i GENETIC ANALYSIS OF RESISTANCE TO ROSETTE DISEASE OF GROUNDNUT (Arachis hypogaea L.) By USMAN ALHASSAN (10293978) 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 SCHOOL OF AGRICULTURE COLLEGE OF AGRICULTURE AND CONSUMER SCIENCE UNIVERSITY OF GHANA LEGON December, 2013 University of Ghana http://ugspace.ug.edu.gh ii 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. .................................................. Usman Alhassan (Student) .................................................. Prof. Eric Yirenkyi Danquah (Supervisor) .................................................. Prof. Kwadwo Ofori (Supervisor) .................................................. Prof. Samuel Kwame Offei (Supervisor) .................................................. Prof. S.G. Ado (Supervisor) University of Ghana http://ugspace.ug.edu.gh iii ABSTRACT Groundnut rosette disease (GRD), transmitted naturally by aphids, Aphis craccivora, is the most destructive viral disease of groundnut (Arachis hypogaea L.) in Nigeria and causes serious yield losses to farmers. The narrow genetic base among groundnuts has impeded efficient utilization for development of host resistance to GRD. Studies were undertaken in Nigeria to: (i) ascertain farmers‘ knowledge of and preferences for rosette resistant genotypes; (ii) assess the genetic diversity among aphid and rosette resistant genotypes using microsatellite markers; (iii) exploit genotype x environment interaction towards improved selection efficiency to obtain high-yielding varieties; and, (iv) determine the mode of inheritance of resistance to groundnut rosette disease. A participatory rural appraisal (PRA) involving 90 farmers was conducted in two groundnut producing communities in Northern Nigeria. Early maturing genotypes and GRD resistance were the most important farmer preferred traits. Farmers ranked insect pests and inadequate rainfall as the most important causes of groundnut rosette disease. Majority of farmers across the study areas were doing nothing to avert the disease. Some farmers however rogue infected plants and use GRD resistant varieties when available. Genetic diversity and association of simple sequence repeat (SSR) markers with GRD resistance were detected in a set of 50 cultivated groundnut genotypes with different levels of resistance to GRD. Out of 170 bands amplified from 36 primers, 166 were polymorphic (97.65%). Each amplified 2 to 12 microsatellite loci, with an average of 4.74 loci per primer. The Polymorphic Information Content value of each marker ranged from 0.19 to 0.82. Average pairwise genetic distance among the 50 genotypes was 0.31. The largest distance was 0.51 (between ICGV – IS – 07812 and RS006F4B1 – 31) and the shortest distance was 0.05 between ICGV – IS – 07865 and ICGV – IS – 07864, all the four lines were GRD-resistant. Cluster analysis revealed seven clusters using disease reaction University of Ghana http://ugspace.ug.edu.gh iv to GRD. The assessment of genetic diversity of GRD-resistant groundnut genotypes will help groundnut breeders to formulate crosses by choosing parents with different genetic backgrounds and in the development of gene-mapping populations with greater marker polymorphism. The 36 F2 populations generated from 9 x 9 half diallel mating scheme were infested with veruliferous aphids, Aphis craccivora and scored three times fortnightly following inoculation. General combining ability (GCA) and specific combining ability (SCA) effects for GRD resistance were highly significant, indicating that both additive and non-additive gene effects govern the inheritance of GRD resistance. Low narrow sense heritability for Area Under Disease Progress Curve (29.29 %) along with high broad sense heritability (94.78 %) further highlight the influence of non-additive gene action in controlling resistance to GRD, suggesting effective selection of superior genotypes at advanced generations when maximum homozygosity is fixed. University of Ghana http://ugspace.ug.edu.gh v DEDICATION To my mother, wife and children, brothers and sisters for their prayers and support throughout this study. University of Ghana http://ugspace.ug.edu.gh vi ACKNOWLEDGMENT I sincerely wish to thank my principal supervisor Prof Eric Yirenkyi Danquah for offering me the opportunity to work on this project at the same time keeping me well on track. I deeply appreciate the efforts and suggestions of my co supervisors Prof. Kwadwo Ofori and Prof. S. K. Offei for their visit to Abuja to thoroughly discuss my first thesis draft despite their busy schedule. My in – country supervisor, Prof. S.G Ado became my true soul mate during the course of the study. He was always open to discussions and gave invaluable advice and contribution on my field work and thesis write up. I am indebted to Dr. Bonney Ntare for reading through all my chapters and for sharing his wealth of experience on groundnut rosette virus with me despite not being one of my supervisory team. I am grateful to WACCI administrative staff for their continued support for ensuring that my stay in Ghana was comfortable, and that the critical resources for the in-country research were received on time. I am grateful to the Alliance for Green Revolution in Africa (AGRA) for the PhD scholarship. They supported this research by making funds available for my research. Many thanks to Ahmadu Bello University, Zaria and Institute for Agriculture in particular for allowing me to pursue the programme. I would like to recognize the encouragement and useful critique provided by the members of my cohort (Cohort 2) – particularly at the proposal and final writing-up stages. Most importantly, special thanks to my wife, Hauwa Ahmed, and my children, Muhammad Sani, Hassana and Hussaina, M. Almustapha, Usman and M. Jaafar for their patience and understanding during the critical times of the research. University of Ghana http://ugspace.ug.edu.gh vii TABLES OF CONTENTS DECLARATION I ABSTRACT III DEDICATION V ACKNOWLEDGMENT VI TABLES OF CONTENTS VII LIST OF FIGURES XI LIST OF TABLES XII LIST OF ABBREVIATIONS XIV CHAPTER ONE 1 1 GENERAL INTRODUCTION 1 CHAPTER TWO 4 2 LITERATURE REVIEW 4 2.1 Groundnut (Arachis hypogaea L.) 4 2.2 Genetic Resources 6 2.3 Groundnut Rosette Disease 7 2.3.1 Origin and occurrence of Groundnut Rosette Virus 8 2.3.2 Symptoms of Groundnut Rosette Disease (GRD) 9 2.3.3 Causal agents of Groundnut Rosette Disease 10 2.3.4 Diagnosis of Groundnut Rosette Disease 13 2.3.5 Epidemiology of Groundnut Rosette Disease 13 2.3.6 Virus-vector interactions and transmission 14 2.4 Progress made in combating Groundnut Rosette Disease in Nigeria 15 2.5 Genetics of Resistance to Groundnut Rosette Disease 16 University of Ghana http://ugspace.ug.edu.gh viii 2.6 Combining Ability for traits in Groundnut 17 2.7 Genotype x Environment Interaction (GEI) in groundnut 19 2.8 Genetic Diversity in Cultivated Groundnut Based on Molecular Markers 21 2.9 Participatory breeding and varietal selection in groundnut 24 CHAPTER THREE 26 3 FARMERS‘ PERCEPTION OF PRODUCTION CONSTRAINTS AND PREFERRED TRAITS FOR RESISTANT GROUNDNUT ROSETTE VARIETIES 26 3.1 Introduction 26 3.2 Materials and Methods 27 3.2.1 Description of the study areas 27 3.2.2 Farmer survey and data analysis 28 3.3 Results 29 3.3.1 Household and demographic information 29 3.3.2 Characteristics of preferred groundnut varieties in Batsari and Nasarawa – eggon districts 35 3.3.3 Preferred groundnut varieties and associated characteristics 38 3.3.4 Perception of farmers on constraints to groundnut production 38 3.4 Discussion 39 3.5 Conclusions and Recommendations 42 CHAPTER FOUR 43 4 ASSESSMENT OF GENETIC DIVERSITY OF GROUNDNUT (ARACHIS HYPOGAEA L.) GENOTYPES FOR RESISTANCE TO ROSETTE DISEASE USING SSR MARKERS 43 4.1 Introduction 43 4.2 Materials and methods 44 4.2.1 Plant material and DNA extraction 44 4.2.2 SSR Analysis 45 4.2.3 Electrophoresis and data collection 47 4.3 Results 48 4.3.1 Allelic variation at SSR loci 48 University of Ghana http://ugspace.ug.edu.gh ix 4.3.2 Comparison of gene diversity 50 4.4 Discussion 55 4.5 Conclusions and Recomendations 57 CHAPTER FIVE 58 5 INHERITANCE ON RESISTANCE TO GROUNDNUT ROSETTE DISEASE 58 5.1 Introduction 58 5.2 Materials and Methods 59 5.2.1 Population Development and Phenotype Evaluation 59 5.2.1 Aphid and Rosette resistance evaluation 60 5.2.2 Agronomic performance 64 5.2.3 Data analysis 64 5.3 Results 71 5.3.1 Variance components and heritability of traits 71 5.3.2 Performance of the groundnut genotypes grown at Samaru and Lafia, 2012 75 5.4 General and specific combining ability for traits 87 5.5 Selection for superior genotypes for resistance to groundnut rosette disease 95 5.6 Discussion 97 5.7 Conclusions and Recommendations 103 CHAPTER SIX 105 6 MOLECULAR CONFIRMATION OF ROSETTE RESISTANCE IN PROMISING GROUNDNUT GENOTYPES BY ONE-STEP REVERSE TRANSCRIPTASE POLYMERASE CHAIN REACTION (RT – PCR) 105 6.1 Introduction 105 University of Ghana http://ugspace.ug.edu.gh x 6.2 Materials and methods 107 6.2.1 Collection of plant samples 107 6.2.2 RNA Extraction and Purification 107 6.2.3 Complementary DNA (cDNA) synthesis and Polymerase Chain Reaction (PCR) 108 6.3 Results 111 6.4 Discussion 114 6.5 Conclusions and Recommendations 116 CHAPTER SEVEN 118 7 GENERAL DISCUSSION 118 7.1 Participatory rural appraisal (PRA) 118 7.1.1 Epidemiology of GRD from the studied genotypes 119 7.1.2 Performance of the genotypes across the two contrasting locations 121 7.1.3 Inheritance of resistance to groundnut rosette disease 123 7.2 Challenges 124 7.3 Conclusions and recommendations 125 BIBLIOGRAPHY 126 APPENDIX 1: PARTICIPATORY RURAL APPRAISALS QUESTIONNAIRE 146 University of Ghana http://ugspace.ug.edu.gh xi LIST OF FIGURES Figure 3. 1: Map of Nigeria showing Batsari (Katsina state) and Nasrawa-Eggon (Nasarawa state) 28 Figure 3. 2: Distribution of traits preferred by farmers 35 Figure 3. 3: Farmers perception on the causes of groundnut rosette disease 36 Figure 3. 4: Farmers perception of yield loss due to groundnut rosette disease 37 Figure 3. 5: Groundnut rosette disease control measures adopted by farmers 38 Figure 4. 1: Hierarchical dendrogram of 50 groundnut genotypes by using similarity coefficients based on the Nei‘s (1983) original genetic distance calculated from data of 166 SSR loci using the UPGMA method: Refer to Table 4.1 for names of the corresponding codes 54 Figure 4. 2: The genetic relationships among the five groundnut populations calculated using UPGMA method based on the Nei‘s (1983) genetic distance 55 Figure 5. 1: Cross over Genotype x Location Interaction for sound kernel yield per plant across Batstari and Lafia Locations 84 Figure 5. 2: Performance of F2 groundnut for sound kernel yield per plant and AUDPC 84 Figure 6. 1: Amplification banding pattern of GRAV – CP (HRP92/93), GRV – CP (OFR3P and 4P) and Sat-RNA markers in 16 groundnut genotypes 113 University of Ghana http://ugspace.ug.edu.gh xii LIST OF TABLES Table 3. 1: Farmer and Household information for Batsari and Nasarawa – eggon LGA in Nigeria for the 2010 growing season 31 Table 3. 2: Pair wise correlation of some farmers‘ level of awareness of the groundnut rosette disease 32 Table 3. 3: Pair-wise ranking of major crops grown by farmers in Batsari and Nasarawa – eggon in 2011 34 Table 3. 4: Pair-wise ranking of the most important constraints in groundnut production in Batsari and Nasarawa – eggon 39 Table 4. 1: Groundnut genotypes with different levels of resistance and susceptibility to groundnut rosette disease (GRD) included in the study 46 Table 4. 2: Primers used in the study, gene bank ID, repeat motif, frequency and number of alleles as well as gene diversity, and polymorphic information contents (PIC) based on the analysis of 50 groundnut genotypes for 35 polymorphic SSR markers 51 Table 4. 