIS AND COMBINING ABILITY FOR DOWNY MILDEW RESISTANCE AND GRAIN YIELD OF PEARL MILLET IN SENEGAL 6.1. Introduction As the world population is continuously increasing and is projected to reach nine billion by 2050, pearl millet (Pennisetum glaucum) is expected to play an important role for achieving food security in West and Central African countries, which have the highest population growth rates in the world (Tan, 2015). However, in this part of the world, yields of pearl millet are very low compared to yields in India and yet African farmers, particularly in Senegal, have not adopted improved varieties in a large scale. In pearl millet growing areas in Africa, adoption rate of improved OPVs varied from 5 to 37% (Christinck et al., 2014). In contrast, most Indian farmers are using improved varieties, particularly hybrids since the 1960s. Indeed, in India hybrids had 25-30% grain yield advantage over OPVs, leading to the rapid adoption of the hybrids whose yield increased from 305 kg ha-1 during 1951-1955 to 998 kg ha-1 during 2008-2012 (Yadav & Rai, 2013). Thus, enhancement of pearl millet production and productivity in Africa, which is a high priority, can be achieved through the identification of elite parent materials. These parents can be used to develop hybrids or synthetic varieties which theoretically have higher grain yield potential compared to local populations. ICRISAT has developed pearl millet inbred lines derived from landraces originating from West and Central Africa which can be useful in developing high yielding pearl millet hybrids and synthetic varieties with considerable adaptation to this pearl millet growing environment. These lines were screened for pearl millet downy mildew resistance in Senegal. Some of them showed good agronomic traits and resistance to the pearl millet downy mildew. However, the per se 60 University of Ghana http://ugspace.ug.edu.gh performance of these pearl millet inbred lines does not predict the performance of hybrids for disease resistance and agronomic traits (Hallauer et al., 2010). Therefore, to make effective use of these pearl millet inbred lines, their general and specific combining abilities need to be elucidated (Falconer, 1989). This genetic information can be obtained by different mating design including line x tester (Kempthorne, 1957). Thus, General Combining Ability (GCA) and Specific Combing Ability (SCA) estimates of pearl millet inbred parents or landraces for different traits such as micronutrients (Govindaraj et al., 2013), grain quality (Parmar et al. 2013), early maturity (Kumhar, 2007) and fodder yield (Chaudhary et al. 2012) have been reported. Drabo (2016) reported that additive genetic action was important in controlling traits such as grain yield, flowering time and panicle length in pearl millet. In Senegal, there is scanty published information on the combining ability of pearl millet lines for disease resistance and agronomic traits. The objectives of this study were to: i. estimate combing ability and heterosis of pearl millet inbred lines for downy mildew, yield and other agronomic traits under downy mildew infested fields, and ii. identify superior pearl millet hybrids for yield, yield components and resistance to downy mildew. 61 University of Ghana http://ugspace.ug.edu.gh 6.2. Materials and methods 6.2.1. Plant material and mating design Seventeen inbred lines and two OPVs (Table 6.1) were used to produce F1 top-cross hybrids. The 17 inbred lines were used as females and crossed each to the two OPVs used as males according to the line x tester mating design (Kempthorne, 1957) to generate 34 F1 hybrids. The OPVs varieties were considered as testers and the inbred lines as lines. The two testers named Souna 3 and SOSAT C 88 are popular varieties adapted to the groundnut agro-ecological zone. The pearl millet inbred lines used for the study were selected from a pool of pearl millet landraces from West and Central Africa converted to inbred lines through successive selfing (Gemenet et al., 2015). These inbred lines were selected through a downy mildew phenotypic evaluation conducted at Bambey and Nioro research stations during the rainy season 2016. They showed less than 10% DMI and were classified as resistant varieties. The seeds of the male parents were planted in 4 different dates in order to synchronise the flowering time of these male parents with the ones of the female parents. Thus, from January 2017, the sowing of the male parents was done every week and seeds of each of the male parents were sown in 5 rows of 15 hills per row. All the female parents were sown in one time during the second sowing date of the male parents in a one row- plot of 15 hills. At the booting stage, at least plant heads of 4 panicles per plant of the male and female parents were covered in order to avoid undesirable pollination. At flowering, each covered panicle of female plant was pollinated with bulk pollen collected from at least 20 different plants of the male parent. At maturity stage, F1 panicles of the female parents were harvested and the lower and upper of each of the panicle were cut before threshing to minimize outcrossing from unknown plants or selfing. Indeed, because of the protogynous nature of the crop the stigmata of a plant are receptive 62 University of Ghana http://ugspace.ug.edu.gh before the shedding of pollen and the flowering starts from the upper to the lower. Then, the upper part of the panicle may be pollinated by unknown plants if not covered on the right time and the lower part of the panicle may be pollinated by the pollen from the same plant. After threshing, F1 seeds from the same female parent were bulked and used as F1 hybrids. The 34 F1 hybrids along with the 17 inbred lines, the two testers and an OPV Thialack II as check, providing 54 genotypes, were used for the evaluation (Table 6.1). Table 6.1: List of parental lines and check used in the study Number Genotype Source Role in crosses Status Response to downy mildew 1 IBL 001-4-1 ICRISAT Line Inbred line Resistant 2 IBL 003-B-1 ICRISAT Line Inbred line Resistant 3 IBL 011-4-1 ICRISAT Line Inbred line Resistant 4 IBL 021-3-1 ICRISAT Line Inbred line Resistant 5 IBL 055-4-1 ICRISAT Line Inbred line Resistant 6 IBL 091-1-1 ICRISAT Line Inbred line Resistant 7 IBL 095-4-1 ICRISAT Line Inbred line Resistant 8 IBL 098-3-1 ICRISAT Line Inbred line Resistant 9 IBL 106-B-1 ICRISAT Line Inbred line Resistant 10 IBL 110-B-1 ICRISAT Line Inbred line Resistant 11 IBL 114-6-1 ICRISAT Line Inbred line Resistant 12 IBL 119-B-1 ICRISAT Line Inbred line Resistant 13 IBL 155-2-1 ICRISAT Line Inbred line Resistant 14 IBL 165-1-1 ICRISAT Line Inbred line Resistant 15 IBL 179-2-1 ICRISAT Line Inbred line Resistant 16 IBL 179-3-1 ICRISAT Line Inbred line Resistant 17 IBL 206-1-1 ICRISAT Line Inbred line Resistant 18 Souna 3 Senegal Tester OPV (improved) Susceptible 19 SOSAT C 88 Senegal Tester OPV (improved) Resistant 20 Thialack II Senegal Check OPV (improved) Susceptible 6.2.2. Study sites, experimental design, and field management The 34 hybrids, together with their parents and the OPV check were evaluated under rainfed conditions during the rainy season of 2017 at two locations in Senegal. The study sites were Bambey (13° 49’ 12’’ North, 13° 55’ 12’’ West) and Nioro (13° 45’ 0’’ North, 15° 48’ 0’’ West) 63 University of Ghana http://ugspace.ug.edu.gh research stations. Both locations are located in the groundnut agro-ecological zone, which is the main pearl millet growing area in Senegal, and were characterized as hotspots for downy mildew in the previous study. The genotypes were arranged in a 9 x 6 alpha lattice design with three replications at each site. Each block was surrounded by a downy mildew infector row consisting of a downy mildew susceptible line, 7042 S, sown 3 weeks before the tested materials. Each plot consisted of one row of 8.1 m length with a spacing of 0.9 m between rows and between plants within a row. All other recommended cultural practices were applied. At least 10 seeds were planted per hole and later thinned to two plants. The fields were weeded two times after sowing. The trials received the recommended 15N-15P-15K basal fertilizer at a rate of 150 kg ha−1 just before sowing. During the crop development a top dressing using area at a rate of 100 kg ha-1 was done in two fractions: 50 kg ha-1 after thinning and 50 kg ha -1 after the second weeding. 6.2.3. Data collection Downy mildew incidence (DMI) was obtained by dividing the total number of infected plants 30 DAS from a plot by the total number of plants. Flowering (FWT) was recorded by counting the total number of days from sowing to when 50% of plants in a plot flowered. Five random plants were selected in each plot just before harvesting to measure the plant height (PH) from the base of the plant to the tip of the panicle, number of productive tillers (PT) by counting the number of tillers per plant which produced productive panicles; panicle length (PL) and panicle diameter (PDIA). Panicles harvested in a plot were weighed to determine panicle yield (PY) and then threshed. Grains obtained in each plot were weighed and used to calculate grain yield (GY) in kg ha-1 using this formula: GY (kg ha-1) = [grain weight (kg plot -1) x 10,000 / plot size m2] 64 University of Ghana http://ugspace.ug.edu.gh For each harvested plot, five random samples of 1000 grains was weighed using a sensitive balance to determine the 1000 grain weight (TGW). 6.2.4. Data analysis Analysis of variance for each experimental site as well as for combined data after the homogeneity test of variance across the two experimental sites was performed using the general linear model (GLM) procedure in SAS version 9.4 (SAS Institute, Cary, NC). The following mathematical linear model was used: 1234 = µ + 62 + 73 + 6723 + :34 + ;34 + 9234 Where: 1234 is the observed value of the variable for the ith entry in the jth location within kth replication; µ is the overall general mean; 62 is the effect of the ith genotype; 73 is the effect of the jth location; 6723 is the interaction effect of the ith entry and the jth location; :34 is the effect of the kth replication within the jth location; ;34 is the effect of the lth block of the kth replication in the jth location; 9234< is the experimental pooled error. For the combining ability, analysis of variance was performed for traits that showed significant differences among hybrids. Thus, the sum of square of hybrids was partitioned into various 65 University of Ghana http://ugspace.ug.edu.gh variations due to lines, testers and their interactions based on the following statistical model described by Singh & Chaudhary (1977): 1234 = µ + =2 + >3 + =>23 + +9234 Where: 1234 kth observation on the ith and jth progeny; µ is the overall general mean; =2 is the effect of the ith male; >3 is the effect of the jth female; =>23 interaction effect; 9234 is error associated with each observation. The values of the general combining ability for both male and female and the specific combining ability effects for all the studied traits were estimated as follows: 6?@A2BC = 7DE9 F9GE HD − IJ9:GKK F9GE (H… ) 6?@MCNMC< = O9PQ9: F9GE HR − IJ9:GKK F9GE (H… ) S?@T2BC U MCNMC< = ?:VPP F9GE HDR − 7DE9 F9GE HD − O9PQ9: F9GE HR + IJ9:GKK F9GE (H … ) Where, X..= Overall mean X thi = mean of all the hybrids containing an i line average over all replications, sites and males Xj = mean of all the hybrids containing a jth tester average over all replications, sites and females Xij = mean of the cross between ith line and jth tester across all replications and sites. 66 University of Ghana http://ugspace.ug.edu.gh The significance of the GCA effects was tested using the formula described by Cox & Frey (1984): QWXT = YZ[ Where, S. e = hC \] (fWX gXTC)^_` (a`bc) − ?ℎ9ty) SQGEuG:u ℎ9Q9:VPDP (Sw) = x H 100 ?ℎ9ty (> −=|) =Du − {G:9EQ ℎ9Q9:VPDP (=|w) = x H 100 =| (>x − w|)}9QQ9: {G:9EQ ℎ9Q9:VPDP (}|w) = H 100 w| Where F1 denotes the mean performance of the hybrid averaged over the two locations. The mean value of the OPV check was used to calculate the standard heterosis. The parent giving the highest mean value was used as better parent in the calculation of high-parent heterosis while the average between the two parents was used for the mid-parent heterosis. 67 University of Ghana http://ugspace.ug.edu.gh 6.3. Results 6.3.1. Performance of hybrids and parents across locations Combined analysis of variance across locations showed highly significant (P < 0.01) genotype effect for all measured traits (Table 6.2). Site effect was also significant for all the traits, except for downy mildew incidence (DMI) and plant height (PH). However, interaction Genotype x Site effect was only significant for flowering time (FWT), thousand grain weight (TGW), panicle yield (PY) and grain yield (GY. Table 6.2: Mean squares for studied traits across locations Source of variation d.f. DMI FWT PH PL PDIA PT TGW PY GY Rep (Site) 4 232.4** 28.4** 746.4 28.5 0.4 2.2 4.9* 2499694* 1258762** Block(RepxSite) 48 85.5 8.2 359.9 32.9 0.14 1.5 2.3 817310 324630 Site 1 84.6 248.9*** 0.1 444.1** 1.6** 29.4** 36.2*** 5279671* 3168503** Genotype 53 156.4*** 82.3*** 3193.5*** 278.6*** 0.46*** 5.1*** 10.4*** 3975982*** 1157443*** Genotype x Site 53 72.7 13.62** 488.1 45.1 0.19 2.0 2.4* 1418904* 603031** Error 164 60.04 7.9 474.7 31.6 0.17 1.9 1.6 997156 297217 DMI=Downy Mildew Incidence; FWT= Flowering Time; PH=Plant Height; PL=Panicle Length; PDIA=Panicle Diameter; PT=Productive tillers; TGW=1000 grain weight; PY=Panicle Yield; GY=Grain Yield; *, **, ***, Significant at 0.05 and 0.01 and 0.001 probability levels, respectively; d.f.= degree of freedom; Rep =replication All genotypes were resistant to downy mildew with a mean DMI of 4%, except for IBL 155-2-1 and its progeny with Souna 3 which displayed both 22% DMI (Table 6.3). Days from sowing to 50% flowering (DAS) of genotypes across the two sites ranged from 50 to 69 DAS with an average of 56 DAS. The genotypes were tall with plant height ranging from 2 to 3.2 m. Panicle length of the pearl millet genotypes varied from 27 to 58 cm with an average of 44 cm while their diameter ranged from 1.2 to 2.7 cm with a mean diameter of 2.1 cm. The number of productive tillers ranged from 2 to 6 tillers per plant with a mean value of 4 productive tillers per plant. The 1000 seeds weight varied from 5 to 12 g with a mean of 9 g. The panicle yield of genotypes across the two sites varied from 376 kg ha-1 for IBL 119-B-1 to 4190 kg ha-1 for IBL 110-B-1 x Souna 3 and their grain yield from 92 kg ha-1 for IBL 119-B-1 to 2024 kg ha-1 for IBL 206-1-1 x Souna 3. 68 University of Ghana http://ugspace.ug.edu.gh Based on grain yield, the F1 hybrids were generally more productive compared to the inbred lines and OPVs. The top five genotypes across sites were hybrids IBL 206-1-1 x Souna 3 (2024 kg ha- 1); IBL 091-1-1 x Sosat C 88 (2019 kg ha-1); IBL 206-1-1 x Sosat C 88 (1988 kg ha-1); IBL 001- 4-1 x Souna 3 (1923 kg ha-1) and IBL 003-B-1 x Sosat C 88 (1883 kg ha-1). Among these top hybrids, two of them involved the inbred line IBL 206-1-1 as parent. The hybrid IBL 179-2-1 x Souna 3 (735 kg ha-1) was the lowest yielding among the tested hybrids. The check, Thialack II was the most productive OPV with an average grain yield of 1694 kg ha-1 and ranked among the ten best genotypes. The best inbred line was IBL 003-B-1 (1340 kg ha-1). 