University of Ghana http://ugspace.ug.edu.gh IDENTIFICATION OF SOURCES AND LOCI FOR DEVELOPMENT OF SOYABEAN [Glycine max (L.) Merrill] CULTIVARS THAT COMBINE BACTERIAL PUSTULE AND SEED DECAY RESISTANCE WITH HIGH POD CLEARANCE BY MICHAEL TEYE BARNOR (10082535) THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE AWARD OF DOCTOR OF PHILOSOPHY DEGREE IN PLANT BREEDING WEST AFRICA CENTRE FOR CROP IMPROVEMENT COLLEGE OF BASIC AND APPLLIED SCIENCES UNIVERSITY OF GHANA LEGON DECEMBER 2018 i University of Ghana http://ugspace.ug.edu.gh DECLARATION I do hereby declare that except for research works that have been duly cited and referenced, this work is my original research and that neither part nor whole has been presented elsewhere for the award of a degree. …………………………………. MICHAEL TEYE BARNOR STUDENT ……………………………… PROF. KWADWO OFORI SUPERVISOR ………………………………… DR. BEATRICE ELOHOR IFIE SUPERVISOR ………………………………… DR. AGYEMANG DANQUAH SUPERVISOR ………………………………… DR. NICHOLAS NINJU DENWAR SUPERVISOR ii University of Ghana http://ugspace.ug.edu.gh ABSTRACT Globally, diseases constitute a major biotic constraint to soyabean production. Although, the list of diseases of soyabean in Ghana has been published, the incidence, severity, and distribution of listed diseases were not indicated. Secondly, the Savannah Agriculture Research Institute (SARI) has accumulated considerable number of soyabean genotypes most of which were sourced from the International Institute of Tropical Agriculture (IITA) and the United States Department of Agriculture (USDA). Efficient use of germplasm depends on knowledge of variation that exists among genotypes for traits of interest coupled with an understanding of genetic diversity. Therefore, it was imperative to morphologically and genetically characterise genotypes in SARI’s germplasm to guide breeding efforts. Lastly, it is important to consider the architecture of released varieties in relation to farming systems including commercial farming. Most of the released soyabean varieties in Ghana belong to the early and medium maturing groups with low pod clearance. Low pod clearance is not a problem with subsistence farming systems where harvesting is manually done, but in commercial farms that use combine harvesters, low pod clearance could be the source of considerable losses. Detailed assessment of soyabean disease situation at production centres and detailed characterisation of genotypes in SARI’s soyabean germplasm have been carried out to identify sources and loci of resistance to predominant diseases of soyabean in Ghana. The objective of this study were to: (1) carry out a survey of soyabean diseases across major production centres in Ghana and their incidence and severity, (2) evaluate SARI’s soyabean germplasm for resistance to bacterial leaf pustule, (3) evaluate SARI’s soyabean germplasm for resistance to soyabean seed decay, (4) assess diversity, population structure and identify lines with high pod clearance among genotypes in SARI’s germplasm, (5) identify Phomopsis seed decay (PSD) and bacterial leaf pustule (BLP) resistant loci through genome-wide association mapping, (6 ) identify loci associated with pod clearance. iii University of Ghana http://ugspace.ug.edu.gh Results of the survey revealed that some of the important diseases of soyabean in Ghana included Cercospora leaf blight (Cercospora kikuuchi), frog eye leaf spot (Cercospora sojina); target spot Corynespora cassiicola), bacterial leaf pustule (Xanthomonas campestris pv. glycine), downy mildew (Peronospor manshurica), leaf blight (Rhizoctonia solani), brown spot (Septoria glycine), soyabean mosaic, alfafa mosaic, and soyabean seed decay due to Diaporthe sp. Bacterial pustule was the most wide spread with incidence ranging from 15.3% at Wenchi to 84.7% at Malzire in the Yendi municipality. Two viral diseases, soyabean mosaic virus and alfalfa mosaic virus were present at all 15 locations but mostly at very low incidence except for Ejura and Wenchi where incidence was very high. Cercospora leaf blight and Frog eye leaf spot both occurred at 13 of 15 locations surveyed. Generally, incidence and severity of identified diseases was higher at Malzire, Karaga and Gushegu than other locations with the exception of soyabean mosaic virus for which the highest incidence occurred at Ejura followed by Wenchi. Furthermore, this study identified soyabean seed decay disease, and molecularly identified the causal agent as Diaporthe phaeolorum var. sojae and tentatively designated the isolate as MTB-1. This is the first report of Diaporthe phaeolorum var. sojae occurrence on soyabean pods and seeds in Ghana. Twenty-eight lines were identified with immune response to Xanthomonas axonopodis pv. glycines which could be used to develop resistant varieties. Area under chlorophyll retention curve had a negative correlation with area under disease progress curve, which translates to a positive correlation with resistance to bacterial pustule. Area under chlorophyll retention curve could replace disease severity score in bacterial pustule screening trials. Though disease pressure was high, eleven lines, PI416806B, PI224271, PI417120 BRS 361, PI417013, TGX 1903-7F, PI594767B, PI423958, TGX 1485-1D, TGX 2006-3F and Liu Yuemang were highly resistant to PSD infection. Genotypes within SARI’s germplasm differed widely for growth and reproductive parameters including plant height, number of branches, pod clearance, flowering and maturity periods, all of which were highly iv University of Ghana http://ugspace.ug.edu.gh heritable except for number of branches. Likewise, high variation among genotypes for yield components, most of which were under genetic control was recorded. Principal component analyses identified number of pods per plants, number of seeds per plant, days to flowering and days to maturity as the main drivers of variation in SARI’s germplasm. Genetic diversity analyses of a 96 representative panel showed high diversity among genotypes. Structure analyses revealed nine sub-populations within the panel. Genetic variation among sub- populations was high as indicated by high allele divergence frequency among sub-population. All genotypes in the 96 representative panel were distinct and most were almost fixed as reflected in high Fst values. Genome-wide association studies identified two SNP markers, S3_14585048 and S3_1991311 with large negative effect on pod clearance. SAUR like-auxin responsive protein known to positively regulate cell expansion to promote hypocotyl growth was identified at the physical position of the SNPs. Associations between genomic locations and resistance to bacterial leaf pustule were identified but the associations were not significant because of low statistical power due to small sample size. Nonetheless, three SNP markers, S16_34242886, S16_35202484 and S1_38672307 were found linked to three candidate genes, Glyma.01go93200, Glyma.g187500 and Glyma.g187400. All three candidate genes code for Leucine-rich repeat receptor like protein kinases known for eliciting immune response to pathogen infection. The SNP, S18_7792728 associated with PSD resistance occurred at a genomic region that harbours the candidate gene AT3G07040.1 (NBARC-domain-containing disease resistance protein) known to play critical roles in plants response to pathogens. This confirms the existence of resistant genes for PSD among the lines in the representative panel. The identified SNP could be validated to aid with marker-assisted selection. This study presents a series of useful information on SARI’s soyabean germplasm, prevailing diseases of soyabean and their distribution, sources of resistance and resistant loci, and the discovery of informative markers that can alter the architecture of future varieties. v University of Ghana http://ugspace.ug.edu.gh DEDICATION This work is dedicated to the rock of my life, my mother, Mary Koryoe Barnor, and to uncle David Akornor and Maafio Marku for their generosity, and to my wife and son, Thamar and Nattey for their extreme patience and support. vi University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENT I am most grateful to Professors Eric Danquah, Kwadwo Ofori, Kwame Offei and E.T. Blay. My life has in one way or the other evolved around you since I joined the faculty of Agriculture in 2001. I am also very grateful to Drs Agyemang Danquah and John Eleblu, classmates who became my friends and then my lecturers. Thank you very much for encouraging me to pursue PhD in Plant breeding. Dr. Beatrice Ife Elehor, I am most grateful for your directives and support. I am also thankful to Dr. Nicholas Denwar, my in-country supervisor for hosting me, my field research and making enormous input into my research. Fredrick Awuku, Baba Kassim and Salim Lamini, I am grateful for your support during my research. My friends, Dr. Richard Oteng Frimpong and Dr. Doris Puoza, I am most grateful to you for making my stay in Tamale a very unique one. Dr. Francis Kwame Padi, I am most grateful to you for ensuring that work on shea continued unabated. My cousins, Eugenia Akornor, Tetteh Akornor, and Eugene Nartey Akornor, I am very grateful for your support throughout the over two decades that we have been together. vii University of Ghana http://ugspace.ug.edu.gh Table of Contents LIST OF TABLES ................................................................................................................................. xi LIST OF FIGURES .............................................................................................................................. xii CHAPTER ONE ..................................................................................................................................... 1 1.0 GENERAL INTRODUCTION ........................................................................................................ 1 2.0 LITERATURE REVIEW ................................................................................................................. 5 2.1 Origin and botany of cultivated soyabean ........................................................................................ 5 2.2 Global distribution and production of soyabean ............................................................................... 6 2.3 Soyabean production in sub-Saharan Africa .................................................................................... 7 2.4 Soyabean production in Ghana ......................................................................................................... 9 2.5 Constraints to soyabean production ................................................................................................ 11 2.5.1 Drought as a production constraint of soyabean .......................................................................... 12 2.5.2 Salinity as constraint to soyabean production .............................................................................. 13 2.6 Important diseases limiting soyabean production ........................................................................... 13 2.6.1 Soyabean seed decay .................................................................................................................... 15 2.6.2 Pathogens within the Phomopsis- Diaporthe complex (DPC) of fungi ....................................... 17 2.6.3 Management of Phomopsis seed decay of soyabean (PSD) ........................................................ 18 2.6.3.1 Cultural approaches to PSD control.......................................................................................... 18 2.6.3.2 Biological control of PSD ......................................................................................................... 19 2.6.3.3 Chemical control ....................................................................................................................... 19 2.6.3.4 The use of resistant genes ......................................................................................................... 19 2.6.3.5 Markers linked to PSD resistant loci ........................................................................................ 20 2.6.3.6 Inheritance of PSD .................................................................................................................... 21 2.6.4 Bacterial leaf pustule .................................................................................................................... 21 2.6.4.1 Management of bacterial leaf pustule ....................................................................................... 22 2.6.4.2 Molecular resources available for management of bacterial leaf pustule ................................. 23 CHAPTER THREE .............................................................................................................................. 26 3.1 Introduction .................................................................................................................................... 26 3.2 Materials and Method ..................................................................................................................... 28 3.2.1 Study area and time ...................................................................................................................... 28 3.2.2 Field selection .............................................................................................................................. 29 3.2.3 Identification of diseases on selected fields ................................................................................. 29 3.2.4 Sampling method ......................................................................................................................... 29 3.3 Confirmation of Diaporthe sp. causing soyabean seed decay in Ghana ......................................... 30 3.3.1 Sampling of diseased tissues or organs ........................................................................................ 30 3.3.2 Isolation of disease-causing organism ......................................................................................... 31 3.3.3 Microscopy .................................................................................................................................. 31 3.3.4 Primer Design .............................................................................................................................. 32 3.3.6 DNA quality test .......................................................................................................................... 32 viii University of Ghana http://ugspace.ug.edu.gh 3.3.7 PCR amplification and gel electrophoresis .................................................................................. 33 3.3.8 Polyacrylamide gel electrophoreses of amplified DNA ............................................................... 33 3.3.9 Sequencing of amplified DNA products ...................................................................................... 33 3.3.10 Data collection and statistical analysis ....................................................................................... 34 3.4 Results ............................................................................................................................................ 34 3.4.1 Results of disease survey ............................................................................................................. 34 3.4.2 Results of pathogen confirmation study ....................................... Error! Bookmark not defined. 3.4.2.1 Colony characteristics ............................................................... Error! Bookmark not defined. 3.4.2.2 Molecular identification of causal pathogen ............................................................................. 38 3.5 Discussion ....................................................................................................................................... 42 3.6 Conclusions .................................................................................................................................... 47 CHAPTER FOUR ................................................................................................................................ 49 4.0 Genetic diversity, population structure, and identification of key phenotypic traits driving variation in the assembled soyabean germplasm .................................................................................. 49 4.1 Introduction .................................................................................................................................... 49 4.2 Materials and Methods ................................................................................................................... 52 4.2.1 Plant materials .............................................................................................................................. 52 4.2.2 Study area..................................................................................................................................... 52 4.2.3 Experimental design and field establishment ............................................................................... 52 4.2.4 Data Collection and Analysis of morphological and phenotypic traits ........................................ 53 4.3 Sampling of plant tissue for genotyping-by-sequencing ................................................................ 54 4.3.2 Dendometric analyses .................................................................................................................. 56 4.4 Results ............................................................................................................................................ 56 4.4.2 Multivariate analysis of eight quantitative traits .......................................................................... 62 4.5 Discussion ....................................................................................................................................... 76 4.5 Conclusion ...................................................................................................................................... 81 CHAPTER FIVE .................................................................................................................................. 83 5.0 Field evaluation of soyabean lines for identification of sources of resistance to bacterial leaf pustule .................................................................................................................................................. 83 5.1 Introduction .................................................................................................................................... 83 5.2 Materials and Methods ................................................................................................................... 86 5.2.1 Materials ...................................................................................................................................... 86 5.2.2 The study area .............................................................................................................................. 86 5.2.3 Experimental design and field establishment ............................................................................... 86 5.2.4 Data collection and analysis ......................................................................................................... 87 5.3 Results ............................................................................................................................................ 89 5.4 Discussion ....................................................................................................................................... 94 5.5 Conclusion ...................................................................................................................................... 99 CHAPTER SIX ................................................................................................................................... 100 6.0 Field evaluation of soyabean genotypes for resistance to Phomopsis Seed decay (PSD) ............ 100 ix University of Ghana http://ugspace.ug.edu.gh 6.1 Introduction .................................................................................................................................. 100 6.2 Materials and Methods ................................................................................................................. 103 6.2.1 Plant materials ............................................................................................................................ 103 6.2.2 Experimental design and field establishment. ........................................................................... 103 6.2.3 Culturing of pathogen to obtain inoculum ................................................................................. 103 6.2.4 Preparation of inoculum and inoculation ................................................................................... 104 6.2.5 Data collection and analyses ...................................................................................................... 104 6.2.6 Plating of seeds and estimation of infection rate ....................................................................... 105 6.3 Result ............................................................................................................................................ 106 6.4 Discussions ................................................................................................................................... 111 6.5 Conclusions .................................................................................................................................. 113 CHAPTER SEVEN ............................................................................................................................ 114 7.0 Genotyping by sequencing, and genome-wide association studies of PSD and BP resistance, and high PLP. ............................................................................................................................................ 114 7.1 Introduction .................................................................................................................................. 114 7.2 Materials and Methods ................................................................................................................. 117 7.2.1 Genotyping-by-sequencing 96 sub-sample of 250 lines ............................................................ 117 7.2.2 Plant materials ............................................................................................................................ 117 7.2.3 Sampling of plant tissue for genotyping-by-sequencing ............................................................ 117 7.2.5 Phenotypic data .......................................................................................................................... 118 7.2.6 Data analysis .............................................................................................................................. 118 7.3 Results ........................................................................................................................................... 119 7.3.1 General properties of SNP markers and individuals in the representative panel ....................... 119 7.3.2 Discovery of SNP markers associated with PLP ....................................................................... 123 7.3.3 Genome-wide scanning for loci associated with bacterial leaf pustule ..................................... 129 7.3.4 Genome-wide scanning for loci associated with PSD resistance ............................................... 132 7.4 Discussion ..................................................................................................................................... 135 7.4.1 Characteristics of samples and GBS generated SNP markers.................................................... 135 7.4.2. Associations studies involving PLP .......................................................................................... 136 7.4.3 Loci associated with bacterial leaf pustule ................................................................................ 138 7.5. Conclusion ................................................................................................................................... 141 CHAPTER EIGHT ............................................................................................................................. 142 8.0 General conclusion and recommendation ..................................................................................... 142 8.1Conclusion ..................................................................................................................................... 142 8.2 Recommendation .......................................................................................................................... 145 BIBLIOGRAPHY AND APPENDICES ............................................................................................ 146 APPENDICES .................................................................................................................................... 164 x University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES Table 3. 1: Incidence and severity of diseases identified during survey ............................................... 37 Table 4. 1: Broad sense heritability estimates for eight quantitative traits……………………………60 Table 4. 2: Eigenvalues, variability (%) and cumulative variability from PCA .................................. 61 Table 4. 3: Contribution of variable of variables (%) to observed variation in each component ..... 63 Table 4. 4: Pearson correlation matrix of eight quantitative traits ........................................................ 64 Table 4. 5: Growth traits, and yield components of lines grouped together by principal components. 68 Table 4. 6: Table 4.6: Origin, and distribution of 96 lines in 9 inferred clusters based on Bayesian clustering analyses ..............................................................................................................71 Table 4. 7: Allele frequency divergence among sub-populations ......................................................... 