3: Pairwise genetic distance coefficients of 50 GRD-resistant genotypes using 36 SSR primer pairs combinations analyzed by PowerMarker software 52 Table 5. 1: Pedigree, source, description and characteristics of parental genotypes used for population development 61 Table 5. 2: Table 5. 2:Format of ANOVA for individual location 65 Table 5. 3: Table 5. 3: Format of ANOVA for the combined locations 66 Table 5. 4: Format of Diallel analysis of variance for model I method II for groundnut progenies evaluated in one location 70 Table 5. 5: Mean squares of measured traits for 9 parents and 36 F2 half diallel progenies of groundnut evaluated over Samaru and Lafia Locations in 2012 73 Table 5. 6: Table 5. 6: Variance components, Heritability estimates and expected gain for groundnut traits over combined Samaru and Lafia location in 2012 74 Table 5. 7: Performance of parents and their F2 progenies for sound kernel weight per plant (g) over Samaru and Lafia environment in 2012 76 Table 5. 8: Performance of parents and their F2 progenies for aphid damage Index (DI) over Samaru and Lafia environment in 2012 79 Table 5. 9: Performance of parents and their F2 progenies for AUDPC over University of Ghana http://ugspace.ug.edu.gh xiii Samaru and Lafia environment in 2012 82 Table 5. 11: Mean squares of combined ANOVA for half 9 x 9 diallel analysis for general and specific combining abilities and their interactions with location for ten morphological traits of groundnut evaluated at two locations in 2012 89 Table 5. 12: Variance component for GCA, SCA and their interactions with location, Bakers ratio, additive and dominance variances considering random effect model for 9 parents and 36 F2 evaluated across Samaru and Lafia Locations in 2012 90 Table 5. 13: Estimates of general combining ability (GCA) effects of 9 parental lines for four important morphological characters of groundnut evaluated across Samaru and Lafia Locations in 2012 91 Table 5. 14: Table 5. 14: Estimates of specific combining ability (SCA) effects measured in the 36 F2 progenies evaluated across Samaru and Lafia Locations in 2012 93 Table 5. 15: The top 10 and 4 poorest performing F2 genotypes selected based on Rank summation Index of SKWT, PWT, DI and AUDPC 96 Table 6. 1: Primers used in amplification of various regions of causal agents of groundnut rosette disease complex primers in the 100 series represent internal primers for specified regions 110 Table 6. 2: Field resistance scored by DI and AUDPC and RT – PCR confirmation of GRD-resistance in some groundnut genotypes 112 University of Ghana http://ugspace.ug.edu.gh xiv LIST OF ABBREVIATIONS 100SKWT One hundred sound kernel weight AFLP Amplified fragment length polymorphism ANOVA Analysis of variance ATA Agricultural Transformation agenda AUDPC Area under disease progress curve BW Bacteria wilt cDNA Complementary Deoxyribo nucleic acid CP Coat protein CTAB Cetyl trimethyl ammonium bromide DAPI diamidino-2-phenylindole DI Aphid damage index DNA Deoxyribo nucleic acid EST Express sequence tag F1 First filial generation F2 Second filial generation FAO Food and Agricultural Organization GCA General combining ability GD Genetic distance GEI Genotype x environment interaction GISH Genomic in situ hybridization GLM Generalized linear model GRAV Groundnut rosette assistor virus University of Ghana http://ugspace.ug.edu.gh xv GRD Groundnut rosette disease GRV Groundnut rosette virus HCL Hydrochloric acid He Heterozygote percentage HS Half – sib IC - RT – PCR Immunocapture-reverse transcriptase-polymerase chain reaction ICRISAT International Crop Research Institute for Semi – Arid Tropics ILRI International Livestock Research Institute IAP Inoculation access period IAR Institute for Agricultural Research LGA Local Government Area LRR Leucine-rich repeats LSD Least significant difference MAb Monoclonal antibody MAS Marker assisted selection MS Mean square MT Metric ton NAICPP National Accelerated Industry Crop Production Program NARP National Agricultural Research Program NBS Nucleotide binding site NBS – LRR Nucleotide-binding-site leucine-rich repeat NCBI National Centre for Biotechnology Information NID Normally and independently distributed NOPP Number of pods per plant University of Ghana http://ugspace.ug.edu.gh xvi ORF Open reading frame PBNV Peanut bud necrosis tospovirus PCR Polymerase chain reaction PIC Polymorphic information content PRA Participatory rural appraisal PK protein kinases PWPT Pod weight per plant PWTON Pod weight in tons per hectare QTL Quantitative trait loci RAPD Random amplified fragment length polymorphism RGA Resistant gene analogue RNA Ribose nucleic acid RSI Rank summation index RT – PCR Reverse transcriptase polymerase reaction SAS Statistical analysis software Sat-RNA Satellite – RNA SCA Specific combining ability SCAR Sequence characterized amplified region SHP Shelling percentage SKWPT Sound kernel weight per plant SKWTTON Sound kernel weight ton per hectare SSA Sub – Saharan Africa SSR Simple sequence repeat Taq Thermos aquaticus University of Ghana http://ugspace.ug.edu.gh xvii TAS – ELISA Triple antibody sandwich - ELISA TIR Toll and interleukin-1 receptor TSWV Tomato Spotted Wilt Virus UPGMA Unweighted pair group method with arithmetic averaging. USDA United States Department of Agriculture UV Ultra violet WACCI West Africa Centre for Crop Improvement WB Wash buffer University of Ghana http://ugspace.ug.edu.gh 1 CHAPTER ONE 1 GENERAL INTRODUCTION Groundnut (Arachis hypogaea L.) is an important food legume highly adapted to tropical and subtropical climates of the world. It is a key crop for small scale farmers especially in Africa and Asia where the crop serve as a valuable source of dietary protein, oil, and fodder for livestock. It contains 48-50% oil and 26-28% protein, and a rich source of dietary fibre, minerals, and vitamins (Janila et al., 2013). In addition, groundnut has the ability to fix atmospheric nitrogen to the soil to help in the maintenance of soil fertility. This crop is cultivated annually on about 24.63 million hectares worldwide with annual production of 41.27 million tons in shell and productivity of about 1.85 t ha –1 (FAO, 2012). The global annual increase in production is 0.4% between 2009 and 2012 was attributed to both, an annual increase in yield by 0.1% and in area by 0.3% (Janila et al., 2013). In West Africa, Nigeria is the largest producer of groundnuts with production of 3.07 million tons on about 2.4 million hectare (FAO, 2012). Groundnut is the fifth most important oilseed in the world in terms of volume of oil production and is widely grown in more than 100 countries of tropical, subtropical, and warm temperate regions of the globe (Upadhyaya et al., 2012) Despite the economic, social and cultural importance of groundnuts, its productivity is severely constrained by several biotic and abiotic factors. Drought is the major abiotic constraint affecting groundnut productivity and quality worldwide. Two thirds of the global production occurs in rain-fed regions of the semi-arid tropics where rainfall is generally erratic and insufficient, causing unpredictable drought stress (Reddy et al., 2003). Groundnut yield and quality are severely constrained by a wide variety of fungal, University of Ghana http://ugspace.ug.edu.gh 2 bacterial, viral, and nematode pathogens. Among the fungal diseases, early leaf spot (Cercospora arachidicola) and late leaf spot (Cercosporidium personatum) are most prevalent, and occur throughout groundnut growing regions (Liu et al., 2013). Late leaf spot and rust (Puccinia arachidis) diseases often occur simultaneously and can cause 50– 70% yield loss in India and some African countries (Khedikar et al., 2010). Groundnut rosette disease (GRD) causes greater yield loss than any other viral disease affecting groundnut in the semi-arid tropics of the world (Naidu et al., 1999). The disease is caused by three interdependent viruses (causal agents) (see 2.3.3). It is the most destructive viral disease of groundnut in Africa (Naidu et al., 1999) resulting in sporadic yield loss of about 30% annually in farmers field. It is endemic to groundnut-growing regions of sub-Saharan Africa (SSA) and Madagascar (Yayock et al., 1976). The most serious yield losses were reported during 1975 when an epidemic in northern Nigeria destroyed approximately 0.7 million hectares of groundnut, with an estimated loss of US$250 million (Yayock et al., 1976). In Nigeria, farmers have identified groundnut rosette disease (GRD) as widespread and devastating, compelling some of them to abandon production in some areas. The disease is characterised by small, chlorotic, twisted and distorted leaflet with shortened internodes and thickened stems. Affected plants especially those infected at a young stage are severely stunted (Bock et al., 1990). The disease also affects both quality of the haulm and the pod. This disease can cause up to 100% yield (Adu-Dapaah et al., 2004; Waliyar, et al., 2007). Previous studies indicate that GRD could be managed by chemical control of the vector (Bock and Nigam, 1998). However, resource poor farmers seldom use chemical control measures due to lack of resources, labour constraints and cost (Dwivedi et al., 2003; Adu-Dapaah et al., 2004). The above factors coupled with health hazards associated with University of Ghana http://ugspace.ug.edu.gh 3 the use of insecticides make the use of host resistance the most cost effective and environmentally friendly alternative. Earlier attempts to control GRD using host resistance resulted in development of several resistant varieties across West Africa (Waliyar et al., 2007). However, these varieties were only tolerant to one of the components of groundnut rosette virus (GRV) but susceptible to groundnut rosette assistor virus (GRAV) indicating lack of resistance to this component of the rosette complex that serves as a helper component for aphid transmission (Subrahmanyam et al., 1998; Olorunju et al., 2001). Furthermore, resistance to GRV is not immunity and seems to be overcome under high inoculum pressure and in adverse environmental conditions (Bock et al., 1990). Therefore development of genotypes that are resistant to GRD virus is the most effective, economic and sustainable method of limiting virus inoculum build- up and spread of both the aphid and the virus. The main objective of this study was to develop groundnut breeding lines with potential for resistance to groundnut rosette disease with acceptable farmers‘ preferred agronomic and yield traits The specific objectives were to: a. ascertain farmers‘ knowledge and preferences of rosette resistant genotypes; b. assess molecular polymorphism among aphid and rosette resistant genotypes; c. asses genotype x environment interaction towards improved selection efficiency to obtain high-yielding varieties; and d. determine the mode of inheritance of resistance to groundnut rosette disease University of Ghana http://ugspace.ug.edu.gh 4 CHAPTER TWO 2 LITERATURE REVIEW 2.1 Groundnut (Arachis hypogaea L.) Groundnut (Arachis hypogaea L.) is an annual or perennial plant that is distinguished from most other species by producing aerial flowers, but fruiting below the soil level. Arachis belongs to the family Fabaceae, tribe Aeschynnomeneae, sub-tribe Stylanthinae. Arachis hypogaea L. is the only domesticated species in the genus (Tillman and Stalker, 2009), and Krapovickas (1969) concluded that A. hypogaea var. hypogaea is likely the most ancient type because it has similar branching patterns as wild species, no compound florets, and a prostrate growth habit. The centre of origin for the genus Arachis is the Mato Grosso area of Brazil, but species evolved over a wide range of habitats in South America (Gregory et al., 1980). Molecular data indicate that the centre of genetic variation also is the Mato Grosso