69 University of Ghana http://ugspace.ug.edu.gh Table 6.3: Performance of tested genotypes for studied traits across sites Genotype DMI FWT PHIG PLEN PDIA PT TGW PY GY IBL 001-4-1 2 60 255 34 2.1 2 10 1104 585 IBL 003-B-1 4 51 231 31 2.5 5 11 1872 1340 IBL 011-4-1 14 59 237 44 1.9 3 7 1164 428 IBL 021-3-1 0 63 223 32 2.1 2 8 1178 419 IBL 055-4-1 6 65 210 37 1.6 2 6 517 286 IBL 091-1-1 0 60 246 47 2.2 4 8 2328 1136 IBL 095-4-1 0 50 243 37 1.9 5 10 2015 947 IBL 098-3-1 11 51 239 31 1.8 4 10 1946 1188 IBL 106-B-1 6 54 278 39 1.8 3 10 1340 710 IBL 110-B-1 0 62 240 45 1.6 2 6 997 404 IBL 114-6-1 0 62 199 27 1.8 2 8 750 324 IBL 119-B-1 0 69 260 41 1.2 2 5 376 92 IBL 155-2-1 22 57 200 35 1.3 3 7 1099 466 IBL 165-1-1 2 64 238 39 1.6 4 7 1349 489 IBL 179-2-1 0 60 272 43 1.9 2 8 850 394 IBL 179-3-1 8 54 233 35 2.0 5 8 1497 886 IBL 206-1-1 0 56 228 38 1.8 4 8 2102 1126 IBL 001-4-1 x Souna 3 17 53 273 43 2.0 5 11 3475 1923 IBL 003-B-1 x Souna 3 8 52 259 44 2.1 4 9 3218 1646 IBL 011-4-1 x Souna 3 5 52 255 50 1.8 4 8 2368 942 IBL 021-3-1 x Souna 3 4 57 210 44 1.9 3 7 1772 852 IBL 055-4-1 x Souna 3 0 58 241 38 2.7 3 8 2232 1397 IBL 091-1-1 x Souna 3 0 54 250 54 2.0 4 8 2369 1211 IBL 095-4-1 x Souna 3 0 52 262 58 2.0 4 9 2975 1641 IBL 098-3-1 x Souna 3 19 54 277 46 1.9 4 9 2326 1289 IBL 106-B-1 x Souna 3 6 56 256 52 1.6 3 9 2462 1351 IBL 110-B-1 x Souna 3 0 56 274 52 2.6 5 7 4190 1493 IBL 114-6-1 x Souna 3 4 57 270 47 2.0 3 8 2991 1181 IBL 119-B-1 x Souna 3 0 56 321 49 2.0 5 9 3586 1530 IBL 155-2-1 x Souna 3 22 56 287 46 1.9 5 9 2562 1101 IBL 165-1-1 x Souna 3 0 59 289 45 2.1 4 8 3420 1612 IBL 179-2-1 x Souna 3 0 59 298 44 2.0 3 9 1906 735 IBL 179-3-1 x Souna 3 2 55 275 46 2.2 4 8 2943 1402 IBL 206-1-1 x Souna 3 2 54 245 44 2.1 5 9 3671 2024 IBL 001-4-1xSosat C88 9 56 284 36 2.4 4 9 2997 1579 IBL 003-B-1xSosat C88 2 50 234 28 2.6 4 12 3182 1883 IBL 011-4-1xSosat C88 0 50 257 39 2.1 5 10 2468 1331 IBL 021-3-1xSosat C88 0 54 284 37 2.3 4 10 3053 1654 IBL 055-4-1 x Sosat C 88 0 57 298 40 2.0 3 9 1824 1131 IBL 091-1-1 x Sosat C 88 0 50 260 41 2.3 4 10 2912 2019 IBL 095-4-1 x Sosat C 88 2 55 281 39 2.3 4 10 2426 1368 IBL 098-3-1 x Sosat C 88 6 57 285 43 2.2 3 9 2722 1381 IBL 106-B-1 x Sosat C 88 0 57 255 37 2.2 3 9 1933 949 IBL 110-B-1 x Sosat C 88 2 55 256 38 2.6 4 9 2827 1433 IBL 114-6-1 x Sosat C 88 0 52 264 31 2.4 5 9 2944 1416 IBL 119-B-1x Sosat C88 2 60 298 41 2.0 4 8 1993 871 IBL 155-2-1xSosat C88 0 51 260 35 1.9 6 10 2414 1215 IBL 165-1-1x Sosat C88 0 60 287 39 2.5 4 10 2785 1114 IBL 179-2-1x Sosat C88 0 54 277 32 2.4 6 9 2920 1586 IBL 179-3-1xSosat C88 0 56 260 29 2.4 4 9 2375 1247 IBL 206-1-1x Sosat C8 0 53 242 34 2.4 5 9 3325 1988 Souna 3 5 56 273 55 2.0 4 8 2918 1268 SOSAT C 88 2 51 251 33 2.3 4 10 2568 1548 THIALACK II 7 53 292 54 1.9 4 8 3380 1694 Mean 4 56 259 41 2.1 4 9 2313 1171 Range 0-22 50-69 200-321 27- 58 1.2-2.7 2-6 5-12 376-4190 92-2024 Standard Deviation 5.6 4.1 25.8 7.3 0.3 0.9 1.4 868.8 483.1 DMI=Downy Mildew Incidence; FWT= Flowering, PH=Plant Height; PL=Panicle Length; PDIA=Panicle Diameter; PT=Productive tillers; TGW=1000 grain weight; PY=Panicle Yield; GY=Grain Yield 70 University of Ghana http://ugspace.ug.edu.gh Genotypes flowered 2 days earlier in Nioro (55 DAS) compared to Bambey (57 DAS) but they performed better under Nioro conditions than under Bambey conditions (Table 6.4). The average TWG in Nioro was 8 g while in Bambey it was 9 g. The panicle yield of genotypes under Nioro conditions ranged from 461 kg ha-1 for the inbred line IBL 110-B-1 to 4647 kg ha-1 for hybrid IBL 110-B-1 x Souna 3 while under Bambey conditions it varied from 251 kg ha-1 for the inbred IBL 119-B-1 to 4660 kg ha-1 for the hybrid IBL 165-1-1 x Souna 3. Grain yield of genotypes under Bambey environment ranged from 61 kg ha-1 for inbred IBL 119-B-1 to 2162 kg ha-1 for hybrid IBL 165-1-1 x Souna 3. In Nioro, the grain yield varied from 87 kg ha-1 for inbred line IBL 110- B-1 to 2966 kg ha-1 for the hybrid IBL 206-1-1 x Sosat C 88. Based on grain yield, the ten best genotypes in Nioro were only hybrids while in Bambey, the three OPV were among the top ten genotypes. The hybrids IBL 091-1-1 x Sosat C 88 and IBL 206-1-1 x Souna 3 performed well under both locations and were among the best ten genotypes across the two environments. 71 University of Ghana http://ugspace.ug.edu.gh Table 6.4: Mean flowering time, yield and related traits of genotypes per site -1 -1 Genotype FWT (das) TGW (g) PY (kg ha ) GY (kg ha ) Bambey Nioro Bambey Nioro Bambey Nioro Bambey Nioro IBL 001-4-1 60 61 9 10 1284 924 768 402 IBL 003-B-1 51 52 11 12 1856 1888 1512 1168 IBL 011-4-1 63 55 7 7 1195 1132 437 419 IBL 021-3-1 62 63 9 7 1423 933 718 119 IBL 055-4-1 68 63 5 7 321 712 139 432 IBL 091-1-1 63 56 8 8 2557 2099 1140 1132 IBL 095-4-1 49 50 10 11 1658 2372 924 969 IBL 098-3-1 49 53 9 11 1687 2206 1057 1318 IBL 106-B-1 55 54 10 10 780 1900 428 992 IBL 110-B-1 61 62 7 5 1533 461 721 87 IBL 114-6-1 65 58 6 9 628 873 267 382 IBL 119-B-1 73 64 3 6 251 500 61 122 IBL 155-2-1 55 58 7 6 1235 962 649 284 IBL 165-1-1 64 63 7 8 1970 727 792 186 IBL 179-2-1 59 60 8 7 1078 622 559 228 IBL 179-3-1 54 53 8 8 1330 1664 746 1025 IBL 206-1-1 60 52 8 8 920 3283 564 1687 IBL 001-4-1 x Souna 3 54 51 10 12 2625 4325 1223 2624 IBL 003-B-1 x Souna 3 52 51 9 10 2758 3677 1011 2281 IBL 011-4-1 x Souna 3 54 50 9 8 2151 2585 923 961 IBL 021-3-1 x Souna 3 58 56 7 7 1672 1872 1113 591 IBL 055-4-1 x Souna 3 59 57 9 8 2140 2324 1609 1185 IBL 091-1-1 x Souna 3 55 53 8 7 2375 2363 1295 1127 IBL 095-4-1 x Souna 3 53 50 9 8 2847 3102 1513 1768 IBL 098-3-1 x Souna 3 56 53 7 10 1983 2669 1018 1559 IBL 106-B-1 x Souna 3 57 55 8 9 1908 3015 1005 1697 IBL 110-B-1 x Souna 3 59 53 7 7 3733 4647 934 2051 IBL 114-6-1 x Souna 3 57 56 7 10 3154 2828 1436 926 IBL 119-B-1 x Souna 3 59 53 10 8 4175 2996 1829 1231 IBL 155-2-1 x Souna 3 59 54 8 9 1913 3211 907 1294 IBL 165-1-1 x Souna 3 60 57 8 8 4660 2180 2162 1061 IBL 179-2-1 x Souna 3 60 58 9 8 1728 2084 617 854 IBL 179-3-1 x Souna 3 56 54 7 9 2031 3855 822 1982 IBL 206-1-1 x Souna 3 56 53 8 11 3285 4058 1609 2438 IBL 001-4-1 x Sosat C 88 57 54 9 10 2503 3490 1298 1860 IBL 003-B-1 x Sosat C 88 48 51 12 12 3562 2802 1959 1806 IBL 011-4-1 x Sosat C 88 49 51 10 10 2380 2555 1259 1403 IBL 021-3-1 x Sosat C 88 53 55 9 10 2526 3580 1304 2005 IBL 055-4-1 x Sosat C 88 58 56 9 9 1630 2017 1007 1255 IBL 091-1-1 x Sosat C 88 48 51 11 10 2563 3262 1976 2061 IBL 095-4-1 x Sosat C 88 56 53 9 10 1963 2889 1080 1657 IBL 098-3-1 x Sosat C 88 59 55 8 11 1260 4184 404 2357 IBL 106-B-1 x Sosat C 88 59 54 8 11 2133 1733 944 953 IBL 110-B-1 x Sosat C 88 56 53 9 10 2367 3288 1032 1835 IBL 114-6-1 x Sosat C 88 53 51 9 10 2917 2971 1662 1171 IBL 119-B-1 x Sosat C 88 59 60 9 7 2992 994 1261 482 IBL 155-2-1 x Sosat C 88 51 51 9 11 1939 2889 801 1630 IBL 165-1-1 x Sosat C 88 60 59 11 9 3801 1770 1450 778 IBL 179-2-1 x Sosat C 88 53 55 9 10 2449 3390 1103 2070 IBL 179-3-1 x Sosat C 88 58 54 9 10 1644 3107 727 1767 IBL 206-1-1 x Sosat C 88 54 52 8 9 2411 4240 1010 2966 Souna 3 55 56 8 8 3465 2372 1731 805 SOSAT C 88 50 52 9 10 2956 2180 1707 1389 THIALACK II 54 52 7 8 3716 3044 1659 1729 Mean 57 55 8 9 2186 2441 1072 1270 SD 5 4 2 2 955 1089 476 703 FWT= Flowering Time; TGW=1000 grain weight; PY=Panicle Yield; GY=Grain Yield 72 University of Ghana http://ugspace.ug.edu.gh 6.3.2. Combining ability analysis across locations The total variation due to crosses was partitioned into line, tester and line x tester interaction (Table 6.5). The mean squares due to hybrids were significant for all the traits except for PY and GY. Line mean squares across the two locations were also significant for all the traits except for PDIA while tester mean squares were not significant for PT, PY and GY. Line x tester mean square were significant for most traits except PDIA, PY and GY. The mean squares due to line x site were significant for TGW, PY and GY whereas the mean squares due to tester x site interaction were significant for FWT and DMI. However, the mean squares due to site x line x tester interaction were not significant for all the traits across the two locations. Table 6.5: Mean squares for combining ability for studied traits across locations Source of variation d.f. DMI FWT PH PLEN PDIA PT TGW PY GY Rep (Site) 2 55.5 12.7 300.9 1.2 0.7 3.1 5.0 316965 25012 Site 1 158.8* 178.8*** 711.6 280.0** 0.2 12.2* 26.1** 9629606** 6762757*** Hybrid 33 188.1*** 45.7*** 2852.4*** 309.4*** 0.5* 4.3** 6.2*** 1962646 688597 Line (GCA) 16 229.6*** 64.9*** 3607.8*** 160.6*** 0.4 5.4** 4.4** 2290926* 775061* Tester (GCA) 1 776.5*** 40.5** 270.7 6226.1*** 1.6* 0.7 64.8*** 1998612 123285 Line x Tester (SCA) 16 110.0** 26.8*** 2258.6*** 88.5*** 0.4 3.4* 4.4*** 1632117 637466 Site x Line 16 52.4 7.5 311.4 18.0 0.3 2.3 3.9** 3292948** 1351732*** Site x Tester 1 184.4* 54.0** 79.1 26.1 1.6 0.9 0.04 95074 444360 Site x Line x Tester 16 64.1 6.0 173.9 28.2 0.3 1.2 1.2 703213 424419 Error 134 39.7 5.3 281.7 25.4 0.3 1.9 1.8 1244846 414329 DMI=Downy Mildew Incidence; FWT= Flowering Time; PH=Plant Height; PL=Panicle Length; PDIA=Panicle Diameter; PT=Productive tillers; TGW=1000 grain weight; PY=Panicle Yield; GY=Grain Yield; GCA=General Combining Ability; SCA=Specific Combining Ability; *, **, ***, Significant at 0.05 and 0.01 and 0.001 probability levels, respectively; d.f.= degree of freedom; Rep =replication 6.3.3. Relative contributions of mean squares for additive and non-additive effects Across the two locations, the relative importance of mean squares, for additive effect (GCAm + GCAf) were higher for all the traits compared to the dominance effect (SCA). GCA effects accounted for most of the variation observed for most of the traits with more than 80% of the total genotypic variation among hybrids except for PH, PT, PY and GY. The overall contribution of GCA sums of squares to the total mean squares across the two locations varied from 58% for GY 73 University of Ghana http://ugspace.ug.edu.gh to 99% for PLEN while SCA varied from 1% for PLEN to 42% for grain yield. The contribution of GCAm was higher than GCAf and SCA for DMI, PLEN, PDIA and TGW while GCAf was larger than GCAm and SCA mean square for FWT, PH, PT, PY and GY. The contribution of GCAf (50%) was slightly higher than SCA (42%) for grain yield. 100% 90% 80% 70% 60% 50% SCA 40% GCAm 30% GCAf 20% 10% 0% DMI FWT PH PLEN PDIA PT TGW PY GY Studied traits Figure 6-1: Proportion of total mean squares of studied traits attributable to GCAm, GCAf and SCA across locations DMI=Downy Mildew Incidence; FWT= Flowering Time; PH=Plant Height; PL=Panicle Length; PDIA=Panicle Diameter; PT=Productive tillers; TGW=1000 grain weight; PY=Panicle Yield; GY=Grain Yield; GCAm=General Combining Ability for male parent; GCAf = General Combining Ability for female parent SCA=Specific Combining Ability: 6.3.4. General combining ability effects The contribution of lines and testers to crosses for traits studied across the two locations is presented in Table 6.6. For female lines, significant GCA effects were observed for most of the traits while for male lines, significant GCA effects were recorded only for TGW. For DMI, the GCA effects varied from -3.2 for IBL 091-1-1 to 9.7 for IBL 001-4-1. Positive and significant GCA effects for DMI were observed on parental lines IBL 001-4-1; IBL 098-3-1 and IBL 155-2- 1. For FWT, GCA effects ranged from -4.2 for IBL 003-B-1 to 4.2 for IBL 165-1-1 and both 74 % contribution of GCAm, GCAf and SCA University of Ghana http://ugspace.ug.edu.gh positive and negative significant GCA effects were observed. Estimates of GCA effects for PH ranged from -25.2 for IBL 206-1-1 to 19.4 for IBL 179-2-1. Out of the 19 parental lines, six showed negative and significant effects whereas four lines exhibited positive and significant effects for PH. GCA effects for PLEN varied from -5.7 for IBL 003-B-1 to 6.7 for IBL 095-4-1 with both positive and negative significant effects whereas GCA effects for PDIA ranged from -0.3 for IBL 011-4-1 to 0.4 for IBL 110-B-1 with no significant effects. The GCA effects due to parental lines for PT across locations varied from -1.2 to 1.3 for IBL 055-4-1 and IBL 155-2-1, respectively. Significant positive GCA effects for PT was observed in lines IBL 155-2-1 and IBL 206-1-1 while significant negative GCA effects were observed in lines IBL 055-4-1 and IBL 106-B-1. Across research stations, the GCA for TGW ranged from -0.9 for IBL 110-B-1 to 1.7 for IBL 003-B-1. The tester SOSAT C 88 and the inbred line IBL 003-B-1 had significant positive GCA effects while the tester Souna 3 showed significant negative GCA effects for TGW. For PY and GY traits, no significant GCA effects were showed. However, among parental lines, inbred lines IBL 206-1- 1, IBL 003-B-1, IBL 001-4-1, IBL 091-1-1, IBL 095-4-1 and IBL 110-B-1 manifested desirable positive GCA effects for GY and most other studied traits under the two research stations. In contrast, inbred lines IBL 011-4-1, IBL 106-B-1, IBL 155-2-1 and IBL 179-2-1 ranked among the worst lines for GY with negative GCA effects. 75 University of Ghana http://ugspace.ug.edu.gh Table 6.6: Estimates of GCA effects of lines and testers evaluated across the two sites Lines DMI (%) FWT (das) PH (cm) PLEN (cm) PDIA (cm) PT TGW (g) PY (kg ha-1) GY (kg ha-1) IBL 001-4-1 9.7*** -0.8 10.1 -2.5 0.1 0.6 1.1 483.9 354.4 IBL 003-B-1 1.6 -4.2*** -21.7*** -5.7*** 0.3 0.2 1.7* 447.9 367.4 IBL 011-4-1 -0.4 -3.8*** -12.0* 3.0* -0.3 0.3 0.0 -334.1 -260.4 IBL 021-3-1 -1.1 0.8 -21.3*** -1.2 -0.2 -0.8 -0.5 -339.1 -143.8 IBL 055-4-1 -3.2 2.6** 0.9 -2.9* 0.3 -1.2** -0.4 -724.2 -132.9 IBL 091-1-1 -3.2 -3.0** -13.2* 5.5*** -0.1 -0.2 0.0 -111.2 217.9 IBL 095-4-1 -2.4 -1.7* 3.0 6.7*** -0.2 -0.2 0.0 -51.7 107.5 IBL 098-3-1 9.0*** 0.8 12.5* 2.8* 0.0 -0.6 -0.1 -227.9 -62.2 IBL 106-B-1 -0.4 1.4 -12.5* 2.3 -0.2 -1.1* 0.1 -554.6 -247.0 IBL 110-B-1 -2.3 0.6 -3.4 3.0* 0.4 0.4 -0.9 756.7 66.2 IBL 114-6-1 -1.4 -0.5 -1.7 -2.7* 0.0 -0.2 -0.3 215.7 -98.2 IBL 119-B-1 -2.4 2.9** 41.2*** 3.3* -0.2 0.3 -0.4 37.4 -196.4 IBL 155-2-1 7.6** -1.0 5.3 -1.2 -0.2 1.3** 0.2 -263.9 -239.0 IBL 165-1-1 -3.2 4.2*** 19.8*** 0.3 0.2 0.0 -0.1 350.7 -34.2 IBL 179-2-1 -3.2 1.6 19.4*** -3.7** 0.0 0.3 -0.1 -339.1 -236.0 IBL 179-3-1 -2.4 0.9 -1.1 -4.3** 0.1 -0.2 -0.3 -92.9 -72.4 IBL 206-1-1 -2.3 -0.9 -25.2*** -2.7* 0.0 0.9* 0.0 746.5 609.0 SE 2.0 0.8 4.9 1.2 0.2 0.4 0.6 508.2 325.6 Testers Souna 3 2.0 0.4 -1.2 5.5 -0.1 -0.1 -0.6* 99.0 -24.6 SOSAT C 88 -2.0 -0.4 1.2 -5.5 0.1 0.1 0.6* -99.0 24.6 SE 1.0 0.5 0.6 0.4 0.1 0.1 0.0 21.6 46.7 DMI=Downy Mildew Incidence; FWT= Flowering Time; PH=Plant Height; PL=Panicle Length; PDIA=Panicle Diameter; PT=Productive tillers; TGW=1000 grain weight; PY=Panicle Yield; GY=Grain Yield; *, **, ***, Significant at 0.05 and 0.01 and 0.001 probability levels, respectively. 