73 Table 5. 1: Fixed effect estimates of AUDPC for soyabean lines in relation to the intercept……….91 Table 6. 1: Pair-wise relation among growth traits, seed quality and seed infection rate……………102 Table 6. 2: Fixed effect estimates of visual scores (VQ) for lines ...................................................... 102 Table 6. 3: Fixed estimates of infection rate (IR) ............................................................................... 104 Table 7. 1: Properties of significant SNP markers and their allelic effects………………………….120 Table 7. 2: PLP random and fixed effect estimates for lines in the representative sample................. 122 xi University of Ghana http://ugspace.ug.edu.gh LIST OF FIGURES Fig 3. 1:Diseases identified during disease survey)............................................................................ 34 Fig 3. 2: Colony characteristics of Diaporthe species isolate MTB-1………………………………...37 Fig 3. 3: Gel electrophoresis of 4 replicates of amplified 28S RNA gene of MTB-1 and the check (C). ............................................................................................................................................................. 38 Fig 3. 4: A phylogenetic tree showing the relatedness of the isolated fungus (MTB-1) to 14 isolates of Diaporthe and Phomopsis species obtained from NCBI .......................................................................40 Fig 4. 1: Distribution of flowering period in assembled germplasm .................................................... 54 Fig 4. 2: Distribution of number of days maturity 230 lines ................................................................. 55 Fig 4 3: Distribution of height at maturity ............................................................................................ 56 Fig 4. 4: Mean number of primary branches of 230 soyabean genotypes ............................................ 56 Fig 4. 5: Plant height at first pod. .......................................................................................................... 57 Fig 4. 6: Number of pods produced per plant by 230 soyabean genotypes .......................................... 58 Fig 4. 7: Number of seeds per pod of 230 soyabean genotypes ............................................................ 58 Fig 4. 8: Average number of seeds per plant ........................................................................................ 59 Fig 4.9: Seed weight measured in 100 lots for 230 soyabean lines ..................................................... 60 Fig 4.10: Loading plot of the first two principal components (F1 &F2) ............................................. 62 Fig 4.11: Clustering of lines based on growth traits, and yield components ....................................... 67 Fig 4.12: Clustering of 96 soyabean lines based on Bayesian clustering analyses ............................... 69 Fig 4.13: A plot of mean Fst and Heterozygosity among soyabean lines within clusters .................... 72 Fig 4.14: Clustering of 96 soyabean lines using mean euclidian distance based on 3200 SNP markers. ............................................................................................................................................................. 74 Fig 5. 1: Grouping of soyabean lines based on fixed effects estimates for AUDPC…………………..89 Fig 5.2: Associations between DTF, AUDPC, and AUCRC ............................................................... 90 Fig 6.1: Distribution of visual assessment score and infection rate of soyabean seeds; ..................... 101 Fig 7.1: Population structure and kinship; A (PC1 VP C2), B(PC1VPC3); C (3D-plot of PC1, PC2 and PC3………………………………………………………………………………………114 Fig 7.2: Extent of heterozygosity of A (individual lines) and B (SNP markers) ............................... 115 xii University of Ghana http://ugspace.ug.edu.gh Fig 7.3: A (SNP marker density), and B (decay of linkage disequilibrium with map distance) ........ 116 Fig 7.4: Estimated variance components of PLP; genetic variance (66.8%) and Residual (33.2%)... 117 Fig 7.5: Cross validation of significantly associated SNPs with PLP through quantile-quantile plot...117 Fig 7.6: The effect of Minor allele frequency on the significance of association between markers and PLP .....................................................................................................................................................118 Fig 7.7: Manhattan plot displaying significant associations between markers and PLP .................. 120 Fig 7.8: Candidate gene, Glyma.0.3G.14800 associated with PLP ................................................... 121 Fig 7.9: SSR markers flanking the identified candidate gene ............................................................ 121 Fig 7.10: Estimate of heritability of reaction of soyabean lines to bacterial pustule infection; genetic variance (50%) and residual variance (50%) ....................................................................... 124 Fig 7.11: Manhattan plot for SNPS associated with bacterial leaf pustule resistance on soyabean chromosomes. ......................................................................................................................124 Fig 7.12: Effect of allele frequencey on false discovery rate, and test of statisitical significance of association between genomic regions and resistance to bacterial pustule ........................ 125 Fig 7.13: Candidate genes associated with BP resistance...................................................................126 Fig 7.14: Genetic and residual variance of components of soyabean seed decay due to Diaporthe sp. ........................................................................................................................................................... 126 Fig 7.15: Manhattan plot depicting associated with Diaporthe sp seed decay resistance on soyabean chromosomes ................................................................................................................... 127 Fig 7.16: Test of significance of observed p-values of association studies ....................................... 127 xiii University of Ghana http://ugspace.ug.edu.gh University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE 1.0 GENERAL INTRODUCTION Soyabean is one of the most important crops worldwide, as its seeds are used for the production of both protein meal and vegetable oil. It is one of few crops cultivated in the tropics, subtropics and temperate regions of the world (Asafo-Adjei, 2015). Soyabean oil is the world’s most extensively utilized edible oil because of its low cholesterol level and a natural taste with odour that is nearly imperceptible (Van Ee, 2009; Gracia et al.,1997). Soyabean is produced more cheaply than other legume crops, and gives the highest oil yield per unit area among leguminous crops (Kawuki et al., 2003). Sub-Saharan Africa (SSA) contributed less than 1% to global soyabean production in the 2017 production season (FAOSTAT, 2018). Meanwhile, demand for soyabean and its products in Africa has been rising steadily due to growing livestock industry that uses soyabean as a major feed input (Joubert et al., 2013). The consumption of soyabean products in the form of soy milk drinks, yoghurt, baked beans and tofu are also on the rise on the African continent (Dlamini et al., 2014; Shannon and Kalala, 1994). Globally, soyabean demand is projected to surge between 2010 and 2050 (Masuda and Goldsmith, 2009; Gasparri et al., 2015). However, land availability for soyabean production will be generally limited by competing needs of land for urbanisation and land conservation (Goldsmith, 2008; Gasparri et al., 2015). The area considered suitable for soyabean production in Sub-Saharan Africa (SSA) is in the range of 140 to 270 Mha (Shurtleff and Aoyagi, 2009; Gasparri et al., 2015). Less than 3% of this area is currently under soyabean production and this makes SSA a likely frontier for future expansion in soyabean production to meet growing global demand ( Hartman et al., 2011; Gasparri et al., 2015). 1 University of Ghana http://ugspace.ug.edu.gh Sinclair et al., (2014) noted that much of West Africa could support soyabean production with the exception of pockets of locations in southern Ivory Coast and Ghana. However, Africa will have to adapt productive germplasm to prevailing abiotic and biotic constraints (Goldsmith, 2014) if it were to realise this potential. Biotic stresses including pest, pathogens, and weeds constitute a major constraint to soyabean production globally (Hartman et al., 2011. Hartman et al., (1999) reported that diseases are by far the most important biotic constraint to soyabean production globally. In Ghana, Offei et al. (2008) reported of bacterial leaf pustule, bud blight, charcoal rot, leaf spot, mosaic, purple seed stain, root rot, and web blight as the predominant diseases of soyabean. Bacterial leaf pustule listed among diseases of soyabean in Ghana is a known major disease in most soyabean producing countries including neighbouring Benin (Wrather et al., 2001; Zinsou et al., 2015). Asante et al. (1998) reported the occurrence of Phomopsis longicolla and other Phomopsis species known for causing soyabean seed decay and as well as Diaporthe phaseolorum var. meridionalis, causal agent of southern stem canker of soyabean in Ghana. Although, Phomopsis seed decay (PSD) caused by Diaporthe/Phomopsis complex of fungi was not listed in subsequent compilation of diseases of soyabean in Ghana (, Offei et al., 2008), warm and wet condition that enhances the development of this disease (Wrather et al., 2003) is common to the major soyabean production areas of the country. Soyabean seed decay is the primary cause of decreased soyabean seed quality in most production areas worldwide (Santos et al., 2011; Li et al., 2011). Additionally, early maturing varieties are said to be more susceptible to PSD infection than late maturing varieties (Suli et al., (2013). The infection and development of both bacterial leaf pustule and phomopsis seed decay is enhanced by warm and humid conditions (Wrather et al., 2003). Unfortunately, Northern Ghana, which accounts for 77% of soyabean (MoFA-SRID, 2016) production, is humid and 2 University of Ghana http://ugspace.ug.edu.gh warm during the production season, a situation that predisposes the country’s soyabean production to these diseases. Varieties developed for the Guinea Savannah agro-ecology are early to medium maturing types. However, early maturing soyabean varieties are generally susceptible to Phomopsis seed decay (Suli et al., 2013). Therefore, developing early maturing varieties for the short rainfall duration that pertains to northern Ghana may inadvertently result in developing phomopsis seed decay susceptible lines. Current soyabean breeding programmes at the Savannah Agriculture Research Institute (SARI) do not have gene deployment for disease control as an objective (Dr. Nicholas Denwar, personal communication). This leaves varieties vulnerable to possible outbreak of disease, especially bacterial leaf pustule and phomopsis seed decay, which the major production environment is suitable for the development of their causative pathogens. Considering the current push to expand soyabean production in Ghana coupled with projected climate change scenarios, it is important to begin to factor disease management, especially, the use of resistant genes to manage important diseases of soyabean into pending breeding programmes. Pre- emptive or anticipatory breeding strategies should be adapted to prepare for diseases that exist in Ghana but had not recorded an outbreak or diseases that thrive under conditions that are similar to those pertaining to soyabean production areas in Ghana. Plant architecture is of major agronomic importance, as it strongly influences the suitability of a plant for cultivation, its yield and the efficiency with which it is harvested (Reinhardt & Kuhlemeier, 2002). Soyabean production in Ghana has largely been dominated by small holder farmers with very limited deployment of machines. Recent emergence of commercial farms which deploy combine harvesters for harvesting soyabean have complained of combined losses during harvest due to the low pod clearance of some released varieties. Farming system in the country is gradually changing. The use of selective herbicide for weed control on soyabean 3 University of Ghana http://ugspace.ug.edu.gh farm is gaining traction. It is therefore important that the architecture of future varieties reflect changing farming systems including mechanisation and herbicide control of weeds. Germplasm is key for crop improvement. An understanding of genetic and morphological diversity within available germplasm ensures its effective use for crop improvement and conservation. The use of new sources of genetic variation to develop improved varieties is enhanced when the selection of parental genotypes is based on both phenotypic and genetic dissimilarity (Bisen et al., 2014). It is therefore important to evaluate SARI’s soyabean germplasm for phenotypic and genetic diversity and identify lines that disease resistant to bacterial leaf pustule and phomopsis seed decay for development of resistant cultivars. It is important to identify lines with high pod clearance so that future cultivars can meet the needs of both peasant and commercial farmers. The goal of this study was to find sources of resistance to Phomopsis/Diaporthe seed decay (PSD) and bacterial leaf pustule leaf (BLP) spot diseases of soyabean towards the development of cultivars that combines high pod clearance with resistance to PSD and bacterial leaf pustule. The specific objectives of this study were to; 1. carry out a survey of soyabean diseases across major production centres in Ghana, and their incidence and severity 2. assess diversity, population structure and identify lines with high pod clearance among genotypes in Savannah Agriculture Research Institute’s soyabean germplasm and identify loci associated with high pod clearance 3. evaluate Savannah Agriculture Research Institute’s soyabean germplasm for resistance to bacterial leaf pustule 4. evaluate Savannah Agriculture Research Institute’s soyabean germplasm for resistance to soyabean seed decay 5. identify loci associated high pod clearance, and PSD and BLP resistance through genome- wide association mapping 4 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO 2.0 LITERATURE REVIEW 2.1 Origin and botany of cultivated soyabean Soyabean [glycine max (L.) Merr.] (2n=2x= 40) is a leguminous plant belonging to the family, fabaceae (Singh, 2017). The crop originated from East Asia where its domestication is said to have occurred some 6000-9000 years ago (Kim et al., 2012). The specific location in East Asia where soyabean is supposed to have originated has been the subject of debates. Whilst Guo et al. (2010) concluded on the Yangtze basin in southern China as the origin of cultivated soyabean, Han et al., (2016) on the other hand pointed to Huang Haui valley in Central China as the origin of soyabean. Meanwhile earlier suggestions based on genetic diversity (Dong et al., 2004; Li et al., 2010) and archaeological evidence (Lee et al., 2011) placed the origin of cultivated soyabean at Yellow River basin in China. The works of Guo et al. (2010); Zhou et al., (2015); and Han et al., (2016) concluded that cultivated soyabean originated from a single domestication event given rise to the single origin hypothesis (Gao et al., 2010). Archaeological evidence dating back to 5000–3000 years showed that Japanese and Korean soyabean seeds were distinct from that of China (Lee et al., 2011). This was supported by molecular analysis which showed that Japanese and Korean soyabean possess distinct gene pools in both their chloroplast and nuclear genome (Xu et al., 2002; Abe et al., 2003; Zhou et al., 2015). Considering the uniqueness of Japanese and Korean wild soyabean in addition to the long divergence time between G. max and G. soja (0.27 or 0.8 Ma), Sedivy et al. (2017) postulated a view on soyabean domestication in which a prolonged period of low-intensity management of wild soyabeans at multiple locations preceded the domestication event, which gave rise to cultivated soyabean. They therefore concluded that cultivated soyabean resulted 5 University of Ghana http://ugspace.ug.edu.gh from independent multiple domestication events at different locations in East Asia, given rise to the complex origin hypothesis about the origin of cultivated soyabean. Cultivated soyabean belongs to the subgenus soja which contains G. soja, its wild progenitor and the cultigen, G. max, both of which contain 2n=40 chromosomes. Both species are cross- compatible with cultivated soyabean (Thseng et al., 1999). Cultivated soyabean [glycine max (L.) Merr.] is an annual plant that exhibits epigeal germination with two cotyledons. Its root system is comprised of taproot and large number of secondary roots with which it establishes symbiotic relationship with nitrogen fixing bacteria. Stem growth may be determinate, semi- determinate or indeterminate with trifoliate leaves (Lersten and Carson, 2004). 2.2 Global distribution and production of soyabean Soyabean is cultivated in various environments throughout the world. Production of soyabean occurs extensively in tropical environments as observed in Brazil, and under sub-tropical climate as pertains in Central America and southern USA. Soyabean production also occurs in temperate climate as seen in northern USA and Canada, and in the sub-arctic climate such as pertains in Sweeden and Siberia (Asafo-Adjei, 2015; USDA/FAS, 2018). Global production of soyabean during the 2017/2018 cropping season stood at 346.919 Mmt (USDA/FAS, 2018). The three top soyabean producing countries, USA (119.518 Mmt), Brazil (112 Mmt), and Argentina (54 Mmt), accounted for 82% of the global soyabean output in the 2017/018 season (USDA/FAS, 2018). In Asia, China, the top producer of soyabean recorded a seasonal output of 14.2 Mmt, which was followed by India, the Asian second top producer with 9.5 Mmt (USDA/FAS, 2018). Globally, China and India ranked fourth and fifth, respectively with China doubling as the world’s largest importer of soyabean (index Mundi, 2018). 6 University of Ghana http://ugspace.ug.edu.gh 2.3 Soyabean production in sub-Saharan Africa Sub-Saharan Africa (SSA) harbours 600 million hectares of arable land, 10% of which is currently under the cultivation of various crops including soyabean (Khojely et al., 2018). Consequently, Sub-Saharan Africa is seen as the last frontier to expand soyabean production to meet the growing global demand. Khojely et al. (2018) described soyabean (Glycine max (L.) Merr.) as non-native and non-staple crop in sub-saharan Africa, a reflection of the origin and the limited role of the crop in traditional food preparation on the continent. South Africa, Nigeria, Zambia, and Uganda are the leading producers of soyabean in SSA (UDS/FAS, 2018). South led soyabean production in Africa with 1,575,000 mt in the 2017/2018 season (UDS/FAS, 2018). Nigeria, which obtained soyabean output of 600,000 mt during the same period, was second to South Africa in production. Introduction and cultivation of soyabean in sub-Saharan is relatively recent with earliest records of cultivation dating back to 1903 in South Africa, 1907 in Tanzania, 1908 in Nigeria, and 1909 in Ghana (Plahar, 2005; Giller and Dashiell, 2006; Shurtleff & Aoyagi, 2009). Soyabean production in sub-Saharan Africa increased gradually, mainly, through increasing the area under production. Area under soyabean cultivation in sub-Saharan Africa rose from 20,000ha in 1978 to 1,500,000ha in 2016 (USDA/FAS, 2017). Correspondingly, soyabean production also rose from 13,000 mt in 1978 to 2,300,000 mt in 2016 (USDA/FASS, 2017). However, yield per unit area remained almost the same at 1.1 t/ha over the period (Khojely et al., 2018). The current yield per unit area of 1.1 t obtained in sub-Saharan Africa in 2016 was far below the global average yield of 2.4 t/ha (USDA/FAS, 2018). Among the top soyabean producing countries in sub-Saharan Africa, Nigeria recorded the largest area under production in 2016 (UDS/FAS, 2017) followed by South Africa. Increases in soyabean production in Nigeria was largely driven by increases in area under production as the country’s yield per unit area remained at 1 t/ha between 2012 to 2104 and reduced to 7 University of Ghana http://ugspace.ug.edu.gh 960kg/ha in the 2015 and 2016 production seasons (Khojely et al., 2018). The third top producer of soyabean in SSA, Zambia was the second in terms of yield per unit area (1.94 t/ha) whilst. Uganda, the fourth leading producer of soyabean in SSA has the least yield per unit area (600kg/ha) (Khojely et al., 2018). Despite the gradual increase in soyabean production over the years, SSA contributes less than one percent (1%) to global production and so is its share of global Farmer Revenue from soyabean FOASTAT, 2018). For instance, in 2012, Farmer Revenue from soyabean globally was USSD126 billion dollars of which Africa accounted for less than 1% of production earned a revenue of $1,012,500,000, representing 0.8% of the global revenue (Goldsmith, 2014). Though, soyabean is rightly described as a non-staple crop in SSA, demand for the crop and its products has seen a steady growth over the years (Kolapo et al., 2012). Demand for soyabean and its products in SSA is largely driven by a growing livestock industry (Joubert et al., 2013) that largely relies on protein meal of which soyabean contributes about 67% of global consumption. Secondly, there is increasing consumption of soyabean oil and other products of soyabean (Dlamini et al., 2014; Jobert & Jooste, 2013; Kolapo, et al., 2012). In addition, the promotion and utilisation of soy-based food to fight malnourishment among children is widely pursued by many countries in SSA (Amagloh et al., 2018). Gradually, soyabean is entering the food chain in sub-Saharan African countries whereby soyabeans enhance either traditional food preparations or original ingredient substituted with soyabean is gaining grounds across Africa. In Ghana, tubaani, an indigenous food originally prepared with cowpea is now prepared with soyabean (Khojelly et al., 2018). 8 University of Ghana http://ugspace.ug.edu.gh 2.4 Soyabean production in Ghana Soyabean cultivation in Ghana started in 1909 during which soyabean was introduced to, and cultivated by farmers at Bimbila, Nakpanduri, Karaga, Tilli, and Bawku in the northern protectorate of then Gold coast (Quarshie and Nsowah, 1975; Plahar, 2006). The introduction of soyabean in Ghana was aimed at making soyabean a staple food, and an export commodity (Quarshie and Nsowah, 1975). Research conducted from 1909 to 1956 focused on understanding the agronomy of soyabean whilst promoting its consumption. The period saw seventeen annual trials conducted over 12 locations stretching from the Asuansi in the south to Tono in the Upper east region (Quarshie and Nsowah, 1975). Amidst initial disappointing results, understanding of the crops lead to improved results from 1956 onwards, and yield up to 1457 kg/ha was recorded for some of the trials. This initial effort at promoting cultivation and consumption of soyabean failed because farmers complained that it took too long a time to cook soyabean, in addition to having unappealing taste (Quarshie and Nsowah, 1975). Scientific research from the 1960s to early 1970s by the Crop Research Institute focused on the utilisation of soyabean for feed meal, and vegetable oil (Mbanya, 2011). The absence of processing companies to absorb what was produced caused a glut. Consequently, there was no market for soyabean and as such, the farmers stopped soyabean production (Plahar, 2006). A more sustained campaign to boost soyabean production began in the 1990s where the Ministry of Food and Agriculture joined efforts with non-governmental agencies (NGOs) including the Adventist Development and Relief Agency. Despite the failure of such programmes to make Ghana a soyabean production hub, soyabean production gradually gained traction as reflected in the adoption of released varieties and production technologies with production rising from 1,000 tons in 1979 to 10,000 mt in 1992 (CRI, 2010). Meanwhile, the 9 University of Ghana http://ugspace.ug.edu.gh country met its need for soyabean and soyabean products through imports, which stood at 198,000 mt in 2009 (MoFA, 2009). Though, several areas in Ghana, including those located in the coastal Savannah, the forest savannah or transitional belt, Guinea and Sudan agro-ecologies have been described as suitable for soyabean production (Sinclair et al., 2014), current production largely occurs in the Northern region which accounts for 70% of the production area, and 77% of the country’s soyabean production (SRID, 2016). Soyabean production in Ghana has seen consistent increases from the 1990s, largely due to increases in area under cultivation supported by marginal increases in average yield. Area under soyabean cultivation rose form 47,000 ha in 2007 to 87, 000ha in 2016 and output rose from 50,000 mt in 2007 to 143, 000 mt in 2016 (MoFA-SRID, 2017). Maximum output of soyabean occurred in 2011 where production reached 165, 000 mt (MoFA-SRID, 2017). Increased soyabean production in Ghana is due to both increases in area under production, and in average yield. Average yield of soyabean in Ghana rose from 1.1t/ha in 2007 to 1.65 in 2016 (MoFA-SRID, 2017). In Ghana, and most countries in sub-Saharan Africa, soyabean production is characterised by poor yielding varieties, limited application of fertilizers and non-inoculation of seed with Bradyrhizobium japonicum (Woomer et al., 2012). Additionally, weak value chain that results in poor market access to farmers leading to very low producer price is reported to be a major hindrance to soyabean production in Ghana with similar reports from Nigeria (Aidoo et al., 2014; Agada, 2015; Dogbe et al., 2013; Mbanya, 2011). Most soyabean varieties released in Ghana were obtained from IITA and consequently, as noted by Khojely (2018), national contribution to soyabean varietal and production technology development, and dissemination has been limited in the past. 10 University of Ghana