area of Brazil to eastern Bolivia (Stalker et al., 1994). Eighty species have been described (Krapovickas and Gregory, 1994; Valls and Simpson, 2005) which have been divided into nine sections based on morphology and cross-compatibility relationships. Smartt and Stalker (1982) proposed that the A and B genomes of section Arachis may be an A1 and A2 rather than being truly different based on chromosome pairing relationships. The species of different sections have overlapping distributions in many areas. Hybrids between species in different sections are difficult to produce and are usually sterile, while intrasectional hybrids can be fertile if they have similar genomic make-up (Stalker et al., 1991). The species A. hypogaea has two subspecies hypogaea and fastigiata which are further divided into six botanical types hypogaea and hirsute, and fastigiata, vulgaris, equatoriana and peruviana botanical varieties. The subspecies are separated morphologically based on presence or absence of flowers on the main stem and regularly alternating vegetative and reproductive University of Ghana http://ugspace.ug.edu.gh 5 nodes on branches. The characteristics of these both botanical varieties shows that hypogaea has no floral axes or branches on main stem; alternating pairs of vegetative and reproductive axes on branches (alternate branching); inflorescence simple; vegetative branches moderate to profuse; primary branches longer than main stem; growth habit spreading, intermediate, or erect; usually two seeds per pod; pod beak not very prominent; seed size medium (runner market type) to large (Virginia market type); testa color generally tan (red, white, purple, or variegated also exist); cured seed dormancy moderate; maturity medium to late (Ntare et al., 2007). The fastigiata has floral axes on main stem; irregular pattern of vegetative and productive branches with reproductive branches predominating on branches (sequential branching); inflorescence usually simple; vegetative branches sparse; primary branches shorter than main stem; growth habit upright; two to four seeds per pod; pod beak absent, slight, or prominent; seed size small to medium; testa color tan, red, white, yellow, purple, or variegated; cured seed dormancy little (Ntare et al., 2007). Cultivated groundnut has an allotetraploid genome (AABB, 2n = 4x = 40). The low level of genetic variation within the cultivated gene pool and its polyploid nature limit the utilization of molecular markers to explore genome structure and facilitate genetic improvement. Nevertheless, a wealth of genetic diversity exists in diploid Arachis species (2n = 2x = 20), which represent a valuable gene pool for cultivated peanut improvement (Guo et al., 2012). Arachis hypogaea is a recent allotetraploid (David at al., 2012), most probably resulting from the hybridization of two wild species followed by natural chromosome duplication (Halward et al., 1991; Young et al., 1996; Seijo et al., 2007). The genome of A. hypogaea is large, being estimated at 2·8 Gb (Grilhuber, 2005) with a large repetitive fraction of approximately 64 % determined by DNA renaturation kinetics (Dhillon et al., 1980) Cytogenetic analyses in A. hypogaea have revealed two types of chromosomes: ten pairs of A-type chromosomes, with strongly 4', 6-diamidino-2-phenylindole (DAPI)-stained (and hence AT-rich) heterochromatin at the centromeres, including the smallest pair of all University of Ghana http://ugspace.ug.edu.gh 6 chromosomes (Robledo et al., 2010), and another ten pairs of chromosomes with more weakly staining centromeric heterochromatin bands, designated B chromosomes (Seijo et al., 2004; Robledo et al., 2010). Studies comparing the chromosomal heterochromatic banding patterns together with evidence from positions of rDNA clusters (Robledo, et al., 2010) and genomic in situ hybridization (GISH) (Seijo et al., 2007) suggest that A. hypogaea A chromosomes are similar to those in the wild diploid A. duranensis, whilst the peanut B chromosomes are similar to those in the wild diploid A. ipaënsis Other evidence such as species geographic distribution (Robledo et al.,2010) and molecular phylogenies (Kochert et al., 1996; Burow et al., 2009; Moretzsohn et al., 2013) corroborates that the most probable A and B genome donors to A. hypogaea are A. duranensis and A. ipaënsis. 2.2 Genetic Resources for Groundnut Rosette Disease Over 3400 germplasms accessions evaluated for reaction to groundnut rosette disease at Chitedze Agricultural Research Station Lilongwe, Malawi, only 89 long duration Virginia types were identified as resistance to the disease. A high percentage (76%) of them originated in West Africa (Nigeria 39%, Burkina – Faso 13.9%, Cote d‘voire 9.9%, Senegal 6.9%, Mali 30%, Gambia 2.0% and Equitoria Guinea 1.0%) and the rest were from Southern Africa (Subrahmanyam et al., 1998). In addition, 11 short duration Spanish types were identified in Africa germplasm originating from West Africa especially Burkina – Faso. Out of a total of over 2000 accession evaluated in a preliminary screening trial in 1994/95 growing season at the same station, 15 were found to be rosette disease resistance (Subrahmanyam et al., 1998). Additional resistant sources were sought from 2,301 accessions from different sources and 252 advanced breeding lines derived from crosses involving earlier identified sources of resistance to rosette (Olorunju et al., 2001). The lines were evaluated in field screening trials using an infector row technique during 1996 and 1997 growing seasons at the Institute for University of Ghana http://ugspace.ug.edu.gh 7 Agricultural Research (IAR), Samaru, Nigeria (Olorunju et al., 2001). Among the germplasm lines, 65 accessions were reported to show high levels of resistance while 134 breeding lines were resistant (Olorunju et al., 2001).The report concluded that all rosette disease resistant lines were susceptible to groundnut rosette assistor virus (GRAV) and that the identified germplasm and breeding lines will contribute to an integrated management of groundnut rosette disease. In addition to A. hypogaea collections, more than 1,300 Arachis species accessions have been collected (Stalker et al., 2002), with about 800 Arachis entries being maintained by the USDA (Stalker and Simpson, 1995). Preservation of wild Arachis species is difficult for most accessions because the long, fragile pegs break during harvest and the soil must be sifted to recover pods. Stalker and Simpson (1995) reported that nearly 25% of the species from which seed can be obtained under nursery conditions have fewer than 50 seeds in storage. Additionally, at least 25% of the Arachis species accessions in germplasm nurseries are maintained vegetatively because of their poor seed production under cultivation. A large number of disease and insect resistance, and other agronomically useful traits are present in accessions of Arachis species, which makes them potentially valuable genetic resources for crop improvement (Stalker and Moss, 1987; Stalker and Simpson, 1995). 2.3 Groundnut Rosette Disease Groundnut rosette disease is the most destructive viral disease of groundnut in Africa and can cause serious yield losses under favourable conditions. Groundnut production is constantly threatened by potential outbreaks of rosette disease epidemics. Improvement of host plant resistance to this disease provides the most effective control strategy (Olorunju et al., 2001; Herselman et al., 2004). ICRISAT and its partners have made significant contributions towards the understanding of the epidemiology of the disease and confirmation based on University of Ghana http://ugspace.ug.edu.gh 8 molecular diagnostic assays. This knowledge has provided a basis for development and utilization of groundnut cultivars with resistance to the groundnut rosette disease and impacted the lives of thousands of farmers in sub-Saharan Africa (Olorunju et al., 2001). 2.3.1 Origin and occurrence of Groundnut Rosette Virus Groundnut is the only known natural host of a complex of three agents of rosette disease (GRV sat-RNA and GRAV). It is likely that the viruses have evolved and survived in the host species native to Africa before the introduction of groundnut (Subrahmanyam et al., 1998). After its introduction in the 16 th century, groundnut became an accidental host of rosette disease representing a case of the ―new – encounter‖ phenomenon (Buddenhagen and Ponti, 1984). It is possible that resistance came to Africa in some of the original introductions from South American centre (s) of origin and due to recurrent epidemics in West Africa (Olorunju et al., 2001). It was concentrated to a greater degree by natural out crossing and recombination. Groundnut rosette disease was first reported in 1907 from Tanganyika (Waliyar et al., 2007), now called Tanzania, and has since been reported in several other African countries south of Sahara. The major areas of disease occurrence include Burkina Faso, Ghana, Nigeria, Malawi, Mozambique and Uganda (Ntare et al., 2002). Symptoms similar to groundnut rosette disease have been reported in some countries of Asia and South America, but diagnostic tests to unequivocally confirm the presence of the disease have not been conducted (Reddy, 1991). Thus, it is generally assumed that groundnut rosette disease is endemic to groundnut growing countries in Africa South of the Sahara and its off-shore islands such as Madagascar (Ntare et al., 2002). During the course of evolution, as these genes did not possess any survival value in the absence of the disease, they may have been altered in the majority of genotypes (Reddy, 1991). One prerequisite for the loss of traits University of Ghana http://ugspace.ug.edu.gh 9 during ‗evolution‘ is their simple inheritance and rosette resistance is governed by two independent major recessive genes (Nigam and Bock, 1990). 2.3.2 Symptoms of Groundnut Rosette Disease (GRD) GRD occurs with two variant symptoms, chlorotic rosette and green rosette, with considerable variation within each type (Murant, 1989; Naidu et al., 1999). Both forms of the disease cause plants to be severely stunted, with shortened internodes and reduced leaf size, resulting in a bushy appearance of plants (Naidu et al., 1999). In chlorotic rosette, leaves are usually bright yellow with a few green islands and leaf lamina is curled. In the green rosette, leaves appear dark green, with light green to dark green mosaic (Naidu et al., 1999). Chlorotic rosette occurs throughout the Sub-Sahara Africa (SSA), whereas green rosette has been reported from East and West Africa (Naidu et al., 1999). A less common symptom variant, mosaic rosette, due to mixed infection of the plants by the Sat-RNA causing chlorotic variant and mottle variant, was reported from East Africa (Waliyar, et al., 2007). Variability in Sat-RNA is mainly responsible for symptom variations (Murant and Kumar, 1990; Taliansky et al., 1997). In addition, differences in genotypes, plant stage at infection, variable climatic conditions and mixed infections with other viruses also contribute to symptom variability under field conditions (Naidu et al., 2007). Infection due to chlorotic or green rosette disease occurring in young plants (prior to flowering) usually results in 100% yield loss. In contrast, plants infected during later growth stages (between flowering and pod setting) may show symptoms only in some branches or parts of branches and yield loss depends on severity of infection (Naidu et al., 2007). Infection after pod setting/maturation causes negligible effects on pod yield (Waliyar et al., 2007). An average annual yield loss due to GRD is estimated to be between 5 and 30% in non-epidemic years and epidemics often result in 100% yield loss (Waliyar et al., 2007). The deleterious impact of GRAV or GRV on University of Ghana http://ugspace.ug.edu.gh 10 host plant together with Sat-RNA in a synergistic manner is not known. Ansa et al. (1991) have reported that stunting is more severe in diseased groundnut plants containing all the three agents than in diseased groundnut plants containing only GRV and Sat-RNA. Other reports have suggested that GRAV or GRV infection alone in groundnut results in transient mottle symptoms with insignificant impact on the plant growth and yield (Taliansky et al., 2000). These results have, however, been contradicted by Naidu and Kimmins (2007) who demonstrated that GRAV infection alone affects plant growth and contributes to significant yield losses in susceptible groundnut cultivars. 