6.3.5. Specific combining ability effects Significant positive and negative SCA effects were recorded for all the observed traits (Table 6.7). The top-cross hybrid IBL 155-2-1 x Sosat C 88 was the only one which exhibited negative and significant SCA effects for DMI. In addition, its SCA effects for FWT and PH were negative and significant while its SCA effects for PT were significant and positive. Amongst 34 top-cross hybrids, six top-cross hybrids had significant SCA effects, of which three were positive. All the Significant and positive SCA effects for PY and GY were recorded in the crosses between SOSAT C 88 with the inbred lines IBL 179-2-1, IBL 091-1-1 and IBL 021-3-1. 76 University of Ghana http://ugspace.ug.edu.gh Table 6.7: Estimates of SCA effects for hybrids evaluated across the two sites DMI FWT PLEN PDIA PY GY Hybrid (%) (das) PH (cm) (cm) (cm) PT TGW(g) (kgha-1) (kgha-1) IBL 001-4-1 x Souna3 2.5 -1.9* -4.7 -1.8 0 0.6 1.2*** 140.3 196.8 IBL 003-B-1 x Souna3 1.1 0.5 13.8*** 2.6 -0.3** 0.2 -0.9** -81.1 -93.6 IBL 011-4-1 x Souna 3 0.8 0.7 0.2 0.2 -0.1 -0.1 -0.1 -148.8 -170.1 IBL 021-3-1 x Souna 3 0.1 1 -35.4*** -2.3 -0.1 -0.4 -0.8* -739.7** -376.5* IBL 055-4-1 x Souna 3 -2 0.2 -27.2*** -6*** 0.5*** 0.4 0.1 105.2 157.6 IBL 091-1-1 x Souna 3 -2 1.8* -3.7 0.9 0 -0.4 -0.9** -370.6* -379.3* IBL 095-4-1 x Souna 3 -2.8 -2** -8.3* 4.2** 0.1 0.4 0.1 175.4 160.8 IBL 098-3-1 x Souna 3 4.4 -1.7* -2.9 -4* -0.1 0.5 0.3 -296.6 -21.5 IBL 106-B-1 x Souna 3 0.9 -0.8 1.4 2.1 -0.1 -0.2 0.1 165.6 226 IBL 110-B-1 x Souna 3 -2.9 0.3 10.3** 1.9 0.2 0.6 -0.5 582.3* 54.2 IBL 114-6-1 x Souna 3 -0.2 1.8* 4.2 2.2 -0.1 -0.6 0 -75.6 -92.9 IBL 119-B-1 x Souna 3 -2.8 -2.3** 12.4** -1 0.1 0.4 0.9** 697.4** 354.1 IBL 155-2-1 x Souna 3 8.8*** 2.1** 14.5*** 0.3 0.1 -0.8* 0 -25 -32.8 IBL 165-1-1 x Souna 3 -2 -0.9 2.3 -2.6 -0.2 0.4 -0.5 218.4 273.5 IBL 179-2-1 x Souna 3 -2 2.1** 11.8** 0.5 0.1 -1.2*** 0.3 -605.7* -400.8* IBL 179-3-1 x Souna 3 -1.1 -0.9 8.5* 3.2* -0.2 0.3 0 184.9 102.3 IBL 206-1-1 x Souna 3 -1 0.1 2.7 -0.4 0.1 -0.2 0.7* 73.9 42.3 IBL 001-4-1 x Sosat C 88 -2.5 1.9* 4.7 1.8 0 -0.6 -1.2*** -140.3 -196.8 IBL 003-B-1 x Sosat C 88 -1.1 -0.5 -13.8*** -2.6 0.3** -0.2 0.9*** 81.1 93.6 IBL 011-4-1 x Sosat C 88 -0.8 -0.7 -0.2 -0.2 0.1 0.1 0.1 148.8 170.1 IBL 021-3-1 x Sosat C 88 -0.1 -1 35.4*** 2.3 0.1 0.4 0.8* 739.7** 376.5* IBL 055-4-1 x Sosat C 88 2 -0.2 27.2*** 6*** -0.5*** -0.4 -0.1 -105.2 -157.6 IBL 091-1-1 x Sosat C 88 2 -1.8* 3.7 -0.9 0 0.4 0.9** 370.6* 379.3* IBL 095-4-1 x Sosat C 88 2.8 2** 8.3* -4.2** -0.1 -0.4 -0.1 -175.4 -160.8 IBL 098-3-1 x Sosat C 88 -4.4 1.7* 2.9 4* 0.1 -0.5 -0.3 296.6 21.5 IBL 106-B-1 x Sosat C 88 -0.9 0.8 -1.4 -2.1 0.1 0.2 -0.1 -165.6 -226 IBL 110-B-1 x Sosat C 88 2.9 -0.3 -10.3** -1.9 -0.2 -0.6 0.5 -582.3* -54.2 IBL 114-6-1 x Sosat C 88 0.2 -1.8* -4.2 -2.2 0.1 0.6 0 75.6 92.9 IBL 119-B-1 x Sosat C 88 2.8 2.3** -12.4** 1 -0.1 -0.4 -0.9** -697.4** -354.1 IBL 155-2-1 x Sosat C88 -8.8*** -2.1** -14.5*** -0.3 -0.1 0.8* 0 25 32.8 IBL 165-1-1 x Sosat C88 2 0.9 -2.3 2.6 0.2 -0.4 0.5 -218.4 -273.5 IBL 179-2-1 x Sosat C 88 2 -2.1** -11.8** -0.5 -0.1 1.2*** -0.3 605.7* 400.8* IBL 179-3-1 x Sosat C 88 1.1 0.9 -8.5* -3.2* 0.2 -0.3 0 -184.9 -102.3 IBL 206-1-1 x Sosat C 88 1 -0.1 -2.7 0.4 -0.1 0.2 -0.7* -73.9 -42.3 SE 2.2 0.7 3.7 1.5 0.1 0.3 0.3 234.8 182.5 DMI=Downy Mildew Incidence; FWT= Flowering Time; PH=Plant Height; PL=Panicle Length; PDIA=Panicle Diameter; PT=Productive tillers; TGW=1000 grain weight; PY=Panicle Yield; GY=Grain Yield; *, **, ***, Significant at 0.05 and 0.01 and 0.001 probability levels, respectively. 6.3.6. Heterosis for grain yield across locations The estimates of best parent, mid-parent and standard heterosis for grain yield are summarized in Table 6.8. The best parent heterosis for grain yield across the two locations varied from -44 to 60% and 17 hybrids displayed positive best parent heterosis. IBL 206-1-1 x Souna 3, followed by IBL 001-4-1 x Souna 3 had the largest best parent heterosis for grain yield and was among the best five hybrids while IBL 119-B-1 x Sosat C 88 had the least best parent heterosis value. The mid-parent heterosis varied from -16% for IBL 106-B-1 x Sosat C 88 to 125% for IBL 119-B-1 x Souna 3 which was not among the ten best hybrids. All the crosses displayed positive mid-parent heterosis 77 University of Ghana http://ugspace.ug.edu.gh for grain yield except IBL 106-B-1 x Sosat C 88 (-16%) and IBL 179-2-1 x Souna 3 (-12%). The standard heterosis values for grain yield across the experimental sites varied from -57% for IBL 179-2-1 x Souna 3 to 20% for IBL 206-1-1 x Souna 3. The crosses IBL 206-1-1 x Souna 3; IBL 091-1-1 x Sosat C 88; IBL 206-1-1 x Sosat C 88; IBL 001-4-1 x Souna 3 and IBL 003-B-1 x Sosat C 88 exhibited positive standard heterosis for grain yield. These hybrids were the top best five and displayed also both positive better and mid-parent heterosis values for grain yield. Table 6.8: Mean grain yield and heterosis of pearl millet hybrid across locations Cross GY (kg ha-1) BPH MPH SH IBL 001-4-1 x Souna 3 1923 52 108 14 IBL 003-B-1 x Souna 3 1646 23 26 -3 IBL 011-4-1 x Souna 3 942 -26 11 -44 IBL 021-3-1 x Souna 3 852 -33 1 -50 IBL 055-4-1 x Souna 3 1397 10 80 -18 IBL 091-1-1 x Souna 3 1211 -4 1 -29 IBL 095-4-1 x Souna 3 1641 29 48 -3 IBL 098-3-1 x Souna 3 1289 2 5 -24 IBL 106-B-1 x Souna 3 1351 7 37 -20 IBL 110-B-1 x Souna 3 1493 18 79 -12 IBL 114-6-1 x Souna 3 1181 -7 48 -30 IBL 119-B-1 x Souna 3 1530 21 125 -10 IBL 155-2-1 x Souna 3 1101 -13 27 -35 IBL 165-1-1 x Souna 3 1612 27 83 -5 IBL 179-2-1 x Souna 3 735 -42 -12 -57 IBL 179-3-1 x Souna 3 1402 11 30 -17 IBL 206-1-1 x Souna 3 2024 60 69 20 IBL 001-4-1 x Sosat C 88 1579 2 48 -7 IBL 003-B-1 x Sosat C 88 1883 22 30 11 IBL 011-4-1 x Sosat C 88 1331 -14 35 -21 IBL 021-3-1 x Sosat C 88 1654 7 68 -2 IBL 055-4-1 x Sosat C 88 1131 -27 23 -33 IBL 091-1-1 x Sosat C 88 2019 30 50 19 IBL 095-4-1 x Sosat C 88 1368 -12 10 -19 IBL 098-3-1 x Sosat C 88 1381 -11 1 -18 IBL 106-B-1 x Sosat C 88 949 -39 -16 -44 IBL 110-B-1 x Sosat C 88 1433 -7 47 -15 IBL 114-6-1 x Sosat C 88 1416 -9 51 -16 IBL 119-B-1 x Sosat C 88 871 -44 6 -49 IBL 155-2-1 x Sosat C 88 1215 -22 21 -28 IBL 165-1-1 x Sosat C 88 1114 -28 9 -34 IBL 179-2-1 x Sosat C 88 1586 2 63 -6 IBL 179-3-1 x Sosat C 88 1247 -19 2 -26 IBL 206-1-1 x Sosat C 88 1988 28 49 17 GY=Grain Yield; BPH=Best Parent Heterosis; MPH=Mid Parent Heterosis; SH=Standard Heterosis. 78 University of Ghana http://ugspace.ug.edu.gh 6.4. Discussion The significant differences observed among the genotypes for all the characters studied indicated the presence of large amount of genetic variability among the inbred lines, the OPVs and their crosses, which is a prerequisite in the establishment of a successful breeding programme. Genetic variability for downy mildew disease and several agronomic traits has been also reported in many studies conducted in West and Central Africa (Ouendeba et al., 1993; Issaka, 2012; Gemenet et al., 2014; Drabo, 2016; Pucher et al., 2016). The results indicated also the influence of the environment in the performance of the genotypes for FWT, TGW, PY and GY traits as their genotype x location interaction effect was significant. The environment effect in the performance of genotypes for flowering time was also reported in Burkina Faso (Drabo, 2016). The mean grain yield at Nioro research station was higher compared to Bambey research station. This could be explained by rainfall pattern and soil texture variability existed between the two locations where the experiments were established. Bambey research station is located in the northern part of the groundnut basin in the Soudano-Sahelian area and the soil texture is sandy while Nioro research station, located in the southern part of the groundnut basin in the Soudanian zone, has sandy-clay soil texture. However, despite the site effect on grain yield and yield related traits, some of the genotypes such as IBL 091-1-1; IBL 091-1-1 x Sosat C 88 and Thialack II have performed well under the two environments. Besides the existence of useful variability, the establishment of a successful breeding programme depends on a depth understanding of the underlying gene action of the traits of interest. Indeed, this genetic information will guide breeders on which breeding methods and lines to use for the development of improved varieties (Falconer, 1989). In this study, GCA and SCA mean squares were significant for all the traits studied except for the SCA of PDIA, PY and GY traits indicating 79 University of Ghana http://ugspace.ug.edu.gh that both additive and non-additive gene actions were important for the inheritance of these traits across the two locations. This result is contrary to the findings of Ouendeba et al. (1993) and Issaka (2012) who reported only significant GCA effects for agronomic traits such as flowering time, downy mildew incidence, plant height and panicle length. However, in the present study, the larger proportion of GCA over SCA mean squares for most of the traits such as DMI, FWT, PL, PDIA and TGW indicated the preponderance of additive gene action over non-additive gene action. This would imply that recurrent selection could be effectively used for improvement of these traits. The result of this study is consistent with that of Drabo (2016) who reported additive gene action to be more important that non-additive gene action in controlling agronomic traits such as grain yield, flowering time and panicle length. Similarly, Jethva et al. (2011) reported the importance of additive gene action over non-additive gene action in the expression of panicle length and diameter. The additive gene action was also reported for other traits in pearl millet such as Fe and Zn densities (Govindaraj et al., 2013). For grain yield, the significance of GCAf and the lack of significance for SCA suggest that grain yield is controlled by additive gene effects as reported by several authors (Ouendeba et al., 1993; Drabo, 2016). However, the slight difference of their mean squares suggests that non-additive gene action is also important in the inheritance of grain yield trait. This study has also provided information on parental effects in controlling the traits studied. The larger GCAm mean squares over GCAf mean squares for DMI, PL, PDIA and TGW displays the role of paternal effects in the control of these traits while the larger GCAf mean squares over GCAm mean squares for FWT, PH, PT and GY suggests the role of maternal effects in the control of these traits across the two locations. Similarly, Drabo (2016) found a paternal effect in controlling PDIA and a maternal effect for FWT and PH under different locations in Burkina Faso. Therefore, when it is intended to breed for highly productive pearl millet varieties adapted to the 80 University of Ghana http://ugspace.ug.edu.gh groundnut basin agro-ecological, the best performing cross for high grain yield and resistance to downy mildew disease may be produced by crossing male parents resistant to the disease with female parents having good yield potential. Inbred lines IBL 001-4-1; IBL 003-B-1; IBL 091-1-1; IBL 095-4-1; IBL 110-B-1 and IBL 206-1- 1 had positive GCA effects for grain yield indicating that these lines contributed favourable alleles for grain yield. They produced hybrids that were among the best 15 across the two locations. Thus, such lines could be used as parents to create high yielding synthetic or F1 hybrid varieties. However, IBL 001-4-1 unlike the other five inbred had positive and significant GCA effect for downy mildew and produced hybrids with a certain level of disease incidence. The other lines showed negative GCA effects and would be good sources of resistance for downy mildew under Senegalese growing conditions. In addition, they had negative GCA effects for flowering time and plant height. Thus, their cross is expected to produce a medium plant height and early maturing synthetic pearl millet varieties, tolerant to the downy mildew disease with improved grain yield. In this study, higher means were observed for hybrids compared with inbred lines and OPVs. The top five genotypes across the two locations were hybrids, showing evidence of heterosis for grain yield in pearl millet which has been also reported previously (Ouendeba et al., 1993; Yadav & Rai, 2013; Drabo, 2016). Grain yield showed a mid-parent heterosis ranging from -16% to 125% and most of the hybrids except IBL 106-B-1 x Sosat C 88 and IBL 179-2-1 x Souna 3 exceeded the parental lines. This finding is consistent with Gemenet et al., (2014) who reported mid-parent heterosis ranged from 1.9 to 98% for top-crosses evaluated under low P conditions. Information about the performance of hybrids compared to the standard check is needed for the farmer to determine the benefit of growing hybrid. In this study, a maximum standard heterosis of 20% for grain yield was observed providing advantage of growing hybrids compared to the local cultivars. 81 University of Ghana http://ugspace.ug.edu.gh Similar standard heterosis for grain yield was also reported in Burkina Faso (Drabo, 2016). The higher mean performance of the crosses compared to their parents and the control check indicate great potential for hybrid pearl millet breeding. Therefore, this technology can be a good strategy to increase national pearl millet production like in India where more than 70% of the pearl millet cultivated area are sown with F1 hybrids (Yadav & Rai, 2013). However, a strong hybrid pearl millet breeding programme need to be established. 6.5. Conclusion The present study revealed that both additive and non-additive gene action were involved in the inheritance of almost all the traits studied. However, the contribution of the additive gene action was higher than that of non-additive gene action for all the traits. Inbred lines IBL 003-B-1; IBL 091-1-1; IBL 095-4-1; IBL 110-B-1 and IBL 206-1-1 exhibited positive GCA effects for grain yield and negative GCA effects for flowering time, downy mildew disease and plant height. The crosses were generally more productive than the inbred lines and OPVs. The crosses IBL 206-1-1 x Souna 3; IBL 091-1-1 x Sosat C 88; IBL 206-1-1 x Sosat C 88; IBL 001-4-1 x Souna 3 and IBL 003-B-1 x Sosat C 88 were the top five hybrids and exhibited positive best parent, mid-parent and standard heterosis for grain yield. 82 University of Ghana http://ugspace.ug.edu.gh CHAPTER SEVEN 7.0. ASSOCIATION ANALYSIS OF DOWNY MILDEW RESISTANCE IN PEARL MILLET UNDER SENEGALESE ENVIRONMENTS 7.1. Introduction To overcome the pearl millet downy mildew disease constraint, many efforts have been made by the national programme including collaboration with international breeding programmes in order to identify new sources of resistance and develop pearl millet varieties resistant to the disease (ISRA et al., 2005; Gupta & Ndoye, 1991). However, this was done mainly through conventional breeding techniques. Nowadays, the pearl millet genome is available and the use of genomic resources in pearl millet improvement has proven to be a powerful tool and could allow rapid progress when coupled with conventional breeding (Hash et al., 2002; Varshney et al., 2017). Based on mapping bi-parental populations, several QTL for pearl millet downy mildew resistance with large effects against the pathogen population from India and several African countries have been identified over the past years (Jones et al., 1995; Hash & Witcombe, 2001; Breese et al., 2002; Jones et al., 2002; Gulia et al., 2004; Supriya et al., 2011). These identified QTL were located on the seven pearl millet linkage groups and the ones against the pathogen population from Senegal were identified on Linkage Group (LG) 2, LG 6 and LG 7. Some of these QTL were successfully transferred to the background of elite parents to improve their resistance to the pearl millet downy mildew through marker assisted backcrossing process. Indeed, a classic and successful example was the introgression through marker-assisted backcrossing selection of pearl millet downy mildew resistance into the background of the male parent of the popular pearl millet hybrid HHB 67 (Hash et al., 2006). 