http://ugspace.ug.edu.gh 2.5 Constraints to soyabean production In Africa, constraints to soyabean are abiotic and biotic, as well as socio-economic challenges (Kawuki et al., 2003). Abiotic factors including weather-related phenomena, soil nutrient, salinity and photoperiod are factors in the physical environment that limit crop production including soyabean (Hartman et al., 2012). Global climate variability and climate change that is causing changes in temperature and rainfall is projected to affect agro-ecological zones (Nelson, et al., 2009) with consequent effects on crop productivity and production. Suggested changes due to climate change includes increased levels of CO2, and extremes of temperature and rainfall. Whilst elevated CO2 levels may increase photosynthetic rate in crops (Cure and Acock 1986; Mendelsohn et al., 1994), temperature and rainfall extremes may have negative effects on soyabean production as observed in China by Zhang et al. (2009). Peralta and Wander (2008) showed that elevated CO2 caused a decrease in soil organic matter due to death of some microbial organisms resulting from elevated CO2 levels. Ghana and many African countries have their soyabean fields dominated by cultivars from IITA which are short- day length, drought tolerance, and photoperiod insensitive cultivars, (Ibrahim, 2011) and so research programmes do not have photoperiodism as a major research goal. However, photoperiodism limits the introduction and use of superior cultivars from other continents. Tropical varieties from USA and Brazil have been found to be highly adapted to growing conditions in SSA but the temperate varieties flowered very early under the prevailing short-day conditions resulting in poor yields due to low vegetative growth (Ibrahim et al., 2012). Two abiotic factors known to dominantly influence soyabean phenology are photoperiod, and temperature (Cregan and Hartwig, 1984; Summerfield and Wilcox, 1978). In soyabean [Glycine max (L) Merr], transition from vegetative to reproductive development is largely 11 University of Ghana http://ugspace.ug.edu.gh influenced by changes in the length of light and dark phase within a day (Langewisch et al., 2017). The response to photoperiod is the major determinant of the maturity groups of soyabean cultivars (Hartman et al., 2012). In the US, soyabean cultivars have been placed in 13 maturity groups ranging from MG00 for groups suitable for growth in northern US with longer day-length to MGX representing cultivars that are suitable for growth near the equator (short day length) (Zhang et al., 2007). Similarly, in China, soyabean cultivars have been placed in 12 maturity groups ranging from MG00 to MGIX (Wang et al., 2018). Photoperiod affects the distribution and performance of soyabean cultivars across the globe. When long day length cultivars are planted in shorter photoperiod areas as pertains in the tropics, they flower early, attaining very little vegetative growth resulting in fewer internodes, which translates into fewer number of pods per plants and lower yield (Hartman et al., 2012) 2.5.1 Drought as a production constraint of soyabean Drought causes yield losses at varying degrees in soyabean when it occurs at germination, vegetative, and reproduction stages. Drought at germination causes uneven and spotty germination whilst drought at the vegetative stage reduces shoot growth and limit sink as reflected in reduced number of internodes at which pods are produced (Mannan et al., 2016). Yield loss from drought stress in soyabean is exacerbated by the lack of nitrogen mineralization and nitrogen fixation that occurs under drought conditions (Kunert, 2016). In dry conditions, nodules cease nitrogen fixation because of a lack of soil moisture and lack of carbohydrate supply from the soyabean plant to Rhizobia sp. (Ku et al., 2010). Argentina, the third largest producer of soyabean in the world and lead exporter of soymeal experienced drought during the 2017/2018 season which lead to a loss of a sixth of the projected yield for the season causing a rise in global soymeal prices (Terazono, 2018). The semi-arid and arid environments that 12 University of Ghana http://ugspace.ug.edu.gh include northern Ghana where most of the country’s soyabean is produced are prone to drought (Hufstetler, 2007). 2.5.2 Salinity as constraint to soyabean production Soil salinity is a term used to describe the salt content within soil water and it affects plant growth and development. The injurious effects of salinity on the growth and yield of crop plants arises from reducing water inside plant cells and the unfavourable effects of sodium (Na+) and chloride (Cl−) which negatively affect plant physiology and morphology (Miransari, 2016). Soyabean is described as being sensitive-to-moderately tolerant to salinity (Kao et al., 2006) and therefore, suffers stress when grown in high saline soils. Different growth stages and different soyabean plant parts are adversely affected by salinity stress including plant height, plant biomass, leaf size, number of branches, internodes, pods and weight of 100 seeds, and plant weight (Miransari, 2016). 2.6 Important diseases limiting soyabean production Biotic constraints including pests, diseases, and weeds pose a major challenge to soyabean production globally (Hartman et al., 2011). Among the biological constraints to soyabean production, diseases are by far the most important (Hartman, 1999) and responsible for an estimated 11% of global yield loss in soyabean (Hartman, 1999). Soyabean diseases including Asian soyabean rust, bacterial leaf pustule, frogeye leaf spot, sudden death syndrome, phomopsis seed decay among others have been reported to cause significant losses to production. Rust caused by phakospora pachyrihzi Syd & P.Syd which causes the most significant yield losses in recent times (Akamatsu et al., 2013). phakospora pachyrihzi is endemic in the eastern hemisphere and has caused significant economic annual yield losses in Africa, and South America (Sohrabi et al., 2014). Globally, soyabean rust has been the subject of research with many sources of resistance identified, and resistant loci mapped with linked 13 University of Ghana http://ugspace.ug.edu.gh markers made available for resistant variety development (Miles et al.,2008; Hyten et al, 2007; Ray et al, 2009; Kim et al, 2012; Gracia et al., 2011, and Yamanaka et al, 2015). In Ghana, however, Asian soyabean rust has not been listed among the predominant diseases of soyabean (Offei et al., 2008), despite the report of its incidence in the country in 2007 (Bandyopadhyay et al., 2007). A disease survey carried out in Ghana in 2006 revealed that soyabean rust was wide spread (Bandyopadhyay et al., 2007), thus officially confirming the presence of soyabean rust disease in Ghana. The survey which was carried out at locations in three regions covering three distinct agro-ecologies; forest, guinea savannah, and Sudan savannah, recorded disease incidence ranging from 50-100% in the farms surveyed (Bandyopadhyay et al., 2007). Subsequently, 34 soyabean lines were screened for resistance to rust, five of which were highly resistant (Karikari and Amponsah-Antwi, 2017). The country has however not recorded soyabean rust outbreak at its major production areas including, Karaga, Gusehegu, and Chereponi in the Northern region where most of the country’s soyabean is produced (MoFA-SRID, 2016). Offei et al. (2008) reported bacterial leaf pustule, bud blight, charcoal rot, leaf spot, soyabean mosaic virus, purple seed stain, root rot, and web blight as diseases of soyabean in Ghana. The said publication did not make mention of rust. Also, there is no literature on the distribution and economic effects of the listed diseases on soyabean production in Ghana. However, considering the current push to expand soyabean production in Ghana coupled with projected climate change scenarios, it is important to begin to factor disease management, especially, the use of resistant genes to manage important diseases of soyabean. Pre-emptive or anticipatory breeding strategies could be adapted to prepare for diseases that exist in Ghana but had not recorded an outbreak or diseases that thrive under conditions that are similar to those pertaining to soyabean production areas in Ghana. To this end, soyabean seed decay due to Diaporthe/ 14 University of Ghana http://ugspace.ug.edu.gh Phomopsis complex of fungi, and bacterial leaf pustules both of which thrives under warm and humid conditions were considered in this study. 2.6.1 Soyabean seed decay Globally, many diseases, especially seed and seed borne diseases have been reported to cause significant yield loss, reduced grain quality, poor seed germination and high seedling mortality (Mengistu et al., 2012). Many of the pathogens of economic importance to soyabean [Glycine max (L.) Merr] production are associated with its seed (Mengistu et al., 2012) with Diaporthe –Phomopsis complex (DPC) of fungi, which causes seed decay, considered the most important. Soyabean seed decay due to Phomopsis/diaporthe species of fungi is considered the most important pre-harvest disorder of soyabean seed worldwide (Sinclair, 1999). Soyabean seed decay has been a serious problem in soyabean production areas with warm and wet weather for the past three decades including southern USA (Zhang et al., 1998), Korea (Oh, 1998), Argentina (Pioli et al., 2003), Italy (Riccioni et al., 2003), Yugoslavia (Medic´- Pap et al., 2007), Brazil (Costamilan et al., 2008), and China (Cui et al., 2009). Seed losses on global scale due to Phomopsis longicolla was approximated at 0.19 million metric tons in 1994 (Kulik & Sinclair, 1999). Effects of Phomopsis seed decay on yields in the United States from 1996 to 2007 ranged from 0.38 to 0.43 Mmt (Wrather and Koenning, 2009). In 2009, an estimated yield loss of 45 million bushels, 23 million of which was due to Phomopsis seed decay because of the hot and humid conditions that characterized pod fill to harvest periods of that season (Koenning and Wrather, 2012). Phomopsis seed infection causes pre- and post- emergence damping off, and in severe conditions, stands maybe affected to the point of reducing yield because of less plant population per unit area. 15 University of Ghana http://ugspace.ug.edu.gh In addition to directly decreasing yield, DPC infection causes seeds to shrivel, elongate, crack or split, mottle become moldy when there is sufficient moisture (Kulik and Sinclair, 1999). Diseased seeds including mottled and moldy seeds are graded lower (Federal Grain Inspection Service, 2004). DPC infection causes off-colour soyabean seeds but off-colour is one of the primary quality-relating factors that negatively influence the market grade of soyabean (Hepperly and Sinclair 1978; Sinclair 1982) and therefore, PDC infection generally reduces market grade and price of soyabean. Furthermore, DPC infected soyabean seeds have reduced germination whilst germinated infected seeds have reduced vigour (Smith et al., 2008; Begum et al., 2008; Li et al., 2011; Gillen et al., 2012; Mayhew and Caviness, 1994). (Mengistu et al., 2009) reported a strong negative association between DPC seed infection rate and seed germination, vigour and seedling emergence. DPC infection of soyabean seeds also affects the quality of products including soyabean flour, meal, and soyabean oil. Soyabean flour from infected seeds is reported to have unpleasant odour as well as being unpalatable (Hepperly and Sinclair, 1978). Similar characteristics are noted for oil obtained from infected seeds rancid seeds with higher peroxide values (Divilov, 2014). Infected seeds also have altered fatty acid, and protein contents. Significant reduction in protein content, and decreased palmitic and oleic acid compositions were found to be significantly correlated with DPC infection (Wrather et al., 2003; Fábrega et al., 2000). Some isolates of Phomopsis longicolla, a member of the DPC of fungi produces the fungal metabolites, cytochalasin compounds that have the ability to bind to actin filaments and block polymerisation, and elongation of actin (Carter, 1967). Infected seeds with Phomopsis longicolla containing moderate to high levels of cytochalasin were found poisonous to sheep and rats (Allen et al., 1992). 16 University of Ghana http://ugspace.ug.edu.gh 2.6.2 Pathogens within the Phomopsis- Diaporthe complex (DPC) of fungi Pathogens that constitute the complex includes Phomopsis longicolla; Diaporthe phaseolorum var. sojae; Diaporthe phaseolorum var. caulivora; and D. phaseolorum var. meridionalis (Sinclair, 1999; Rossman et al., 2007). The members of Phomopsis/ Diaporthe complex of fungi belong to the order Diaporthales and Diaporthaceae. The members of the complex are responsible for a number of diseases on soyabean. Phomopsis longicolla is largely responsible for all reported cases of soyabean seed decay, accounting for about 85% of Diaporthe/Phomopis infection of soyabean seeds (Hobbs et al., 1985; Zimmerman and Minor, 1993; Nevena et al., 1997; Zhang et al., 1999; and Suli et al., 2013). Other diseases due to fungi in the complex include pod and stem blight, caused primarily by D. phaseolorum var. sojae; northern stem canker and top dieback, caused primarily by D. phaseolorum var. caulivora; and southern stem canker, caused primarily by D. phaseolorum var. meridionalis (Sinclair, 1999; Rossman et al., 2007). However, these pathogens overlap in the type of diseases that they cause, often infecting several parts of the soyabean plants ranging from the roots to the seeds. P. longicolla, and D. phaseolorum var. sojae have been shown to cause stem cankers, and P. longicolla has been shown to cause pod and stem blight (Cui et al., 2009; Kmetz et al., 1979; Vidić et al., 1998). P. longicolla, D. phaseolorum var. sojae, D. phaseolorum var. caulivora, and D. phaseolorum var. meridionalis can all infect soyabean seed, and the first three of these fungi have been shown to cause seed rotting and molding with P. longicolla being the most virulent pathogen among the three (Asante et al., 1998; Kmetz et al. 1974; Kmetz et al., 1979; Sun et al., 2012). 17 University of Ghana http://ugspace.ug.edu.gh 2.6.3 Management of Phomopsis seed decay of soyabean (PSD) Many management strategies including cultural, biological, the use of chemicals, and resistant genes have been used to prevent, manage and control PSD. 2.6.3.1 Cultural approaches to PSD control Planting of high-quality seeds that are free of PSD infection is the first step to ensuring reduced field infection and development of PSD. Therefore, the use of certified seeds is highly recommended. This may pose a challenge in Ghana where the use of farmers’ own seed is more prevalent than the use of certified seed. Field preparation is another cultural approach that could be used to reduce the incidence and severity of pod and stem blight, and PSD (Pathan et al., 2009). Deep ploughing of previously infected fields to bury the debris or soyabean crop residue before the next planting helps to eliminate or reduce the source of inoculum and thereby reducing the risk of PSD infection (Roy et al., 1994). Another cultural approach to reducing the incidence and severity of PSD is crop rotation with non-legume crops. Continuous cropping of soyabean on the same piece of land enhances the build-up of Phomopsis/Diaporthe inoculum that in turns increases the incidence and severity of PSD. Rotating soyabean with cereals crops disrupts the build-up of inoculum and therefore, reduces the incidence and severity of pod and stem blight (Sun et al., 2013). Soil nutrients especially potassium (K) and dolomite are said to hinder the development of PSD. Ito et al. (1994) reported of an inverse relationship between the concentration of potassium (K) in soil and leaves, and PSD incidence. Whereas Dolomitic lime which decreases the concentration of Potassium, increased the incidence of PSD whilst Jeffers et al., (1982) observed that application of potassium (K) consistently lowered the incidence of mouldy seed but did not lower the incidence of PSD. 18 University of Ghana http://ugspace.ug.edu.gh 2.6.3.2 Biological control of PSD Biological control is environmentally sustainable and often, its inclusion in an integrated pest management strategy is promoted. Divilov, (2014) reported that the mycelial growth of IL12- DS-1, an isolate of Diaporthe sojae Lehman was inhibited by an Isolate of Acremonium strictum, and observed that Acremonium strictum parasited Diaporthe sojae isolate, L153. Acremonium strictum W. G though, a plant pathogen (Racedo et al., 2013), has been reported to be an endophyte, a mycoparasite, and a saprobe (McGee et al., 1991; Domsch et al., 2007; Rivera et al., 2007). 2.6.3.3 Chemical control Chemical control, mainly the application of fungicide can be used to control of PSD. Benomyl (Methyl (1-9butylamino) carbonyl)-1H-benzimidazol-2-yl can effectively control PSD when applied late pod developmental stage (Ellis et al., 1974). Other fungicides including Dithane M-45, Zineb, Perrazole, Polyram combi, Mertect, and Afugan could be used to control PSD (Wu and Lee, 1985). Lee et al., (1994) reported that spraying Benomyl at the R6 stage gave the highest reduction in Phomopsis longicolla and Diaporthe Phaseolorum var. sojae seed infection. Whilst Oh, (1998) reported that whole plant spraying is required to obtain significant seed infection reduction with benomyl. 2.6.3.4 The use of resistant genes Finding resistance to pest and diseases is an important objective in breeding programmes and cultivar development. The use of resistant cultivars is considered the most effective means of disease, nematode and insect control (Fehr, 1991). The deployment of resistance to control diseases in plants is considered the most economically viable, and environmentally and ecologically friendly, and therefore, the most sustainable form of disease control. 19 University of Ghana http://ugspace.ug.edu.gh A number of sources of resistance to PSD have been identified from several resistant screening trials. These include Taekwamgkong (Sun et al., 2012); PI417479 (Brown et al. 1987; Zimmerman and Minor, 1993); PI424324B (Smith and Nelson, 2011); Aksoy, PI209908, PI22787 and PI229358 (Minor et al., 1995); PI82264 (Walters et al., 1973); PI181550 (Athow, 1987); PI200501 (Ross, 1986); PI360841 (Brown et al., 1987); PI42324B, PI458130 and PI 567381B (Smith et al., 1993). Additionally, Li et al., (2015) reported 15 accessions including PI 189891, PI 398697, PI 417361, PI 504481, PI 504488, and PI 88490 in MG III; PI 158765, PI 235335, PI 346308, and PI 416779 in MG IV; and PI 381659, PI 381668, PI 407749, PI 417567, and PI 476920 in MG V to be resistant to PSD. 2.6.3.5 Markers linked to PSD resistant loci A considerable number of genetic studies including mapping of resistant loci conferring resistance to PSD from a number of resistant sources have been carried out. The resistant loci of PI417479 was mapped to linkage group F using RLFPs markers (Berger and Minor, 1999). The mapping studies carried out using Random fragment length polymorphism (RFLP) markers had PI417479’s resistance linked to the markers A708, and A162 on chromosomes 13 and 12, respectively (Berger and Minor 1999). Suli et al., (2013) mapped the resistance of the Korean variety, Taekwang to two loci, PSD-6-1 and PSD-10-2 which occurred on chromosomes 6 and 10, respectively. The resistant loci, PSD-6-1 is flanked by the SSR markers Satt 100 and Satt 460 whilst the loci, PSD-10-2 is flanked by the SSR markers, Sat_038 and Satt 243 (Suli et al., 2013). It is noteworthy that the SSR marker, Satt460 is also linked to Asian soyabean rust resistant gene Rpp3 on the same linkage group, C (Monteros et al., 2006). Also, Suli et al., (2013) proved that the linkage between PSD resistance and days to maturity where PSD resistance and days to maturity were mapped to the same chromosomal location flanked by the same markers. The quantitative trait loci (QTL), Mat-10-3 associated with days 20 University of Ghana http://ugspace.ug.edu.gh to maturity was mapped to chromosome 10 and flanked by the SSR markers Sat_038 and Satt 243, the same markers that flanked the QTL, PSD-10-2, conferring resistance to PSD (Suli et al., 2013). Similarly, Mat-6-2 also associated with days to maturity was mapped to chromosome 6, and flanked by Satt134 and Satt100. The same markers at the same chromosomal region harbouring the QTL PSD-6-1 which confers resistance to PSD (Suli et al., 2013). Based on qualitative analysis, the resistance in PI80837 was mapped to chromosome 14 flanked by the SSR markers, Sat_177 and Sat_342, whereas the PSD resistance loci in MO/PSD-0259 was mapped to chromosome 13, flanked by the SSR markers, Sat_317 and Sat_120 (Jackson et al., 2009). 2.6.3.6 Inheritance of PSD PSD is a highly heritable trait recording broad sense heritability of 0.44 and 0.83, and narrow sense heritability of 0.42 and 0.49 (Anderson and Buzell, 1985; Zimmerman and Minor, 1993). Resistance to PSD is conferred by a dominant gene with sources of resistance harbouring a single or two genes. Allelsim test showed that PI417479 habours two dominant complimentary genes which confers strong resistance to PSD (Zimmerman and Minor, 1993; Minor et al., 1995). Studies carried out by Smith et al. (2008) confirmed that PI360841 also harbours two dominant complimentary genes. However, PSD resistant line, MO/P SD-259 harbours a single dominant gene (Jackson et al., 2005). 2.6.4 Bacterial leaf pustule Bacterial leaf pustule, caused by Xanthomonas axonopodis pv. glycine, is among the major diseases in several major soyabean (Glycine maxL.) producing countries (Wrather et al., 2001). It is one of eight documented diseases affecting soyabean in Ghana and the only bacterial disease among the eight (Offei et al., 2008). On susceptible cultivars, bacterial leaf pustule is reported to cause yield loss in the range of 15-50% (Prathuangwong and Amnuaykit, 1987). 21 University of Ghana http://ugspace.ug.edu.gh Yield loss due to bacterial leaf pustule depends on the level of incidence and severity. Bacterial leaf pustule severity of 10.1-25, 25.1-50, and 50.1-75% could cause corresponding yield losses of 15-21, 38 and 53, respectively (Shukla, 1994). Yield loss due to bacterial leaf pustule results from degradation of chlorophyll content of leaves as lesions (yellow pale spot) coalesces to form bigger spot and under severe infection may result in premature senesces and defoliation of leaves (Hartman and Murithi, 2017). When infection occurs at the early stages of plant growth, susceptible genotypes may suffer premature defoliation and death. In which case, seed weight, a key component of yield is reduced. The causal agent of bacterial leaf pustule, Xanthomonas axonopodis pv. glycine survives or overwinters on the debris; leaves, pods and stems, as well as the seeds of infected plants (Khaeruni et al., 2007). These attributes have the propensity to make bacterial leaf pustule endemic in areas of continuous soyabean production on the same piece of land whilst spreading to new areas through the distribution of untreated seeds (Khaeruni et al., 2007). On infected soyabean fields, the pathogen spreads from one plant to another or to nearby fields through rain splashes and wind (Goradia et al., 2004). 2.6.4.1 Management of bacterial leaf pustule Cultural practises including crop rotation through which the overwintered inoculum (pathogen) is avoided, and incorporating residue of infected plants through tillage into the soil to minimise the amount of inoculum available to infect new soyabean crops could be used to manage the bacterial leaf pustule disease. Chemicals have been extensively used to control bacterial leaf pustule. These include the chloromycitin, streptocycline and fytolan (Copper oxychloride) (Singh and Jain, 1988). Induction of resistance to bacterial leaf pustule susceptible soyabean lines is another approach that has been developed for the control of bacterial leaf pustule (Khaeruni et al., 2010) as an 22 University of Ghana http://ugspace.ug.edu.gh alternative to chemical control. Biological control is deemed ecologically and environmentally friendly as opposed to chemical control which causes damages to the environment and less sustainable because, the persistent use of chemicals leads to the emergence of resistant Xanthomona axonopodis pv. glycine strains (Khaeruni et al., 2010). Khaeruni et al., (2018) evaluated biological fertilizers (biofresh) composed of three rhizobacteria types, Bacillus ceresus ST21b, B. subbtilis ST2le, Serratia sp 2229A in different formulations and concluded that a solid formulation of biological fertilizer in combination of compost made from soyabean residue was effective at increasing the resistance of soyabean plants to bacterial leaf pustule. Increased resistance to bacterial leaf pustule arose from increased salicylic acid and peroxidase content of plants treated with the biological fertilizer (Khaeruni et al., 2018). The use of resistant genes for the control of plant diseases is the most efficient approach and the control of bacterial leaf pustule is not an exception. Several sources of resistance to bacterial leaf pustule of soyabean has been reported. Some of which include Danbaekong, SS2-2, and Pi96188 (Kim et al, 2010; Kim et al., 2011) as well as CNS soy (Hartwig, 1951). 2.6.4.2 Molecular resources available for management of bacterial leaf pustule A number of mapping studies have identified markers linked to the bacterial leaf pustule resistant gene that could be used for Marker Assisted Selection to facilitate bacterial leaf pustule resistant cultivar development. Deploying SSR markers, Narvel et al. (2001) mapped the bacterial leaf pustule resistant gene, Rxp to linkage group D2 flanked by the SSR markers, Satt372 and Satt014. On other hand, Kim et al. (2010) using RIL and NIL population fined mapped the Rxp gene resistant gene on chromosome 17 to 33 kb region flanked by the SSR markers, SNUSSR17_ and SNUSSR17_12. They found three genes within the 33 kb region and suggested two genes which codes for membrane protein and zinc finger protein as candidate genes offering resistance to bacterial leaf pustule (Kim et al., 2010). 