2.3.3 Causal agents of Groundnut Rosette Disease Groundnut rosette disease is a viral disease, transmitted by an aphid, Aphis craccivora Koch (Insecta: Homoptera) in a persistent calculative manner (Waliyar, et al., 2007). Three causal agents are involved in GRD etiology: Groundnut rosette assistor virus (GRAV), Groundnut rosette virus (GRV) and a Satellite-RNA (Sat-RNA) (Reddy et al., 1985; Murant et al., 1988; Taliansky et al., 2000). The intimate interaction between GRAV, GRV, and sat-RNA is crucial to the development of the disease. GRV, a member of the genus Umbravirus, has a single-stranded, positive-sense RNA genome of 4,019 nt (Talianskyet al., 1996) that contains four large open reading frames (ORFs). ORF 2 is a putative RNA-dependent RNA polymerase and is likely expressed as a fusion protein with the product of ORF1 by a –1 frameshift mechanism. The 3¢ ORFs (Bock et al., 1990); Deom et al., 2000) are almost completely overlapping. The protein encoded by ORF 3 was shown to be a trans-acting long- distance movement protein that can traffic nonrelated viral RNA systemically (Ryabov et al., 1999), while analysis of the ORF 4 putative amino acid sequence suggests that it may be involved in cell-to-cell movement (Taliansky, et al., 1996). University of Ghana http://ugspace.ug.edu.gh 11 GRAV is a member of the family Luteoviridae. It was first recognized as a component of groundnut rosette disease by Waliyar et al., (2007). Casper et al. (1983) and Reddy et al. (1985) characterized the virus and identified it as a luteovirus. The virus replicates autonomously in the cytoplasm of phloem tissue. GRAV is transmitted by A. craccivora in a persistent manner, and experimentally by grafting, but not by mechanical sap inoculation, seed, and pollen or by contact between the plants (Taliansky et al., 2000). Groundnut is the only known natural host of GRAV is reported to occur wherever GRD has been reported (Waliyar et al., 2007). The virus on its own causes symptomless infection or transient mottle, and can cause significant yield loss in susceptible groundnut cultivars (Naidu et al., 1999). There are no reports on occurrence of strains of GRAV causing the disease (Waliyar et al., 2007). GRV belongs to the genus Umbravirus. It was first isolated and characterized by Reddy et al. (1985). The virus replicates autonomously in the cytoplasm of the infected tissues (Taliansky et al., 2000). GRV on its own causes transient symptoms, but a Sat-RNA associated with GRV is responsible for rosette disease symptoms. GRV depends on GRAV for encapsulation of its RNA and transmission by A. craccivora in a persistent mode (Robinson et al., 1999). Groundnut is the only known natural host, but several experimental hosts in the family Chenopodiaceae and Solanaceae have been reported (Murant et al., 1998). No strains of GRV have been reported (Waliyar et al., 2007) and the virus is restricted to SSA and its offshore islands. GRV acts as a helper virus for replication of sat-RNA. The Sat-RNA (sub-viral RNAs) of GRV belongs to the Subgroup-2 (small linear) satellite RNAs. It is a single-stranded, linear, non-segmented RNA of 895 to 903 nucleotides (Murant et al., 1988; Block et al., 1994; Taliansky et al., 2000). It totally depends on GRV for its replication, encapsulation and movement, both within and between the plants. Sat-RNA is responsible for rosette symptoms and plays a critical role in helper virus dependent University of Ghana http://ugspace.ug.edu.gh 12 transmission of GRV. Different variants of Sat-RNA have been shown to be responsible for different rosette symptoms, such as green rosette and chlorotic rosette (Murant and Kumar, 1990; Taliansky et al., 1997). It is mechanically transmissible along with the GRV and is also transmitted by aphids in the presence of GRV and GRAV. A single aphid vector acquires GRAV, GRV, and sat-RNA; however, it does not always transmit the three disease agents together to a host plant (Naidu et al., 1999). GRAV or GRV plus sat-RNA can be transmitted separately. However, for the disease to perpetuate in nature, all three agents must be transmitted by the aphid vector to a plant (Deom et al., 2000). Phylogenetic analysis of the overlapping ORFs 3 and 4 shows that the GRV isolates cluster according to the geographic region from which they were isolated, indicating that Malawian GRV isolates are distinct from Nigerian GRV isolates (Deom et al., 2000). Phylogenetic analysis also showed clustering within the sat-RNA isolates according to country of origin, as well as within isolates from two distinct regions of Malawi Deom et al., 2000). Because the GRAV CP sequence is highly conserved, independent of the geographic source of the GRAV isolates, the GRAV CP sequence represents the most likely candidate to use for pathogen-derived resistance in groundnut and may provide effective protection against groundnut rosette disease throughout SSA (Deom et al., 2000). Groundnut rosette disease has been reported in Angola, Burkina Faso, Côte d‘Ivoire, Gambia, Ghana, Kenya, Madagascar, Malawi, Niger, Nigeria, Senegal, South Africa, Sudan, Swaziland, Tanzania, Uganda, and Zaire (now DR Congo) (Gibbons, 1977; Naidu et al., 1999). The agents of GRD have not been detected elsewhere in the world, despite the fact that groundnut is grown in more than 100 countries around the world (Upadhyaya et al., 2012) and A. craccivora is found in almost all these groundnut growing regions (Naidu et al., 1999). University of Ghana http://ugspace.ug.edu.gh 13 2.3.4 Diagnosis of Groundnut Rosette Disease Groundnut rosette disease can be tentatively diagnosed in the field based on the characteristic symptoms in groundnut or by mechanical inoculation onto a suitable indicator host such as Chenopodium amaranticolor. Symptom development on C. amaranticolor indicates the presence of GRV, but this test is not always reliable when the indicator plants are subjected to the widely fluctuating temperatures of SSA (Naidu et al., 1999). Improved diagnostic methods include a triple antibody sandwich enzyme-linked immunosorbent assay (TAS- ELISA) for detection of GRAV (Rajeshwar et al., 1987) and reverse transcription- polymerase chain reaction (RT-PCR) that allows detection of each of the three agents (Naidu et al., 1999). The advantage of the RT-PCR method is that it concurrently detects all three groundnut rosette disease agents in plants and aphids (Naidu et al., 1999). 2.3.5 Epidemiology of Groundnut Rosette Disease The epidemiology of GRD is complex, involving interactions between and among two viruses and a Sat-RNA, the vector, and the host plant and environment (Naidu et al., 1998). Since none of the causal agents is seed-borne, primary infection of crops depend on the survival of infected plants (virus sources) and vectors (aphids) (Naidu et al., 1998). In the West, East and Southern Africa, A. craccivora maintains itself successfully through the dry and wet seasons on some crops and wild host plants. In Nigeria these hosts are found in the Amaranthaceae, Asteraceae, Caesalpinaceae, Compositae, Euphorbiaceae, Fabaceae, Moringaceae, Nyctaginaceae, Papilionaceae, Portulacaceae, Solanaceae and Verbenaceae (Alegbejo, 2000). The epidemics of groundnut rosette virus disease that occurred in the main groundnut producing areas of Nigeria was speculated to be due to unusual combination of weather and groundnut sowing dates, which lead to massive build-up, early dispersal and survival of University of Ghana http://ugspace.ug.edu.gh 14 aphids in the wet season (Misari et al.,1988). Also intermittent wet and dry spells in the early part of the season, without heavy rainfall, were probably responsible for the development and successful dispersal of alate aphids (Yayock et al., 1976). 2.3.6 Virus-vector interactions and transmission Aphis craccivora, commonly known as the cowpea aphid is the principal vector involved in the transmission of all the GRD agents in a persistent and circulative manner (Waliyar, 2007). GRV and Sat-RNA must be packaged within the GRAV coat protein to be aphid transmissible. Studies have shown that all the GRAV particles whether they contain GRAV RNA or GRV RNA and Sat-RNA are acquired by the aphid vector from phloem sap in 4h and 8h acquisition access feeding for chlorotic and green rosette, respectively (Misari et al., 1988). Then, there is a latent period of 26h 40min and 38h 40min for chlorotic and green rosette, respectively, and the inoculation access feeding period of 10min for both forms (Misari et al., 1988). Once acquired, aphid can transmit virus particles for up to two weeks and beyond. All stages of the aphid can acquire and transmit the disease agents. Transmission rates of 26-31% have been reported with one and two aphids per plant, and 49% with five aphids per plant (Misari et al., 1988). Aphid vector does not always transmit all the three agents together (Naidu et al., 1999). Under natural conditions, some GRD-affected plants (GRV and Sat-RNA positive) were found to be free from GRAV, and GRAV was detected in some non-symptomatic plants (no GRV and Sat-RNA) (Naidu et al., 1999). This situation was due to difference in inoculation feeding behaviour of the vector leading to transmission of (i) all the three agents together, (ii) only GRAV or (iii) GRV and Sat-RNA, as demonstrated by the electrical penetration graph (EPG) studies of aphid stylet activities (Naidu et al., 1999). This showed that during short inoculation feeding (test probe or stylet pathway phase) vector aphids probe groundnut leaves University of Ghana http://ugspace.ug.edu.gh 15 without reaching the phloem, transmitting only GRV and Sat-RNA, which multiply in the epidermal and mesophyll cells. Even if GRAV particles are deposited in the mesophyll cells, they cannot replicate, as they can replicate only in the phloem cells (Naidu et al., 1999). However, vector aphids can transmit GRAV, and GRV, Sat-RNA when the stylets penetrate sieve elements (salivation phase) of the phloem cells. Therefore, the success of transmitting all the three agents together is high when inoculation feeding period is longer or increasing the number of aphids per plant (Misari et al., 1988). Vector aphids fail to acquire or transmit GRV and Sat-RNA from diseased plants lacking GRAV and such plants become dead-end sources. However, if such plants receive GRAV later due to vector feeding, the plants again serve as source of inoculum (Waliyar et al., 2007). 