83 University of Ghana http://ugspace.ug.edu.gh However, the resolution of identified QTL through linkage mapping is limited due to the relatively few recombinants generated from the two parents used to develop mapping population. This resolution can be improved now by the use of Genome Wide Association Study (GWAS) approach. The main advantage of this approach is that it exploits phenotypic and genetic variations present across a natural population with a high resolution and then draws inferences on the basis of past recombination events. In addition, this technique is cost effective and does not require the development of a mapping population (McCarthy et al., 2008; Korte & Farlow, 2013). In recent years, GWAS has been applied in many crops such as maize, rice, sorghum, cassava and foxtail millet to identify markers associated with a range of traits of interest. In pearl millet, there are few reports of the use of GWAS to map traits of interest such as resistance to downy mildew (Drabo, 2016), tolerance to low phosphorus (Gemenet et al., 2014), tolerance to drought (Sehgal et al., 2015) and high grain iron and zinc content (Satyavathi & Srivastava, 2017). For resistance to pearl millet downy mildew, Drabo (2016) has detected eight SSR markers associated with resistance to three different pathotypes from Burkina Faso. Nevertheless, to the best of our knowledge, there is no study on identification of favourable alleles mapping for resistance to downy mildew in pearl millet through association mapping using high resolution SNPs. Nowadays, with the pearl millet genome sequenced and published (Debieu et al., 2017; Varshney et al., 2017), it is possible to use genotyping-by-sequencing (GBS) approach (Elshire et al., 2011) and GWAS to precisely identify SNPs associated with traits of interest. In this work, an association analysis in pearl millet inbred lines from West and Central Africa based on GBS data was conducted. Specially, this study was initiated to identify genomic regions associated with natural variations for downy mildew in pearl millet as basis for breeding varieties resistant to downy mildew under Senegal environments. 84 University of Ghana http://ugspace.ug.edu.gh 7.2. Material and methods 7.2.1. Plant materials The set of lines assessed for downy mildew (chapter 4) during the rainy season of 2016 at Bambey and Nioro research stations was used for this study. It consisted of 101 entries and included an array of 99 inbred lines generated from landraces of diverse genetic backgrounds collected in West and Central Africa (Haussmann et al., 2006). 7.2.2. Phenotypic data Downy mildew assessment for the 99 association panel along with SOSAT C-88 and 7042 S as tolerant and susceptible checks, respectively was conducted during the rainy season of 2016 between August and October at two locations in Senegal: Bambey and Nioro. These two sites have been identified as hotspots for downy mildew. At each site, the trial was laid out in an augmented design with nine tested lines along with the two checks in a block as described in chapter 4. The downy mildew incidence recorded at 30 DAS was used as phenotypic data for this study. 7.2.3. Genotypic data Genomic DNA on individual pearl millet line was extracted from fresh leaves sampled on young plants as described in Mariac et al. (2006). Leaves were frozen in liquid N2 just after sampling and ground with Tissue Lyser II (Qiagen). Ground leaves (approximately 0.2 g) were re-suspended in 1 mL extraction buffer (Tris (100 mM), NaCl (1.4 M), EDTA (20 mM), CTAB (2%) and DTT (0.2%), pH = 8) and incubated at 65 oC for 3 hours. Samples were centrifuged 1 minute at 8 000 rpm and supernatants were washed (1:1) with phenol-chloroform-isoamyl alcohol (25:24:1). Samples were centrifuged for 10 minutes at 10000 rpm at 10 oC and supernatants were retrieved and incubated with 10 µl RNase A (Qiagen) for 30 minutes at 37 oC. A second washing step with phenol-chloroform-isoamyl alcohol was then applied. DNA was precipitated with isopropanol (0.6 85 University of Ghana http://ugspace.ug.edu.gh volume) and sodium acetate (3 M, pH=5, 1/10) at 4 oC for 20 minutes. Samples were centrifuged 10 minutes at 10 000 rpm at 10 oC and supernatants were removed. DNA pellet were washed two times with ethanol (70%) and centrifuged 5 minutes at 10 000 rpm at 10 oC. DNA pellets were re- suspended in 100 µl TE buffer and stored at -20 oC. DNA concentration and purity were checked with Spectrophotometer ND-1000 (Nanodrop). DNA quality was assessed on a 1% agarose gel. An aliquot of 10% of the samples was digested with EcoRI or HindIII (1 unit/reaction; Promega) for 2 hours at 37 oC to check DNA accessibility to restriction. Genotyping was performed at the Genetic Diversity Facility of Cornell University (Ithaca, USA) using the GBS approach as described by Elshire et al. (2011). For each pearl millet line, two runs were performed using the HiSeq2500 (Illumina). ApeK1 restriction enzyme was used to cut the genomic DNA and sequencing was done with single-end-reads. The Genetic Diversity Facility provided the fastq file. This raw sequence data was de-multiplexed using barcode to determine which reads belong to which samples and then filtered. These filtered sequence reads were aligned to the pearl millet genome (Varshney et al., 2017) using the Burrows-Wheeler Alignment tool (BWA) and SNPs were called with the TASSEL 5.2 pipeline. To ensure high quality SNPs for subsequent analyses, markers that may cause false associations in GWAS were removed using vcftools software. Furthermore, inbred lines which showed missing data and/or > 50% of heretozygosity and the SNPs which have missing data were removed. The genotypic data were further filtered for minor allele frequency (MAF) less than 5% to remove uninformative sites. Finally, 77 genotypes were considered and the filtered dataset represented a reduced marker set with 18,591 SNPs with high quality. These SNPs were used for estimating population structure and subsequent statistical analysis. 86 University of Ghana http://ugspace.ug.edu.gh 7.2.4. Analysis of data The analysis of variance (ANOVA) on the DMI phenotypic data for each location was performed using the SAS macro for analysis of data from augmented block designs. Based on the general ANOVA as described by You et al. (2016), the variance components were derived (Table 7.1) Table 7.1: Analysis of variance for augmented design Source df MS Expected mean square Genotype g - 1 MSG σ2e+ σ2G Check c-1 MS σ2C e+ rσ2C Error (c-1) (r-1) MSE σ2e Where: ~ is the genetic variance; ~9 is the error variance; : is the number of replications for checks, g the number of tested genotypes and c number of checks. Broad sense heritability for DMI based on the ANOVA result was calculated using the equation formulated by Allard (1960) which is given as follows: Ä ~ Ä w = ~Ä + ~Ä9/: To identify the best model for genetic association analysis, the population structure of the pearl millet lines was studied. This population structure was assessed using a discriminant principal component analysis (DAPC) method implemented in the R package “adegenet” (Jombart, 2008). The GWAS was performed using the compressed mixed linear model (MLM) (Zhang et al., 2010) implemented in GAPIT (Lipka et al., 2012). To identify significantly associated SNPs, the 87 University of Ghana http://ugspace.ug.edu.gh Bonferroni correction test for false positives at 5% was found to be too stringent for this study. Therefore, a less stringent approach to determine the threshold level to declare a marker to be significantly associated to downy mildew was considered, since population structure effects and the SNPs, which may cause false associations in GWAS were previously removed. In this approach, which was proposed and applied in Arabidopsis (Chan et al., 2010) and spring barley (Pasam et al., 2012), the bottom 0.1 percentile distribution of the p values is considered as significant. In this study, this approach resulted in a threshold levels of –log (p values) > 3. 7.3. Results 7.3.1. Phenotypic variation Results from the analysis of variance and the calculated broad sense heritability (H2) for DMI at Bambey and Nioro at 30 das are presented in Table 7.2. Analysis of variance revealed highly significant effects (P < 0.01) among pearl millet inbred lines, the two controls and their interactions for the disease incidence at both locations. DM is a highly heritable phenotype, with broad sense heritability of 0.9 both at Bambey and Nioro research stations. Table 7.2: Mean square and broad sense heritability from ANOVA for DMI traits under artificial downy mildew infestation at Bambey and Nioro during the rainy season 2016 DMI Source DF Bambey Nioro Tested lines 98 546.3*** 605.5*** Controls 1 45818.9*** 24689.5*** Test vs Controls 2 16320.2*** 14589.4*** Error 10 15.8 21.8 Heritability 0.9 0.9 DF= Degree of freedom; DMI = Downy mildew incidence; ***- Significant at 0.1% 88 University of Ghana http://ugspace.ug.edu.gh 7.3.2. Marker distribution on the linkage group, linkage disequilibrium and population structure The number of SNPs resulting from the cleaning pipeline and imputation was composed of 18, 591 SNPs covering all the seven Linkage Groups (LG) of the pearl millet genome, with an average density of one SNP per 84 kbp. The average number of SNPs per LG was 2, 656. The LG2 had the highest number of SNPs (3, 379) while the LG4 encompassed the lowest number (1, 895). Based on the genotypic data, a rapid linkage disequilibrium (LD) decay was observed in the pearl millet lines. Indeed, the R2 estimate of LD ranged between 0 and 1 and declined below 0.2 at a distance of about 3 kb. Figure 7-1: Linkage disequilibrium (LD) decay plot of the pearl millet genotypes The DAPC method revealed a clear genetic separation among the 77 pearl millet genotypes and divided them into three sup-populations (Figure 7-2). Cluster 1 (Q1) encompassed only four genotypes while Cluster 2 (Q2) and 3 (Q3) had 38 and 35 genotypes, respectively. 89 University of Ghana http://ugspace.ug.edu.gh Figure 7-2: Estimated population structure of 77 pearl millet genotypes It appeared that these lines have been grouped mainly based on panicle length rather than their origin. Indeed, Q2 was positively correlated with panicle length (r = 0.47), and Q3 was negatively correlated with panicle length (r = 0.40), while Q1 did not show significant correlation with the panicle length. Table 7.3: Pearson correlation coefficients between observed variables and membership probability Q1 Q2 Q3 DMI -0.10 0.09 -0.04 Severity -0.05 -0.06 0.09 Flowering -0.25 * -0.05 0.16 Plant Height -0.14 0.25 * -0.18 *** ** Panicle length -0.17 0.47 -0.40 Productive tillers 0.02 -0.05 0.04 DMI = Downy mildew incidence 7.3.3. Association mapping Association mapping analysis was conducted with the 77 genotypes, which had both phenotypic and genotypic data using the compressed mixed linear model implemented in GAPIT. The 90 University of Ghana http://ugspace.ug.edu.gh threshold p values 10-3 was considered as a threshold to declare significant SNPs associated to the downy mildew disease. Normal distributions were observed for DMI phenotypic values and the model used showed almost a perfect fit between the observed and the expected p-values, expect for few values both at Bambey and Nioro research stations (Figure 7-3). Figure 7-3: Manhattan plot, phenotypic distribution and quantile-quantile plots for downy mildew at Bambey (a) and Nioro (b) research stations A total of 16 markers were found to be significantly associated with downy mildew at both locations (Table 7.4), each individually explaining about 14-20% of the observed variation for downy mildew resistance. Four of these markers (S6_43687524, S6_43689389, S6_80112161 and S6_80112341) located on LG6 were found to be significantly associated with downy mildew in both locations, while the remaining 10 were found to be significantly associated with downy mildew in a single environment. At the Bambey research station, the MLM analysis revealed 13 SNPs that were significantly associated with resistance to downy mildew. These candidates SNPs covered LG 1, LG 3 and LG 91 University of Ghana http://ugspace.ug.edu.gh 6 and explained 14 to 19% of the phenotypic variation for resistance to S. graminicola population from Bambey. The most significant SNP (S6_80112161) increased downy mildew resistance up to 20.5%. At the Nioro research station, a total of seven markers, located on LG3, LG4 and LG6 were found to be significantly associated with resistance to pearl millet downy mildew and the most significant SNPs were detected on LG6 (Table 7.4). These SNPs explained 15 to 20% of the phenotypic variance for resistance in the Nioro S. graminicola population. The most significant marker at Bambey (S6_80112161) was found to be the most significant at Nioro also with the same allelic effect at both location and increased resistant against S. graminicola populations from Nioro up to 22.5%. Table 7.4: Markers significantly associated with downy mildew resistance at Bambey and Nioro research stations Allele Location SNP Chr Position P Value MAF PVE Allelic Major Minor Effect S1_174774804 1 1.75E+08 9.90E-04 0.27 T C 0.14 -13.13 S1_266879682 1 2.67E+08 3.10E-04 0.48 A G 0.17 13.13 S1_267555181 1 2.68E+08 2.20E-04 0.27 A G 0.18 13.85 S2_21493058 2 21493058 5.70E-04 0.38 G T 0.16 12.62 S2_21624130 2 21624130 5.70E-04 0.29 T G 0.16 13.78 S3_292483735 3 2.92E+08 5.80E-04 0.26 G A 0.16 13.79 Bambey S3_292483759 3 2.92E+08 7.60E-04 0.25 G C 0.15 13.56 S6_43687524 6 43687524 8.00E-04 0.25 G A 0.15 -13.76 S6_43689389 6 43689389 8.00E-04 0.25 A C 0.15 13.76 S6_80112161 6 80112161 1.80E-04 0.10 T C 0.15 20.50 S6_80112174 6 80112174 8.70E-04 0.09 T C 0.19 19.40 S6_80112341 6 80112341 7.10E-04 0.13 G A 0.15 16.22 S6_214422694 6 2.14E+08 6.50E-04 0.42 C G 0.15 -12.06 S3_293109187 3 2.93E+08 7.21E-04 0.42 T G 0.15 -11.99 S4_169567419 4 1.70E+08 3.93E-04 0.23 A G 0.17 -13.93 S4_169567480 4 1.70E+08 3.93E-04 0.23 C T 0.17 -13.93 Nioro S6_43687524 6 43687524 1.69E-04 0.25 G A 0.19 -16.64 S6_43689389 6 43689389 1.69E-04 0.25 A C 0.19 16.64 S6_80112161 6 80112161 1.33E-04 0.10 T C 0.20 22.15 S6_80112341 6 80112341 6.82E-04 0.13 G A 0.15 17.23 PVE= Phenotypic Variation Explained. In bold are the consistent markers observed across the two locations 92 University of Ghana http://ugspace.ug.edu.gh 7.4. Discussion The population size used in the study was small compared to other used in several pearl millet association studies (Gemenet et al., 2015; Sehgal et