23 University of Ghana http://ugspace.ug.edu.gh In another study, a source (P196188) resistant to bacterial leaf pustule, had its resistant gene mapped to chromosome 10 linked to the SSR marker, Sat_108 (Kim et al. 2011). Yang et al., (2011) used available genomic data on the bacterial leaf pustule resistant gene, Rxp to develop markers and identified the marker Rxp17-700 which was expected to be tightly linked to Rxp gene. This presents a list of markers that could be assessed, validated and used for the development of bacterial leaf pustule resistant cultivars. Some level of the understanding of the nature of inheritance of the Rxp resistant gene has been reported. The nature of inheritance of resistance harboured by PI96188, and CNS soy were reported to be single recessive genes (Hartwig-1951; Kim et al., 2011). It is note-worthy that with exception of Yang et al., (2011) who developed Rxp markers using Rxp genomic sequence data, all other mapping studies stated here were carried out using bi-parental populations. 2.7 Genome-wide association studies (GWAS) and Genotype-by-sequencing (GBS) Marker identification is key to marker-assisted selection which aids with the speed and accuracy of selection in breeding programmes. Conventionally, marker identification is achieved through quantitative trait loci (QTL) mapping based on populations. However, population- based mapping is time consuming, expensive and has limited resolution due to limited number of meiotic events accumulated over the period of population development (Jannink and Walsh, 2002). The limitations associated with population or interval-based QTL mapping are overcome with genome-wide association studies (GWAS) which utilises natural occurring variation in assembled germplasm to detect loci associated with desired traits was invented (Hall et al., 2010). Assembled germplasm are often a collection of unrelated genotypes which come along with large number of meiotic (crossing over) events since their last common progenitor (Sonah et al., 2013). Since development of a population is not a 24 University of Ghana http://ugspace.ug.edu.gh requirement for genome-wide association studies, time, and cost of developing population are saved (AlMarskri, 2012). The advent of high-through put genotyping including genotyping-by-sequencing (GBS) provides high marker coverage on individuals in an assembled germplasm. Genome-wide association studies use the high maker coverage to scan the entire genome of individuals in search of genomic regions in linkage disequilibrium that are associated with traits of interests. In soyabean, many lines have been genotyped at higher marker coverage with the 50K SNP array and GBS (Song et al., 2013; Sonah et al., 2013). With GWAS, SNP markers linked to some important diseases of soyabean were identified including bacterial leaf pustule (Chang et al., 2016); Sclerotinia sclerotiorum (Wei et al., 2017), and sudden death syndrome (Wen et al., 2014). Markers linked to other traits including 100-seed weight (Contreras-Soto et al., 2017), seed oil content, and seed protein (Li et al., 2018) were also identified through GWAS. Genotyping-by-sequencing (GBS) offers a more flexible and an efficient approach to genome- wide level identification of genetic variation in crops (Fu and Peterson, 2011; Peterson et al., 2014). GBS yields high-density genotypic information (Poland, 2012) on samples which is suitable for genome-wide association studies, and genomic selection (Poland and Rife, 2012; He et al., 2014; Gore et al., 2009; Yu and Buckler, 2006). GBS method provides the numerous generations of ancestral recombination events required for direct mapping of variation in natural populations without generating populations through crosses (Yu and Buckler, 2006). Harnessing the advantages of GBS and GWAS for identification and use of markers will bring efficiency to soyabean breeding programmes in Ghana. This will translate into development and release of cultivars that combine high productivity with resistance to prevailing abiotic and biotic stresses. 25 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE 3.0 A survey of soyabean diseases across production centres in the Transition, Guinea and Sudan Savannah ecologies of Ghana 3.1 Introduction Soyabean cultivation in Ghana has seen consistent rise largely due to increases in area under production (MoFA-SRID, 2016). Additionally, the ongoing “Feed the Future” programme in which soyabean is being targeted to alleviate poverty and eradicate malnutrition (https://www.feedthefuture.gov/country/ghana) is promoting the cultivation and utilisation of soyabean. Globally, diseases are a major constraint to soyabean production (Hartman, 1999). Offei et al. (2008) published a catalogue of diseases of soyabean in Ghana. This compilation of diseases did not indicate distribution of the stated diseases across the soyabean production centres. Also, the incidence and severity of diseases catalogued were not indicated, making it difficult to assess their relative importance. In the past, Ghana largely relied on IITA for soyabean germplasm. However, with the inception of the Soyabean Innovation Laboratory (SIL) project in Ghana, the USA has also become a direct source of soyabean germplasm to the country. Myriad of soyabean diseases have been reported in the USA that have not been observed in Ghana. The spread of soyabean production to new areas, receiving germplasm from IITA and the USDA, in addition to projected climate change scenario creates a ripe condition for the introduction of new pathogens in to the country whilst existing ones may become more virulent. It is therefore important to routinely survey soyabean production centres in the country for diseases, their incidence and severity. Knowledge of crop diseases and their distribution in a country or across production zones is key to developing effective strategies for their control. Information on the distribution, 26 University of Ghana http://ugspace.ug.edu.gh incidence and intensity of crop diseases is adduced through disease surveys (Elizondo et al., 2014; Kebede and Gidesa, 2017). Therefore, surveying soyabean production centres for diseases will aid in formulating appropriate disease management strategies including breeding for resistance. The goal of this research activity was to identify dominant diseases of soyabean to inform the development of management strategies including the use of resistant genes. The specific objectives were to; 1. identify diseases on cultivated soyabean fields in major production areas in Northern, Upper East, Upper West, Ashanti, and Brong Ahafo regions 2. estimate the incidence and severity of identified diseases using visual assessment; and 3. identify new and unidentified diseases under laboratory conditions. 27 University of Ghana http://ugspace.ug.edu.gh 3.2 Materials and Method 3.2.1 Study area and time A survey was undertaken in the major soyabean production areas in Ghana, most of which are in the Northern region. Also, production areas were selected to account for some of the different ecological zones in which soyabean is cultivated in Ghana. The study was carried out from the third week in September to the first week of October 2016. This period was chosen to ensure that plants were in advanced reproductive stage. Most soyabean fields in the three regions of northern Ghana are established between the last week of June and the first week of July. This sowing period ensures that maturity and harvesting occur in the months of October and November. Districts of interest were pre-selected based on knowledge of soyabean production adduced from scientists at CSIR-SARI. However, Ejura and Wenchi were selected because they occur within a different agro-ecology. Both Wenchi and Ejura occur in the forest savannah transition belt and therefore have bimodal rainfall pattern whilst the remaining locations are found either in the Guinea savannah or the Sudan savannah, both of which have monomodal rainfall pattern. These two districts have higher annual rainfall and lower mean annual temperature than the remaining districts surveyed. In total, fifteen (15) locations in nine districts/ municipalities were visited for the selection of soyabean farms for disease survey. Most of the locations including Malzire, Sang, Karaga, Gushegu, Akukayili, Tingoli, Seripe, and Damango constituting eight out of a total of fifteen locations surveyed were located in the Northern region. Wa and Tumu were the only two locations surveyed in the Upper west. Three locations, Navrongo, Binduri and Manga were surveyed for soyabean diseases in the Upper east region. In the Ashanti region, only one location, Ejura, was surveyed. Similarly, Wenchi was the only the location surveyed in the Brong Ahafo. The selection of Ejura and 28 University of Ghana http://ugspace.ug.edu.gh Wenchi was because they represent an entirely different agro-ecology. Also, scientists at SARI have begun multi-location evaluation trials in these areas with the hope of promoting soyabean production among farmers at Ejura and Wenchi 3.2.2 Field selection Selection of locations and farms within districts were carried out with the assistance of staff of the Ministry of Food and Agriculture (MoFA) within the selected districts. Prior arrangements were made with district officers of MoFA who know soyabean farmers and their soyabean fields. Upon arrival at a location, soyabean farms were randomly selected with the assistance of MoFA officers. At every location, at three farms at 1 km or more apart were selected for the survey. 3.2.3 Identification of diseases on selected fields Selected fields were first traversed to identify the prevailing disease(s). The compendium of soyabean diseases and pests (Hartman et al., 2015), and a hand lens (Opticron, 23 mm doublet) were used to identify diseases on the field through visual assessment. Pictures displaying the symptoms and signs of various diseases in the compendium of soyabean diseases and pest were compared to those observed on the field. A hand lens of 10x magnification was used to magnify signs and symptoms for confirmation of diseases on the field. For diseases that could not be identified on the field, plant parts harbouring symptoms and or signs of such unidentified diseases were harvested into zip lock bags and placed in an ice chest for transportation to the pathology laboratory at SARI, Nyankpala, for identification. 3.2.4 Sampling method Sampling units within farms was based on the method described by James (1971) and modified by Elizondo et al., (2011). Two diagonal transects was made across each selected farm. Plants for disease assessment were selected at 5-10 m interval along the diagonal transects based on 29 University of Ghana http://ugspace.ug.edu.gh the size of the farm. At each interval, 20 plants were randomly sampled for estimation of disease incidence 5 of which were assessed for disease severity. Disease incidence was estimated by counting the number of diseased plants and expressed as a percentage of the 20 randomly selected plants at each 5-10 m interval along the diagonal. The estimation of disease incidence is mathematically written as indicated: Incidence (%) = Severity of identified diseases were visually scored using standard diagram displaying severity on a1-9 scale (Subrahmanyam et al., 1982) where 1=no lesions; 2= few lesions in the lower canopy, 3= lesions on most leaves in the lower canopy; 4= lesions on most leaves on lower canopy and few lesions on upper canopy; 5= Both lower and upper canopy are completely covered by disease; 6= lesions cover leaves, petioles and pods with 5; 7= 25% defoliation; 8= 50% defoliation; 9=complete defoliation/ death of plant. 3.3 Confirmation of Diaporthe sp. causing soyabean seed decay in Ghana 3.3.1 Sampling of diseased tissues or organs Blighted pods and stems showing pycnidia were harvested into Ziploc bags wherever they were encountered during the survey. Matured pods were manually threshed for visual assessment of seeds. The seeds were visually assessed for symptoms and signs of seed decay including wrinkling, shrivelling, molding, mottling, discolouration, cracking and splitting. Seeds with any of the above symptoms were selected from the seed lot and cultured. At SARI’s experimental field in Nyakpala, soyabean fields were monitored until maturity. Upon physiological maturity, pods were sampled from plants in both experimental and non- experimental soyabean fields. Pods obtained from Nyankpala were also manually threshed and the resulting seeds visually assessed as those obtained from disease incidence survey. In all, 30 University of Ghana http://ugspace.ug.edu.gh 150 pods yielding 282 seeds were visually assessed for signs and symptoms of seed decay including moldiness, mottling, distortion, shrinkage, and discolouration. Seeds and pods with signs and symptoms reminiscent of seed decay due to Diaporthe sp. were cultured for the development and isolation of pathogens for further analysis. 3.3.2 Isolation of disease-causing organism Potato dextrose agar (HIMEDIA) was modified with lactic acid to inhibit bacteria growth during pathogen isolation from soyabean seeds. PDA of 15g was suspended in 1litre of distilled water. The resultant was heat sterilised by autoclaving at 121°C for 15 minutes. The prepared medium was acidified by amending it with 50% lactic acid (VWR International) to adjust the pH to 4.8. The acidified PDA (APDA) was poured into 9 cm- diameter petri dishes. In some cases, diseased seeds were first cultured on water agar HIMEDIA and fungal growth on water agar was sub-cultured on APDA. Symptomic seeds were surface disinfected with 70% ethanol, sterilised in 1 % sodium hypochlorite for 2 minutes and then rinsed in distilled water twice. Sterilised seeds were plated on the prepared APDA medium and incubated at a temperature range of 25-30 ℃ and 12 hr light/ 12hr dark. Cultures were monitored regularly for fungal growth. Mycelia emerging out of seeds was sub-cultured to obtain pure isolates of various fungi growing from the plated seeds. Pure isolates were incubated for 15 days with regular examination of morphology (colour, macrostructures/fruiting bodies) and microscopic features (spores). 3.3.3 Microscopy Cultures which morphology conformed to fungi of interest were separated from other plates and further confirmed through microscopy for mycelia and spore architecture. PDA microslides were prepared with mycelia of isolated pathogen and incubated at 28 ± 2 oC for 7 days. Cover slides of microslides and media were examined for microscopic feature using 31 University of Ghana http://ugspace.ug.edu.gh Leica DM 1000 microscope at a magnification of 40 × and 60 ×. 3.3.4 Primer Design The National Centre for Biotechnology Information (NCBI) database (https://www.ncbi.nlm. nih.gov/nucleotide/) was searched for deposited sequences of members of the Diaporthe- Phomopsis complex of fungi. Partial sequences of 28S Ribosomal RNA genes of isolates of Diaporthe phaseolorum var sojae, Diaporthe phaseolorum var caulivora and Phomopsis longicolla. Nucleotide sequences were used to design primers for each tentatively identified fungi by loading obtained sequences into the software, primer3 (Untergasser et al., 2012). Parameters including annealing temperature range (57-60 ℃), desired number of nucleotides (18-25) and GC content (40-60 %) were set for the software for the generation of primers from nucleotide sequences of ribosomal DNA of suspected fungi. The designed primers were sent for synthesis at Inqababiotec in South Africa. 3.3.5 DNA extraction Each tentatively identified pathogenic fungus was cultured in four petri dishes representing four replicates of each fungus. Matured fungi mycelium of 15 days old were used for the DNA extraction. The fungi mass was scraped off the media using a spatula. The scraped fungi mass was grounded in a mortar with a pestle and subjected to DNA extraction using the protocol by Doyle & Doyle, (1990) with slight modification. 3.3.6 DNA quality test The quality of extracted DNA of each sample was quality checked by performing electrophoresis in a 2 % agarose gel. Samples with no bands, smears or very faint bands were discarded and extractions carried out on them again to obtain very good bands for further analysis. 32 University of Ghana http://ugspace.ug.edu.gh 3.3.7 PCR amplification and gel electrophoresis Designed primers synthesised by Inqaba biotec were used to amplify extracted DNA of fungi. PCR amplification was carried out in a 10µL reaction volume, consisting of 5µL of 2X one taq PCR master mix, 3µL of nuclease free water, 1 µL of forward and reverse primer and 1 µL of DNA. The PCR was carried out using the ABI thermal cycler (Applied Biosystems by life Technologies) with the conditions; initial denaturation at 94 °C for 3min, 35 cycles of denaturation at 94 °C for 30 secs, annealing at X °C for 30 secs (depending on primer) and extension at 72 °C for 30 secs. It was finally extended at 72 °C for 7 min and held at 4 °C. To test the fidelity of primers designed and synthesised for each fungus, soyabean DNA was taken through the amplification process using designed fungi primers. Additionally, primers designed for a specific fungus were tested on other suspected fungi as a way of determining if DNA amplification of primers were specific to the fungi to which they were designed. 3.3.8 Polyacrylamide gel electrophoreses of amplified DNA Amplified products were electrophoresed using a 6 % polyacrylamide gel on a horizontal PAGE system (Cleaver Scientific). Band separation was carried out at a voltage of 120 V for three hours. A peristaltic pump (Williamson manufacturing company LTD, 201-AQUA- SC050) was used with the PAGE system to enhance even distribution of running buffer and to maintain equilibrium in conductivity of the buffer. After electrophoresis, the gel was stained with ethidium bromide for 30min and visualized using a UV trans-illuminator (UVP, Benchtop UV transilluminator). A photograph of the gel was then taken for further analysis. 3.3.9 Sequencing of amplified DNA products Photographed gels were scored for bands and position of bands for each fungus. Fungi which DNA amplification by its primer was specific to it alone and yielded the same amplicon size for all its four replicates were selected as putatively confirmed. The PCR product (amplicon) and primer sequences of confirmed pathogenic fungi were sent to Inqaba biotec for sequencing 33 University of Ghana http://ugspace.ug.edu.gh of the amplicon in order to obtain partial sequences of the confirmed pathogens using the Sanger sequencing method. Nucleotide sequences of the amplicon were loaded into the nucleotide Basic Alignment Search Tool (BLAST) within the online database, NCBI in search for similar sequences for the identification of the fungus isolated from infected soyabean seeds. Sequences of isolates, mainly, Diaporthe and Phomopsis species obtained from NCBI were aligned with amplicon sequence and used to construct a phylogenic tree based on the weighted neighbour joining tree in the Darwin software (Perrier and Jacquemoud-Collet, 2006). Sequence dissimilarities were calculated at 1000 bootstraps or iterations. 3.3.10 Data collection and statistical analysis Disease incidence and severity scores recorded for each farm were tabulated for each location. For each disease identified, the data from farms were used as replicates for location. The data was then subjected to analysis of variance (ANOVA) using the GenStat software (version 18). Where differences in diseases incidence, and severity differed significantly between locations, means were separated using Student-Newman-Keuls (SNK) method. 3.4 Results 3.4.1 Results of disease survey A total of 60 fields were surveyed at 15 locations spread across five regions and three agroecologies. Ten diseases were identified during field survey (Fig 3.1). These diseases were cercospora leaf blight (A) due to Cercospora kikuchii; frog eye leaf spot due (F) cause by Cercospora Sojina; target spot (G) due to the pathogen, Corynespora cassiicola; bacterial leaf pustule (B) due to Xanthomonas campestris pv. glycine; downy mildew (H) due to Peronospor manshurica; leaf blight (D) due to Rhizoctonia solani; brown spot (E) due to Septoria glycine; soyabean mosaic (C) due to virus, soyabean seed decay(I) and alfafa mosaic virus (J). 34 University of Ghana http://ugspace.ug.edu.gh Additionally, there were evidence of sudden death syndrome, root rot, charcoal rot, and web blight diseases but they were localised on isolated fields at isolated locations. Fig 3. 1: Diseases identified during disease survey: A (Cercospora leaf blight); B (Bacterial leaf pustule); C (Soyabean mosaic virus); D (Rhizoctonia leaf blight); E (Septoria brown spot); F (Frog eye leaf spot); G (Target spot); H (Downy mildew); I (Seed decay); J (Alfalfa mosaic virus). The distribution, incidence and severity of four diseases that were most consistent across locations are indicated in Table 3.1. Incidence of bacterial leaf pustule due to Xanthomonas campestriz pv. glycine ranged from a locational average of 15% for Wenchi to 84% for Malzire in the Yendi Municipality. The incidence of bacterial leaf pustule in Malzire was significantly higher than that of Sang in the Yendi municipality; Akukayili in the Tolon district; SARI research station in the Wa municipality; Tumu in the Sissala west district, and Navrongo. However, the 84% incidence of bacterial leaf pustule recorded in Malzire was statistically comparable to that of Tingoli in the Tolon district; Gushegu and Karaga in the Gushegu-Karaga district; Seripe in the Bole-Bamboi district; Damongo in the Gonja west district, and Manga in the Bawku district. The severity of bacterial leaf pustule was highest in Malzire, and lowest in Wenchi. Malzire, Karaga, Gushegu, Tingoli and Akukayili, recorded statistically comparable levels of bacterial leaf pustule severity. 35 University of Ghana http://ugspace.ug.edu.gh The incidence of Cercospora leaf blight ranged from 22.5% to 80.1%. Karaga recorded Cercospora leaf blight incidence of 80.1% which was statistically similar to that of Gushegu, Akukayili and Navrongo but higher than the remaining locations. The least incidence of Cercospora leaf blight was 22.5% which occurred at Tingoli. Similarly, severity of Cercospora leaf blight was highest at Karaga which recorded severity was higher than the remaining 14 locations. The next location with high severity was Akukayili. Binduri, Gushegu Navrongo, and Manga were similar in Cercospora leaf blight severity. Cercospora leaf blight due to Cercospora kikuuchi did not occur at Wenchi and Ejura. Karaga, Gushegu, Damongo, and Binduri recorded frog eye leaf spot disease incidence of 82%, 79.7%, 80%, and 72.3%, respectively. Although, Karaga recorded the highest incidence of Frog eye leaf spot, it was statistically comparable to that of Gushegu, Damongo and Binduri. There was no occurrence of frog eye leaf spot at Seripe in the Bole-Bamboi district, and at Tingoli in the Tolon district. Damongo, and Gushegu recorded frog eye leaf spot severity of 4.4 and 4.2, respectively. The two locations were similar in frog eye leaf spot severity and statistically higher than all other locations surveyed in this study. The least incidence and severity of frog eye leaf spot occurred at Ejura and Wenchi. Though it occurred at lower levels at most locations, viral diseases of soyabean occurred at all 15 locations that were surveyed. Ejura recorded the highest incidence of soyabean mosaic virus (8.40%). This was followed by Wenchi where average incidence of viral diseases of soyabean was 38.2%. The least incidence of soyabean mosaic virus occurred at Damongo. 36 University of Ghana http://ugspace.ug.edu.gh Table 3.1:Incidence and severity of diseases identified during survey Bacterial leaf pustule Cercospora leaf blight Frog eye leaf spot District Location (Xanthomonas campestris Mosaic virus (Cercospora kikuchii) (Cercosporasojina) pv. Glycine) DI (%) DS DI (%) DS DI (%) DS DI (%) Yendi Malzire 84.7a 4.9a 61.1e 3.5c 50.1d 2.8cd 9.57 Yendi Sang 46.27c 2.0bc 32.1g 1.7f 50.8d 2.7cd 11.8 Karaga/Gushegu Karaga 66.6abc 3.4abc 80.6a 5.3a 82.3a 3.5b 5.4 Karaga/Gushegu Gushegu 60.1abc 3.9abc 75.2c 3.4c 79.7a 4.4a 10.4 Tolon Akukayili 49.33bc 3.9ab 80.1ab 4.5b 13.2f 1.4h 21.6 Tolon Tingoli 65.8abc 3.3abc 22.5h 1.9f - - 5.2 Bole Bamboi Seripe 63.2abc 2.9bc 67.9d 2.3e - - 4.4 Gonja west Damango 70.1abc 1.83bc 55.3f 1.87ef 80.8a 4.2a 3.0 Wa Mun. Wa 53.0bc 2.3bc 39.6g 1.6f 42.3e 1.7gh 2.38 Sissala east Tumu 43.3c 1.9bc 39.4g 1.5f 38.8e 2.1fg 6.2 69.10 abc 2.1bc 3.3c Binduri Binduri 65.1e 72.3a 3.2bc 7.46 Kassana Nakana Navrongo 52.0bc 1.7c 77.2ac 3.3c 49.2d 2.4df 5.0 Bawku Manga 74.7ab 2.7bc 74.3c 2.8c 63.c 2.0fg 9.0 Ejura Ejura 18.3d 1.5c - - 4h 1.3h 80.4 Wenchi Wenchi 15.3d 1.4c - - 7h 1.2h 38.17 Note: numbers in the same column followed by the same alphabet are statistical similar at p=0.05 DI=disease incidence; DS=disease severity 37 University of Ghana http://ugspace.ug.edu.gh 3.4.2 Results of pathogen confirmation study 3.4.2.1 Colony characteristics From the 282 symptomatic seeds plated from the Nyakpala seed lot, 40% o had fungal growth by what was morphologically determined to be Diaporthe sp. [Lehman]. White aerial mycelium emerged from plated seeds as indicated in plates D&E in Fig 3.2. The growth was not fluffy, but slightly raised and yellowish reverse plate colour. The growth was slow, taking more than 15 days to completely cover the 9cm-diameter petri dish. A matured pure culture had a whitish mycelium with dark visible oblong fruiting structure forming after 10 days known as pycnidia. The pycnidia emerged from the aged culture with shot necks. Colony colour changes from white to gray which darkens with time as indicated in plate F of (Fig 3.1). the reverse plate however showed a yellowish colour in the middle of the culture. Microscopic examination showed that both alpha and beta conidia were produced as shown in plate B of (Fig 3.1). Puncturing pycnidia with a needle on a moist microslide under microscope showed oozing of abundant spores. F Fig 3.1: Pycnidia on soyabean stem(A); (B)Microscopic view of punctured pycnidia oozing spores; (C) culturing diseased seeds in water agar; (D) soyabean seeds plated on APDA with initial growth of Diaporthe sp; (E) A plate showing the distinction between diaporthe sp colony as indicated by the two arrows from fusarium sp colony which is the cotton like growth between the two diaporthe sp colonies (F) Displays the changes in colour of diaporthe sp as the culture ages (G) Pycnidia producing culture. 