2.4 Progress made in combating Groundnut Rosette Disease in Nigeria In Nigeria, research on the development of groundnut cultivars with resistance to rosette was initiated in 1986 by Institute for Agricultural Research (IAR), Ahmadu Bello University, Zaria in collaboration with ICRISAT (Olorunju et al., 1992). Concerted efforts were made to improve resistance to groundnut rosette viruses in the existing locally grown varieties. These earlier attempts resulted in the development of a number of rosette resistant varieties such as SAMNUT 10 (RMP 12), SAMNUT 11 (RMP 91), SAMNUT 16 (M554.76), SAMNUT 20 (M412.80I), and SAMNUT 21 (MDR-8-19).SAMNUT 22 and SAMNUT 23 and SAMNUT 24. Popularization of these varieties and availability of seed to farmers were made possible by the then Federal government projects (NAICPP and NARP), collaborative work between IAR and ICRISAT, ILRI and GGP/CFC using farmer participatory approach. There was a dramatic increase in production from 1.4 million MT to over 2 million MT from 1994 to 2003 in Nigeria because of the well-coordinated collaborative work between NARJs and IARCs in combating the groundnut virus disease in Africa (Ntare et al., 2002). Despite these University of Ghana http://ugspace.ug.edu.gh 16 attempts, sporadic occurrences of the disease of about 30 % from year to year and among fields were still observed in farmer‘s field (Waliyar et al., 2007). Control strategies adopted by most farmers have traditionally emphasized on vector control mainly by pesticide and cultural practices such as manipulating sowing dates and plant density (Subrahmanyam et al., 1998). Chemical control methods have been only partially effective, since aphid populations can reach very high numbers, leading to intensive pesticide application in an attempt to eliminate the vector, and when accompanied with drought may lead to epidemics. Furthermore, there are concerns that the vector may develop pesticide resistance and the intense application may have deleterious effect on the environment. Therefore the development of genotypes that are resistant to both aphids and the virus is the most effective, economic and sustainable method of limiting virus inoculums build-up (Herselman et al.,2004). 2.5 Genetics of Resistance to Groundnut Rosette Disease Breeding for resistance to groundnut rosette disease demands a good knowledge of the breeding methodologies as well as a good understanding of the disease and its causal organisms. Identification of sources of resistance and its efficient utilization require an understanding of the genetic control of resistance and knowledge of the amount of genetic variability available for selection. Determining the suitable parents to use for development of resistant genotype is particularly important. Early genetic studies on groundnut rosette disease showed that resistance was effective against GRV and its sat-RNA and was governed by 2 independent recessive genes (de Berchoux, 1960). He also stated that resistant lines were not immune and that individual plants could become infected when subjected to inoculation by massive number of aphids. This resistance was reported to operate equally against both chlorotic rosette (de Berchoux, 1960) and green rosette (Harkness, 1977). He attributed the University of Ghana http://ugspace.ug.edu.gh 17 low recovery of resistant plants from Virginia x Spanish crosses to heavy inoculum pressure at early stage of growth and suggested occurrence resistance breakdown from generation to generation. Nigam and Bock (1990) studied the inheritance of resistance to chlorotic rosette (GRV and its sat-RNA) in crosses involving botanical varieties of groundnut in Malawi and confirmed the findings of de Berchoux (1960) of two recessive genes governing the resistance in all the backgrounds. In resistant plants, the presence of GRAV was detected. Gene conferring resistance to GRV and its sat-RNA did not confer resistance to GRAV (Bock and Nigam, 1988; Bock et al., 1990). Similar findings on the inheritance of resistance to green rosette using mixed infection in the field (GRV + and its sat-RNA + GRAV) and single GRV infection under greenhouse conditions were reported from Nigeria by Olorunju et al. (1992). There was one exception from RMP12 x M124.781 crosses, where in F2 generation, the plant segregated into 1 susceptible: 3 resistant, suggesting dominant gene action governing rosette resistance (Olorunju et al., 1992). Amin (1985) reported high level of resistance to A. cracivora in some crosses under greenhouse conditions. Progenies of A. chacoense and A. villas interspecific derivatives with cultivated groundnut also showed high resistance to A. crracivora. Resistance to aphid vector identified in cultivated groundnut ICG 5240 [EC36892] (Padgham et al., 1990) is reported to be controlled by single a recessive gene (van de Merwe, 2001; Herselman et al., 2004) 2.6 Combining Ability for traits in Groundnut Cultivar improvement for yield and stress resistance requires availability of genetic resources that could act as sources of genes conferring the desire traits that could be introgressed into the present cultivars (Sitaresmi et al., 2010; Nsabiyera et al., 2013). Gene introgression could be achieved through a combination of desired traits into target genotypes using recombination breeding under local conditions. This is essential to the generation of genetic diversity and University of Ghana http://ugspace.ug.edu.gh 18 fixing genes in the progeny (Marame et al., 2009; Zecevic et al., 2011). This, however, involves a lengthy and costly process of identifying and combining superior parents into superior hybrids (Rego et al., 2009). Diallel mating systems provide plant breeders with estimates for general combining ability (GCA) and specific combining ability (SCA). The GCA effects reflect the parent‘s genetic ability to influence all of its progeny for a specific trait, which is an expression of additive genetic effects (Griffing, 1956). The SCA effects represent non-additive genetic effects such as intra-allelic (dominance) or inter-allelic (epistasis) interactions or multiplicative gene action, which can be viewed as a departure from performance, can be predicted in simple additive models (Henderson, 1952; Griffing, 1956). Breeders have largely used the diallel mating scheme to estimate the potential value of genotypes per se, their combining ability effects for resistance to foliar disease in groundnut from a fixed or randomly chosen set of parental lines (Adamu et al.,2008). The studies of combining ability provide a guideline for selecting elite parents or combiners which may later be hybridized to accumulate fixable genes through selection. Both SCA and GCA have been reported to be significant in conditioning resistance to foliar disease in groundnut (Vishnuvardhan et al., 2011). Pensuk et al. (2002) from a 6 x 6 diallel cross of resistance to peanut bud necrosis tospovirus (PBNV) reported highly significant GCA effects for PBNV incidence in F2 and F3 generations. SCA was also significant, but the relative contribution to variation among crosses was much less than those of GCA effects. In an earlier study, Anderson et al. (1990) reported significant GCA and SCA effects for peanut stripe virus (PStV) and rust incidence from a study of diallel in groundnut. Makne (1992) and Dwivedi et al. (1994) found significant SCA for seed weight per plant, number of pods per plant and pod weight per plant and concluded that these traits are controlled by non-additive gene action. Adamu et al. (2008) recommended that selection for pod yield and resistance to groundnut rosette disease should be done among progenies from RMP12/ICGV87281 and University of Ghana http://ugspace.ug.edu.gh 19 RMP12/ICGV87018 since they depicted best general combiners for these traits. He also suggested that the significance of SCA mean squares for some of the traits is an indication that non-additive gene effects played an important role in their inheritance. SCA mean square was much smaller than GCA mean squares, which indicates that additive genetic variance was more important than non-additive genetic variance for these traits. Studies on combining ability in F2 and F3 crosses of Spanish and Virginia groundnut have shown that GCA and SCA were significant for almost all traits (Ali, et al., 2001) with preponderance of SCA which implies that selection for pod yield would be more effective in later generations. However, greater magnitude of GCA effect over SCA have been reported (Dwivedi et al., 1998) indicating the importance of additive genetic variance over non-additive variance. From the available reports it is evident that information on the precise nature of genetic control of GRD in groundnut is still lacking. Appropriate experimental design that includes the GRD resistant lines should provide additional information on the gene action involved in the expression of resistance. The knowledge on combining ability and type of gene action responsible for regulation of expression of GRD would certainly help in planning appropriate breeding strategies. 2.7 Genotype x Environment Interaction (GEI) in groundnut To design an appropriate breeding program, it is important to know the proportion of phenotypic variation of a trait that is heritable (Kearsey and Pooni, 1996), since the efficiency of a selection program is mainly dependent on the magnitude of genetic variation and heritability of a trait (Falconer and Mackay, 1996). Apart from high haulm and kernel yield in groundnut, adaptation to specific environments has also been a major breeding goal for groundnut breeders. GEI is a major problem involving quantitative traits, complicates the interpretation of genetic experiments, makes predictions difficult, and reduces the efficiency University of Ghana http://ugspace.ug.edu.gh 20 of selection. For quantitative traits, this interaction can be caused by genotypic rank change or by changes in the absolute differences between genotypes without rank change (Cooper and DeLacy, 1994). Therefore, knowledge about the magnitude of GEI is important to develop cultivars with higher yields and stable performance over a wide range of environmental conditions. Studies and interpretations of GEI range from simple analysis of variance to more specific analyses of genotype performance (Amini et al., 2013). The existence of GEI in groundnut breeding has been reported by Bentur et al. (2004); Senapathi et al. (2004) and Hariprasanna et al. (2008). The expectation has been the identification of suitable genotypes having maximum GEI with moderate level of resistance or susceptible to disease would be of immense benefit to improve the production of groundnut (Mothilal et al., 2010). They further reported significant linear component of GEI for kernel yield and concluded that genotypes differed for their linear response to fluctuations in environments. The magnitude of variation due to environment for kernel yield was higher than G x E (linear) for the same trait which depicted the major part of the total variation and was considered a linear function of environment only (Mothilal et al., 2010). In an earlier study of G x E interaction for PBNV, Buiel et al. (1995) reported that Genotype x environment interaction variance was significant but small. The field resistance of the genotypes studied was equally effective in all environments. Selection in any of these environments may be possible, but is more effective in environments which are favorable for disease development. Yan and Kang (2003) described the different types of G x E interactions and highlighted the implications of these in plant breeding and crop production. Crossover interactions (change in rankings of varieties across environments) are of greatest interest to breeders as these directly affect genotype selection in specific environments. Consequently, promising selections in one environment may perform poorly in another. Such crossover interactions often compel breeders to implement multiple selection programs within industries based on the University of Ghana http://ugspace.ug.edu.gh 21 homogeneity of regions, thereby utilizing greater resources. Ignoring significant G x E in favour of resource savings can lead to reduced genetic gains from selection (Ramburan et al., 2011). Inaccurate characterization of genotype adaptability may lead to poor productivity in environments that interact negatively with specific genotypes and this has implications on industry sustainability. With regards to genetic gains from selection, large G x E interactions, as components of total phenotypic variance, affect heritability (proportion of total phenotypic variance that is due to genetic variance) negatively. The larger the G x E interaction component, the smaller the heritability estimate; thus, progress from selection would be reduced as well (Yan and Kang, 2003). 