al., 2015; Satyavathi & Srivastava, 2017) and may have influenced the power of detecting significant associations (Zondervan & Cardon, 2004). In fact, the higher the population size is, the higher the power of detecting QTL is. This is because effects due to rare alleles cannot be mapped easily in small population size. Therefore, increasing population size, increase the frequency of rare alleles and consequently allow a better QTL detection. However, the numbers of genotypes used in the present study was larger than the 60 rice cultivars used for genome wide association mapping of resistance to blast disease (Shinada et al., 2015) and was comparable with the 81 spring wheat genotypes used in mapping resistance to four major wheat diseases (Perez-Lara et al., 2017). Hence, present size of the panel used for this study can be considered as adequate for association studies. Furthermore, with the high broad sense heritabilty noted at both locations, much of the observed variation in the downy mildew reaction phenotype can be attributed to genetic variation. Therefore, QTL with major effects can be easily detected, even in a small population size. In association mapping, presence or absence of population structure is another important aspect, which has to be considered. Indeed, it may result in false positives when it is not adequately controlled for in a GWAS (Zhu et al., 2008). For example, Drabo (2016) using a general linear model approach discovered eight markers associated to pearl millet downy resistance and this number decreased to four when he applied the mixed linear model approach. In the current study, the pearl millet lines were subdivided into three sub-populations. It appeared that this classification was mainly based on the panicle length. This finding is consistent with previous results where structuration of pearl millet lines from West and Central Africa was not based on neither country 93 University of Ghana http://ugspace.ug.edu.gh of origin nor agro-ecological zone (Tostain & Marchais, 1989; Oumar et al., 2008; Stich et al., 2010; Bashir et al., 2015; Hu et al., 2015) but based on agronomic traits such as flowering time. Moreover, Gemenet et al. (2015) in an association study using a set of pearl millet lines from West and Central Africa, including these lines, observed three sub-populations mainly based on flowering time. Phenotypic-genotypic association study requires also abundant markers in the study population to achieve high resolution and this marker density is determined by the extent of the LD within the population. In this current study, LD decay rate observed in the out-crossing pearl millet crop is lower compared to those of self pollinated rice (~ 100 kb) (Huang & Han, 2014) and sorghum (~ 150 kb) (Morris et al., 2012) crops but similar to that of the outcrossing maize crop (~ 2 kb) (Li et al., 2013). This rapid LD decay in pearl millet was also noted by Hu et al. (2015) in Senegalese pearl millet genotypes. Therefore, this rapid LD decay combined to the weak population observed in the panel, make it a good population for genome wide association study with high density markers. However, by employing a moderately high marker density (84 SNPs/kb) in the current study, significant SNPs associated to downy mildew were identified. Given this existing population structure observed among the pearl millet genotypes, as well as the kindship among these lines, GWAS using MLM approach which takes into account population struture and kindship (Yu et al., 2006) was carried out. Based on this model, no significant deviation between the observed and expected p values for the downy mildew incidence trait was observed. The analysis revealed a total of 14 SNPs that were significantly associated with downy mildew resistance at a treshold of –log (p values) > 3 in the two locations. A lowest treshold value (-log(p values) > 2) have been considered by Gemenet et al. (2015) and they identified several markers associated with phosphorus efficiency-related traits in pearl millet. 94 University of Ghana http://ugspace.ug.edu.gh The 16 identified markers are located on four out of the seven linkage groups (LG1, 3, 4 and 6). This finding is consistent with previous results where several QTL involved in the resistance againts different pathotypes of S. graminocola located on these linkage groups have been reported (Jones et al., 1995; Jones et al., 2002; Breese et al., 2002; Gulia et al., 2004; Drabo, 2016). In this study, four of these markers located on the same genomic region (LG6) have been constantly significantly associated with downy mildew resistance across the two locations with the same allelic effects at both location. These four consistent markers explained 15 to 20% of downy mildew incidence variation. These markers could be considered as the most reliable markers associated to downy mildew resistance and may be grouped into two QTL located on the same linkage group. Indeed, physical position between S6_43687524 and S6_43689389 and between S6_80112161 and S6_80112341 is less than 2 kb while distance between S6_43687524 and S6_80112341 is more than 36 Mbp. Interestingly, Jones et al. (1995) detected also a QTL against pathogen population from Senegal on the same linkage group which explained 12% of the downy mildew phenotypic variation. Therefore, these four SNPs markers may be located on the same position that the RFLP marker M713 identify previously. Drabo (2016) also reported two significant markers against two S. gramincola isolates from Burkina Faso located on LG6 using genome wide association study. However, none of these 14 identified markers were located on the LG2 and LG7 where QTL against the pathogen population from Senegal were identified previously (Jones et al., 1995). Thus, the identified markers located on LG1, LG3 and LG4 could be considered as new markers associated to downy mildew resistance under Senegalese environments. 95 University of Ghana http://ugspace.ug.edu.gh 7.5. Conclusion The first results on marker-trait associations for pearl millet downy mildew resistance using high resolution SNPs markers were reported here. Based on the mixed linear model, 16 SNPs significantly associated to downy mildew resistance under Senegalease environments and located on LG1, LG3, LG4 and LG6 were identified. Four of these SNPs, located on LG6, were found to be consistent over the two locations and explained 15 to 20% of the downy mildew resistance phenotypic variation. These four markers could be considered as the most reliable markers associated to downy mildew resistance and are positioned on the same LG that were identified previously using RFLP markers. The SNPs located on LG1, LG3 and LG4 were identified for the first time under Senegalease environments and could be considered as new markers. With the pearl millet genome recently sequenced and published, the integration of GWAS and biparental linkage mapping approach and an ever increasing number of SNPs on these identified genomic regions will faciliate the identification of important genes associated to downy mildew resistance of pearl millet. 96 University of Ghana http://ugspace.ug.edu.gh CHAPTER EIGHT 8.0. GENERAL CONCLUSION AND RECOMMENDATIONS 8.1. General conclusion - In the groundnut agro-ecological zone of Senegal, the occurrence of the pearl millet downy mildew disease was noted by the farmers and almost all the interviewed farmers recognized the disease symptoms, particularly the panicle transformation. However, they were not aware of the damages that the disease caused and did not rank it as the main production constraints. Based on their knowledge, they considered Striga and low soil fertility as the main pearl millet production constraint. When selecting a pearl millet variety to grow in their fields, farmers considered high grain yield potential as the most important trait. - A spatio-temporal pathogenic variation within the main Senegalese pearl millet production areas was noted. Indeed, the downy mildew of the differential resistance lines changed over the locations and years. The pearl millet downy mildew incidence was higher in Kolda, Bambey and Nioro compared to Sinthiou Maléme. - A large phenotypic variation for downy mildew disease parameters and other agronomic traits was observed among the inbred lines derived from landraces originated to several West and Central African countries. Some of these lines such as IBL 055-4-1; IBL 095-4-1; IBL 098-1-1; IBL 106-B-1; IBL 111-3-1; IBL 160-1-1; IBL 161-1-1; IBL 186-1-1 were resistant to the disease. The lines were grouped into 3 clusters with the downy mildew disease parameters and plant height as the most discriminated factors. - The assessment of 34 crosses and their parents revealed that the performances of the crosses were generally higher than those of the inbred lines and OPVs. For instance, the crosses IBL 206-1-1 x Souna 3; IBL 091-1-1 x Sosat C 88; IBL 206-1-1 x Sosat C 88; IBL 001-4-1 x 97 University of Ghana http://ugspace.ug.edu.gh Souna 3 and IBL 003-B-1 x Sosat C 88 were the top best five genotypes. It has been also showed that both additive and non additive gene action were involved in the inheritance of almost all the studied traits. However, the additive gene action was more important than the non-additive gene action for all the traits. Inbred lines IBL 003-B-1; IBL 091-1-1; IBL 095-4-1; IBL 110-B-1 and IBL 206-1-1 showed positive GCA effects for grain yield and negative GCA effects for flowering time, downy mildew disease and plant height. - Sixteen SNPs located on LG1, LG3, LG4 and LG6 were identified as significantly associated to downy mildew resistance at Bambey and Nioro research stations. The ones located on LG6 (S6_43687524, S6_43689389, S6_80112161, S6_80112341) could be considered as the most reliable markers because they were consistent over the two locations and are located on the same linkage group that a previous QTL identified under Senegalese environment. Furthermore, these SNPs explained a large downy mildew incidence phenotypic variation (15 to 20%). The SNPs markers located on LG1, LG3 and LG4 were identified for the first time under Senegalease conditions and, therefore, could be considered as new markers. 8.2. Recommendations - Despite the fact that farmers recognized the disease symptoms, farmers’ awareness about damages caused by the disease should be created through training sessions. - To sustain pearl millet production in Senegal, particularly in the groundnut basin agro- ecological zone, the national pearl millet breeding programme should consider these identified principal farmers’ preferred traits and constraints when developing new pearl millet variety. - Breeding materials should be screened for downy mildew reaction under field conditions at Bambey, Nioro and/or Kolda research stations. However, this screening may be more accurate when conducted under greenhouse condition. 98 University of Ghana http://ugspace.ug.edu.gh - The tremendous phenotypic variation observed in the inbred lines should be exploited when improving lines for different traits. Furthermore, contrasting lines can be used to develop mapping populations for genetic study. - Since inheritance of traits like high yield, downy mildew resistance and earliness are under additive gene action, recurrent selection should be used for the improvement of these traits. - The resistant inbred lines IBL 003-B-1; IBL 091-1-1; IBL 095-4-1; IBL 110-B-1 and IBL 206-1-1 should be used as parents by the pearl millet improvement programme in order to create early high yielding synthetic or hybrid varieties resistance to the disease or improve the farmers’ preferred pearl millet varieties. - Genes associated with downy mildew resistance of pearl millet located around the significant SNPs should be identified since the pearl millet genome has been published. The identification of these genes identification should be done through the integration of GWAS and biparental linkage mapping approcahs. 99 University of Ghana http://ugspace.ug.edu.gh REFERENCES Abakemal, D., Hussein, S., Derera, J., & Laing, M. (2013). Farmers’ perceptions of maize production systems and breeding priorities, and their implications for the adoption of new varieties in selected areas of the highland agro-ecology of Ethiopia. Journal of Agricultural Science, 5(11), 159–171. Anandajayasekeram, P., Martella, D.R., Sanders, J., & Kupfuma, B. (2007). Impact of pearl millet research and development in Namibia: the case of Okashana 1. In Anandajayasekeram, P., Rukuni, M., Babu, S. Liebenberg F., & Keswani, C.L. (eds.) Impact of science on African agriculture and food Security. CABI Publishing, Oxfordshire, UK, pp. 68–73. Agrios, G. N. (2005). Plant pathology, Fifth Edition. (A. press Elsevier, Ed.) (5th ed., Vol. 40). Dana Dreibelbis. Aliyu, B., & Hati, S. (2011). Comparative assessment of metalaxyl enhanced protection of pearl millet varieties in the control of downy mildew. Journal of Cereals and Oilseeds, 2(2), 26– 31. Allard, R. (1960). Principles of Plant Breeding (John Wiley). New York. Allouis, S., Qi, X., Lindup, S., Gale, M. D., & Devos, K. M. (2001). Construction of a BAC library of pearl millet, Pennisetum glaucum. Theoretical and Applied Genetics, 102, 1200–1205. Anaso, A. B., & Anaso, C. E. (2010). Pearl millet downy mildew as affected by seed dressing and cultivar in the sahel and sudan savanna of Nigeria. Archives of phytopathology and plant protection, 43(4), 404–411. Andrews, D. J., King, S. B., Witcome, J. R., Singh, S. D., Rai, K. N., Thakur, R. P., … Singh, P. (1985). Breeding for disease resistance and yield in pearl millet. Pearl millet, 11(0), 241–258. Angarawai, I., & Kadams, A. M. (2008). Quantitative nature of downy mildew resistance in Nigerian elite pearl millet lines. Journal of SAT Agricultural Research, 6, 1-3. 100 University of Ghana http://ugspace.ug.edu.gh Angarawai, I., Kadams, A. M., & Bello, D. (2008). Gene Effects Controlling Heritability of Downy Mildew Resistance in Nigerian Elite Pearl Millet Lines. World Journal of Agricultural Sciences, 4(5), 545-549. ANSD, agence nationale de la statistique et de la demographie. (2014). Recensement Général de la Population et de l’Habitat , de l’Agriculture et de l’Elevage. Asante, M. D., Asante, B. O., Acheampong, G. K., Offei, K., Gracen, V., Adu-dapaah, H., & Danquah, E. Y. (2013). Farmer and consumer preferences for rice in the Ashanti region of Ghana : Implications for rice breeding in West Africa. Journal of Plant Breeding and Crop Science, 5(12), 229–238. Badiane, M. (1999). Gestion integree des ennemis du mil dans des champs paysans a Bambey serere: Test des methodes de controle du mildiou. Ecole Nationale Superieure d’Agriculture, Thies/Senegal. Ball, S. L. (1983). Pathogenic variability of downy mildew ( Sclerospora graminicola ) on pearl millet: Statistical techniques for analysis of data. Annals of Applied Biology, 102, 257–264. Ball, S. L., Pike, D. J., & Burridge, C. Y. (1986). Characterisation of populations of Sclerospora graminicola. Annals of Applied Biology, 108, 519–526. Bashir, E. M. A., Ali, A. M., Ali, A. M., Mohamed, E. T. I., Melchinger, A. E., Parzies, H. K., & Haussmann, B. I. G. (2015). Genetic diversity of Sudanese pearl millet (Pennisetum glaucum (L.) R. Br.) landraces as revealed by SSR markers, and relationship between genetic and agro- morphological diversity. Genetic Resources and Crop Evolution, 62(4), 579–591. Bationo, A., & Mokwunye, A. U. (1991). Alleviating soil, water and nutrients management constraints to increased crop production in West Africa. Fertilizer Research, 29, 95–115. Bhat, S. S. (1973). Investigations on the biology and control of Sclerospora graminicola on Bajra. 