38 University of Ghana http://ugspace.ug.edu.gh 3.4.2.2 Molecular identification of causal pathogen SSR primer (MB-DpS1) with the pair, F-GGCCCGAGTTGTAATTTGCA; RCGGTCTCTGGCCGGTATTTA designed from the partial sequence of 28S Ribosomal RNA gene of Diaporthe phaseolorum var sojae isolate, PS/KR3 successfully amplified the DNA of the pathogen that infected 40% of the seeds plated with the same band size, an indication of the high fidelity of the primer, MB-DpS1. The Genebank ID of isolate PS/KR3 is JF704169.1(https://www.ncbi.nlm.nih.gov/nuccore/JF704169.1). The Primer pair of MBDpS1, amplified all 4 replications of the pathogen that was morphologically characterised as Diaporthe sp. but not the check DNA (Fig 3.3). L R1 R2 R3 R4 L 1 2 4 Fig 3. 2: Gel electrophoresis of 4 replicates of amplified 28S RNA gene of MTB-1 and the check (C). The sequenced amplicon yielded 164 base pairs (bp) in size. At a query coverage of 99% and expected value (E-value) of 8e-74, the DNA sequence of the amplicon had a 99% match with more than 25 Diaporthe/Phomopsis species including Diaporthe isolates CBS.180.55; Dioporthe sojae isolate CBS.179.5; Diaporthe Sp.NVL.2017; Diaporthe longicolla isolate ER1679 and Diaporthe sp. HRP-2018b. 39 University of Ghana http://ugspace.ug.edu.gh Phylogenetic analysis placed the 15 isolates of Diaporthe/Phomopsis species into three major clades (Fig 3.4). The constructed phylogenetic tree placed the Ghanaian isolate of Diaporthe sp (MTB-1) in major clade I, made up of three P. longicolla isolates, two of which were not genetically distinct and two Diaporthe species, one of which is a Diaporthe sojae (CBS 179.55 ) which occurred in the same subclade with MTB-1(Fig 3.4). The only other Diaporthe sojae isolate, CBS 180.55 occurred in major clade II with three isolates simply labelled as Phomopsis species, and one Phomopsis longicolla isolate (Fig 3.4). The major clade III contains four isolates two of which were not genetically distinct and one Diaporthe phaseoolorum var caulivor isolate (720) which was the most genetically distant isolate in major clade III (Fig 3.4). In all, no major clade was made of isolates of a single species. Most major clades contain sub-clades contain both Phomopsis/Diaporthe species. 40 University of Ghana http://ugspace.ug.edu.gh I II III Fig 3. 3: A phylogenetic tree showing the relatedness of the isolated fungus (MTB-1) to 14 isolates of Diaporthe and Phomopsis species obtained from NCBI 41 University of Ghana http://ugspace.ug.edu.gh 3.5 Discussion Ten diseases of soyabean including, cercospora leaf blight (Cercospora kikuuchi), frog eye leaf spot (Cercospora sojina); target spot Corynespora cassiicola), bacterial leaf pustule (Xanthomonas campestris pv. glycine), downy mildew (Peronospor manshurica), leaf blight (Rhizoctonia solani), brown spot (Septoria glycine), soyabean mosaic virus, alfafa mosaic, and soyabean seed decay due to Diaporthe sp. of fungi were revealed by the survey. Among the diseases of soyabean identified, only Bacterial leaf pustule and soyabean mosaic virus occurred at all locations surveyed. The incidence of bacterial leaf pustule recorded in this study ranged from 15.3% to 84.7%. This result was slightly higher than what was reported by Zinsou et al., (2015) who recorded bacterial leaf pustule incidence in the range of 15.8% to 70% during a survey of bacterial leaf pustule in the Guinea savannah agro- ecology of Benin. The occurrence and severity of bacterial leaf pustule at Malzire, Karaga, Gushegu, Tingoli and Akuyili were higher than all other locations surveyed. Although, Damongo, Binduri and Manga recorded high incidence of bacterial leaf pustule, the corresponding severity scores were lower. All the locations with high bacterial leaf pustule severity occurred in the Guinea savannah agro- ecology whilst the locations that had high incidence but lower severity occurred in the Sudan savannah agro-ecology except Damongo, which also lies in the Guinea savannah zone. The infection and development of Xanthomonas campestris pv. glycine, causal agent of bacterial leaf pustule is favoured by humid and warm environmental conditions (Bradbury, 1986; Heikmap et al., 2014). Bacterial leaf pustule has therefore been described as an important disease to soyabean production in the tropics (Hokawat, 1993) especially in the Savannah regions where high humidity and temperature characterise the production season. For instance, rainfall, mean temperature and relative humidity of 846.8mm, 28.1 ºC and 78%, 42 University of Ghana http://ugspace.ug.edu.gh respectively were recorded for the 2016 cropping season (June to October) in the Tolon district which harbours Akukayili and Tingoli. The high incidence but lower severity scores obtained at Damongo, Binduri and Manga may be due to late onset of the disease or absence of conducive environmental conditions arising from the different times of planting to support development of the pathogen. Resistant and tolerant cultivars could have equally been the reason for the high bacterial leaf pustule incidence but lower severity at Damongo, Binduri and Manga but most farmers across the surveyed locations claimed to be cultivating the same varieties obtained from the Savannah Agriculture Research with Jenguma being the most cultivated variety. The least bacterial leaf pustule incidence and severity scores were recorded at Ejura (18.3% & 1.5) and Wenchi (15% &1.4) which occurs in the Forest transition zone of Ghana with bimodal rainfall, and therefore has two cropping seasons in a year. Most of the fields surveyed at Wenchi and Ejura were planted in the second cropping season, and therefore most plants were at the vegetative state or at the R2 stage. The stage of soyabean plants at Wenchi and Ejura stood in sharp contrast to stage of soyabean plants surveyed from remaining locations which ranged from R5-R6 growth stages. For instance, at Ejura, of the five soyabean fields surveyed, only two fields were at the R5 stage. The two farms at the R5 stage were Genotype-Environmental trial plot carried out by SARI’S soyabean unit at Ejura. The remaining three farms were ADRA soyabean –cereal intercropping demonstration trial in which soyabean plants were at yet to reach the flowering stage. Also, there were fewer soyabean farms at Wenchi and Ejura compared to the other locations surveyed. Cercospora leaf blight caused by Cercospora kikuuchi occurred at 13 of 15 locations surveyed. There was no incidence of Cercospora leaf blight at farms surveyed in Wenchi and Ejura. The highest incidence and severity of Cercospora leaf blight occurred in farms at 43 University of Ghana http://ugspace.ug.edu.gh Karaga, Akukayili Gushegu and Malzire all in the Northern region. The wide variation between Malzire and Sang in Yendi municipality, and Tingoli and Akukayili in the Tolon district in terms of cercospora leaf blight incidence and severity is a demonstration of fluctuations among farms within locations for the occurrence of diseases. The three locations, Navrongo, Binduri and Manga in the Upper east region had similar level of cercospora leaf blight severity. Cercospora kikuchii was considered the fifth most important yield– depressing pathogen in the world (Sinclair and Hartman, 1999). Another disease with wide distribution across the surveyed locations is the frog eye leaf spot disease due to Cercospora sojina, a pathogen that is closely related to Cercospora kikuuchi. Frog eye leafspot is an important constraint to soyabean production world-wide, especially in warm and humid production centres (Ma, 1994; Mian et al., 2008; Pharm et al., 2015). The occurrence of frog eye leaf spot on soyabean in Ghana has been reported in previous publication on soyabean diseases (Offei et al., 2008). Like bacterial leaf pustule, warm and humid conditions heighten the development of frog eye leaf spot. However, warm and humid conditions largely defines the soyabean production season in Ghana, which largely occurs in the Northern region where 77% of the country’s soyabean is produced. The incidence and severity of frog eye leaf spot disease was high at Karaga, Gushegu and Damongo. Among the three locations surveyed in the Upper east region, frog eye leaf spot incidence was high at Navrongo but not at Manga and Biduri. Although, Navrongo, Binduri and Manga are found in the Upper east region, they occur in different agro-ecological zones. Navrongo is located in the Guinea savannah ecology whilst Binduri and Manga are in the Sudan Savannah ecological zone. Wenchi and Ejura recorded the least incidence and severity of frog eye leaf spot disease among the 15 locations surveyed whilst Tingoli and Seripe had no incidence of frog eye leaf spot. 44 University of Ghana http://ugspace.ug.edu.gh Known sources of resistance to Frog eye leaf spot have been reported including PI594774 and PI594891 (Pham et al., 2015). Also, the Brazillian cultivars, Bossier, Cristalina, Davis, Kent, Lincoln, Paraná and Uberaba are known sources of resistance to frog eye leaf spot (Gravina et al., 2003). Pham et al., (2015) fine mapped the resistance to frog eye leaf spot in PI594774 to 72kb on chromosome 13 and developed Kompetitive allele specific PCR(KASP) for detection of the frog eye leaf spot resistant allele on chromosome 13. These lines could be evaluated at the locations with high incidence and severity of frog eye leaf spot identified in this survey for their suitability for use as sources of resistance for cultivar development in Ghana. Two types of viral diseases, soyabean mosaic and alfafa mosaic virus were present at all 15 locations surveyed. However, their incidences were very low at most locations except for Ejura and Wenchi. Ejura recorded the highest incidence of soyabean mosaic virus followed by Wenchi. Viral diseases are responsible of significant yield losses in major soyabean production centres such as the United States of America. For instance, an estimated 35 million United States dollars was lost to viral diseases in 2010 (Wrather and Koenining, 2010; USDA 2013). However, Hill and Whithman, (2014), suggested that losses due to viral diseases might be underestimated since viruses may be latent and still cause losses. Therefore, it is important to initiate management of virus caused diseases even if the incidence and severity are as low as recorded at most locations surveyed. Fungi constitute the second largest group among eukaryotic organisms (Raja et al., 2017) and responsible for most diseases of plants worldwide (Knogge, 1996). Three approaches including morphological characterisation, characterisation of secondary metabolite, and molecular characterisation are used separately, or in combinations for fungal identification (Raja et al., 2017). In this study, I used to some limited extent morphology characteristics, 45 University of Ghana http://ugspace.ug.edu.gh and to a large extent, molecular characteristics to identify a fungal species causing seed decay of soyabean in Ghana. Colony morphology of the fungus isolated in this study bear key characteristic known to members of the diaporthe/phomopsis complex of fungi which includes white areal mycelia that changes colour as the culture ages, and the presence of black pycnidia from which conidia is produced. The isolate produced both alpha and beta conidia which is characteristic of Diaporthe phaseolorum var. sojae and Phomopsis longicolla but not D. phaseolorum var. caulivora. Pyncnidia produced has short neck which is common to Diaporthe phaseolorum var. sojae but not to Phomopsis longicolla (Petrović et al., 2016). Mengitsu et al. (2014) noted that morphological characteristics were not sufficient to make distinction between members of the diaporthe/ phomopsis complex of fungi. Alignment of 28S RNA partial sequence of MTB-1 in BLAST yielded the same level of similarity to two isolates of Diaporthe sojae and five isolates of Phomopsis longicolla. Similarly, the phylogenetic tree based on 28S RNA gene sequence placed the newly isolated fungus in a major clade comprising of three isolates of Phomopsis longicolla, and one isolate of Diaporthe sojae. It is noteworthy that the Ghanaina isolate MTB-1 occurred in the same sub-clade with Diaporthe sojae isolate CBS 179.55, an indication that isolate MTB-1 is genetically closely related to Diaporthe sojae than other Diapothe/Phomopsis species used in this study. The pairing of Diaporthe/Phomopsis isolates in the same major and sub-clades in some cases as indicated in the phylogenetic tree may be an indication of the confusion that dogged identification members of the Diaporthe/ Phomopsis complex of fungi (Santos et al., 2011; Rensburg et al., 2006). Alignment and clustering make it certain that newly identified pathogen is a member of the Diaporthe sojae species. 46 University of Ghana http://ugspace.ug.edu.gh 3.6 Conclusions Disease survey identified ten diseases seven of which were fungal diseases. Two viral diseases and one bacterial disease were identified in addition to the seven fungal diseases. Of ten diseases identified during the survey, bacterial leaf pustule, cercospora leaf blight, frog eye leaf spot and two viral diseases were present at varying levels of incidence and severity at almost all 15 locations surveyed. Bacterial leaf pustule and soyabean mosaic virus occurred at all 15 locations surveyed whilst frog eye leaf spot and cercospora leaf blight occurred at 13 of 15 locations surveyed. The severity of identified diseases was higher at Malzire, Karaga and Gushegu than other locations with the exception of soyabean mosaic virus for which the highest incidence occurred at Ejura followed by Wenchi. Severity of bacterial leaf pustule, cercospora leaf blight and frog eye leaf spot was generally lower at the three locations surveyed in the Upper east region than locations surveyed in the Northern region. The two locations surveyed from the Transitional belt had the least prevalence and severity of diseases encountered during the survey except for soyabean mosaic virus for which incidence was highest at two locations occurring in the transitional belt. Furthermore, this study identified soyabean seed decay disease, and successfully identified the causal agent as Diaporthe sojae and tentatively designated the isolate as MTB-1. This is the first report of soyabean seed decay due to Diaporthe sojae in Ghana. Asante et al. (1998) reported the occurrence of Phomopsis longicolla and other Phomopsis species known for causing soyabean seed decay and as well as Diaporthe phaseolorum var. meridionalis, causal agent of southern stem canker of soyabean in Ghana but not Diaporthe sojae. 47 University of Ghana http://ugspace.ug.edu.gh Diseases of soyabean are widespread at major production centres for which detail evaluations including assessment of economic damage is required for formulation of management strategies such as the use of resistant genes. This research also provides knowledge of locations where specific diseases are prevalent. This will be useful for multi-location evaluation of sources of resistance. It could also serve as a guide for the extension service division of the Ministry for Food and Agriculture to direct resources at specific areas for famer education on management of specific diseases of soyabean. 48 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR 4.0 Genetic diversity, population structure, and identification of key phenotypic traits driving variation in the assembled soyabean germplasm 4.1 Introduction Genetic variability is useful for successful crop improvement programme. Thus, an understanding of prevailing genetic diversity within a population or a collection is key for planning an efficient plant breeding programme (Chandra et al. 2014). Soyabean production in Ghana has seen a steady rise from 50,000 MT in 2005 to 142, 000 MT in 2015 (MoFA-SRID, 2007 & 2016). However, the national average yield of 1.65 t/ha (MoFA-SRID, 2016) is far lower than the average of 3.38 t/ha for Brazil (Statista, 2018), and the 3.34 t/ha obtained in the US (USDA-NASS, 2018). Production and utilization of soyabean in Ghana received a boost through the ongoing Soyabean Innovation Lab project funded by USAID - Feed the Future Programme. Under this project, the Council for Scientific and Industrial Research-Savanna Agricultural Research Institute (CIR- SARI) received soyabean germplasm as part of its collaboration with the United States Department of Agriculture and the University of Illinois as part of an effort to genetically improve the crop. SARI, normally receives limited amounts of germplasm from only the International Institute of Tropical Agriculture (IITA), hence its gene pool was narrow with limited diversity. With the arrival of additional germplasm, it is important to re-evaluate the current germplasm holding for phenotypic and genetic diversity, and suitability for addressing abiotic and biotic challenges in major growing areas of the country (Buri et al., & Wakatsuki, 2010; Goldsmith, 2017). The usage of new sources of genetic variation to develop improved varieties is enhanced when the selection of parental genotypes is based on both agronomic value and genetic dissimilarity 49 University of Ghana http://ugspace.ug.edu.gh (Bisen et al., 2014). Therefore, estimation of variation and dissimilarity among soyabean germplasm assembled in Ghana will be required to effectively select parental lines for hybridization and development of improved varieties. Conventionally, dissimilarity among genotypes was estimated based on their uniqueness in traits such as flower colour, seed shape, and colour, growth habit and so forth referred to as morphological markers (Holbrook & Stalker, 2003). However, the methods for evaluating some of these traits for differentiating genotypes are subjective, making the morphological approach to diversity study less efficient. More so, the influence of environmental factors on gene expression and its consequent effects on the stability of morpho-agronomic traits affects the repeatability of experimental results and thus, the inferences drawn from such experiments (Holbrook and Stalker, 2003). The advances in molecular biology and biotechnology ushered in the era of molecular markers, especially DNA based markers which more efficient compared to morphological markers in the estimate of dissimilarity. According to Mullis & Faloona, (1987), the invention of the polymerase chain reaction (PCR) led to the development and utilization of PCR based markers such as simple sequence repeats (SSRs), single nucleotide polymorphism (SNP), random amplified polymorphic DNA (RAPDs) and amplified fragment length polymorphisms (AFLPs). SSR markers have been extensively used in genotyping and characterizing of soyabean (Bisen et al., 2014; Guan et al., 2010; Kuroda et al., 2006; Li et al., 2010). More recently, the soySNP50K iSelect BeadChip that contains 50,000 SNP markers have been developed for characterization of soyabean genotypes, and quantitative mapping of desired loci (Song et al., 2013). Increasingly, automated genotyping platforms and genotyping service providing companies are moving towards the use of SNP markers. 50 University of Ghana http://ugspace.ug.edu.gh Among SNP genotyping techniques, genotyping-by-sequencing offers an efficient method for genetic diversity studies, and estimation of population structure based on genomewide scale variation among test subjects (Poland and Rife, 2012). Bara et al., (2018) used 45, 507 SNP markers generated with GBS to decipher the relatedness among crested wheatgrass lines. In order to effectively utilize assembled germplasm in SARI’s gene bank, understanding of the population structure, genetic diversity, and key phenotypic trait driving variation among the assembled germplasm is required. The objectives of this study were to; • examine the grouping of the 230 lines based on growth and agronomic parameters and identify lines with high pod clearance • identify the most important traits based on contribution to total variation • determine the structure and genetic diversity using SNP markers generated through genotype- by-sequencing (GBS) 51 University of Ghana http://ugspace.ug.edu.gh 4.2 Materials and Methods 4.2.1 Plant materials Two hundred thirty (230) soyabean lines comprising of 126 genotypes introduced from the United States Department of Agriculture (USDA); 56 advanced breeding lines sourced from the International Institute for Tropical Agriculture (IITA); 29 Zimbabwean accession; 13 Brazilian accessions; 3 Chinese soyabean lines obtained from the US, and 3 Ghanaian varieties. 4.2.2 Study area The study was conducted at the experimental fields of the Savanna Agriculture Research Institute, Akukeyili in the Tolon district. The area is characterised by Ferric Luvisols of the Tingoli series which is well drained, devoid of concretions, and brown in colour (Atakora and Kwakey, 2016). The experiment was conducted in two seasons, 2016 and 2017. The year 2016 recorded a total of 846.8mm of rains spread across 54 days within the trial period which began in June and ended in October. Mean temperature of 28.1 ºC and mean relative humidity of 78% with average sunshine of 6 hours were also recorded for the same period. 2017 recorded 795.6mm of rainfall, mean temperature of 27.9, mean relative humidity of 75% and sunshine hours of 6.8 during the five months in which the trial took place. 4.2.3 Experimental design and field establishment The trial was laid in an alpha lattice design in three replicates with each replication comprising 23 blocks of 10 lines per block laid out in two replications. Each line was planted in a single row of 5 m long with a plant spacing of 5cm. Rows were planted at 50 cm apart. Phosphorus in the form of Triple Superphosphate (TSP) was applied at a rate of 60 kg of P2O5 at emergence. Weed control was manually done with hoes. 52 University of Ghana http://ugspace.ug.edu.gh 4.2.4 Data Collection and Analysis of morphological and phenotypic traits • The number of days to flowering was recorded for the number of days from planting to when 50% of record plants for each line had at least one flower opened. • The number of pods per plant was obtained by averaging the number pods counted on 20 plants for each line. • The number of days to maturity was recorded at physiological maturity (R8). • The number of seeds per pod was taken on whole plant bases where all filled pods of 20 plants were counted, manually threshed and seeds obtained. • The number of seeds per plant was obtained from a sample of 20 plants of each line. The values of yield component obtained were analysed using the R statistical software (version 3.5.1) (R Core Team, 2018). Genotypic effects wereestimated using a linear mixed- effect model (LMEM) with restricted maximum likelihood (REML) based on the equation below: 𝑦𝑖𝑗𝑘 = 𝑔𝑖 + 𝑟𝑗 + 𝑏𝑘 + 𝜀𝑖𝑗𝑘 … 𝑒𝑞𝑛 4. 1 Where yijk, gi, rj, bk and Ɛijk denote observation on genotype i in block k nested within replication j, the effect of genotype i, effect of replication j, effect of block k nested within replication k and the effect of random error [~N (µ = 0, σ2 = 1)]. The combined results from year 1 and 2 was first subjected to analyses of variance which result showed that year to year variation was not significant. Also, interaction between year and lines was not significant. Therefore, the model was fitted using year as replication, given four replications to increase statistical power. The model was fitted using the lme4 package (Bates et al., 2015) of R. 53 University of Ghana http://ugspace.ug.edu.gh To estimate the heritability of the various traits, eqn 4.1 was refitted with all terms considered random. Based on the criteria proposed by Yan (2014), heritability was estimated using the equations below: …. eqn. 4.2 2 𝜎𝑟 2 𝜎𝑏2 𝜎𝜀2 4σp2 = 𝜎 + + + … eqn. 4.3 nj nj nj Where 𝐻2 is the heritability estimate, is the genotypic variance, 𝜎 2𝑝 is the phenotypic variance, is the replication variance, 𝜎 2𝑏 is the block nested within replication variance, 𝜎 2Ɛ is the residual variance, and 𝑛𝑗 number of replications. Percent quantile plot was performed in XLSTAT (version 2018.1). Likewise, multivariate analysis including factor analysis, principal component analysis, and cluster analysis were performed using the XLSTAT software (version 208.1). 4.3 Sampling of plant tissue for genotyping-by-sequencing Ninty-six (96) lines were selected out of the 230 lines for genotyping. Selection was based on flowering and maturity groups, plant size, seed size and colour, and reaction to both bacterial leaf pustule and Phomopsis seed decay. Genotyping was done in 2018 after data from phenotypic and disease trials were used to select the 96 representative sample. Sample collection kit comprised a collection box, a cutting mat, tubes with perforated strip caps, a cutting tool and a bag of silica gel (desiccant). 54 University of Ghana http://ugspace.ug.edu.gh For each of the 96 selected lines, leaf tissues were obtained by cutting 8 leaf discs with the cutter. These were placed in a labelled tube, covered with perforated strip cap and a desiccant-containing bag fastened to it. The collected tissues were sent to LGC genomics in Berlin, Germany for genotyping-by-sequencing. 4.3.1 Diversity, and population structure analysis 158,587 SNPs across all 96 samples were received from LGC-genomic. This was filtered in TASSEL (version 5.2.50) to obtain 32,638 SNP marker. These markers fully covered at least 66% of the 96 samples with allele frequency at or above 10% and a minimum read count of 8 and were used to examine the structure and diversity of the 96 representative samples. The genotypes were assigned population IDs (1-6) based on their origin. Bayesian distribution analysis that deploys Markov chain Monte Carlo (MCMC) simulation was carried out in the STRUCTURE programme (Pritchard et al., 2000) to determine the number of subpopulation (structure) in the representative sample of 96 lines. Three independent runs were performed for each number of assumed subpopulations (K) ranging from 1 to 12, using the admixture model assuming that allele frequencies were correlated. A burn-in period of 50,000 and an MCMC run length of 100,000 for each run was performed according to the Evanno’s method (Earl and vonHoldt, 2012; Evanno et al., 2005) to estimate the best number of subpopulations (ΔK) that explains the structure of the genotypes. The run with the maximum log likelihood was used to assign genotypes into subpopulations based on their membership probability. Genotypes with membership probability ≥ 0.25 were put into the same subpopulation. The STRUCTURE software was used to estimate the fixation index (Fst) for each subpopulation based on the estimated “K”. 