2.8 Genetic Diversity in Cultivated Groundnut Based on Molecular Markers DNA-based markers provide a reliable means for estimating the genetic relationships among genotypes or taxonomic groups as compared to the morphological markers (Sajib et al., 2012). Precise understanding of the degree of genetic relationships among genotypes, botanical varieties of peanut, and Arachis species could provide insights into the domestication and evolution of this crop. Furthermore, it would have a valuable impact on peanut improvement, through identification of appropriate parents, to ensure a broad genetic base by inter-variety and inter-species crosses. DNA-markers, such as, randomly amplified polymorphic DNA (RAPD), amplified fragment length polymorphism (AFLP), and simple sequence repeats (SSR) have been used for cultivar discrimination and to study the botanical relationships among the cultivated peanut varieties (Subramanian, et al., 2000; Raina et al., 2001; He and Prakash, 2001; He et al., 2003). AFLP and SSR techniques can be used to detect DNA polymorphism in the cultivated peanut (He and Prakash, 1997; Tang et al., 2007). AFLP and SSR are two powerful DNA fingerprinting techniques. A number of loci can be analysed in an experiment and there is a higher reproducibility of banding patterns by University of Ghana http://ugspace.ug.edu.gh 22 AFLP. SSR markers have several advantages over other molecular markers for their codominant inheritance, large number of alleles per locus, and abundance in genomes (Sajib et al., 2012). These characteristics have promoted the application of SSR as molecular markers in fingerprinting (Wang and Du, 2005), genome mapping (Yin et al., 2006), phylogenetic and genetic relationship studies (Duan et al., 2006), and marker assisted breeding (Yi et al., 2004) in many crops. However, there are few reports concerning SSR and AFLP for evaluation of genetic diversity and relationships among the Arachis species, and much remains to be discovered (Tang et al., 2007). He and Prakash (1997) reported considerable DNA polymorphism in A. hypogaea revealed by the AFLP approach, this assay has been used for molecular diversity studies in peanut by several researchers (Dwivedi et al., 2001; He et al., 2003; Jiang et al., 2007). Comparing SSR and AFLP primers Jiang et al. (2007) reported that SSR primers amplified 91 polymorphic loci in total with an average of 3.14 alleles per primer, and the AFLP primers amplified 72 polymorphic loci in total with an average of 2.25 alleles per primer. Four SSR primers (14H06, 7G02, 3A8, 16C6) and one AFLP primer (P1M62) were found to be most efficient in detecting diversity. They also noted that genetic distance between pairs of Bacteria Wilt (BW) genotypes ranged from 0.12 to 0.94 with an average of 0.53 in the SSR data and from 0.06 to 0.57 with an average of 0.25 in the AFLP data. The SSR-based estimates of the genetic distance were generally larger than that based on the AFLP data. The genotypes belonging to subsp. fastigiata possessed wider diversity than that of subsp. hypogaea. The clustering of genotypes based on the SSR and AFLP data were similar but the SSR clustering was more consistent with morphological classification of A. hypogaea (Jiang et al., 2007). Optimum diverse genotypes of both subsp. hypogaea and subsp. fastigiata can be recommended based on this analysis for developing mapping populations and breeding for high yielding and resistant cultivars. University of Ghana http://ugspace.ug.edu.gh 23 In a study of Phylogenetic Relationships in Genus Arachis based on SSR and AFLP markers, Tang et al. (2008) found genetic distance detected by the SSR markers ranged from 0.09 to 0.95, and the mean was 0.73; and the genetic distance detected by the AFLP markers ranged from 0.01 to 0.79 with an average of 0.42. They also reported that in all the tested BW resistant peanut genotypes, SSR primer pairs were multilocus ones, and the amplified fragments per SSR marker in each peanut genome ranged from 2 to 15 with a mean of 4.77. The peanut cultivars were closely related to each other, and shared a large number of SSR and AFLP fragments. Jiang et al. (2007) partitioned the BW resistant peanut genotypes into two main groups and four subgroups at the molecular level, and that A. duranensis is one of the wild ancestors of A. hypogaea. The lowest genetic variation was detected between A. cardenasii and A. batizocoi, and the highest was detected between A. pintoi and the species in the section Arachis (Tang et al., 2007). Distinct clustering pattern of wild and cultivated genotypes was also reported in genetic diversity studies through SSR and EST – derived SRR maker systems (Moretzshon et al.,2005; Kottapalli et al., 2007; Koppolu et al., 2010). In a related study using single nucleotide polymorphism–based genetic diversity in the reference set of peanut (Arachis spp.) Khera et al. (2013) reported high level of diversity between wild and cultivated peanut and affirmed that grouping pattern exhibited discrete clustering of genotypes based on subspecies, botanical varieties and genome types. Mean genetic similarity between genotype pairs was found to be 0.13 and maximum between ICG 8200 and ICG 8206 at 0.4. They also reported the average major alleles was maximum in AA genome (0.81) and minimum in EE genome (0.56) while for BB and AABB genomes, it was found to be 0.71 and 0.63, respectively. The average PIC ranged from 0.21 (AA genome) to 0.38 (EE genome) while BB and AABB genomes recorded 0.31 and 0.32 respectively (Khera et al., 2013). The narrow genetic base variation observed in cultivated tetraploid groundnut may be attributed to its very recent origin in its evolutionary time as compared to other crops University of Ghana http://ugspace.ug.edu.gh 24 and is a serious genetic bottle neck towards modern breeding effort (Khera et al., 2013). Hence tapping the maximum genetic variation in the primary gene pool is vital to groundnut improvement. From the literature reviewed so far, the genetic background of parents in breeding programs is still narrow, which may have impeded the progress of breeding (Mace et al., 2007). Hence, a better understanding of the genetic diversity amongst the available GRD resistant germplasm is a prerequisite for further efficient improvement of GRD resistance. 2.9 Participatory breeding and varietal selection in groundnut Over the three past decades groundnut production in Nigeria has declined in importance both as food and cash crop for household and National economies (Ndjeunga et al.., 2010). Prospects for regaining production and market share lie in the adoption of improved varieties and crop management techniques that will significantly increase productivity and the improvement of quality standards. The key factors that constrain farmers‘ adoption of technologies are inappropriateness of the technologies, unavailability of required inputs and farmers‘ socio-economic conditions (KARI, 1996; Martins et al., 2002). Technologies that do not meet farmers‘ preferences, objectives and conditions are less likely to be adopted (Upton, 1987). During priority setting within the KARI-Kisii mandate region, groundnut was ranked fourth in importance for arid and semi-arid areas (Andima et al., 2006). Reasons for this included, lack of improved high yielding disease tolerant varieties, organized seed production system, poor agronomic practices, pests and diseases, low producer prices, lack of markets and market information and low adoption of developed technologies (Rachier et al., 2006, Okoko et al., 1998). The matrix ranking of the groundnut varieties conducted by 20 farmers in Suba District of Kenya indicated ICGV-SM 99568 was ranked first because it is early maturing, it is easy to shell, has tasty big seeds with good colour that has high market demand University of Ghana http://ugspace.ug.edu.gh 25 (Okoko et al., 2010). They further concluded that although ICGV-SM 12991 had the highest yield it was ranked second because of its small seed size leading to low marketability. Homabay local and ICGVSM 90704 were ranked last because of their poor germination, growth vigour, late maturing. In a PRA study conducted on Bambara groundnut Alhassan and Egbe (2013) indicated that more males (52.91%) than females (47.08%) were engaged in the production of bambara groundnut. This contrasted the works of Gibbon and Pain (1985) and Mkandawire and Sibuga (2002).These reports indicated that bambara groundnut production was done mainly by female subsistence farmers. Many men might have gone into the production of bambara groundnut because the crop fetches higher income now than it did previously (Tanimu and Aliyu, 1995). Alhassan and Egbe (2013) also indicated that 83.33% of farmers in the study area planted Bambara groundnut on ridges and 65.83% of farmers intercropped cowpea and cassava. The growing of crops in mixed cropping is consistent with the goal of food security (Alhassan and Egbe, 2013). However, apart from participatory variety selection of tropical legume II project by ICRISAT, there is no information at the Institute for Agricultural Research on PRA to assess farmers‘ preferences with the view of involving them in groundnut improvement program. Thus, there is need for concerted efforts to study the problems through research and social motivation for improving sustainability of cropping system and for meeting the challenges of low adoption of improved varieties. University of Ghana http://ugspace.ug.edu.gh 26 CHAPTER THREE 3 FARMERS’ PERCEPTION OF PRODUCTION CONSTRAINTS AND PREFERRED TRAITS FOR RESISTANT GROUNDNUT ROSETTE VARIETIES 3.1 Introduction The groundnut improvement programme at the Institute for Agricultural Research, Samaru is currently developing high yielding, GRD resistant groundnut varieties that are acceptable to farmers using the participatory variety selection approach. However, this conventional breeding procedure has been cited to be more formal concentrating on researchers‘ objectives of solving problems without considering farmers‘ preferences and opinion (Assefa et al., 2005). The consequences of neglecting farmers in setting up research and policy agenda are well documented (Gupta and Lagoke, 1999; Bänziger and Cooper, 2001; Snapp, 2002; Danial, 2003; Kamara et al., 2006; Derera et al., 2006; Ceccarelli and Grando, 2007). It is important for plant breeders to understand how and why farmers choose varieties of their crops, because this will ultimately determine whether a new or improved variety will be useful. Understanding farmers‘ choice and selection practices, their knowledge and goals underlying them, and the similarities and differences with plant breeders provides a means for the two groups to work together more effectively. This understanding and collaboration is critical for supporting long-term global food security (Makanda et al., 2009). In order to ensure sustainable groundnut production, there is a need for combining farmers‘ and researchers‘ objectives. These combinations have significantly contributed to agricultural development (Bucheyeki et al., 2011). Gathering groundnut production constraints and farmers‘ varietal preferences with the view to incorporating them into breeding objective was expected to contribute to increased rate of adoption, improved food security, and reduced poverty. The objectives of the participatory rural appraisal (PRA) were to: University of Ghana http://ugspace.ug.edu.gh 27 a. identify groundnut production constraints. b. assess farmers‘ knowledge of groundnut rosette disease c. appraise farmers‘ preference for rosette resistant varieties. 