101 University of Ghana http://ugspace.ug.edu.gh University of Mysore, Mysore, India. Bidinger, F. R. (1998). Farmer participation in pearl millet research in Namibia. In Participatory Plant Improvement (pp. 21–30). Chenna, India. Breese, W. A., Hash, C. T., Devos, K. M., & Howarth, C. J. (2002). Pearl millet genomics: An overview with respect to breeding for resistance to downy mildew. Sorghum and Millets Diseases, 243–246. Brunken, J.N. (1977). A systematic study of Pennisetum (Gramineae). American Journal of Botany, 64(2), 161-176. Butler, E. J. (1907). Some diseases of cereals caused by Sclerospora graminicola. Memoirs of the Departement of Agricultural in India, Botanical Series, 2(1), 1-24. Camara, Y., Bantilan, M. C. S., & Ndjeunga, J. (2006). Impacts of sorghum and millet research in West and Central Africa (WCA): a synthesis and lessons learnt. ICRISAT. Ceccarelli, S., Grando, S., & Baum, M. (2007). Participatory plant breeding in water-limited environments. Experimental Agriculture, 43(4), 411–436. Chan, E. K., Rowe, H. C., & Kliebenstein, D. J. (2010). Understanding the evolution of defense metabolites in Arabidopsis thaliana using genome-wide association mapping. Genetics, 185(3), 991–1007. Chaudhary, V. P., Dhedhi, K. K., Joshi, H. J., & Mehta, D. R. (2012). Combining ability studies in line × tester crosses of pearl millet [Pennisetum glaucum (L.) R. Br.]. Research on Crops, 13(3), 1094-1097. Christinck, B. A., Diarra, M., & Horneber, G. (2014). Innovations in Seed Systems—lessons from the CCRP funded project Sustaining farmer-managed seed initiatives in Mali, Niger and Burkina Faso. 102 University of Ghana http://ugspace.ug.edu.gh Collard, B. C. Y., & Mackill, D. J. (2008). Marker-assisted selection: an approach for precision plant breeding in the twenty-first century. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 363(1491), 557–572. Cox, D. J., & Frey, K. J. (1984). Combining ability and the selection of parents for interspecific oat matings Crop. Crop Science, 24(5), 963–967. DAPSA. (2016). Resultats definitifs de la campagne agricole 2013/2014. Ministre de l’Agriculture et de l’equipement rural, Senegal. Debieu, M., Kanfany, G., & Laplaze, L. (2017). Pearl millet genome: Lessons from a tough crop. Trends in Plant Science, 22(11), 911-913. Deepak, S., Shailasree, S., Kini, R. K., Hause, B., Shetty, S. H., & Mithöfer, A. (2007). Role of hydroxyproline-rich glycoproteins in resistance of pearl millet against downy mildew pathogen Sclerospora graminicola. Planta, 226(2), 323–333. Deswal D.P., Govila O.P. (1994) Genetics of resistance to downy mildew (Sclerospora graminicola) in pearl millet (Pennisetum glaucum). Indian Journal of Agricultural Science, 64, 661–663. Drabo, I. (2016). Breeding pearl millet (Pennisetum glaucum (L) R. BR.) for downy mildew resistance and improved yield in Burkina Faso. University of Ghana. Elshire, R. J., Glaubitz, J. C., Sun, Q., Poland, J. A., Kawamoto, K., Buckler, E. S., & Mitchell, S. E. (2011). A robust, simple genotyping by sequencing (GBS) approach for high diversity species. PLos One, 6(5), 1–10. Esuma, W., Herselman, L., Tine, M., Ramu, P., Lu, F., Baguma, Y., & Buckler, E. S. (2016). Genome-wide association mapping of provitamin A carotenoid content in cassava. Euphytica, 212 (1), 97-110. 103 University of Ghana http://ugspace.ug.edu.gh Falconer, D. S. (1989). Introduction to Quantitative Genetics. London. UK.: Oliver and Boyd,. FAO. (2015). FAO Database. Retrieved May 26, 2015, from http://faostat.fao.org Gamero, A., Buser, A., & Kappeler, P. M. (2013). Characterization of 10 polymorphic microsatellite loci for White-breasted mesites (Mesitornis variegata). Conservation Genetics Resources, 5(2), 553–554. Geetha, N. P., Amruthesh, K. N., Sharathchandra, R. G., & Shetty, H. S. (2005). Resistance to downy mildew in pearl millet is associated with increased phenylalanine ammonia lyase activity. Functional Plant Biology, 32(3), 267–275. Gemenet, D. C., Hash, C. T., Sy, O., Zangre, R. G., Sanogo, M. D., Leiser, W. L., … Haussmann, B. I. G. (2014). Pearl millet inbred and testcross performance under low phosphorus in West Africa. Crop Science, 54(6), 2574–2585. Gemenet, D. C., Leiser, W. L., Zangre, R. G., Angarawai, I. I., Sanogo, M. D., Sy, O., … Haussmann, B. I. G. (2015). Association analysis of low-phosphorus tolerance in West African pearl millet using DArT markers. Molecular Breeding, 35(8), 1–20. Girard, J. C., & Delassus, M. (1978). Les maladies parasitaires. Centre Nationale de Recherches Agronomiques / Bambey, Senegal. Govindaraj, M., Rai, K. N., Shanmugasundaram, P., Dwivedi, S. L., Sahrawat, K. L., Muthaiah, A. R., & Rao, A. S. (2013). Combining ability and heterosis for grain iron and zinc densities in pearl millet. Crop Science, 53(2), 507–517. Gulia, S. K., Hash, C. T., Thakur, R. P., Breese, W. A., & Sangwan, R. S. (2004). Mapping new QTL for improvement of downy mildiew resistance in pearl millet. In Crop Production in Stress Environments: Genetic and Management Options (Agrobios I, pp. 373–386). Jodhpur, Rajasthan, India. 104 University of Ghana http://ugspace.ug.edu.gh Gupta, G. K., & Singh, D. (1996). Studies on the influence of downy mildew infection on yield and yield contributing plant characters of pearl millet in India 42: 89–93. International Journal of Pest Management, 42, 89–93. Gupta, S. C. (1986). The pearl millet improvement programme in Senegal. Bambey. Gupta, S. C., & Ndoye, A. T. (1991). Yield stability analysis of promising pearl millet genotypes in Senegal. Maydica, 36, 83.86. Gwary, D., Garba, I., & Gwary, S. (2007). Evidence of pathogenic variation in Sclerospora graminicola populations from pearl millet regions of Nigeria. International Journal of Agricultural and Biology, 9(1), 94–97. Hallauer, A. R., Carena, M. J., & Miranda-Filho, J. B. (2010). Quantitative genetics in maize breeding (Springer S). New York. Hash, C. T., & Bramel-Cox, P. J. (1999). Marker applications in pearl millet. Application of molecular markers in plant breeding.Training manual for a seminar held at IITA, Ibadan, Nigeria, from 16-17 August 1999. International. Ibadan, Nigeria. Hash, C. T., Schafferfl, R. E., & Peecock, J. M. (2002). Prospects for using conventional techniques and molecular biological tools to enhance performance of “orphan” crop plants on soils low in available phosphorus. Plant and Soil, 245, 135–146. Hash, C. T., Sharma, A., Kolesnikova-Allen, M. A., Singh, S. D., Thakur, R. P., Raj, A. B., … Sagar, P. (2006). Teamwork delivers biotechnology products to Indian small-holder crop- livestock producers: Pearl millet hybrid “HHB 67 Improved” enters seed delivery pipeline. Journal of SAT Agricultural Research, 2(1), 1–3. Hash, C. T., Singh, S. D., Thakur, R. P., & Talukdar, B. S. (1999). Pearl millet breeding. In Pearl millet breeding. New Delhi, India: Oxford & IBH Publishing Co. Pvt. Ltd. 105 University of Ghana http://ugspace.ug.edu.gh Hash, C. T., Witcome, J. R., Thakur, R. P., Bhatnagar, S. K., Singh, S. D., & Wilson, J. P. (1996). Breeding for pearl millet disease resistance. In International Conference on Genetic Improvement of Sorghum and Pearl Millet (pp. 337–372). Hash, C., & Witcombe, J. (2001). Pearl millet molecular marker research. International Sorghum and Millets Newsletter, 1(1), 1–8. Haussmann, B. I. G., Hess, D. E., Welz, H., & Geiger, H. H. (2000). Improved methodologies for breeding Striga-resistant sorghums. Field Crops Research, 66(3), 195–211. Haussmann, B. I. G., Boubacar, A., Boureima, S. S., & Vigouroux, Y. (2006). Multiplication and preliminary characterization of west and central african pearl millet landraces. International Sorghum and Millets Newsletter, 47, 110–112. Howarth, C. J., & Yadav, R. S. (2002). Successful marker assisted selection for drought tolerance and disease resistance in pearl millet. IGER Innovations, 6, 18–21. Hu, Z., Mbacké, B., Ramasamy, P., Gueye, M. C., Sy, O., Bouchet, S., … Morris, G. (2015). Population genomics of pearl millet (Pennisetum glaucum (L.) R. Br.): Comparative analysis of global accessions and Senegalese landraces. BMC Genomics, 16, 1–12. Huang, X., & Han, B. (2014). Natural variations and genome-wide association studies in crop plants. Annual Review of Plant Biology, 65(1), 531–551. ISRA, ANCAR, & RESOPP. (2013). Diffusion à grande échelle de nouvelles variétés de mil et sorgho dans le Bassin Arachidier. ISRA, ITA, & CIRAD. (2005). Bilan de la recherche agricole et agroalimentaire au Sénégal. Institut sénégalais de recherches agricole. Issaka, A. (2012). Development of Downy Mildew Resistant F1 pearl millet Hybrids in Niger. University of Ghana. 106 University of Ghana http://ugspace.ug.edu.gh Jamil, M., Kanampiu, F. K., Karaya, H., Charnikhova, T., & Bouwmeester, H. J. (2012). Striga hermonthica parasitism in maize in response to N and P fertilisers. Field Crops Research, 134, 1–10. Jeger, M. J., Gilijamse, E., Bock, C. H., & Frinking, H. D. (1998). The epidemiology, variability and control of the downy mildews of pearl millet and sorghum, with particular reference to Africa. Plant Pathology, 47(5), 544–569. Jethva, A. S., Raval, L., Madriya, R. B., Mehta, D. R., & Mandavia, C. (2011). Combing ability over environments for grain yield and its related traits in pearl millet. Crop Improvement, 38(1), 92–96. Jombart, T. (2008). adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics, 24(11), 1403–1405. Jones, E., Liu, C., & Gale, M. (1995). Mapping quantitative trait loci for downy mildew resistance in pearl millet. Theoretical and Applied Genetics, 91, 448–456. Jones, E. S., Breese, W. A., Liu, C. J., Singh, S. D., Shaw, D. S., & Witcombe, J. R. (2002). Mapping quantitative trait loci for resistance to downy mildew in pearl millet : Field and glasshouse screens detect the same QTL. Crop Science, 42, 1316–1323. Kannan, B., Senapathy, S., Bhasker Raj, A. G., Chandra, S., Muthiah, A., Dhanapal, A. P., & Hash, C. T. (2014). Association analysis of SSR markers with phenology, grain, and stover-yield related traits in pearl millet (Pennisetum glaucum (L.) R. Br.). The Scientific World Journal, 2014, 1–14. Kempthorne, O. (1957). An Introduction to Quantitative Genetics. Longman, New York and London. Korte, A., & Farlow, A. (2013). The advantages and limitations of trait analysis with GWAS. Plant 107 University of Ghana http://ugspace.ug.edu.gh Methods, 9(1), 1-29. Kountche, B. A., Hash, C. T., Dodo, H., Laoualy, O., Sanogo, M. D., Timbeli, A., … Haussmann, B. I. G. (2013). Development of a pearl millet Striga-resistant genepool: Response to five cycles of recurrent selection under Striga-infested field conditions in West Africa. Field Crops Research, 154, 82–90. Krall S., Youm O., Kogo, S.A. (1995). Panicle insect pest damage and yield loss in pearl millet. In Panicle insect pests of sorghum and pearl millet. International Consultative workshop, 4- 7 Oct 1993, ICRISAT Sahelian Center. 135-145. Kumar, K. A., Rai, K. N., Andrews, D. J., Talukdar, B. S., Rao, A. S., Babu, P. P., & Reddy, B. P. (1995). Registration of ICMP 451 parental line of pearl millet. Crop Science, 35(2), 605-605. Kumar, K., & Anand, M. (1993). Pearl millet in West Africa. In Millet and Sorghum commodity and Research Environment (pp. 1–16). Patancharu, A. P. India: ICRISAT. Kumari, J., Bag, M. K., Pandey, S., Jha, S. K., Chauhan, S. S., Jha, G. K., … Dutta, M. (2017). Assessment of phenotypic diversity in pearl millet [Pennisetum glaucum (L.) R. Br.] germplasm of Indian origin and identification of trait-specific germplasm. Crop and Pasture Science, 67(12), 1223–1234. Li, H., Peng, Z., Yang, X., Wang, W., Fu, J., Wang, J., … Liu, J. (2013). Genomewide association study dissects the genetic architecture of oil biosynthesis in maize kernels. Nature Genetics, 45, 43–50. Lipka, A. E., Tian, F., Wang, Q., Peiffer, J., Li, M., Bradbury, P. J., … Zhang, Z. (2012). GAPIT: Genome association and prediction integrated tool. Bioinformatics, 28(18), 2397–2399. Liu, C. J., Witcombe, J. R., Pittaway, T. S., Nash, M., Hash, C. T., Busso, C. G., & Gale, M. D. (1994). An RFLP-based genetic map of pearl millet (Pennisetum glaucum). Theoretical and 108 University of Ghana http://ugspace.ug.edu.gh Applied Genetics, 89, 481–487. Lubadde, G., Tongoona, P., Derera, J., & Sibiya, J. (2015). Pearl Millet Socioeconomic and Production Characteristics in Uganda. Kampala, Uganda: National Semi Arid Resources Research Institute of the National Agricultural Research Organisation-Uganda. Lubadde, G., Tongoona, P., Derera, J. &, Sibiya, J. (2016). Combining ability and heterosis for grain yield and rust resistance in pearl millet. Journal of Agricultural Science, 8(7), 80-96. MAER, ministere de l’agriculture et de l’equipement rural. (2012). Catalogue officiel des especes et des varietes cultivees au Senegal. Dakar, Senegal. Mahatma, M. K., Bhatnagar, R., Mittal, G. K., & Mahatma, L. (2011). Phenol metabolism in downy mildew resistant and susceptible genotypes of pearl millet. Archives Of Phytopathology And Plant Protection, 44(7), 623–636. Mangione, D., Senni, S., Puccioni, M., Grando, S., & Ceccarelli, S. (2006). The cost of participatory barley breeding. Euphytica, 150(3), 289–306. Manning, K., Pelling, R., Higham, T., Schwenniger, J. L., & Fuller, D. Q. (2011). 4500-year old domesticated pearl millet (Pennisetum glaucum) from the Tilemsi Valley, Mali: new insights into an alternative cereal domestication pathway. Journal of Archaeological Science, 38(2), 312-322. Mariac, C., Luong, V., Kapran, I., Mamadou, A., Sagnard, F., Deu, M., … Vigouroux, Y. (2006). Diversity of wild and cultivated pearl millet accessions (Pennisetum glaucum [L.] R. Br.) in Niger assessed by microsatellite markers. Theoretical and Applied Genetics, 114, 49–58. Matanyaire, C. M. (1998). Potential contribution of the pearl millet (Pennisetum glaucum) variety Okashana-1 to household food security in the northern communal areas of Namibia. South African Journal of Science, 94, 387–390. 