55 University of Ghana http://ugspace.ug.edu.gh 4.3.2 Dendometric analyses The relatedness of individuals within the 230 soyabean accessions was visualised in a neighbour joining hierarchical dendrogram in Darwin V6 (Perrier and Jacquemoud-Collet, 2006) based on molecular data. A phylogenetic tree was constructed using the neighbour- joining method based on a dissimilarity matrix generated using Euclidian distance. The analysis was performed at 2000 bootstraps with an average edge distance between bootstrapped trees being 0.6589. 4.4 Results 4.4.1 Distribution of quantitative traits The number of days to flowering ranged from 23 to 51 days with a mean of 38 days (Fig 4.1). 25% of the 230 lines evaluated flowered within 34 days upon planting whilst 75% reached the reproductive stage within 42 days. It took 47 days for most (95%) of the soyabean lines to flower whilst the latter 5% took from 47 to 51 days to move from vegetative growth into the reproductive stage of development (Fig 4.1). The number of days to flowering was largely controlled by genetic effect as reflected in a broad sense heritability estimate of 0.9 (Table 4.1). Fig 4.1: Distribution of 230 lines based on mean number of days to flowering 56 University of Ghana http://ugspace.ug.edu.gh The earliest maturing soyabean genotype attained physiological maturity in 82 days after planting whereas it took 138 days for the latest maturing genotype to attain physiological maturity (Fig 4.2). The mean number of days to maturity for the 230 genotypes was 108 days with a broad sense heritability estimate of 0.8 (Table 4.1). The lower 25% maturing lines matured between 82 to 100 days whilst 50% of soyabean genotypes matured in the range of 82 - 109 days. The upper 25% took between 115 to 138 days after planting to attain physiological maturity (Fig 4.2). Fig 4. 2: Frequency distribution of number of days to maturity of 230 lines The mean height at maturity of the 230 soyabean genotypes evaluated was 52.8cm. The shortest genotype had a height of 14cm whilst the tallest genotype was 96.8 cm (Fig 4.3). In all, the lower 25% of the genotypes had heights within the range of 14 to 37.1cm. The upper 25% of soyabean genotypes had heights between 64.8cm to 96cm (Fig 4.3). Broad-sense heritability estimate (H2) for plant height at maturity was 0.49 (Table 4.1) 57 University of Ghana http://ugspace.ug.edu.gh Fig 4.3: Distribution of plant height at maturity of 230 lines The number of primary branches per plant ranged from 2-to- 21 with a mean of 7 branches per plant. 75% of the 230 soyabean genotypes studied had less than 8 primary branches per plant. Only 5% of the 230 lines had greater than 10 primary branches per plant (Fig 4.3). The number of primary branches per plant recorded a low broad sense heritability of 0.1 (Table 4.1). Fig 4: 4: Distribution of mean number of primary branches of 230 soyabean genotypes Height at first pod was measured as the height of plant from the base to the first pod on the main stem. It ranged from 1.7 to 53 cm with a mean of 10.cm (Fig 4.5). 50% of the genotypes studied had a height at first pod of less than 10 cm. 25% recorded between 11 to 53cm in height from the base to the first pod on the main stem with 5 % of them recording a height of 58 University of Ghana http://ugspace.ug.edu.gh greater than 20cm from the base to the first pod on their main stem (Fig 4.5). The heritability estimate for plant height at first pod was 0.76 (Table 4.1). Fig 4.5: Distribution of plant height at first pod of 230 lines The 230 soyabean genotypes investigated in this study produced pods per plant ranging from 17 to 290 pods, and a mean value of 72 pods (Fig 4.6). The lower 25% of the genotypes bore less than 50 pods per plant whereas the top 25 % produced more than 86 pods per plant. The top 10% recorded more than 120 pods per plants. On the other hand, the lower 10% of the 230 soyabean produced from 17 to 35 pods per plant (Fig 4.6). Number of pods per plant had a broad sense heritability of 0.7 (Table 4.1). 59 University of Ghana http://ugspace.ug.edu.gh Fig 4.6: Distribution of mean number of pods per plant of 230 lines The number of soyabean seeds in a pod ranged from 1.4 to 2.6 seed and a mean of 1.9 seeds in a pod with a broad sense heritability estimate of 0.6 (Fig 4. 7 & Table 4.1). The lower 25% of the genotypes recorded from 1.4 to 1.7 seeds in each pod whilst the upper 25% of the genotypes possessed from 2.1 to 2.6 seeds per pod. The top 10% in terms average number of seeds in a pod recorded from 2.2 to 2.7 seeds in a pod. On another hand, the bottom 10% recorded on the average 1.4 to 1.6 seeds per pod. 50% of genotypes evaluated had on the average 1.8 seeds per pod whereas 75% recorded an average of 2.1 seeds in a pod (Fig 4.7). Fig 4: 7.Distribution of 230 lines based on mean number of seeds per pod 60 University of Ghana http://ugspace.ug.edu.gh The average number of seeds per plant ranged from 36 to 484 seeds with an average of 129 seeds for the 230 soyabean lines studied (Fig 4.8). The lower 10 percent of the lines, in terms of the number of seeds per plant, had from 36 to 59 seeds whereas the top 10% yielded between 200 and 484 seeds (Fig 4.8). The top 25% produced more than 140 seeds whilst the lower 25% recorded in a range of 36 to 84 seeds per plant (Fig 4.8). The number of seeds per plant was moderately heritable as reflected in the broad sense heritability estimate of 0.7(Table 4.1). Fig 4. 8: Percentile plot showing distribution of number of seeds per plant of 230 lines Seed weight of soyabean weighed in a lot of 100 seeds (100-seed weight) ranged from 6.7 to 24 g with a mean of 14.8 g for the 230 genotypes investigated (F4.9). 25, 50, and 75% of the materials studied had a seed weight of 12.8, 14. 8 and 16.6, respectively (F.9). The lower 10% of the genotypes had seed 100 seed weight of less than 12g whilst the top 10% had seeds weighing from 18.6 to 24.9 g. The lower 5% of the studied plant materials had 100-seed weight in the range of 6.7 to 10.5g whereas the upper 5% had the weight of 100 seeds in the range of 19.6g to 24.6g. (Fig 4.9). Estimated broad sense heritability for 100- seed weight was 0.8 (Table 4.1) 61 University of Ghana http://ugspace.ug.edu.gh Table 4.1: Variance of components and broad sense heritability estimates for eight quantitative traits Trait Line Rep:Block Rep Residual VG VP H2 DTF 36.4 2.2 0.0 19.6 36.4 41.9 0.9 DTM 112.6 14.6 0.0 90.4 112.6 138.9 0.8 100-SW 6 1 0.0 5.8 6.0 7.7 0.8 NPB/P 4.5 0 0.0 215.4 4.5 58.4 0.1 NPD/P 1157.8 103.8 37.2 1757.8 1157.8 1632.5 0.7 NS/P 3677.7 546.4 137.0 5220.8 3677.7 5153.8 0.7 NS/PD 0.1 0 0.0 0.2 0.1 0.15 0.7 PHT 275.9 3.2 0.0 1105.6 275.9 552.7 0.5 PLP 26 0.04 0.0 32.6 26.0 34.2 0.8 Fig 4: 9: Percentile plot showing the distribution of 100-seed weight of 230 lines Number of days to flowering (DTF); number of days to maturity (DTM); hundred seed weight(100-SW); number of primary branches per plant (NPB/P), number of pod/ plant (NPD/P); number of seed per plant (NS/P); number of seed per pod (NSPD/P), and plant height at maturity (PHT). 4.4.2 Multivariate analysis of eight quantitative traits Multivariate analyses based on principal components decomposed eight quantitative traits into eight factors (F) or components (Table 4.1). The first three components had eigenvalues of greater than one where PC1, PC2 and PC3 had eigenvalues of 2.69, 1.39 and 1.06, respectively (Table 4.1). The first component (PC1) accounted for 33.35% of the total variation whilst the second component (PC2) explained 17.32%, and cumulatively accounted for 50.67% of total variation. The third component (PC3) with an eigenvalue of 1.05, accounted for 13.210% of 62 University of Ghana http://ugspace.ug.edu.gh total variation and in addition to the first two components, 63.9% of total variation was explained. The cumulative of eight principal components explained 100 percent of the inherent variation in the data (Table 4.1). Table 4.2: Eigenvalues, variability (%) and cumulative variability from PCA PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 Eigenvalue 2.7 1.4 1.1 0.9 0.8 0.8 0.3 0.0 Variability (%) 33.4 17.3 13.2 11.7 10.4 9.7 4.2 0.2 Cumulative % 33.4 50.7 63.9 75.56 85.9 95.6 99.8 100.0 A loading plot of the first two components showed the association of the eight quantitative traits, number of days to flowering (DTF); number of days to maturity (DTM); number of pods per plant(NPD/P); plant height at maturity (PHT); number of primary branches per plant (NPB/P); number of seeds of per pod (NS/PD); number of seeds per plant (NSP); and hundred seed weight (100SW) and the first two components (Fig 4.10). The first principal component has large positive associations with the number of pods per plant (0.840), the number of seeds per plant (0.837), days to flowering (0.691) and days to maturity (0.650) (Table 4.2). Also, the number of primary branches per plant (0.42), plant height (0.39), and a hundred seed weight (0.18) positively associated positively with the first principal component (F1). Among the eight quantitative traits, only number of seeds per pod (-0.05) associated negatively with principal component one (F1). Similarly, DTF, DTM, PHT, NPB/P and 100SW associated positively with principal component two (F2) whilst NPD/P, NS/PD, and NS/P negatively associated with principal component two (F2) (Fig 4.10 & Table 4.2). With regards to principal component three (F3), 100SW (0.726) had the largest positive association (Fig 4. 10). This was followed by NS/PD (0.623). DTF, DTM and NPB/P had a positive but low association with PC3. On the other 63 University of Ghana http://ugspace.ug.edu.gh hand, PHT (-0.201), NPD/P (-0.225), and NS/P (-0.017) associated negatively with PC3 (Fig 4.10 & Table 4.1) Fig 4. 10: Loading plot of the first two principal components (F1 &F2). Number of seed per pods (NS/PD), 100 seed weight (100SW), plant height (PHT), days to maturity (DTM), days to flowering (DTF), number of pod per plant (NPD/P), number primary branches per plant (NPB/P), and number of seeds per plant (NS) The contribution of each variable in percentages to principal components is as indicated in (Table 4.3.). Principal component one (PC1) which accounted for 33.4% of observed variation had very high contributions from NPD/P, NS/P, DTF, and DTM which contributed 26.4%, 26.3%, 17.9&, and 15.8%, respectively. PHT, NPB/P and 100SW also contributed 5.8, 6.54 and 1.15% respectively, to observed variation in F1. NS/PD had the least contribution of 0.08 to the total variation in component one (PC1). With regards to principal component two (PC2), DTM (20.7%), NS/PD (16.3%), NS/P (15.31%), and DTF (12.8%) gave the most contribution to observed variation in PC2 which constitutes 17.3% of the total variation (Table 4.3). 100 SW (6.5) contributed the least to total variation in PC2. 100SW contributed 49.9% of the observed 64 University of Ghana http://ugspace.ug.edu.gh variation in F3. This was followed by NS/PD which contributed 36.8% to recorded variation in PC3. Table 4. 3: Contribution of variable of variables (%) to observed variation in each component Variable PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 DTF 17.90 12.81 0.04 7.25 8.08 6.81 47.08 0.02 DTM 15.84 20.67 3.30 0.18 6.48 5.09 48.36 0.07 PHT 5.80 7.75 3.87 26.03 32.39 23.71 0.45 0.00 NPB/P 6.54 12.38 1.29 2.37 35.24 41.99 0.13 0.06 NPD/P 26.43 8.25 4.79 8.87 3.84 0.89 0.02 46.90 NS/PD 0.08 16.32 36.76 30.84 2.94 8.94 0.21 3.91 NS/P 26.27 15.31 0.03 2.06 4.96 2.25 0.12 49.01 100SW 1.15 6.51 49.91 22.42 6.07 10.30 3.62 0.03 Table 4.4. showed the correlation between the eight studied quantitative traits. The number of days to flowering (DTF), a growth parameter had a significant positive association (0.64) with the number of days to maturity (DTM). Also, how long it took soyabean genotypes studied to flower (DTF) had a significant positive association with plant height (PHT), number of pods per plant (NPD/P), number of seeds per plant (NSP). There was also a weak but significant positive association between days to flowering (DTF) and the number of primary branches per plant (NPB/P) (Table 4.4). How long it took soyabean genotypes to attain physiological maturity (DTM) positively and significantly correlated with the number of pods per plant (NPD/P), a number of seeds per plant (NS/P), and 100 seed weight (100SW) (Table 4.4). Primary branches produced per plant (NPB/P) significantly and positively associated (0.16) with the length of a juvenile phase of soyabean lines (DTF) evaluated. The number of primary branches per plant (NPB/P) had a significant positive association with the number of pod/ plant (NPD/P) and the number of seeds per plant (NS/P). There were also positive but non-significant association (0.10) between the 65 University of Ghana http://ugspace.ug.edu.gh number of primary branches per plant (NPB/P), and the number of seeds per plant (NS/P). A similar positive non- significant association (0.003) was recorded for the number of primary branches/plant and 100 seed weight (Table 4.4). The strongest association (0.94) occurred between the number of pods per plant and number of seeds per plant. Number of pods per (NPD/P) plant had significant positive correlation with all quantitative traits evaluated except for number of seeds per pod (NS/PD) with which it recorded weak negative non-significant association (-0.12), and 100 seed weight with which it recorded a positive but weak association (0.01). The number of seeds per plant (NS/P) also recorded significant positive associations with all traits evaluated except for 100-seed weight (100SW) with which it had a positive but weak non-significant correlation (0.065). On the other hand, the number of seeds per pod (NS/PD) had a positive significant association with the number of seeds per plant (NS/P) but its association with other traits was not significant, mostly negative except for number of pod/plant (0.10) and 100 seed weight (0.01). Table 4. 4:Pearson correlation matrix of eight quantitative traits Variables DTFF DTM PHT NPB/P NPD/P NS/PD NS/P 100SW DTFF 0.636 0.286 0.155 0.335 -0.07 0.318 0.064 DTM 0.636 0.224 0.098 0.283 -0.119 0.276 0.239 PHT 0.286 0.224 0.045 0.179 -0.067 0.162 0.028 NPB/P 0.155 0.098 0.045 0.278 0.096 0.333 0.003 NPD/P 0.335 0.283 0.179 0.278 -0.121 0.941 0.031 NS/PD -0.07 -0.119 -0.067 0.096 -0.121 0.16 0.014 NS/P 0.318 0.276 0.162 0.333 0.941 0.16 0.065 100SW 0.064 0.239 0.028 0.003 0.031 0.014 0.065 Values in bold are different from 0 with a significance level alpha=0.05 The generated biplot from the first two principal components based on eight quantitative traits grouped the 230 soyabean lines into two major groups, I & II (Fig 4.11). Table 4.5 shows a sub-sample of the 230 as grouped together by principal component analyse. 66 University of Ghana http://ugspace.ug.edu.gh Most of the tropical glycine lines prefixed (TGX) obtained from IITA occurred in group I. Also, released varieties including Jenguma, Songdaa, Afayak, and TGX 1844-22E, recently released as “Favour” by SARI, all occurred in group I. The grouping of released varieties with other TGX lines obtained from IITA was not surprising. Most soyabean varieties released in Ghana were obtained from IITA as advanced breeding lines and subsequently adapted to Ghana’s production conditions. Group, I separated into two sub-groupings, IA and IB. Sub-group IB comprised of soyabean lines possessing a high number of pods, seeds, and primary branches per plant (Table 4.6). The number of seeds per plant had a significant positive association with the number of pods per plant (Table 4.4). Also, the number of seeds per plant had a weak but positive association with the number of primary branches per plant. Most of the lines in sub-group IB including the released varieties, Jenguma, Afayak, Songdaa, and TGX 1844-22E belonged to the medium maturity (100-to 120 days). Sub-group IA of group I is largely comprised TGX lines and the lines obtained from Brazil with the prefix BRS. Lines sub-group IA flowered and matured later than lines in the other groups. These lines were also taller with higher pod clearance (PLP) than lines in sub-group IB and other groups (Table 4.5). Most lines in group II flowered and matured earlier than lines in group I. Also, most lines grouped in together in I had higher seed weight than most line grouped together to form group I. lines in sub-group IIA (Fig 4.11) possessed higher plant height with high pod clearance (PLP). They were early maturing but had high seed weight. They, however, had a lower number of pod and seed per plant. Lines that clustered to form sub-group IIB (Fig 4.11) were generally short with lower pod clearance values (PLP). The earliest flowering and maturing lines among the 230 lines evaluated occurred in this group. 67 University of Ghana http://ugspace.ug.edu.gh Fig 4 .11: Clustering of lines based on growth traits, and yield components 68 University of Ghana http://ugspace.ug.edu.gh Table 4.5: Growth traits, and yield components of lines grouped together by principal components Line CPA Group PHT PLP PB/P DTF DTM NPD/P NS/PD 100-SW NS/P TGX 1740-2F IA 95.2 19.2 11 43 120 110 1.7 14.1 185 TGX 1799-8F IA 65.6 9.5 8 43 125 115 1.8 14.9 196 TGX 1805-8F IA 65.3 12.9 7 44 120 73 2 13.5 143 SARISOY 26 IA 60.3 5.8 12 47 128 118 1.6 14 185 PI416873B IA 41.6 39.5 3 44 117 45 2 17.5 92 PI605823 IA 82 28.7 5 44 138 53 1.7 11.8 91 TGX 2004-10F IA 87.5 19.7 4 46 118 49 1.4 15.0 69 TGX1989- IA 52.6 12.7 4 43 121 46 1.8 15.6 82 48FN BRS-326 IA 50.2 12 7 44 123 46 1.6 18.5 71 Afayak IB 61.7 9.4 12 47 116 101 1.7 14.0 176 Songdaa IB 49.5 11.3 16 4 119 165 2 19.2 332 TGX 1844-22E IB 35.2 15.8 19 47 119 290 1.8 15.3 510 PI398836B IB 35 3 7 31 105 96 1.8 11.5 173 PI567123A IB 68.2 7.5 13 38 100 180 1.5 9.9 279 ISS-8203 IB 41.3 10.4 10 33 115 147 2 12.9 285 BRS 8660 IB 41.5 8.1 10 41 109 117 1.9 14.7 212 PI507004 IB 64.1 5.9 7 31 105 170 1.7 12.5 285 TGX 1990-52F IB 61.1 4.3 9 41 116 175 1.9 18.2 321 PI468919 IB 66.1 8 13 36 114 248 2 12.9 484 TGX 1989-11F IB 56 7 6 43 121 61 1.4 12.6 86 PI605824A-1 IIA 63.8 18.9 4 38 98 25 1.8 17.4 37 Wamini IIA 75.8 10.3 4 37 101 64 2.1 14.0 127 PI594760B IIA 45.2 14 3 37 106 47 1.8 16.2 81 OZARK IIA 24.5 8.4 5 27 108 38 1.5 18.2 54 PI417013 IIA 66.5 11.2 2 28 98 24 2.3 15.8 55 PI274454 IIA 21.9 6.2 6 27 93 33 2.2 19.5 69 PI506938 IIA 83 11.2 7 38 93 67 1.7 14.8 107 PI594760B IIA 45.2 14 3 37 106 47 1.8 16.2 81 PI605824A-1 IIA 63.8 18.9 4 38 98 25 1.8 17.4 37 ISS-8203A IIB 17 6.2 3 29 93 27 1.7 13.6 36 PI605829 IIB 59 8.5 6 34 91 74 1.6 12.2 110 PI476905A IIB 26 3.8 6 34 89 54 2.1 12.2 113 PI507004A IIB 38 8.1 9 32 102 115 1.6 15.8 169 SARISOY 11 IIB 32.7 8.2 5 31 93 48 2.2 16.0 99 PI587886 IIB 29.5 7.1 3 31 108 41 2.1 16.1 87 PI587886A IIB 24.1 1.6 9 33 103 53 2.4 16.1 122 PI81042A IIB 47.3 3.6 8 32 86 81 1.8 12.0 130 NE 3400 IIB 26.4 2.9 2 23 83 33 2.3 13.2 75 PI398543 IIB 14.3 5.9 4 26 86 38 1.6 8.6 59 PI398512 IIB 31 4.8 5 85 33 1.8 40 12.1 49 Mean 49.8 8.5 7.1 36 107 84 1.84 14.7 148 S.E.M 1.3 3.2 0.6 1.3 2.2 9.9 0.04 0.4 18 69 University of Ghana http://ugspace.ug.edu.gh 4.4.3 Genetic Diversity and Population Structure of 96 representative panel The population structure of the 96 soyabean lines was inferred using Bayesian clustering analyses (Fig 4.12). The highest values of ΔK was recorded when 9 distinct sub-populations (clusters) were identified (Fig 4 .12) based on genetic variation. The nine inferred clusters did not correspond to the six origins (Table 4.4) with which the 96 lines were identified before performing structure analyses. However, lines from some origins were dominant in some clusters but lines from each of the six origins occurred in more than 1 cluster with the exception of cluster 9 (Table 4.6 & Fig 4.12). A B C Fig 4 .12: Clustering of 96 soyabean lines based on Bayesian assignment. A (mean likelihood [L (k)± SD] over 12 runs assuming K= 1-10. B (ΔK), where the modal value of the distribution is considered as the highest level of structuring. C assignment of individual lines to the K= 9 where each individual is represented by a bar, with colours indicating the likelihood of assignment to the corresponding clusters(sub-population) 70 University of Ghana http://ugspace.ug.edu.gh The distributions of the 96 lines in 9 inferred clusters or sub-population are presented in Table 4.6. Cluster 1 consisted of 11 lines, 7 of which were plant introductions (PIs) obtained from the USDA through the University of Illinois. Three of the eleven lines that formed this cluster were sourced from Zimbabwe whilst one “TGX” prefixed line obtained from IITA joined the others to form cluster 1. Cluster 2 comprised of six lines made up of 5 plant introductions (PIs) sourced from the USDA, and one Zimbabwean line whilst cluster 3 contains only two PIs. Cluster 4 is composed of 14 lines comprising 9 PIs, 7 IITA lines and one line each from Ghana and Zimbabwe. Cluster 5 made of 11 USDA lines (PIs), 2 IITA lines and the only line sourced from China. Cluster 6 which was the largest cluster was also the most diverse cluster. Cluster 6 had lines from 5 of the six origins including 7 lines, 11lines, 3 lines, 2-line and1line from Zimbabwe, USA, Brazil, IITA and Ghana, respectively. Six IITA lines, 4 PIs from the USDA, 4 varieties released in Ghana and one line sourced from Zimbabwe formed the 15 lines that occurred in Cluster 7. Cluster 8 consisted of 5 lines made up of 2 PIs from the USDA, 2 IITA lines and 1 Brazilian whilst cluster 9 had only one line, TGX1990-46F sourced from IITA. Lines from the 51 plant introductions obtained from the USDA occurred in 8 of the 9 inferred clusters with most occurring in cluster 4 (11 lines), and cluster 6 (11 lines). Thirteen lines from Zimbabwe were used in this study, seven of them grouped together in Cluster 6.3 of 4 Brazilian lines evaluated grouped together in Cluster 7 whereas 4 of 5 Ghanaian varieties occurred in Cluster 7. The 21 IITA lines occurred in all clusters except cluster 2. The 21 IITA lines were widely distributed with the highest number (7) grouping in cluster 7. 71 University of Ghana http://ugspace.ug.edu.gh Table 4.6: Origin, and distribution of 96 lines in 9 inferred clusters based on Bayesian clustering analyses Line Origin Cluster Line Origin Cluster Line Origin Cluster SARI-S11 Zimb 1 TGX-1990-57 IITA 4 PI423960A USDA 6 SARI-S-1 Zimb 1 TGX1990-40F IITA 4 PI423958 USDA 6 SAR1-S4 Zimb 1 TGX1989-48F IITA 4 PI417013 USDA 6 PI567123A USDA 1 TGX-1989-45 IITA 4 PI399096 USDA 6 PI567104B USDA 1 SARI-4DA Ghana 4 PI200526 USDA 6 PI567099A USDA 1 PI606397B-2 USDA 5 TGX-2011-3F IITA 6 PI567068A USDA 1 PI606397B USDA 5 TGX-1834-5E IITA 6 PI567053 USDA 1 PI605891B USDA 5 SARI-5FK Ghana 6 PI476905A USDA 1 PI594767B USDA 5 BRS-7980 Brazil 6 PI398512 USDA 1 PI594760B USDA 5 BRS-360 Brazil 6 TGX2008-4F IITA 1 PI588053A USDA 5 BRS-326 Brazil 6 SARI-S-13 Zimb 2 PI567516C USDA 5 SARI-S-14 Zimb 7 PI605891A USDA 2 PI567129 USDA 5 PI224271 USDA 7 PI605791A USDA 2 PI506938 USDA 5 PI398477 USDA 7 PI518295 USDA 2 PI417134 USDA 5 PI506491 USDA 7 PI416778 USDA 2 PI398836 USDA 5 PI567046A USDA 7 PI200466 USDA 2 TGX1989-40F IITA 5 TGX1835-10E IITA 7 SARI-6FT USDA 3 TGX-1988-3F IITA 5 TGX-1844-22 IITA 7 PI274954 USDA 3 SARI-9LMG China 5 TGX1845-10E IITA 7 SARI-17 Zimb 4 SARI-S-9 Zimb 6 TGX1988-3F IITA 7 PI657753 USDA 4 SARI-S-5 Zimb 6 TGX-1990-11 IITA 7 PI605854B USDA 4 SARI-S30 Zimb 6 TGX1990-46F IITA 7 PI605824A USDA 4 SARI-S29 Zimb 6 Sal-1 Ghana 7 PI567076 USDA 4 SARI-S27 Zimb 6 SARI-3Jng Ghana 7 PI567069A USDA 4 SARI-S-19 Zimb 6 SARI-7QS Ghana 7 PI416873B USDA 4 SARI-18 Zimb 6 SARI-8SG Ghana 7 PI416810A USDA 4 SARI-2ZK USDA 6 PI462312-2 U 8 PI398825 USDA 4 PI606440A USDA 6 PI567034 USDA 8 PI3984-77A USDA 4 PI588053B USDA 6 TGX-1844-19 IITA 8 TGX2006-3F IITA 4 PI561381 USDA 6 TGX1990-49F IITA 8 TGX-2004-10 IITA 4 PI-534545 USDA 6 BRS-360-2 Brazil 8 TGX-1990-80 IITA 4 PI462312 USDA 6 TGX1990-46F IITA 9 72 University of Ghana http://ugspace.ug.edu.gh Mean Fst was generally high for the all inferred clusters except cluster 8 and 9 (Fig 4.13.) Mean Fst for lines in inferred clusters ranged from 0.007 to 0.9154. Most of the clusters had mean Fst above 0.5. Conversely, mean heterozygosity ranged from 0.048 to 0.446. Cluster 2, 3, 4, and 6 had heterozygosity of approximately 0.1. Heterozygosity was generally low except for cluster 8 and 9 which had heterozygosity of almost 0.5 (50%). 