3.2 Materials and Methods 3.2.1 Description of the study areas The study was conducted in Batsari Local Government Area (LGA) of Katsina State (12°45 ′ 10N 7° 14 ′ 31). Batsari LGA occupies a land area of 1,107 km 2 with population of 208, 978 people (Censor, 2006). Major Socio – economic activities of the inhabitants of this area are farming and livestock keeping. Majority of them are Hausa and Hausa – Fulani. The second location was Nasarawa – eggon LGA of Nasarwa state (8° 32 ′ N, 8 0 17 ″ 58. 78 ′ E,). The LGA occupy land mass of 1,208 km 2 with population of 149, 129 inhabitants (Census, 2006). Major Socio – economic activities here are farming and trading. Majority of them are Eggon and Mada. Maps of these areas are shown in Figure 3.1. These areas represent the groundnut growing regions in sudano – Sahelian and Northern guinea savannah of Nigeria, respectively. These areas are characterised by mono-modal type of rainfall that falls between June and October in Batsari and April to October in Nasarawa eggon. University of Ghana http://ugspace.ug.edu.gh 28 Figure 3. 1:Map of Nigeria showing Batsari (Katsina state) and Nasrawa-Eggon (Nasarawa state) 3.2.2 Farmer survey and data analysis Preliminary visits were made to the two locations to discuss with farmers prior to the study. First visit was made on 15 th and16th January 2010 to Bastari in Kastina state which was followed by visit to Nasarawa Eggon (Nasarawa state) on the 21 st and 22 nd February 2010. The visit provided opportunities for informal interactions with groundnut farmers and processors. During these initial visits, secondary data on groundnut production and Nas-Eggon Batsari University of Ghana http://ugspace.ug.edu.gh 29 constraints were obtained from local extension officers. In addition, enumerators, who spoke the local languages, were identified, trained and made to pre-test the questionnaires. The farmers within villages were randomly selected at different strata. In order to obtain information on specific issues covered under the PRA, a formal survey was then conducted during January to March 2011 using a structured questionnaire, and other participatory rural appraisal tools including focus group discussions and observations made during transects walks across the areas. In both Batsari, and Nasarawa – eggon, 50 farmers were interviewed. Demographic information such as general household structure, education level, wealth status (as judged by property owned), cropping enterprises, and production constraints were obtained using the structured questionnaire (Appendix 1). Group discussions were done on completion of the questionnaire interviews to confirm data obtained and to solicit new information that was not captured during the formal process. Both qualitative and quantitative data were subjected to statistical analysis using the statistical Package for Social Science (SPSS) version 15 (SPSS Inc., Chicago IL). Frequencies were determined for quality questions. Associations and t – test for comparison were determined for quantitative variables. Graphs were used to present results. Preferred and unfavoured traits, as well as the importance and severity of rosette diseases were ranked to highlight farmers‘ perceptions. 3.3 Results 3.3.1 Household and demographic information Majority (36 %) of the farmers in Batsari were above 40 years while those of Nasarawa – eggon were younger (61 %) and ranged between 20 – 35 years (Table 3.1) and majority of them were married. The sex ratios of the household were not significantly different (p ≤ 0.05) University of Ghana http://ugspace.ug.edu.gh 30 in the two districts (t = 6.31, P =0.33). The number of farmers who had tertiary education or belonged to an association did not differ in the two districts. However, the age of farmers was significantly different and so were married and single farmers across the districts. The numbers of farmers with primary and secondary education were significantly different (Table 3.1). In both areas, at least 80% of the farmers had lived in a village for more than 10 years practicing small – scale farming and had acquired experience in farming. Ninety – eight per cent of the farmers at Batsari village were males and in Nasarawa – eggon 7% were female. The majority of farmers in Batsari village were illiterate while 50% of the farmers in Nasarawa – eggon were literate. A proportion of the farmers (8%) in Batsari and 32 % in Nasarawa – eggon had primary education (Table 3.1). Farmers belonging to associations were 12.8% (Batsari) and 9.1% in Nasarawa eggon. Farmers‘ experience (farming for more than 10 years) significantly (P < 0.05) correlated with level of education (r = 0.51517**) and level of awareness of rosette diseases (r = 0.45602*) (Table 3.2) University of Ghana http://ugspace.ug.edu.gh 31 Table 3. 1:Farmer and Household information for Batsari and Nasarawa – eggon LGA in Nigeria for the 2010 growing season Variable Local Government Area t – value Probability Batsari Nasarawa Average Gender of household head (%) (%) (%) 6.31 0.50 Male head 98 93 95 Female head 2 7 5 Age of farmers 1.90 0.49 20 – 25 0 18 9 26 – 30 0 23 12 31 – 35 0 20 10 36 – 40 16 14 15 41 – 45 36 14 25 46 – 50 24 7 16 51 – 55 16 5 11 > 56 8 0 4 Marital status 2.35 0.03 Married 100 88.6 94 ** Single 0 6.8 3 Widowed 0 1 1 Divorce 0 1 1 Over 10 year experience in farming 94 84 89 Level of education acquired 2.35 0.50 Illiterate 88 50 69 Primary 8 32 20 Secondary 4 16 10 Tertiary 0 2 1 Membership of association Member 12.8 9.1 11 6.31 0.33 Non member 76.2 90.9 84 * and ** significant and P < 0.05 and P < 0.01, respectively University of Ghana http://ugspace.ug.edu.gh 32 Table 3. 2: Pair wise correlation of some farmers’ level of awareness of the groundnut rosette disease Gender Age Marital Status Famer Experience Membership of association Level of education Rosette disease awareness Gender 0.202 0.992 ** 0.462 -0.897 ** 0.998 ** 0.999 ** Age 0.321 0.962 ** -0.615 * 0.262 0.196 Marital Status 0.568 -0.945 ** 0.998* 0.992 * Famer Experience -0.807 ** 0.515 ** 0.456 * Membership of association -0.922 -0.894 ** Level of education 0.562 * University of Ghana http://ugspace.ug.edu.gh 33 Maize, sorghum, groundnut, yam, cowpea and sesame were mentioned as the most important crops grown for cash income and sources of food security, while cassava, tomatoes, water melon, onion and pepper were less cultivated by the sampled farmers. Groundnut was ranked as the most important cash crop in both Batsari and Nasarawa – eggon areas (98% and 82%, respectively). While sesame was ranked as the second most important crop in Batsari (83.7%), yam was ranked the second in Nasasrawa – eggon (86.4%) (Table 3.3). Ranking of groundnut as the most important cash crop showed farmers‘ strong interest in the crop. This was probably because of the increasing demand for cultivation to target markets, as well as alleviating poverty and food shortage at household level. The results of the study showed that farmers produced groundnut in association with other crops especially maize and sorghum. Groundnut is intercropped by most farmers and with only few farmers growing it as a sole crop. University of Ghana http://ugspace.ug.edu.gh 34 Table 3. 3: Pair-wise ranking of major crops grown by farmers in Batsari and Nasarawa – eggon in 2011 Crop Districts Batsari (%) Nasarawa – eggon (%) Average (%) Cash crop Food crop Cash crop Food crop Rainy season Groundnut 98 (1) 40 (5) 86 (2) 43 (7) 92.0 Groundnut var: SAMNUT 21 and 23 Maiyado‖ RRB Cowpea 75.5 (3) 30 (6) 64.6 (5) 51 (5) 70.1 Maize 63.4 (4) 100 (1) 75.8 (3) 93 (1) 59.6 Millet 75.3 (3) 98 (2) 53.7 (7) 23 (8) 64.5 Sesame 83.7 (2) 32 (7) 89.4 (1) 47 (6) 86.6 Yam — 86.4 (2) 82 (3) 86.4 Sorghum 54.5 (5) 74 (3) 67.3 (4) 87 (2) 50.9 Cassava 25.2 (6) 51(4) 63.0 (6) 23 (8) 44.1 Rice — 58.9 (7) 58 (4) 58.9 Dry season Tomatoes 56.8 56.5 56.7 Onion 78.0 53.6 65.8 Water – melon 78.5 45.3 61.9 Pepper 75.0 34.3 54.7 The figures are percentage responses, number in parenthesis = rank and — = crop not reported University of Ghana http://ugspace.ug.edu.gh 35 3.3.2 Characteristics of preferred groundnut varieties in Batsari and Nasarawa – eggon districts The choice of a groundnut variety in rural areas is determined by some characteristics of the plant and the environment. Although farmers‘ criteria in choosing varieties were similar across the two study areas, there were marked differences in the characteristics of the varieties preferred by farmers. These differences varied from site to site (Fig. 3.3). Most of the major traits preferred were those associated with yield, market value and those that enabled the crop to escape or produce yield, even when attacked by pests and diseases. Pest and disease tolerance and high oil content (Fig. 3.3) were considered to be the most important desired characteristics in Batsari and Nasarwa – eggon with 21% and 28% of the respondents respectively while early maturity was most important in Batsari. The other important traits preferred by famers in both locations were the pod yield. In Batsari, 14% of the interviewed farmers considered haulm to be an important trait. This is not quite an important trait in Nasarawa – eggon as only 6% use haulm. Figure 3. 2:Distribution of traits preferred by farmers University of Ghana http://ugspace.ug.edu.gh 36 The results revealed that farmers were aware of groundnut rosette disease which was commonly known by various names such as ―Kuturtan gyada‖ emphasizing the predominance of the disease in the area. The majority of the farmers reported the disease to be associated with insects and few of them recognized aphid as being responsible. Eighteen per cent of the respondents associated the disease with inadequate rainfall (Fig. 3.4). Figure 3. 3:Farmers perception on the causes of groundnut rosette disease Most farmers recorded moderate (20 – 40%) to high (50 %) yield loss due to GRD (Fig. 3.5). A total yield loss (100 %) had been experienced by some farmers University of Ghana http://ugspace.ug.edu.gh 37 Figure 3. 4:Farmers perception of yield loss due to groundnut rosette disease About 34 – 40 % of the respondents in both Batsari and Nasarawa – eggon did nothing to combat the menace of groundnut rosette disease, while many other farmers rogued out infected plants with some utilizing rosette resistant groundnut varieties (Figure 3.6) Figure 3. 5: Groundnut rosette disease control measures adopted by farmers University of Ghana http://ugspace.ug.edu.gh 38 3.3.3 Preferred groundnut varieties and associated characteristics Farmers indicated that they selected groundnut varieties for commercial production based on consumer preference. Farmers also chose groundnut varieties for production on the basis of potential pod and haulm yield, oil quality and content, and market price. For instance, in Nasarawa – eggon, farmers ranked ‗Maiyado‘ as the most preferred variety because of market demand, and high yield potential, although it is susceptible to groundnut rosette diseases and other foliar diseases, while ‗SAMNUT21‘ and ‗SAMNUT23‘ varieties were preferred in Batsari because of high pod yield, earliness, seed colour, market acceptance and tolerance to pest and foliar diseases. 