109 University of Ghana http://ugspace.ug.edu.gh Mayee, C., & Siraskar, R. (1982). Relationship between downy mildew evaluation parameter and pearl millet productivity. Current Science, 52(992), 1-3. Mbaye, D. F. (1982). Les maladies parasitaires du mil, etat actuel des connaissances. Centre Nationale de Recherches Agronomiques / Bambey, Senegal. Mbaye, D. F. (1986). Recherches sur les maladies du mil. Centre Nationale de Recherches Agronomiques / Bambey, Senegal. Mbaye, D. F. (1994). Une etude du pathosysteme Pennisetum glaucum - Sclerospora graminicola. Application a la gestion du mildiou du mil au Senegal. Ecole Nationale Supérieure Agronomique de Montpellier, Montpellier, France. McCarthy, M. I., Abecasis, G. R., Cardon, L. R., Goldstein, D. B., Litlle, J., Ioannidis, J. P. A., & Hirschhorn, J. N. (2008). Genome-wide association studies for complex traits: consensus, uncertainty and challenges. Nature Reviews Genetics, 9(5), 356–369. Melvin, P., Prabhu, S. A., Veena, M., Shailasree, S., Petersen, M., Mundy, J., … Kini, K. R. (2015). The pearl millet mitogen-activated protein kinase PgMPK4 is involved in responses to downy mildew infection and in jasmonic- and salicylic acid-mediated defense. Plant Molecular Biology, 87(3), 287–302. Michelmore, R. w., Pawar, M. N., & Williams, R. J. (1982). Heterothallism in Sclerospora Graminicola phytopathology. Phytopathology, 12, 1368–1372. Monyo, E. S., Ipinge, S. A., Geoffrey, M. H., & Chinhema, E. (2001). Participatory breeding: does it make a difference? Lessons from Namibian pearl millet farmers. In Assessing the Impact of Using Participatory Research and Gender / Stakeholder Analysis. Morris, G. P., Ramu, P., Deshpande, S. P., Hash, C. T., Shah, T., Upadhyaya, H. D., … Kresovich, S. (2012). Population genomic and genome-wide association studies of agroclimatic traits in 110 University of Ghana http://ugspace.ug.edu.gh sorghum. Proceedings of the National Academy of Sciences, 110(2), 453–458. Mulatu, E., & Belete, K. (2001). Participatory varietal selection in lowland sorghum in eastern Ethiopia: Impact on adoption and genetic diversity. Experimental Agriculture, 37(2), 211– 229. Murty, B. R., Thakur, S. R., Prakash, N., Mehta, S. L., & Bhakta, S. T. (1983). Mutation breeding for downy mildew resistance in pearl millet. Nucleo-cytoplasmic interactions in disease- resistant lines. International Atomic Energy Agency (IAEA). Nambiar, V. V. S., Dhaduk, J. J., Sareen, N., Shahu, T., Desai, R., Shah, H., & Desai, R. (2011). Potential Functional Implications of Pearl millet (Pennisetum glaucum) in Health and Disease. Journal of Applied Pharmaceutical Science, 1(10), 62–67. Ndoye, M., & Gahukar, R. (1986). Insect pests of pearl millet in West Africa and their control. In Proceedings of the International Pearl Millet Workshop (pp. 7–11). Nene, Y. L., & Singh, S. D. (1976). Downy mildew and ergot of pearl millet. PANS, 22(3), 366- 385. Njoku, D., Egesi, C., & Gracen, V. (2014). Identification of pro-vitamin A cassava (Manihot esculenta Crantz) varieties for adaptation and adoption through participatory research. Journal of Crop Improvement, 28(3), 361–376. Omanya, G. O., Weltzien-Rattunde, E., Sogodogo, D., Sanogo, M., Hanssens, N., Guero, Y., & Zangre, R. (2007). Participatory varietal selection with improved pearl millet in West Africa. Experimental Agriculture, 43(1), 5-19. Ouendeba, B., Ejeta, G., Nyquist, W. E., Hanna, W. W., & Kumar, A. (1993). Heterosis and combining ability among African pearl millet landraces. Crop Science, 33(4), 735–739. Oumar, I., Mariac, C., Pham, J. L., & Vigouroux, Y. (2008). Phylogeny and origin of pearl millet 111 University of Ghana http://ugspace.ug.edu.gh (Pennisetum glaucum [L.] R. Br.) as revealed by microsatellite loci. Theoretical and Applied Genetics, 117, 489–497. Owere, L., Tongoona, P., Derera, J., & Wanyera, N. (2014). Farmers’ perceptions of finger millet production constraints, varietal preferences and their implications to finger millet Breeding in Uganda. Journal of Agricultural Science, 6(12), 126–138. Parmar, R. S., Vala, G. S., Gohil, V. N., Dudhat, A. S. (2013). Studies on Combining Ability for Development of New Hybrids in Pearl Millet [Pennisetum gaucum (L.) R. BR.]. International Journal of Plant Sciences, 8(2), 405-409. Pasam, R. K., Sharma, R., Malosetti, M., van Eeuwijk, F. A., Haseneyer, G., Kilian, B., & Graner, A. (2012). Genome-wide association studies for agronomical traits in a world-wide spring barley collection. BMC Plant Biology, 12(1), 1-16. Perez-Lara, E., Semagn, K., Tran, V. A., Ciechanowska, I., Chen, H., & Iqbal, M. (2017). Population structure and genome-population structure and genomewide association analysis of resistance to disease and insensitivity to Ptr toxins in Canadian spring wheat using 90K SNP array. Crop Science, 57, 1–18. Placide, R., Shimelis, H., & Laing, M. (2015). Farmers ’ perceptions , production and productivity constraints , preferences , and breeding priorities of sweetpotato in Rwanda. HortScience, 50(1), 36–43. Prabhu, S. A., Wagenknecht, M., Melvin, P., Kumar, B. S. G., Veena, M., Shailasree, S., … Kini, K. R. (2015). Immuno-affinity purification of PglPGIP1, a polygalacturonase-inhibitor protein from pearl millet: studies on its inhibition of fungal polygalacturonases and role in resistance against the downy mildew pathogen. Molecular Biology Reports, 42(6), 1123– 1138. 112 University of Ghana http://ugspace.ug.edu.gh Prabhu, S., & Kini, K. (2012). Polygalacturonase-inhibitor proteins in pearl millet: possible involvement in resistance against downy mildew. Acta Biochimica, 44(5), 415–423. Pray, C., & Nagarajan, L. (2009). Pearl millet and sorghum improvement in India (Vol. 919). International Food Policy Research Institute. Pucher, A., Sy, O., Sanogo, M. D., Angarawai, I. I., Zangre, R., Ouedraogo, M., … Haussmann, B. I. G. (2016). Combining ability patterns among West African pearl millet landraces and prospects for pearl millet hybrid breeding. Field Crops Research, 195, 9–20. Pushpavathi, B., Thakur, R. P., & Chandrashekara Rao, K. (2006). Inheritance of avirulence in Sclerospora graminico. Journal of Plant Pathology, 5(1), 54–59. Qi, X., Pittaway, T. S., Lindup, S., Liu, H., Waterman, E., Padi, F. K., … Devos, K. M. (2004). An integrated genetic map and a new set of simple sequence repeat markers for pearl millet, Pennisetum glaucum. Theoretical and Applied Genetics, 109(7), 1485–1493. Radhouane, L. (2009). La photosynthèse du mil (Pennicetum glaucum (L.) R.Br.) en présence de contrainte hydrique et saline. Journal of Agriculture and Environment for International Development, 103 (3), 185-200. Rajaram, V., Nepolean, T., Senthilvel, S., Varshney, R. K., Vadez, V., Srivastava, R. K., … Hash, C. T. (2013). Pearl millet [Pennisetum glaucum (L.) R. Br.] consensus linkage map constructed using four RIL mapping populations and newly developed EST-SSRs. BMC Genomics, 14(1), 159. Satyavathi, C. T., & Srivastava, R. K. (2017). Deciphering genomic regions for high grain iron and zinc content using association mapping in pearl millet. Frontiers in Plant Science, 8(5), 1–17. Sehgal, D., Rajaram, V., Armstead, I. P., Vadez, V., Yadav, Y. P., Hash, C. T., & Yadav, R. S. 113 University of Ghana http://ugspace.ug.edu.gh (2012). Integration of gene-based markers in a pearl millet genetic map for identification of candidate genes underlying drought tolerance quantitative trait loci. BMC Plant Biology, 12(1), 1-9. Sehgal, D., Skot, L., Singh, R., & Srivastava, R. K. (2015). Exploring potential of pearl millet germplasm association panel for association mapping of drought tolerance traits. PloS One, 10(5), 1–28. Sharma, R., Gupta, S. K., Kadvani, D. L., Shivpuri, A. S. H. A., & Rai, K. N. (2014). New virulent pathotypes of Sclerospora graminicola and resistance sources in pearl millet for A1 zone in India. Indian Journal of Agricultural Research, 84(6), 707-710. Sharma, R., Upadhyaya, H. D., Manjunatha, S. V, Rai, K. N., Gupta, S. K., & Thakur, R. P. (2013). Pathogenic variation in the pearl millet blast pathogen Magnaporthe grisea and identification of resistance to diverse pathotypes. Plant Disease, 97(2), 189–195. Sharma, R., Upadhyaya, H. D., Sharma, S., Gate, V. L., & Raj, C. (2015). New sources of resistance to multiple pathotypes of Sclerospora graminicola in the pearl millet mini core germplasm collection. Crop Science, 55, 1–10. Shetty, N. V, Wehner, T. C., Thomas, C. E., Doruchowski, R. W., & Shetty, K. P. V. (2002). Evidence for downy mildew races in cucumber tested in Asia , Europe , and North America. Scientia Horticulturae, 94, 231–239. Shiferaw, B., Bantilan, M. C., Gupta, S. C., & Shetty, S. (2004). Research spillover benefits and experiences in inter-regional technology transfer: An assessment and synthesis. Patancheru 502 324, Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid Tropics. Shinada, H., Yamamoto, T., Sato, H., Yamamoto, E., Hori, K., Yonemaru, J., … Fujino, K. (2015). 114 University of Ghana http://ugspace.ug.edu.gh Quantitative trait loci for rice blast resistance detected in a local rice breeding population by genome-wide association mapping. Breeding Science, 65(5), 388–395. Shivakumar, P., Geetha, H., & Shetty, H. (2003). Peroxidase activity and isozyme analysis of pearl millet seedlings and their implications in downy mildew disease resistance. Plant Science, 164(2003), 85–93. Singh, S. D. (1995). Downy mildew of pearl millet. Plant Disease, 79(6), 545–550. Singh, R. B., & Chaudhary, B. D. (1979). Biometrical methods in quantitative genetic analysis. Ludhiana, India: Kalyani. Singh, S. D., & King, S. B. (1988). Recovery resistance to downy mildew in pearl millet. Plant Disease, 72(5), 425–428. Singh, S. D., King, S. B., & Malla Reddy, P. (1990). Registration of five pearl millet germplasm sources with stable resistance to downy mildew. Crop Science, 30(5), 1164-1164. Singh, S. D., King, S. B., & Werder, J. (1993). Downy mildew disease of pearl millet. Patancheru, Andhra Pradesh, India: International Crops Research Institute for the Semi-arid Tropics. Singh S.D. & Talukdar B.S. (1998). Inheritance of complete resistance to pearl millet downy mildew. Plant Disease, 82, 791-793. Singh, S. D., Singh, P., Rai, K. N., & Andrews, D. J. (1990). Registration of ICMA 841 and ICMB 841 pearl millet parental lines with A1 cytoplasmic-genic male sterility system. Crop Science, 30 (6), 1378-1378. Singh, S. D., Wilson, J. P., Navi, S. S., Talukdar, B. S., Hess, D. E., & Reddy, K. N. (1997). Screening techniques and sources of resistance to downy mildew and rust in pearl millet (48th ed.). Patancheru, 502 324, Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid Tropics. 115 University of Ghana http://ugspace.ug.edu.gh Sivaramakrishnan, S., Thakur, R. P., Kannan, S., & Rao, V. P. (2003). Pathogenic and genetic diversity among Indian isolates of Sclerospora graminicola from pearl millet. Indian Phytopathology, 56(4), 392–397. Sow, M., Seck, P. A., Maiga, I. M., Laing, M., Ortiz, R., & Ndjiondjop, M.-N. (2015). Farmers’ rice knowledge and adoption of new cultivars in the Tillabéry region of western Niger. Agriculture & Food Security, 4(1), 1-17. Spencer-Phillips, P., & Jeger, M. J. (2004). Advances in downy mildew research. Dordrecht: Kluwer Academic Publishers. Stich, B., Haussmann, B. I. G., Pasam, R., Bhosale, S., Hash, C. T., Mel- chinger, A. E., & Parzies, H. K. (2010). Patterns of molecular and phenotypic diversity in pearl millet [Pennisetum glaucum (L.) R. Br.] from West and Central Africa and their relation to geographical and environmental parameters. BMC Plant Biol, 10, 1–10. Sudisha, J., Ananda Kumar, S., & Shekar, H. (2008). Characterization of downy mildew isolates of S clerospora graminicola by using differential cu ltivars and molecular markers. Journal of Cell and Molecular Biology, 7(1), 41–55. Supriya, A., Senthilvel, S., Nepolean, T., Eshwar, K., Rajaram, V., Shaw, R., … Narasu, M. L. (2011). Development of a molecular linkage map of pearl millet integrating DArT and SSR markers. Theoretical and Applied Genetics, 123(2), 239–250. Tan, S. L. (2015). Cassava silently, the tuber fills: the lowly cassava, regarded as a poor man’s crop, may help save the Euphytica world from the curse of plastic pollution. Utar Agriculture Science Journal, 1, 12–24. Tako, E., Reed, S. M., Budiman, J., Hart, J. J. & Glahn, R. P. (2015). Higher iron pearl millet (Pennisetum glaucum L.) provides more absorbable iron that is limited by increased 116 University of Ghana http://ugspace.ug.edu.gh polyphenolic content. Nutrition Journal, 14, 1-11. Thakur, R. P., Rao, V. P., & Sharma, R. (2007). Evidence for temporal virulence change in pearl millet downy mildew pathogen populations in India. Journal of SAT Agricultural Research, 3(1), 1–3. Thakur, R. P., Rao, V. P., & Sharma, R. (2009). Temporal virulence change and identification of resistance in pearl millet germplasm to diverse pathotypes of Sclerospora graminicola. Journal of Plant Pathology, 91(3), 629–636. Thakur, R. P., Rao, V. P., Wu, B. M., Subbarao, K. V., Shetty, H. S., Singh, G., … Panchbhai, S. D. (2004). Host resistance stability to downy mildew in pearl millet and pathogenic variability in Sclerospora graminicola. Crop Protection, 23(10), 901–908. Thakur, R. P., Sharma, R., & Rao, V. P. (2011). Screening Techniques for Pearl Millet Diseases. (No. 89). Thakur, R. P., Shetty, H. S., & Khairwal, I. S. (2006). Pearl millet downy mildew research in India: progress and perspectives. International Sorghum and Millets Newsletter, 47, 125–130. Thakur, R., Sharma, R., Rai, K. N., Gupta, S. K., & Rao, V. P. (2009). Screening techniques and resistance sources for foliar blast in pearl millet. SAT Agricultural Research, 7, 1-5. Tostain, S., & Marchais, L. (1989). Enzyme diversity in pearl millet (Pennisetum glaucum). Theoretical and Applied Genetics, 77, 188–193. Uno, D. (2005). Farmer’s selection of local and improved pearl millet varieties in Ovamboland, northern Namibia. African Study Monographs, 30, 107–117. Upadhyaya, H. D., Vetriventhan, M., Deshpande, S. P., Sivasubramani, S., Wallace, J. G., Buckler, E. S., … Ramu, P. (2015). Population genetics and structure of a global foxtail millet germplasm collection. The Plant Genome, 8(3), 1–13. 117 University of Ghana http://ugspace.ug.edu.gh Van Ginkel, M., Sayer, J., Sinclair, F., Aw-Hassan, M. ., Bossio, D., Craufurd, P., … Ortiz, R. (2013). An integrated agroecosystem and livelihoods approach for the poor and vulnerable in dry areas. Food Security, 5, 751–767. Varshney, R. K., Shi, C., Thudi, M., Mariac, C., Wallace, J., Qi, P., … Xu, X. (2017). Pearl millet genome sequence provides a resource to improve agronomic traits in arid environments. Nature Biotechnology, 35(10), 969–976. Weltzien, E., & King, S. B. (1995). Recurrent selection for downy mildew resistance in pearl millet. Plant Breeding, 114(3), 308–312. Williams, R. J., & Andrews, D. J. (1983). Breeding for disease and pest resistance in pearl millet. International Pearl Millet Improvement, 31(4), 136-158. Williams, R. J., Singh, S. D., & Pawar, M. N. (1981). An improved field screening technique for downy mildew resistance in pearl millet. Plant Disease, 65(3), 239-241. Wilson, J. P., Sanogo, M. D., Nutsugah, S. K., Angarawai, I., Fofana, A., Traore, H., … Muuka, F. P. (2008). Evaluation of pearl millet for yield and downy mildew resistance across seven countries in sub-Saharan Africa. African Journal of Agricultural Research, 3(5), 371–378. Winkel, T., Renno, J.F., Payne, W. (1997). Effect of the timing of water deficit on growth, phenology and yield of pearl millet, Pennisetum glaucum (L.) R. Br., grown in sahelian conditions. Journal of Experimental Botany, 48, 1001-1009. Xu, Y. (2010). Molecular Plant Breeding. (T. Hardwick, S. Hulbert, S. Gehrig, C. Parfitt, M. Caroll, & T. Head, Eds.). Oxfordshire, UK: CABI. Yadav, N. K., Thakur, D. P., & Rathi, A. S. (1997). Morphological alterations in pearl millet due to downy mildew disease. Haryana Agricultural University Journal of Research, 27(2), 137– 138. 