1 0.9154 0.9129 0.9 0.8136 0.8231 0.8035 0.8 0.6889 0.7 0.6 0.5355 0.5 0.4457 0.4459 0.4 0.3 0.2344 0.1979 0.2 0.1052 0.1075 0.1022 0.1 0.048 0.0573 0.0074 0.0128 0 Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 Cluster 6 Cluster 7 Cluster 8 Cluster 9 Heterozygosity Mean Fst Fig 4 .13: A plot of mean Fst and Heterozygosity among soybean lines within clusters Allele divergence frequency among clusters showed that soyabean lines in cluster 2 were more genetically distant from lines in other clusters with a value of 0.346 with cluster 3, 0.360 with cluster 7, 0.327 with cluster 5, 0.258 with cluster 4, 0.239 with cluster 7 (Table 4.7). It, however, had lower with cluster 8 (0.165) and Cluster 9 (0.159). The lines in cluster 1 also recorded pairwise values of (0.313) with those in cluster 2, 0.307 with cluster 3, 0.273 with cluster 4, 0.269 with cluster 5, 0.219 with cluster 6 and 0.269 with cluster 7, and could, therefore, be said to be more distant genetically to lines in these clusters. The genetic distance 73 University of Ghana http://ugspace.ug.edu.gh between lines in cluster 1 and those in cluster 8 and 9 was low. Also, lines in clusters 3, 4, 5 & 6 were relatively distant from lines in other clusters except for lines in cluster 8 and 9 which recorded lower pairwise values with lines in all other clusters due to their high heterozygosity. There were no genetic differences between the 5 soyabean lines that constituted cluster 8, and the line in cluster 9. Table 4.7: Allele frequency divergence among sub-populations 1 2 3 4 5 6 7 8 9 1 - - 2 0.3126 3 0.3071 0.3462 - 4 0.2725 0.2579 0.3195 - 5 0.2689 0.3276 0.317 0.3155 - 6 0.2199 0.2389 0.2693 0.233 0.2682 - 7 0.2685 0.3601 0.3305 0.3103 0.3064 0.2557 - 8 0.1115 0.16 0.1645 0.1324 0.1505 0.0823 0.1438 - 9 0.1112 0.1593 0.1639 0.1321 0.1499 0.0816 0.1439 0 - Hierarchical clustering was largely influenced by the source of germplasm. The 96 lines were placed in two major groups both of which separated into several sub-clusters. Of 52 plant introductions obtained from the USDA, 38 were grouped in major cluster I which separated into many sub-clusters. The sub-cluster IA consisted of 19 lines, all of which were PIs. The sub-cluster IB separated into sub-cluster IC & ID. ID is solely made up of PI. In IC, the four Brazilian lines grouped with PIs, three which occurred in a single clade whilst the remaining one (BRS-7980) occurred in another clade, closely related to PI-534545. All 21 lines obtained from IITA occurred in sub-cluster IIA. Sub-cluster IIB further separated into two sub-clusters, Bi & Bii. Bi consisted of only plant introductions whilst Bii further into clusters, one of which is made up of only PI lines whilst the other is made up of three lines from Zimbabwe and one PI. Most of the lines obtained from Zimbabwe occurred in Cluster IIC, 74 University of Ghana http://ugspace.ug.edu.gh though, they occurred in several subclusters. The varieties released in Ghana, SARI-3Jng, SARI-8SG, SAL-1, SARI-5FK, and SARI-7QS all occurred in IIC. SAR-7QS and SARI-8SG were most related to SAL-1 being the most distant among the five released varieties. SARI- 3jng and SARI-FK also occurred in the same sub-clade. IIC Bii IIB Bi IIA I D I C Fig 4. 14: Clustering of 96 soyabean lines using mean euclidian distance based on 3200 SNP markers. Red (Plant introduction, PIs sourced from the USDA), Blue (Lines from IITA, Nigeria), Green (lines from Brazil), Black (Released varieties in Ghana), Pink (lines from Zimbabwe) 75 University of Ghana http://ugspace.ug.edu.gh 4.5 Discussion Germplasm collections remain valuable tools for crop improvement. This could either be through direct cultivation of genotypes adapted to farmers’ conditions in multi-location tests or hybridization to generate new varieties. To fully utilize such introductions, detailed information is required about the amount and nature of the genetic diversity available within each source of introduced germplasm. As noted by Ali et al. (2017), breeding for changes in phenology and stress tolerance to meet challenges of climate change require access to elite germplasm from other regions, multi-location testing systems that adequately sample the target population of environments and continually adjusting allele frequencies among a set of selected breeding lines Flowering and maturity periods, the two key determinants of soyabean productivity (Jiang et al., 2013; Zhai et al., 2014) varied widely among the 230 soyabean lines evaluated. The number of days to flowering of lines in the germplasm assembly ranged from 23-51 days. Some of the lines obtained from the USDA were photoperiod sensitive and flowered in the third week of planting. They contributed to the wide variations. Broad-sense heritability estimate of 0.9 recorded for the number of days flowering was an indication that differences among 230 lines were largely due to genetic effect. Preetak et al. (2018) recorded broad-sense heritability of 0.93, which was comparable to the 0.9 recorded in this study. However, Desissa (2017) recorded a broad sense heritability of 0.59 for days to flowering which was relatively lower than what was recorded in this trial. The number of days to maturity was as low as 83 days for NE3400 and as high as 138 days for PI605823. The number of days to maturity recorded a moderately high heritability estimate of 0.8 in the broad sense which was lower than 89.5% reported by Preetak (2018). The heritability estimates of 0.8 recorded in this study were also lower than the 91.9% recorded by Preetak (2017). Heritability estimates are population specific 76 University of Ghana http://ugspace.ug.edu.gh and therefore differences between what was recorded in this study and that of other studies could be due to nature of population used for the study The number of days to flowering and maturity are key drivers of differentiation in cultivated soyabean leading to the creation of maturity groups (Zhang et al., 2007; Wang et al., 2018) which is largely influenced by photoperiodism. Photoperiod sensitivity of soyabean lines has limited the use of productive temperate varieties in tropics. Of the 13 maturity groups of soyabeans, only members of the group (MG X) are suitable for cultivation in the tropics. The high variation, and as well as high genetic variance recorded for days to flowering and maturity can be exploited to develop varieties for the different agro-ecologies suitable for soyabean production in Ghana and other African countries. For instance, the line Paranagoina reported to harbouring the long juvenile trait which delays vegetative growth in short-day genotypes (Lu et al., 2017) was part of the lines evaluated. It could be used to adapt some of the productive but photoperiod sensitive temperate lines to the short-day period that characterises our production system as suggested by Miranda (2018). Growth parameters, such as plant height at physiological maturity, and plant height at first pod (pod clearance) varied among the lines studied. Of the 230 genotypes studied, 50% (H2=0.5) of the observed variation in heights among the lines was due to the genetic effect. In soyabean, height is closely related to yield (Zhang et al., 2018). Early maturing and short plants are said to have reduced vegetative growth leading to fewer nodes, few pods and lower yield (Ibrahim et al., 2012). Height at plant first pod (pod clearance) was highly heritable with 80% of observed variation accounted for by the genetic variance. Height at first pod or pod clearance is very important when combine harvester is being deployed for harvesting. Lines that pods closer to ground are difficult to feed into the combine which may result in yield loss (https://www.cornandsoyabeandigest.com/soyabeans/11-combine-adjustments-soyabeanharv 77 University of Ghana http://ugspace.ug.edu.gh est). For instance, SARI has a combine harvester for soyabean but it cannot be used for harvesting because of the generally lower pod clearance of current soyabean varieties (Dr. Nicholas Denwar, 2017, personal communication). The 230 lines possessed wide variation for pod clearance most of which (80%) was due to genetic variance. This could be used to develop cultivars with higher pod clearance for commercial farmers who normally deploys machinery including combine harvesters. lines in the sub-groups IA and IIA based on PCA both have relatively higher clearance. Lines in IIA were late maturing with higher clearance whilst lines in the sub-group IA were early maturing with relatively high pod clearance. Parents could be selected from these groups for cultivar development that combines high yield, earliness and high pod clearance since early to medium maturing lines are most desired in Ghana (Denwar and Wohor, 2011). Heritability estimates of 0.7 for NPD/P recorded in this study was slightly lower than the 89.4 and 86.90 reported by Desissa (2017) and Prateek (2014) but comparable to 74% reported by Ali et al., (2014). The 230 lines varied widely for 100 seed weight most of which was due to genetic effect and culminated in high heritability estimates. This result agreed with the observation by Zhang et al. (2013) that 100 seed weight was less affected by the environment. NS/P and NPD/P were strongly correlated (r=0.94) and had the same level of heritability. Silva et al., (2014) reported that the number of pods per plant had a strong and positive correlation with productivity. NS/PD was moderately heritable but it was negatively correlated with all the growth parameters, and yield components except NS/P with which it had a significant positive association, and NPB/P with which it had a positive but very weak association. Principal component analyses grouped the 230 lines into four sub-groups, mainly driven by the number of seeds per plant and number of pods per plant which accounted for 26.3 and 26.4, respectively and days to flowering and days to maturity also contributed to 17% and 15%, 78 University of Ghana http://ugspace.ug.edu.gh respectively. Number pods per plant which had the highest contribution to observed variation are the key determinant of soyabean productivity (Silva et al., (2014), an indication that the assembled germplasm varied highly in the most important yield determinant. Nine sub-populations (clusters) were inferred from a 96-representative sample from the 230 lines genotyped with 32000 SNP markers. Lines obtained from USDA were most diverse as they occurred in 8 of nine sub-populations. Mean Fst values were high most sub-population with the exception of sub-population 8 and 9. Hipparagi et al. 2018 recorded mean Fst values in the range of 0.464 to 0.688 for 5 sub-populations inferred from 75 soyabean accessions based on SSR marker. This study recorded mean Fst values in the range of 0.0074 to 0.915 for nine clusters based SNP markers. Generally, mean Fst was for most clusters were high, an indication of high inbreeding coefficient for lines evaluated. High differentiation or inbreeding is expected for a self-pollinated crop like soyabean. However, two clusters recorded very high heterozygosity values which showed that lines in those clusters were still segregating. SARI normally receives advanced breeding lines from IITA. Some of which may still be segregating Also, cluster 8 and 9 with high heterozygosity could be containing early breeding lines which mistaken got mixed up with the advanced breeding lines and therefore few generations of advancement would be required before they are placed in the germplasm pool for utilisation. Hierarchical clustering grouped the 96 accessions into two major clusters with accession sourced from the USDA occurring in both clusters. Plant introductions obtained from the USDA distributed into all subpopulations except two IITA lines were closely related and grouped into two sole sub-populations. Lines from Zimbabwe occurred in two populations mixed with Ghanaian released varieties and few plant introductions from the USDA. Brazilian accessions were close to USDA accessions. This is a reflection of the extent of germplasm sharing between the two countries. IITA lines and USDA accessions were very distinct, and so 79 University of Ghana http://ugspace.ug.edu.gh the two sources could serve as different pools from which materials could be assembled for varietal development. Although they occurred in the same major clusters which were similar to the pattern inferred from cluster analysis, the Ghanaian varieties were closely related to Zimbabwe accessions than to IITA accessions as revealed by hierarchical clustering. 80 University of Ghana http://ugspace.ug.edu.gh 4.5 Conclusion The assembled germplasm possessed high variation in most of the morphological traits evaluated. Heritability estimates in the broad sense revealed that variation between individuals for these traits were largely due to genetic effect. Most of Ghana's soyabean production occurs in the Guinea Savannah ecology with monomodal rainfall pattern. Early to medium maturing lines are more suited for this ecology for which reason, varietal development has been targeted at combining high productivity with earliness. However, a larger area of the country, spanning the Transition belt and Coastal savannah ecologies are deemed suitable for soyabean production and these areas with bimodal rainfall pattern can support different crop growth duration as compared to Guinea and Sudan ecologies. Fortunately, days to flowering and days to maturity, key determinants of growth duration, both varied widely and were highly heritable. This could be exploited to develop cultivars for the different agro-ecologies with regards to maturity period The number of pods per plant, number of seeds per plant, and seed weight, which are the main determinants of soyabean productivity were also very variable and mainly under genetic control. The number of pods per plant, number of seeds per plant, and days to flowering were the main drivers of variation among the assembled germplasm. This understanding will not only guide the formulation of utilisation and conservation strategies, but also the decision on what characteristics to consider for future collections. Principal component analysis clustering showed that some accessions combined earliness with high pod and seed production. This could be exploited to develop early maturing but high-yielding cultivars for the Guinea and Sudan savannah ecologies. Structure and diversity analyses proved that all the lines were distinct individuals and subsequently placed them into sub-populations which were genetically distant as revealed by 81 University of Ghana http://ugspace.ug.edu.gh allele frequency divergence and mean Fst values. Grouping by PCA based on phenotypic traits could be combined with grouping based on SNP markers to aid in parental selection, and as well conservation efforts. The distinctiveness of IITA accessions from the USDA collection proved the need to maintain and characterise these lines that SARI has accumulated over the years. More so, the IITA accessions are TGX lines which are adapted to the short-day length and high temperature in the tropics. Their utilisation for varietal development comes with little or no linkage drag as compared to the USDA collection in which some accessions were photoperiod sensitive, high pod shattering and had green, black and brown seed coat colours. 82 University of Ghana http://ugspace.ug.edu.gh CHAPTER FIVE 5.0 Field evaluation of soyabean lines for identification of sources of resistance to bacterial leaf pustule 5.1 Introduction Bacterial leaf pustule, caused by Xanthomonas axonopodis pv. Glycines is among the major diseases of soyabean in several major soyabean producing countries (Wrather et al., 2001; Kim et al., 2008; Kime et al., 2011). It was the only bacterial disease among eight documented diseases affecting soyabean in Ghana (Offei et al., 2008). Bacterial leaf pustule is reported to cause yield loss in the range of 15-50% (Rahayu, 2007). The main characteristic of bacterial leaf pustule disease is brown lesions with yellow halos on adaxial side of leaves with raised pustules at the abaxial side of leaves (Sinclair, 1999a). The yellow-brown lesions on the adaxial side of the leaves coalesces into bigger spots, depleting the chlorophyll content of leaves, and under severe conditions this may results in premature leave senescence, defoliation and death of plants (Hartman and Murithi, 2017). Xanthomona axonopodis pv. glycine the causal agent of bacterial pustule, survives or overwinters on the debris; leaves, pods, and stems, as well as the seeds of infected plants (Sinclair, 1999b). On infected soyabean fields, the pathogen spread from one plant to another or to nearby fields through rain splashes and wind (Goradia et al., 2004). Infected seed is one of the important means of spreading bacterial leaf pustule disease (Khaeruni et al., 2007). The use of farmers’ own seed, and sharing of such seeds among farmers is a common practise in Ghana like in many countries in sub-Sahara Africa. This practise facilitates the spread of seed borne diseases including bacterial pustule. Cultural control including crop rotation, tillage and altering planting dates are some of the strategies used to controlling bacterial leaf pustule (Houng et al., 2012). For instance, Houng 83 University of Ghana http://ugspace.ug.edu.gh et al., (2012), observed that delaying of planting for 15-30 days relative to the traditional planting date at Suwon and Naju in Korea reduces incidence of a bacterial pustule significantly for susceptible and moderately susceptible cultivar. The brevity of the cropping season in the Guinea and Sudan savannah ecologies, where most of Ghana’s soyabean is produced leaves very little room for delaying the planting date. The use of agro-chemicals is the predominant approach to controlling bacterial pustule but excessive use of chemicals often gives rise to resistant strains in addition to ecological and environmental damages (Acuna et al., 2009). Therefore, much emphasis is being placed on the need for a more sustainable, and ecological friendly approaches (Khaeruni et al., 2010). The use of biological agents to control pests and diseases including bacterial leaf pustule has been suggested as a more sustainable and environmentally friendly than agro-chemicals (Bale et al., 2007; Khaeruni et al., 2010; Bommarco, et al., 2013). Khaeruni et al., (2018) evaluated biological fertilizers (biofresh) composed of three rhizobacteria types, Bacillus ceresus ST21b, B. subbtilis ST2le, Serratia sp 2229A in different formulations and concluded that a solid formulation of biological fertilizer in combination of compost made from soyabean residue was effective at increasing the resistance of soyabean plants to bacterial leaf pustule. Increased resistance to bacterial leaf pustule arose from increased salicylic acid and peroxidase content of plants treated with the biological fertilizer (Khaeruni et al., 2018). Biological control, though very sustainable, its application is not popular due to lack of investment (Bale, 2007). Although cultural and biological control of plant diseases are known to be sustainable compared to agro- chemicals, the use of resistant genes remains the most effective approach to controlling pests and diseases. The use of resistant genes for the control of plant diseases is cost effective and environmentally friendly (Matsuo, 2014). Identification of sources of resistance presents the first step towards deploying resistant genes for disease control (Lovato, 2011). Identification of sources of resistance to pests and diseases is achieved through disease screening. Screening 84 University of Ghana http://ugspace.ug.edu.gh of crops for resistance to pest and diseases is performed under natural infestation, mostly on the field or through artificial inoculation. As noted by Abawi (2010), field screening for resistance provides for evaluation under a variety of interacting pressures including pathogenic and non-pathogenic organisms. Additionally, screening crops for diseases provides useful information on released varieties needed by farmers to minimise losses by choosing varieties suitable for a given area based on knowledge of prevailing pests and diseases (Lovato 2011). Several sources of resistance to bacterial leaf pustule have been reported including Danbaekong, SS2-2, and Pi96188 (Kim et al., 2011), and CNS soy (Hartwig, 1951). However, race specificity of resistant genes and environmental effects on gene expression often results in the breakdown of resistance with the emergence of new race of the pathogen (Brun et al., 2010). Therefore, it is important that sources of resistant are evaluated in a target environment to determine their effectiveness before deployment in breeding programmes. Additionally, there is the need to avoid linkage drag in order to speed up varietal development and release. Most soyabean cultivars released in Ghana and other African countries are tropical glycine lines which were bred for short day length and high temperatures that characterised production centres on the continent (Ibrahim, 2011). Thus, introducing resistant gene from poorly adapted sources may delay varietal release as loci associated with undesirable characteristics such as photoperiod sensitivity, pod shattering, and undesirable seed coat colour may take a while to eliminate. Therefore, it is important to find sources of resistance to bacterial leaf pustule in the Guinea savannah ecology where most of the country’s soyabean is produced. The objectives of this study were to 1. evaluate diverse lines for resistance to bacterial leaf pustule under field conditions 2. select bacterial leaf pustule resistant lines for cultivar development 85 University of Ghana http://ugspace.ug.edu.gh 5.2 Materials and Methods 5.2.1 Materials Plant materials used in this study involved a 250 soyabean lines comprising of 131 plant introductions (PI) obtained from the USDA; 66 advanced breeding lines collected from the International Institute for Tropical Agriculture (IITA); 31 soyabean lines obtained from Zimbabwe in SARI’s holding; 14 Brazilian accessions, 3 Chinese accession lines, and 5 Ghanaian released varieties. 5.2.2 The study area The study was conducted at the experimental fields of the Savannah Agriculture Research Institute, Akukeyili in the Tolon district. The area is characterised by Ferric Luvisols of the Tingoli series which is well drained, devoid of concretions, and brown in colour (Atakora and Kwakye, 2016). 5.2.3 Experimental design and field establishment The experiment was laid at the same site (N09°24.388; W000°59.281) for two cropping seasons, 2017 and 2018. The field trials were established on the 24th of June and 18th of June for 2017 and 2018, respectively. In both years, the trial was established on the same plot. The trial was laid in an alpha lattice design in two replicates with each replication comprising 25 blocks of 10 lines per block. Each line was planted in single a single row of 5m long with a plant spacing of 5cm. Rows were planted at 50cm apart. A spreader row using the susceptible genotype, PI81042 a highly susceptible genotype which susceptibility was confirmed in 2016 during seed multiplication on the field was planted between lines under evaluation. Phosphorus in the form of Triple Superphosphate (TSP) was applied at a rate of 60kg of P2O5 at emergence. Weed control was manually done with hoes. 86 University of Ghana http://ugspace.ug.edu.gh 5.2.4 Data collection and analysis The number of days to 50% flowering was recorded as the time each line took from planting to when 50% of its stands had at least one flower opened. At flowering, fifty plants of each line were tagged and used to estimate the incidence of bacterial leaf pustule, twenty of which were scored for severity. Severity was scored once in every 14 days for four consecutive times. The scoring for severity occurred over a period of 56 days upon flowering. Disease severity score recording was based on 1-9 scale (Subrahmanyam, 1990) where a score of 1 denotes no spots and 9 denotes complete defoliation or death of the infected plant. Concurrently, four chlorophyll content readings were taken from 20 record plants per lines for disease severity using SPAD meter. Chlorophyll content reading also started at flowering. The four severity score values were used to estimate the area under disease progressive curve (AUDPC) according to the formula prescribed by Emge and Shrum (1979). AUDPC Where n= number of observations, yi= severity score at the ith observation, and ti= time at the ith observation. The area under disease progress curve obtained for each line was used to represent the overall reaction of the line to bacterial leaf pustule infection. Using equation 1 above, the area under the chlorophyll retention curve (AUCRC) was estimated for each line. The AUCRC value obtained for each was used to represent the depletion of chlorophyll content due to bacterial leaf pustule infection and development on each line. Values of AUDPC and AUCRC obtained were analysed using the R statistical software (version 3.5.1) (R Core Team, 2018). 87 University of Ghana http://ugspace.ug.edu.gh Analyses of AUDPC was first carried out on yearly bases and ranking of genotypes compared. When genotypes ranked the same in both years, combined analyses of variance was carried out to assess the effect of year, and interaction effect of year and genotypes. Both year, and interaction between year and genotypes were not significant. The AUDPC data for both years were then standardised as described by Forbes, et al. (2014) and combined for analyses where years were considered as replications and thereby increasing the total number of replications to four to increase statistical power. Genotypic effects were estimated using a linear mixed-effect model (LMEM) with restricted maximum likelihood (REML) based on the equation below: 𝑦𝑖𝑗𝑘 = 𝑔𝑖 + 𝑟y𝑗 + 𝑏y𝑘 + 𝜀𝑖𝑗𝑘 𝑒𝑞𝑛 2 Where yijk, gi, rj, bk and Ɛijk denote observation on genotype i in block k nested within replication j, an effect of genotype i, effect of replication j, effect of block k nested within replication k and the effect of random error [~N (µ = 0, σ2 = 1)]. The model was fitted using the lme4 package (Bates et al., 2015) of R. Soyabean line, PI606440A described as resistant with few lesions was used as a check (intercept) for pairwise comparison and identification of resistant lines. The difference between the best linear unbiased estimate of AUDPC for each line and the intercept were subjected to pairwise t-test using the Satterthwaite’s method in R. 88 University of Ghana http://ugspace.ug.edu.gh 5.3 Results Best linear unbiased estimates (BLUES) of the area under disease progress curve for the 250 lines as compared to that of the intercept (PI606440A) is presented in the (Appendix 2). Table 5.1 presents the result of a sub-sample of 35 of the 250 lines evaluated for bacterial leaf pustules disease under field condition. The reaction of the line, PI606440A used as intercept, to bacterial leaf pustule under field condition recorded over time and presented in the form of AUDPC (61.50) was statistically higher than zero (Table 5.1). The check, PI606440A recorded a maximum score of 2.5 on a 1- 9 severity score scale and considered as resistant, but not highly resistant (Appendix 1). The lines with estimated AUDPC values less or equal to that of that of the intercept were considered resistant. The lines with lower AUDPC values than the intercept recorded negative fixed effect estimates and were therefore considered to be more resistant than the check (Table 5.1) whilst lines that had positive values that were statistically similar to the intercept were considered to have the same of level of resistance as the intercept PI606440A). 