3.3.4 Perception of farmers on constraints to groundnut production Several constraints were mentioned and ranked by the farmers (Table 3.4). The most important constraints reported were pests and diseases and poor quality seeds and drought. The farmers recounted these limitations to severely reduced yields. Price fluctuation for groundnut was reported to constitute a problem. Whenever there was a bumper harvest groundnut prices dropped, so they kept their produce in storage, until such a time that the price increased in the off-season. Farmers differentiated between threats due to diseases from those caused by weeds and drought. The ranking of the frequent diseases or pests across villages revealed that GRD was the most threatening constraint. The symptoms observed by respondents to describe GRD were stunting, bushy and yellowing of leaves, where groundnut plants were not suffering from water shortage. University of Ghana http://ugspace.ug.edu.gh 39 Table 3. 4:Pair-wise ranking of the most important constraints in groundnut production in Batsari and Nasarawa – eggon Constraint Score by farmers Batsari Nass – eggon Total Score Ranking Drought 2 4 6 3 Pest and Diseases 1 2 3 1 Weeds 3 4 7 4 High cost of insecticides 5 3 8 5 Poor quality seed 2 1 3 1 Price fluctuation 3 2 5 2 1 = very serious problem and 5 = minor 3.4 Discussion A participatory rural appraisal was conducted to understand farmers‘ production systems and perceptions on groundnut rosette diseases across two locations in Nigeria. The PRA helped to obtain information on auxiliary data on socio – economic aspect of farmers. Farmers faced several constraints from seed through crop production, crop protection and marketing in groundnut production. The majority of farmers were males and had been farming for more than 10 years. Most of them were in the age range of 25 – 30 years in Nasarawa-eggon but older in Batsari. The major source of income was from crop growing that accounted for more than 80% of their household income However, most farmers recorded low yield of groundnut crop owing to several constraints that called for intervention and strategies to enhanced productivity. University of Ghana http://ugspace.ug.edu.gh 40 Most farmers at Batsari location were aware of improved groundnut varieties and some grow them together with their landrace varieties. Nasarawa – eggon completely relies on their landrace variety ‗Maiyado‘. Reasons for poor adoption rate of improved varieties could be due to limitations of certified seeds, high prices and inadequate information on the improved varieties. The seeds of landrace varieties with farmers‘ preference traits were sold at reasonably affordable prices. The farmers faced similar groundnut production constraint across Batsarit and Nasarawa – eggon despite differences in geographical locations. The major challenges were poor quality seeds and prevalence of groundnut rosette disease. The findings from this study further showed that groundnut rosette disease was probably associated with insects and drought. Adu – Dapaah et al. (2004) associated insect and drought as the favourable conditions for groundnut rosette disease from the PRA studies of groundnut. The Agricultural Transformation Agenda (ATA) programme of the Federal Government of Nigeria was designed to enhance the livelihood of farming community through improving productivity and hence raising their income through supply of agricultural inputs; fertilizers, good quality seeds and credit facility. These did not reach resource poor farmers because of the requirements that include detailed information on farming activities, and marketing. For farmers with limited education and understanding of the process, the requirements are unobtainable. Furthermore, majority of farmers were not members of any cooperative society. To derive maximum benefit from the ATA programme, the Federal government should improve the extension services to facilitate the formation of farmers‘ cooperative societies at grass root levels for coordinated agricultural activities. University of Ghana http://ugspace.ug.edu.gh 41 Genotype plays a very significant role in achieving higher productivity. In general, across the locations it was noticed that there was no efficient seed system or replacement mechanism for penetration of improved cultivars of groundnut. Most of the farmers use very old landrace ‗Maiyado‘, demonstrating the poor rate of seed replacement in these parts of the country. The risk taking ability and openness of Batsari‘s farmers to new technologies (SAMNUT21 and SAMNUT23) made a big difference to their achieving high productivities, an approach that was relatively lacking in Nasarawa - eggon. ‗Maiyado‘ was preferred by majority of farmers at both locations because of its colour, high oil contents and resistance to foliar disease. These are traits that most farmers and consumers look for in groundnut (Ndjuenga et al., 2010). Until recently, researchers at IAR/ICRISAT focused mainly on earliness, yield and resistance to foliar diseases (Olorunju et al., 2001) to improve groundnut production in sub Saharan Africa. The new focus is involvement of farmers through PRA to final production of improved seed which will ultimately, enhance rapid adoption. This agrees with the findings of Nkonya and Featherstone (2001) who found that varieties with farmers preferred traits were easily adopted. This was evident with SAMNUT21 and SAMNUT23 with a high adoption rate (> 70%) by farmers at Katsina state, Kano, Jigawa and Kaduna states because the background parents of these varieties are local varieties with farmers preferred traits (Ndjeunga et al., 2003). However, the nonavailability of the two varieties in Nasarawa state, compelled the farmers to use the landrace they have and available in market as recycle seeds. The high price of certified seed if available was another reason for low adoption. University of Ghana http://ugspace.ug.edu.gh 42 3.5 Conclusions and Recommendations This study identifies groundnut varieties grown by the farmers, criteria for choice of the varieties and constraints in production, thus providing the basis for formulation of farmer-oriented groundnut breeding programme. Farmers have diverse perceptions and complex combinations of criteria they use in selecting groundnut varieties. The key criteria include high yields, early maturity, tolerance to groundnut rosette disease drought and insect pests. Groundnut production in both Batsari and Nasarawa-Eggon is constrained by related factors. The most important constraints perceived by farmers are pest and rosette diseases, poor quality seeds and drought. Farmers in both locations were aware of groundnut rosette disease as it is called by various local names. For instance ―Kutrtan- gyada‖ was the name given to GRD in Batsari. To increase groundnut production, research should take into consideration the farmers‘ circumstances and preferences and develop varieties and crop management packages meet farmers demands. Incorporation of farmers‘ preferences in selection of groundnut varieties in breeding process would increase likelihood of adoption of the varieties. Whereas groundnut breeding cannot incorporate all the desired attributes, the key attributes should be included in particular varieties and many varieties should be bred focusing the demands of different groups of farmers. Considering that farmers prefer saved seeds of local varieties as a strategy for coping with cash flow constraints, effort should be made to breed varieties that are resistant to insect pests and disease. Such varieties are likely to be highly adopted by smallholder farmers, especially when the other key criteria they apply in variety selection are also incorporated. University of Ghana http://ugspace.ug.edu.gh 43 CHAPTER FOUR 4 ASSESSMENT OF GENETIC DIVERSITY OF GROUNDNUT (ARACHIS HYPOGAEA L.) GENOTYPES FOR RESISTANCE TO ROSETTE DISEASE USING SSR MARKERS 4.1 Introduction Breeding for foliar disease resistant genotypes is the ideal solution for reducing the crop losses. Identification and utilization of a broad spectrum of genetically diverse sources of GRD resistance is critical for the development of a new generation of broad-based high-yielding GRD- resistant groundnut cultivars. Limited knowledge about the genetic diversity of the GRD- resistant germplasm and deleterious linkage drag has impeded the utilization of a wide spectrum of GRD resistance donors. Diversity studies in groundnut have generally revealed extensive phenotypic variation amongst varieties (Upadhyaya al., 2001, 2003) yet limited variation at the molecular level (Subramanian et al., 2000; Moretzsohn et al., 2004). Several approaches including molecular (Jiang et al., 2007; Milla-Lewis et al., 2010; Khera et al., 2013) and morphological characterization have been used in assessing the genetic diversity of groundnut germplasm but results of morphological characterization are highly influenced by environmental factors (Shoba et al., 2010). Molecular marker technologies are playing an increasingly important role in conservation and use of plant genetic resources in plant breeding programmes (Varshney et al., 2009). Among the DNA markers, simple sequence repeat (SSR) markers are more preferable as it is more variable within genomes than other marker types (Belaj et al., 2003). Additionally, SSRs have the advantage of being co-dominant, only requiring very small amounts of DNA and hence have been widely applied in many plant genetics studies, e.g. for evaluating genetic diversity (Zhebentyayaeva et al., 2003; Fahima et al., 1998). University of Ghana http://ugspace.ug.edu.gh 44 In Nigeria, extensive efforts have been made in groundnut breeding for GRD-resistance and several resistant cultivars have been released. However, these cultivars have been released 10 years ago have begun to show resistance breaking that is influenced by genetic variability in the pathogen population (Legrève and Duveiller, 2010). Only a few sources of GRD-resistance have been successfully used in breeding programs at the Institute for Agricultural Research (IAR) even though several resistant genotypes are available (Olorunju et al., 2001). Most GRD- resistant cultivars released in IAR are based on just three sources of resistance (RMP12, MDR-8- 19 and UGA2). Obviously, the genetic background of parents in IAR groundnut breeding programs is still narrow, which may have impeded the progress of breeding. Therefore, a better understanding of the genetic diversity amongst GRD-resistant germplasm is a prerequisite for further efficient improvement of GRD-resistance. The objectives of the present study are to use SSR markers to detect DNA polymorphism among cultivated groundnut genotypes with differential levels of GRD resistance and for selecting parents for further breeding programmes. 4.2 Materials and methods 4.2.1 Plant material and DNA extraction Fifty groundnut genotypes obtained from the IAR and International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) Mali, consisting of aphid and rosette resistant genotypes were assayed in this study (Table 4.1). Total genomic DNA was isolated from young leaves of 15 – 20 days old seedlings. Each sample consisted of about 5g of leaves pooled from 2-3 seedlings and DNA was extracted using a CTAB-based procedure, with 3% (v/v) b-mercaptoethanol in a 3% (w/v) CTAB buffer (Mace et al., 2003). The quantity and quality of DNA were determined University of Ghana http://ugspace.ug.edu.gh 45 electrophoretically through comparison with known concentrations of uncut l DNA standards and spectrophotometric analysis at 260/280nm, and subsequently diluted to 5ng/ml. Laboratory analysis was done at Generation Challenge Program (GCG) Kenya between August, 2012 – November 2012. 4.2.1 SSR Analysis Forty SSR primer pairs (Table 4.2) were used to amplify the genomic DNAs. PCR reactions were carried out in 10μL reaction volume using a GeneAmp PCR System 9700 (Applied Biosystems). The PCR reaction mixtures contained between 5 and 15ng of genomic DNA, 10–30 pmol of each primer, 100–125mM of dNTP, 0.6–1.2U/ml of Taq DNA polymerase (Amersham), 1 PCR buffer (10mM Tris–HCl pH 8.3, 50mM KCl) and 0.5–2.5mM MgCl2. The fixed- temperature PCR programmes consisted of an initial denaturation step for 2 min