118 University of Ghana http://ugspace.ug.edu.gh Yadav, O. P., & Rai, K. N. (2013). Genetic improvement of pearl millet in India. Agricultural Research, 2(4), 275–292. Yadav, R., Hash, C. T., Catherine, H., Witcombe, J. R., & Khairwal, S. I. (2013). Successful marker-assisted selection for disease resistance and drought tolerance in pearl millet in IndIa. In Biotechnologies at work for smallholders: case studies from developing countries in crops, livestock and fish. Rome: Food and Agriculture Organization of the United Nations. Yadav, R. S., Hash, C. T., Bidinger, F. R., Devos, K. M., & Howarth, C. J. (2004). Genomic regions associated with grain yield and aspects of post-flowering drought tolerance in pearl millet across stress environments and tester background. Euphytica, 136(3), 265–277. You, F. M., Song, Q., Jia, G., Cheng, Y., Duguid, S., Booker, H., & Cloutier, S. (2016). Estimation of genetic parameters and their sampling variances for quantitative traits in the type 2 modified augmented design. The Crop Journal, 4(2), 107–118. Yu, J., Pressoir, G., Briggs, W. H., Vroh Bi, I., Yamasaki, M., Doebley, J. F., … Buckler, E. S. (2006). A unified mixed-model method for association mapping that accounts for multiple levels of relatedness. Nature Genetics, 38(2), 203–208. Zarafi, A. B. (2005). Early sowing and metalaxyl seed treatment reduced incidence of pearl millet downy mildew [Sclerospora graminicola (Sacc.)] in Samaru, Nigeria. Journal of Plant Protection Research, 45(3), 163–169. Zarafi, A. B. (2007). Assessment of pearl millet genotypes for resistance to downy mildew (Sclerospora graminicola) under field conditions in Nigeria. Journal of Sustainable Agriculture, 29, 155–171. Zarafi, A., Emechebe, A., Akpa, A., & Alabi, O. (2004). Effect of Metalaxyl on the incidence and severity of pearl millet downy mildew disease in northern Nigeria. Archives Of 119 University of Ghana http://ugspace.ug.edu.gh Phytopathology And Plant Protection, 37(4), 261–268. Zhang, Z., Ersoz, E., Lai, C.-Q., Todhunter, R. J., Tiwari, H. K., Gore, M. A., … Buckler, E. S. (2010). Mixed linear model approach adapted for genome-wide association studies. Nature Genetics, 42(4), 355–360. Zhu, C., Gore, M., Buckler, E. S., & Yu, J. (2008). Status and prospects of association mapping in plants. The Plant Genome Journal, 1(1), 5–20. Zondervan, K. T., & Cardon, L. R. (2004). The complex interplay among factors that influence allelic association. Nature Reviews Genetics, 5(2), 89–100. 120 University of Ghana http://ugspace.ug.edu.gh APPENDICES APPENDIX 3.1.: Participatory rural appraisal questionnaire 1. Interview date: 2. Name of interviewer: 3. Starting time: 4. Ending time: A. Identification of the locality 5. GPS information: 6. Region 1. Kaffrine 2. Thies 3. Diourbel 4. Fatick 5. Kaolack. 7. Rural community: 8. Village: B. Farmer’s identification 9. Name: 10. Surname: 11: Phone Number: 12. Age: 1. 18-24 2. 25-34 3. 35-49 4. 50-64 5. 65+ 13. Sex 1. Male 2. Female 14. Ethnic group 121 University of Ghana http://ugspace.ug.edu.gh 1. Wolof 2. Serere 3. Al pulaar 4. Diolas 5. Mandingue 6. Others 15. Are you the household head? 1. Yes 2. No 16. Your marital status 1. Single 2. Married 3. Divorced 4. Widowed 5. Mandingue 6. Others 17. Level of education 1. Illiterate 2. Adult literacy 3. Koranic study 4. Primary Study 5. Secondary Study 6. University Level 7. Others 18. Is farming your main sources of income? 1. Yes. 2. No 19. If no, which are others sources of income? 20. Are you a member of any farmers’ association? 21. If yes, specify the name of the farmers’ association: C. Pearl millet cropping system 22. How long have you been growing pearl millet? : 23. In average, how many hectares do you grow per year with pearl millet? 24. On which type of soil do you grow pearl millet? 25. How do you plough your pearl millet field? 26. When pearl millet is sown? 1. Before the rain 2. After the rain: 27. Spacing used when sowing the pearl millet: 28. Do you treat your seeds before sowing? 122 University of Ghana http://ugspace.ug.edu.gh 1. Yes. 2. No 29. If yes, with which pesticide? 30. What is your appreciation about the fertility of the soil where pearl millet is grown? 31. Do you apply mineral fertilizer? 1. Yes 2. No 32. If yes, which type of mineral fertilizer? 33. on which rate? 34. Do you apply organic fertilizer? 1. Yes 2. No 35. If yes, which type of mineral fertilizer and rate of application? 36. Do you thinned your pearl millet plot? 1. Yes 2. No 37. If yes, how many plants/hole do you leave? 38. Do you control weeds on your pearl millet plot? 1. Yes 2. No 39. If yes, how many time and when the weeding is done? 40. Can you estimated your pearl millet production during the last season? 41. Which are the others major crops you grow and specify the area? 1. Groundnut 2. Sorghum 3. Maize 4. Rice 5. Fonio 6. Sesame 7. Cowpea 8. Others 42. Are you doing crop rotation? 1. Yes 2. No 43. If yes, what are the alternate crop of sorghum in crop rotation system? 123 University of Ghana http://ugspace.ug.edu.gh D. Constraints to pearl millet production and farmer’s perception on downy mildew 44. Based on your experience, what are the main pearl millet production constraints? 1. Agricultural input 2. Animal rambling 3. Birds 4. Insects 5. Land access 6. Soil fertility 7. Striga 8. Diseases 9. Drought 10. Others 45. How do you manage these constraints? 46. Do you know downy mildew (used pictures). 1. Yes 2. No 47. On which growing stage do you observe it? 1. Growing stage 2. Reproductive stage 3. Maturity stage 48. On which part of the plant the disease is observed? 1. Leaves 2. Stem 3. Panicle 49. Describe the symptoms? 50. Did you observed downy mildew in your field during the past years? 1. Yes 2. No 51. Based on your experience, how can you judge the severity of the disease? 52. Do you do preventive treatment? 1. Yes 2. No 53. If yes, with which pesticide? 54. What is your immediate action when the disease appear in your pearl millet field? 124 University of Ghana http://ugspace.ug.edu.gh E. Farmers’ preferred traits 55. Which pearl millet varieties are you using in your field? 56. Why did you choose these varieties? 57. Where do you get the seeds of these varieties? 1. Own seeds 2. NGOs 3. Cooperative 4. Market. 5. Relatives 6. Research Center 7. Others 58. Based on your experience, which traits are important when selecting a pearl millet variety to grown in your field? 59. Do you want to have another pearl millet variety different to what you have now? 1. Yes 2. No 60. Which traits do you want to have on this new variety? 125 University of Ghana http://ugspace.ug.edu.gh APPENDIX 3.2.: Pearl millet downy mildew symptoms showed to farmers during the survey 126 University of Ghana http://ugspace.ug.edu.gh APPENDIX 5.1: Performance of tested lines at Bambey Genotypes DMI (%) Severity (%) Flowering (das) Plant Height (cm) Panicle length (cm) Productive tillers SOSAT C-88 3 3 44 199 29 2 7042 S 94 90 41 144 13 1 IBL 001-4-1 0 0 57 202 36 2 IBL 003-B-1 0 0 49 223 32 2 IBL 011-4-1 7 2 44 229 32 3 IBL 012-1-1 43 25 60 160 26 1 IBL 012-2-1 100 25 62 134 38 3 IBL 015-1-1 0 4 48 200 47 3 IBL 021-3-1 8 6 60 249 35 2 IBL 023-2-1 0 8 62 155 24 2 IBL 024-3-1 85 36 62 177 25 2 IBL 026-1-1 67 17 55 173 36 2 IBL 026-2-1 17 6 57 191 46 1 IBL 028-B-1 50 56 61 152 40 2 IBL 033-1-1 14 7 60 140 31 1 IBL 037-4-1 10 28 40 225 27 1 IBL 037-5-1 0 10 38 67 13 1 IBL 040-1-1 0 0 43 247 38 5 IBL 040-5-1 0 0 44 174 24 2 IBL 047-1-1 0 19 58 138 17 4 IBL 050-1-1 14 5 55 221 39 2 IBL 053-2-1 0 0 60 233 27 1 IBL 053-3-1 0 4 58 209 33 1 IBL 055-4-1 0 0 49 203 42 1 IBL 056-2-1 9 2 46 159 23 2 IBL 058-5-1 8 10 45 146 22 1 IBL 061-1-1 57 14 61 201 31 2 IBL 064-1-1 91 23 60 161 22 1 IBL 065-B-1 0 6 65 171 27 6 IBL 066-3-1 8 4 42 94 15 1 IBL 066-4-1 67 8 44 113 28 2 IBL 067-2-1 0 4 49 176 27 2 IBL 067-B-1 14 9 47 177 41 2 IBL 069-4-1 75 29 51 158 40 2 IBL 070-1-1 67 21 45 140 37 0 IBL 071-4-1 17 6 62 158 26 4 IBL 073-B-1 43 11 48 217 35 2 IBL 077-1-1 23 6 52 238 35 3 IBL 079-B-1 25 6 49 152 27 2 IBL 081-2-1 46 12 67 193 23 3 IBL 082-B-1 20 8 59 181 42 1 IBL 084-1-1 0 0 47 169 31 1 IBL 091-1-1 0 4 55 218 43 1 IBL 092-3-1 0 4 50 257 38 1 IBL 093-1-1 15 6 46 218 22 2 IBL 094-2-1 21 0 52 252 29 2 IBL 095-1-1 8 2 55 101 36 1 IBL 095-4-1 0 0 44 211 39 1 IBL 098-1-1 0 0 51 199 40 3 IBL 098-3-1 0 2 55 229 39 2 IBL 099-3-1 0 11 45 165 32 2 IBL 100-5-1 25 6 46 233 47 5 IBL 101-3-1 8 2 62 180 28 2 IBL 102-3-1 13 3 51 171 28 3 IBL 105-3-1 0 8 59 152 40 1 IBL 105-B-1 55 2 55 152 36 2 IBL 106-B-1 0 0 51 233 47 2 IBL 107-B-1 31 21 67 211 36 1 IBL 110-B-1 0 4 47 225 40 2 IBL 111-3-1 0 0 42 155 28 2 IBL 114-6-1 0 0 60 231 30 1 IBL 117-2-1 63 19 46 202 20 1 IBL 119-B-1 0 0 60 239 44 2 IBL 121-2-1 67 63 60 60 10 2 IBL 125-B-1 33 10 51 185 47 1 IBL 131-6-1 8 17 42 162 25 2 127 University of Ghana http://ugspace.ug.edu.gh Genotypes DMI (%) Severity (%) Flowering (das) Plant Height (cm) Panicle length (cm) Productive tillers IBL 133-2-1 0 0 71 169 32 2 IBL 138-B-1 30 13 62 169 27 2 IBL 141-B-1 0 0 44 190 23 1 IBL 143-1-1 0 0 41 154 24 3 IBL 143-2-1 0 0 47 168 18 1 IBL 149-1-1 0 9 53 108 26 3 IBL 150-B-1 0 2 60 132 27 2 IBL 151-2-1 7 2 38 204 47 3 IBL 155-2-1 17 25 41 179 33 1 IBL 160-1-1 0 0 58 175 28 3 IBL 161-1-1 0 0 67 195 10 2 IBL 165-1-1 14 4 50 237 45 3 IBL 167-5-1 8 2 44 218 33 3 IBL 170-1-1 22 31 67 206 22 1 IBL 170-B-1 36 16 41 124 30 1 IBL 173-1-1 20 5 61 200 55 1 IBL 173-3-1 0 0 48 167 34 2 IBL 174-3-1 25 6 69 55 13 0 IBL 179-2-1 0 0 52 190 28 1 IBL 179-3-1 0 2 44 222 24 3 IBL 180-2-1 31 12 46 194 29 1 IBL 181-2-1 9 9 46 167 37 2 IBL 183-4-1 0 0 48 173 27 2 IBL 183-5-1 57 13 44 140 25 1 IBL 185-3-1 10 8 48 149 34 2 IBL 186-1-1 0 0 50 198 26 3 IBL 188-1-1 0 0 42 113 38 1 IBL 198-1-1 29 9 42 119 29 2 IBL 198-2-1 14 4 42 217 37 2 IBL 200-3-1 0 20 42 180 32 2 IBL 206-1-1 0 8 46 200 30 1 SL 2-B-1 0 8 67 128 31 4 SL 4-3-1 91 32 62 159 45 1 SL 5-1-1 7 4 58 193 21 1 SL 5-4-1 21 5 44 189 42 2 128 University of Ghana http://ugspace.ug.edu.gh APPENDIX 5.2: Performance of tested lines at Nioro Genotypes DMI (%) Severity (%) Flowering (das) Plant Height (cm) Panicle length (cm) Productive tillers SOSAT C-88 11 9 47 222 24 4 7042 S 78 79 49 175 15 1 IBL 001-4-1 0 0 65 213 30 1 IBL 003-B-1 8 5 47 222 27 3 IBL 011-4-1 0 0 47 233 37 4 IBL 012-1-1 64 4 64 158 24 1 IBL 012-2-1 25 0 65 177 35 1 IBL 015-1-1 5 7 57 198 37 4 IBL 021-3-1 0 0 65 263 40 1 IBL 023-2-1 50 80 64 220 25 0 IBL 024-3-1 100 36 55 208 43 3 IBL 026-1-1 8 4 65 247 42 2 IBL 026-2-1 15 2 53 185 65 1 IBL 028-B-1 0 0 62 172 38 1 IBL 033-1-1 9 0 53 187 32 1 IBL 037-4-1 0 92 47 220 22 2 IBL 037-5-1 25 38 58 65 17 0 IBL 040-1-1 18 30 47 227 35 3 IBL 040-5-1 0 0 47 162 20 1 IBL 047-1-1 10 0 63 205 26 3 IBL 050-1-1 0 5 63 207 22 1 IBL 053-2-1 8 0 63 305 38 2 IBL 053-3-1 77 0 64 262 30 1 IBL 055-4-1 0 0 51 210 30 2 IBL 056-2-1 0 19 50 155 17 1 IBL 058-5-1 8 0 49 192 23 3 IBL 061-1-1 21 2 74 258 28 2 IBL 064-1-1 82 95 76 161 22 0 IBL 065-B-1 0 0 75 195 27 1 IBL 066-3-1 0 20 49 137 18 2 IBL 066-4-1 0 10 47 120 20 2 IBL 067-2-1 58 0 44 190 37 2 IBL 067-B-1 14 0 47 178 25 3 IBL 069-4-1 83 23 63 202 30 2 IBL 070-1-1 0 9 44 168 38 2 IBL 071-4-1 0 0 64 195 15 5 IBL 073-B-1 14 0 51 248 40 2 IBL 077-1-1 0 0 53 225 25 4 IBL 079-B-1 15 0 51 138 17 3 IBL 081-2-1 0 0 55 218 25 1 IBL 082-B-1 23 0 57 178 20 2 IBL 084-1-1 0 0 51 197 20 3 IBL 091-1-1 8 0 64 233 28 1 IBL 092-3-1 20 5 62 247 38 2 IBL 093-1-1 8 4 51 203 30 2 IBL 094-2-1 8 5 53 272 37 2 IBL 095-1-1 58 48 60 203 42 2 IBL 095-4-1 0 0 44 250 55 3 IBL 098-1-1 0 0 64 173 20 5 IBL 098-3-1 7 0 51 248 47 3 IBL 099-3-1 14 0 47 165 30 2 IBL 100-5-1 0 0 64 260 23 2 IBL 101-3-1 0 0 51 200 20 3 IBL 102-3-1 0 0 51 197 27 2 IBL 105-3-1 25 0 63 175 37 3 IBL 105-B-1 0 2 54 180 25 2 IBL 106-B-1 0 0 58 283 45 3 IBL 107-B-1 31 12 74 182 33 1 IBL 110-B-1 6 0 51 222 33 2 IBL 111-3-1 0 0 55 167 20 4 IBL 114-6-1 0 0 79 190 30 2 IBL 117-2-1 0 0 79 202 20 2 IBL 119-B-1 15 0 60 295 43 1 IBL 121-2-1 89 100 83 65 10 0 IBL 125-B-1 18 0 58 153 13 3 IBL 131-6-1 0 0 49 163 20 4 129 University of Ghana http://ugspace.ug.edu.gh Genotypes DMI (%) Severity (%) Flowering (das) Plant Height (cm) Panicle length (cm) Productive tillers IBL 133-2-1 0 0 79 223 33 1 IBL 138-B-1 100 100 72 175 23 0 IBL 141-B-1 0 0 51 198 18 2 IBL 143-1-1 0 0 49 187 20 4 IBL 143-2-1 0 0 47 205 25 5 IBL 149-1-1 2 4 55 160 26 3 IBL 150-B-1 8 0 79 223 32 1 IBL 151-2-1 0 5 51 147 25 2 IBL 155-2-1 0 4 51 230 27 2 IBL 160-1-1 0 0 77 177 22 2 IBL 161-1-1 0 0 65 228 23 1 IBL 165-1-1 0 0 53 268 42 3 IBL 167-5-1 20 0 49 182 33 3 IBL 170-1-1 0 34 64 202 20 1 IBL 170-B-1 15 10 53 193 25 2 IBL 173-1-1 8 0 83 157 23 1 IBL 173-3-1 9 0 51 205 32 3 IBL 174-3-1 100 100 79 60 12 0 IBL 179-2-1 0 0 55 228 23 2 IBL 179-3-1 8 0 50 243 22 1 IBL 180-2-1 7 5 51 187 17 3 IBL 181-2-1 0 15 49 167 28 3 IBL 183-4-1 0 0 49 190 27 4 IBL 183-5-1 14 7 53 150 25 1 IBL 185-3-1 10 0 53 197 27 2 IBL 186-1-1 0 0 47 250 25 4 IBL 188-1-1 0 0 55 193 38 2 IBL 198-1-1 20 10 63 177 28 1 IBL 198-2-1 18 0 49 220 20 3 IBL 200-3-1 7 16 58 185 32 1 IBL 206-1-1 17 0 58 210 23 2 SL 2-B-1 50 0 77 148 20 2 SL 4-3-1 100 37 64 226 42 2 SL 5-1-1 0 8 61 140 17 1 SL 5-4-1 0 5 49 192 38 3 130 University of Ghana http://ugspace.ug.edu.gh APPENDICE 7.1.: List of lines used for the GWAS, their origin and membership Number or Nom Or Origin cluster Number Nom Or Origin cluster 1 IBL 125-B-1 BENIN 2 40 IBL 091-1-1 NIGER 2 2 IBL 077-1-1 BURKINA FASO 3 41 IBL 092-3-1 NIGER 2 3 IBL 079-B-1 BURKINA FASO 3 42 IBL 093-1-1 NIGER 2 4 IBL 081-2-1 BURKINA FASO 3 43 IBL 094-2-1 NIGER 2 5 IBL 160-1-1 BURKINA FASO 2 44 IBL 095-1-1 NIGER 2 6 IBL 161-1-1 BURKINA FASO 3 45 IBL 098-1-1 NIGER 2 7 IBL 165-1-1 BURKINA FASO 2 46 IBL 098-3-1 NIGER 2 8 IBL 167-5-1 BURKINA FASO 3 47 IBL 099-3-1 NIGER 2 9 IBL 170-B-1 BURKINA FASO 3 48 IBL 100-5-1 NIGER 2 10 IBL 179-2-1 BURKINA FASO 3 49 IBL 105-3-1 NIGER 3 11 IBL 047-1-1 CAMEROUN 2 50 IBL 105-B-1 NIGER 3 12 IBL 050-1-1 CENTRAFRIQUE 2 51 IBL 106-B-1 NIGER 2 13 IBL 053-2-1 CENTRAFRIQUE 3 52 IBL 110-B-1 NIGER 2 14 IBL 053-3-1 CENTRAFRIQUE 3 53 IBL 111-3-1 NIGER 3 15 IBL 055-4-1 MALI 2 54 IBL 114-6-1 NIGER 3 16 IBL 056-2-1 MALI 2 55 IBL 117-2-1 NIGER 2 17 IBL 058-5-1 MALI 3 56 IBL 082-B-1 SENEGAL 2 18 IBL 061-1-1 MALI 2 57 IBL 084-1-1 SENEGAL 3 19 IBL 064-1-1 MALI 3 58 IBL 121-2-1 SENEGAL 3 20 IBL 065-B-1 MALI 3 59 SL 2-B-1 SENEGAL 3 21 IBL 066-3-1 MALI 1 60 SL 4-3-1 SENEGAL 2 22 IBL 131-6-1 MALI 3 61 SL 5-1-1 SENEGAL 3 23 IBL 133-2-1 MALI 3 62 SL 5-4-1 SENEGAL 2 24 IBL 138-B-1 MALI 2 63 IBL 001-4-1 unknown 3 25 IBL 141-B-1 MALI 1 64 IBL 003-B-1 unknown 3 26 IBL 143-2-1 MALI 2 65 IBL 012-1-1 unknown 3 27 IBL 149-1-1 MALI 3 66 IBL 012-2-1 unknown 3 28 IBL 150-B-1 MALI 3 67 IBL 021-3-1 unknown 2 29 IBL 155-2-1 MALI 3 68 IBL 023-2-1 unknown 3 30 IBL 186-1-1 MALI 3 69 IBL 024-3-1 unknown 2 31 IBL 188-1-1 MALI 2 70 IBL 026-1-1 unknown 2 32 IBL 180-2-1 MAURITANIE 3 71 IBL 026-2-1 unknown 2 33 IBL 181-2-1 MAURITANIE 2 72 IBL 028-B-1 unknown 2 34 IBL 183-4-1 MAURITANIE 1 73 IBL 033-1-1 unknown 2 35 IBL 067-2-1 NIGER 1 74 IBL 037-4-1 unknown 3 36 IBL 067-B-1 NIGER 2 75 IBL 040-5-1 unknown 3 37 IBL 069-4-1 NIGER 2 76 IBL 198-1-1 unknown 3 38 IBL 070-1-1 NIGER 2 77 IBL 206-1-1 unknown 2 39 IBL 071-4-1 NIGER 2 131