28 of 250 soyabean lines evaluated recorded negative fixed effect estimates ranging from -28.8 to - 24.8 relative to the intercept (PI606440A) (Table Appendix 2) constituting 10.7% of lines screened. All the 28 lines recorded severity score of “1” (Appendix 1), a reaction described as resistant without lesions by Grant et al., (2010). As indicated in Table 5.1., SARISOY 5(-28.8) and SARISOY 20B (-28.57) from SARI's soyabean germplasm assembly recorded the least estimates of AUDPC relative to the intercept and even though, their fixed effect estimates were not statistically different from the intercept, they were highly resistant than intercept as they recorded no lesions (Appendix 2) dues to Xanthomona axonopodis pv. glycine infection. These were followed by the plant introductions, BRS 8660A, BRS-7980B and PI274954B (Table 5.1). The tropical glycine lines TGX-1990-21F (-26.94) and TGX-1989-48FN (-26.90) from IITA were among the most resistant lines (Table 5.1) 89 University of Ghana http://ugspace.ug.edu.gh Two released soyabean varieties in Ghana, Afayak (-24.8) and Sal-1(-24.8) were among the highly resistant lines. Jenguma, a widely cultivated variety was moderately resistant having recorded a fixed effect estimate of -3.7 relatives to the intercept and a severity score of 3. Two varieties, Quarshie (83.84) and Songdaa (97.82) also released in Ghana were highly susceptible with AUDPC fixed effect estimates significantly (p<0.01) higher than the intercept (Table 5.1). The lines PI468919, SARISOY 10, TGX 1805-8F, 1990-80F, TGX 19988-3F and BRS 8660 with fixed estimates of 1.11, 1.60, 3.27, 4.70, 5.06, 29.48, respectively were statistically similar to the intercept in their reaction to bacterial leaf pustule disease. However, PI605791A (158.97), PI417120A (162.6), PI81042 (166.19), PI417120 ((176.5), PI605854B (177.7), PI606405 (183.3), PI398825 (190.5), PI506938 (190.5), and PI567516C (209.8) were among the most susceptible of the 250 lines evaluated (Table 5.1). The area under disease progress curves of these lines was higher than that of the intercept and the differences between them and the intercept was statistically significant (p<0.001) (Table 5.1) 90 University of Ghana http://ugspace.ug.edu.gh Table 5.1: Fixed effect estimates of AUDPC for soyabean lines in relation to the intercept Std. Line Estimates Error Df t-value Pr (>|t|) SARISOY 5 -28.8 31.66 211.18 -0.91 0.364 SARISOY 20B -28.57 31.66 211.18 -0.9 0.368 BRS 8660A -28.23 31.66 211.18 -0.89 0.374 BRS-7980B -28.2 31.66 211.18 -0.89 0.374 PI274954B -28.2 31.66 211.18 -0.89 0.374 TGX-1990-21F -26.94 25.84 210.99 -1.04 0.298 TGX-1989-48FN -26.9 25.84 211 -1.04 0.299 PI506677 -26.77 25.84 210.99 -1.04 0.301 ISS-8203 -26.49 25.84 211 -1.03 0.306 Afayak -24.8 25.84 211 -0.96 0.338 Sal-1 -24.875 31.66 211.18 -0.78 0.435 TGX 1990-57F -15.987 25.84 211 -0.61 0.54 JENGUMA -3.7 25.84 211 -0.14 0.886 TGX 1989-20FA -2.02 25.84 210.99 -0.08 0.938 PI468919 1.11 25.84 211 0.04 0.966 SARISOY 10 1.6 25.84 211 0.06 0.951 TGX 1805-8F 3.27 31.66 211.18 0.1 0.918 TGX 1990-80F 4.7 25.84 211 0.18 0.856 TGX 1988-3F 5.06 25.79 199.7 0.2 0.845 BRS 8660 29.48 25.84 211 1.14 0.255 (Intercept) 61.5 18.38 205.69 3.35 0.001** Quarshie 78.47 25.84 211 3.04 0.003** PI159925 83.84 25.84 211 3.25 0.001** Songdaa 97.82 25.84 211 3.79 0.000*** TGX-1989-68FN 102.82 25.84 211 3.98 1E-04*** PI605791A 158.97 25.84 210.99 6.15 3.8E-09*** PI417120A 162.6 25.84 210.99 6.29 1.8E-09*** PI81042 166.19 25.84 211 6.43 8.3E-10*** PI417120 176.5 25.79 199.8 6.84 9.3E-11*** PI605854B 177.7 31.66 211.18 5.61 6.2E-08*** PI606405 183.3 25.84 210.99 7.09 1.9E-11*** PI398825 190.49 25.84 211 7.37 3.8E-12*** PI506938 197.5 25.84 211 7.64 7.3E-13*** PI567516C 209.58 25.84 211 8.11 4.1E-14*** Note: *, **, and *** represent significant at 5, 1 and 0.1 percent, respectively. Grouping of the lines based on AUDPC estimates placed the lines into two major groups, I and II (Fig 5.1). Group I consist of a subset of the susceptible lines whilst group II is comprised of resistant lines. Resistance to bacterial leaf pustule could be placed into three categories including highly resistant, resistant, and moderately as indicated in Fig 5.1. The highly resistant 91 University of Ghana http://ugspace.ug.edu.gh category was made up of lines that recorded no lesions, and therefore had an immune response to the pathogen, Xanthomona axonopodis pv. glycine under natural infestation conditions on the field. Of 250 lines evaluated, 28 lines comprising of 10 plant introductions (PI587886, PI274954B, PI398512, PI398874, PI416778A, PI416810A, PI416939, PI417126A, PI506677, ISS-8203, PI506939); three Brazilian lines (BRS 8660A, BRS-7980B, PARANGOINA); five SARI’s germplasm holding (SARISOY13A, SARISOY15, SARISOY20B, SARISOY2A and SARISOY5); eight advanced breeding lines from IITA (TGX 1989-20FC, TGX 1990-80FA, TGX-1835-10E, TGX-1989-40FB, TGX-1989-48FN, TGX-1989-48FNA, TGX-1990-21F and TGX1989-68FN); and one variety released in Ghana, Sal-1 had severity score of 1 and were classified as highly significant (Fig 5.1 & Appendix 2).The resistant category was composed of lines that had a negative fixed effect estimate relative to the intercept with a mean severity score in the range of 1.5 to 2. I I I Susceptible Resistant Moderately Highly resistant Resistant Fig. 5: 1: Grouping of soyabean lines based on fixed effects estimates for AUDPC 92 University of Ghana http://ugspace.ug.edu.gh The number of days it took soyabean lines from emergence to flowering (DTFR) had a significant (P<0.001) negative correlation (r=-0.45) with the area under disease progress curve (AUDPC) (Fig 5.2) recorded for the lines. On other hand, DTFR had a strong (P value<0.001) positive association (r= 0.64) with area under chlorophyll reduction curve (AUCRC). Also, the area under chlorophyll retention (AUCRC) had a significant (P<0.001) but negative association (r=-0.52) with AUDPC (Fig 5.2). Fig 5.2: Associations between DTF, AUDPC, and AUCRC 93 University of Ghana http://ugspace.ug.edu.gh 5.4 Discussion The result of the field evaluation revealed reactions of the lines to bacterial leaf pustule disease. This was through natural infestation for two consecutive years. Disease pressure was high in both years which enabled successful characterisation of the lines with regards to the pathogen, Xanthomonas axonopodis pv. glycine. Broadly, lines were grouped into two classes, resistant and susceptible. The resistant class was further separated into highly resistant or immune, resistant, and moderately resistant or tolerant. 28 lines had an immune response to the pathogen Xanthomonas axonopodis pv. glycine infection, and were therefore classified as highly resistant. Of the 28 highly resistant lines, eight including TGX 1989-20FC, TGX 1990-80FA, TGX-1835-10E, TGX-1989-40FB, TGX1989-48FN, TGX-1989-48FNA, TGX1989-68FN, TGX-1990-21F were advanced breeding lines sourced from IITA. Five Zimbabwean accessions, SARISOY 13A, SARISOY 15, SARISOY 20B, SARISOY 2A and SARISOY 5 were also part of the 28 highly resistant lines. Only one Ghanaian genotype Sal-1(Salintuya I), a variety released in 1985 (CSIR, 2015) was part of the 28 highly resistant lines identified in this study. Most of the lines in the 28 highly resistant lines were plant introductions sourced from the USDA such as PI506677, PI506939, PI398512, and PI398874. These lines had a smaller area under the disease progress curve and therefore recorded negative fixed effect estimates relative to the intercept (check), although, their AUDPC scores did not differ significantly from that of the intercept The second class of resistant lines simply designated as resistant, as grouped by AUDPC estimates were synonymous to lines that had a mean severity score of 1.5 to 2 on a 1-9 severity score scale. This tallies with lines classified as resistant with few lesions in the GRIN database (GRIN, 2018). Lines classified as moderately resistant based on AUDPC estimates had severity 94 University of Ghana http://ugspace.ug.edu.gh score ranging from 2.5 to 4. Lesions were conspicuous in the lower canopy of lines in this category but only a few lesions were observed in their upper canopy. Lines in this category were not significantly different from the intercept (PI606440A) according to a pairwise T-test based on AUDPC estimates. Jenguma, the most cultivated Ghanaian variety released in 2003 (National Seed Council, 2015) was grouped in this category based on AUDPC estimates. The susceptible group could be further divided into susceptible and highly susceptible. Lines in the susceptible category such as PI230970, PI398387, and LG 04-600 had severity scores of 5. Both lower and upper canopy of lines in this group were covered by lesions of bacterial leaf pustule. Lesions were observed on petioles and pods but no leaf defoliation occurred as reported by Hartman and Murithi (1999). These lines differed significantly from the intercept in the pair-wise T-test. On the other hand, leaf defoliation varying from few leaves to complete defoliation were observed, and in some lines, premature maturation and death occurred in a the highly susceptible category. Lines in this category recorded severity score from 6 to 9. AUDPC estimates for lines in this class differed significantly from the intercept. In this study, natural infestation provided the needed disease pressure which led to the successful characterisation of the 250 lines with regards to their reaction for a bacterial pustule. This result agrees with Abawi (2010) who noted that field screening for resistance provides for evaluation under a variety of interacting pressures including pathogenic and non-pathogenic organisms. Narvel et al., (2001) successfully mapped bacterial leaf pustule resistant gene (Rxp) in a segregation population through natural disease infestation under field condition. It was observed that some of the plant introductions (PIs) from the USDA recorded different reactions to bacterial leaf pustule as compared to their classification based on the disease as stated in The Germplasm Resources Information Network” (GRIN)database 95 University of Ghana http://ugspace.ug.edu.gh (https://npgsweb.ars-grin.gov/gringlobal/descriptors.aspx). For instance, PI417120 was classified as resistant with few lesions according to information on the GRIN database, but it was classified as susceptible and per its high AUDPC estimate, it actually belonged to the very susceptible group with a mean severity score of 8.5 in this study. However, some other lines recorded similar reaction and rating in this trial as their previous reactions stated on GRIN database. This included PI506677 classified as resistant with no lesion which also recorded immune response and was classified as highly resistant in this study. Also, PI81042 classified as completely covered was also classified as susceptible in this trial with a severity score of 8.5. The observed changes in the reaction to a bacterial pustule recorded for some lines in this trial as compared to previous reports could be due to break down of resistance. Pathogens including Xanthomonas axonopodis pv. glycine has the inherent ability to evolve to overcome the resistance offered by plants. The breakdown of resistance is often induced by the emergence of a new race or physiological form of the pathogen through mutation or introduction of a different race of the pathogen from another area (Rimbaud, 2018). Resistance breakdown is frequent with resistance conferred by a single or few genes as compared to resistance due to many genes (Brun et al., 2010). Single recessive genes were reported to be the source of resistance to bacterial leaf pustule in PI96188, and CNS soy (Hartwig-1951; Kim et al., 2011). It is therefore not surprising that some of the lines had a complete breakdown of resistance in this study. It was observed that most of the lines classified as resistant in this study flowered after 40 days. This resulted in a significant negative association between AUDPC and days to flowering (DTF). Higher values of AUDPC connotes susceptibility, hence the negative correlation between DTF and AUDPC implies that increases in DTF resulted in decreases in AUDPC. 96 University of Ghana http://ugspace.ug.edu.gh Most of the lines classified as resistant in the GRIN database belonged to maturity groups MGV, MGVI, MGVII, and MGVII. Lines in these maturity groups require a higher number of days to mature and this generally agreed with results obtained in this study. However, there were few lines that had resistant disease score but flowered earlier than 40 days. These lines could be evaluated and used to develop a bacterial pustule resistant early maturing cultivar for the Guinea and Sudan savannah ecologies of Ghana. SPAD readings taken during disease evaluation were used to estimate the area under chlorophyll retention (AUCRC). AUCRC had a significant but negative association with AUDPC. Chlorophyll depletion increased with disease severity and therefore, resistant lines retained higher chlorophyll content in their leaves than susceptible lines. Also, AUCRC had a positive association with DTF. This observation agrees with the association between DTF and resistance to bacterial pustule. In general, late maturing lines were resistant to bacterial pustule infection and therefore, there was less depletion of chlorophyll content of their leaves. Thus, a higher number of days to flowering generally resulted in higher chlorophyll content recorded over time which resulted in the positive association between the two traits. Foliar diseases generally degrade leaf tissues including chlorophyll as observed in yellowing or browning of leaves as disease severity increases. Ghose et al. (2010) reported a reduction of 53.24, 51.65, 56.55, and 58.04% for total chlorophyll, chlorophyll a, chlorophyll b and βcarotene, respectively in mulberry leaves due to blight infection. Due to the arbitrariness and errors associated with the visual assessment of disease severity, more accurate methods such as the light photography, digital imagery and hyperspectral techniques (Bock, 2010) are commonly used in advance research programmes for screening genotypes for resistance to diseases. However, the high cost and sophistication of some of this equipment is out of reach to most research programmes. Chlorophyll content, a trait easily measured with simple 97 University of Ghana http://ugspace.ug.edu.gh equipment such as SPAD, could be a more accurate trait to indirectly access the reaction of plants to foliar diseases like a bacterial pustule. 98 University of Ghana http://ugspace.ug.edu.gh 5.5 Conclusion Natural infestation under field conditions provided an effective environment for characterising the reactions of the 250 lines to bacterial leaf pustule. The 1-9 severity rating scale covered the different classes of response to the disease including defoliation and death which the 1-5 scale compounded with other reactions. The area under disease progress curve resolved the ordinal data into a quantitative one. Grouping of lines based on AUDPC estimates was efficient and resulted in the identification of resistant lines. Twenty-eight highly resistant lines were identified. Fourteen of the twentyeight highly resistant lines were tropical glycine lines obtained from IITA and Zimbabwe. These lines could readily be deployed in breeding programmes since they are adapted to the short-day length and high temperatures that characterises major soyabean production centres in Ghana. This study also identified a bacterial pustule resistant plant introductions (PIs) obtained from the USDA that had their resistance broken down and will therefore not be useful for developing bacterial pustule resistant cultivars for northern Ghana. The area under chlorophyll retention curve, a trait computed in this study, had a strong association with AUDPC and could, therefore, be further developed for use to indirectly screen soyabean lines for resistance to bacterial pustule. This result presents a useful information on SARI’s soyabean germplasm assembly including released varieties for scientists and farmers. It is highly recommended that the 28 highly resistant lines are further evaluated across bacterial pustule prevalent locations reported in Chapter three to identify stable sources of resistant for cultivar development. 99 University of Ghana http://ugspace.ug.edu.gh CHAPTER SIX 6.0 Field evaluation of soyabean genotypes for resistance to Phomopsis Seed decay (PSD) 6.1 Introduction The current soyabean production in Ghana stands at 143, 000 t from a total cultivated area of 87,000ha (MoFA-SRID, 2016). However, efforts are underway to make soyabean a major staple as well as a cash crop for the country (https://www.feedthefuture.gov/country/ghana). One of the eight key objectives of SIL was the utilization of soyabean for human nutrition through the provision of the high protein content of soyabean supplement for the eradication of malnutrition among children (Goldsmith, 2014). The high protein, carbohydrates, and vitamin A food complement is targeted at babies of six months to augment breastfeeding. Availability of high-quality grains, free of infection will be required for the realisation of this objectives. Though many biotic stresses including pests, diseases, and weeds constitute major production constraint to soyabean world-wide, diseases are by far the most important biotic constraint (Harman, 1999). Globally, many diseases, especially seed, and seed-borne diseases have been reported to cause significant yield loss, reduced grain quality, poor seed germination and high seedling mortality (Mengistu et al., 2012). Many of the pathogens of economic importance to soyabean are associated with its seed (Mengistu et al., 2012). Fungi in Diaporthe –Phomopsis complex (DPC) causes Phomopsis seed decay (PSD) and was considered the most important pre-harvest disorder of soyabean seed worldwide (Sinclair, 1999). PSD causes significant yield loss, and reduced grain quality in many countries (Riccioni et al., 2003; Medic´-Pap et al., 2007; Cui et al., 2009). Many control and management strategies have been explored to limit the effects of PSD. However, the use of resistant genes remains the most cost effective and environmentally sustainable approach to controlling crop diseases including PSD of soyabean (Ali, et al., 2013 100 University of Ghana http://ugspace.ug.edu.gh & Li et al., 2015). Many sources of resistance to PSD have been reported (Sun et al., 2013; Brown et al. 1987; Zimmerman and Minor, 1993; Li et al., 2011; Minor et al., 1995; Walters et al., 1973; Athow, 1987; Ross, 1986; Brown et al., 1987; Li et al., 2015). However, frequent breakdown of disease resistance requires continuous search for new sources of resistant for disease control. The first step toward deploying resistant genes is the identification of sources of resistance. The two exploited methods are, artificial inoculation of plants, and natural infestation of pathogens under field conditions. In soyabean, inoculating the stem of seedlings has been a method of screening for resistance to stem canker (Cui et al., 2009; Li et al., 2010) caused by members in the DPC of fungi. A modification of the inoculation of the immature stem of seedlings method has been developed by Li et al., (2018) for rapid evaluation of soyabean lines for resistance to PSD called the cut-seedling assay method. On the other hand, Li et al., (2015) successfully screened and recorded differences among 135 soyabean lines by spraying them with conidiospores to inoculate the lines under field conditions. Abawi (2010) noted that field evaluation of crops for resistance provides for evaluation under a variety of interacting pressures including pathogenic and non-pathogenic organisms. Thus, field evaluation mimics what pertains of on field conditions. Classifying lines as resistant or susceptible to PSD could be based on visual assessment of seeds of PSD infected plants for symptoms and signs of the pathogen on a 1-5 scale (Li et al., 2015). The other approach involves plating seeds of infected plants on agar medium and culturing it. The growth of the pathogen on seed is recorded and used to estimate the infection rate (Sun et al., 2013). Li et al., (2015) evaluated 135 soyabean lines at three locations in three years, and recorded a differential reaction of the lines to PSD infection between locations and years. This was 101 University of Ghana http://ugspace.ug.edu.gh attributed to differences in the pathogens across trial locations. The existence of different physiological races of pathogens has been linked to differential response of sources of resistance often leading to a breakdown in resistance as observed with Asian soyabean rust (Maphosa et al, 2012). In order to find stable sources of resistance to soyabean seed decay, it is imperative to screen assembled germplasm using prevailing strains of PSD in a target environment where the identified resistant genes will be deployed. It is therefore important to identify sources of resistance to Phomopsis seed decays (PSD) under prevailing conditions of northern Ghana where most of the country’s soyabean is produced. The goal of this study was to find sources of resistance to soyabean seed decay for development of PSD resistant cultivars. Objectives of the study were to; 1. Screen and identify PSD resistant lines for varietal development 2. determine the relationship between growth traits and PSD infection 102 University of Ghana http://ugspace.ug.edu.gh 6.2 Materials and Methods 6.2.1 Plant materials Plant materials used in this study involved a 250 soyabean lines comprising 131 plant introductions (PI) obtained from the USDA; 66 advanced breeding lines collected from the International Institute for Tropical Agriculture (IITA); 31 soyabean lines obtained from Zimbabwe in SARI’s holding; 14 Brazilian accessions, 3 Chinese accession lines, and 5 Ghanaian released varieties 6.2.2 Experimental design and field establishment. The trial was laid in an alpha lattice design in three replicates with each replication comprising of 25 blocks of 10 lines per block. Each line was planted in a single row of 5m long with a plant spacing of 5 cm. Rows were planted at 50cm apart. Phosphorus in the form of Triple Superphosphate (TSP) was applied at a rate of 60kg of P2O5 at emergence. Weed control was manually done with hoes. Planting was staggered based on the number of days to flowering data obtained when the lines were grown in the 2016 cropping season. This was done to ensure lines flowered at the same time to facilitate inoculation of lines with spores of Diaporthe sp. 6.2.3 Culturing of pathogen to obtain inoculum Stored Diaporthe sp. isolate MTB-1 were cultured on Acidified Potato Dextrose Agar (APDA) prepared. Media was prepared by suspending 15g of Agar in 1litre of distilled water. The resultant was heat sterilised by autoclaving at 121°C for 15 minutes. The prepared medium was acidified by amending it with 50% lactic acid (VWR International) and pH adjusted to 4.8. The acidified PDA was poured into 9cm- diameter petri dishes, cooled to room temperature and inoculated with stored mycelia of MTB-1 strain. The inoculated media was incubated at a temperature range of 25-30℃ and subjected to 12hr light/12hr dark phase. Several cultures 103 University of Ghana http://ugspace.ug.edu.gh were incubated simultaneously to obtain sufficient pycnidia from which to collect conidia. Cultures normally produced pycnidia between 15 to 25 days. 6.2.4 Preparation of inoculum and inoculation Culture plates bearing pycnidia were flooded with sterile water. Suspending pycnidia were agitated with a sterilised ceramic pestle. The resulting suspension of mycelia and conidia were filtered through a cheese cloth into a 1-liter conical flask. Hemocytometer readings and sterile water were used to adjust the concentration of conidia to 1x 105. Field inoculation was carried out on rainy days late in the afternoon between 5 and 6:30 pm, using flashlights. Three inoculations were carried out. The first inoculation was carried out at full flowering (R2) stage. The second inoculation was carried out during pod formation (R4) stage and the last inoculation carried at the seed development stage (R5). Five plants of each line were inoculated. Inoculations were targeted at the lower canopy. based on the planting date, flowering occurred in July, the rainiest month of the year during which relative humidity ranged from 72-93% with a mean of 83%. The mean temperature was 27.1 ºC, providing a good condition for infection and development of pathogen. The 12 hours between inoculation and sunshine also provided good conditions for infection whilst dense planting of 50 cm x5 cm ensured canopy closure, giving high humidity in the lower canopies of the plants. 6.2.5 Data collection and analyses Data were taken on the number of days to flowering and number of days to maturity for each line. Plants were harvested at the R8 stage. Pods were manually plucked and threshed. The seeds were visually assessed and rated on a 1-5 scale as described by Li and Chen (2013) using seed quality parameters including wrinkling, mottling, molding, discoloration, and shrinking where 1= excellent (no bad/infected seed); 2= good (less than 10%/infected seed); 3 =fair 104 University of Ghana http://ugspace.ug.edu.gh (1130% bad/infected seed); 4= poor (31-50% bad/infected seeds); and 5= very poor (> 50% bad/infected seeds). Visual assessment was carried out with the aid of a 100 watts LED bulb (Philip). 6.2.6 Plating of seeds and estimation of infection rate Plating was done according to the general procedure as described in Li and Chen (2013). Twenty-five seeds were randomly counted from each line. Seeds were surface disinfected with 70% ethanol, sterilised in 1 % sodium hypochlorite for 2 minutes and then rinsed in distilled water twice. Sterilised seeds were plated on the prepared APDA medium and incubated at a temperature range of 25-30℃ and 12hr light/ 12hr dark phase for 7 days. 25 seeds for each line were plated at a rate of 5 seeds per plate in 5 replications. Seeds with diaporthe sp. growth were counted and expressed as a percentage of the number of seeds plated. The scores from the 5 replications of seeds from each line were averaged to represent the score for the line. 105 University of Ghana http://ugspace.ug.edu.gh 6.3 Result Most of the lines had visual assessment score of 3. Forty lines had a score of 2 with 2 lines having a score of 1 (Fig 6.1). Infection rate Diaporthe MTB-1 range from 0 to 93 % with a mean of 53% and a median of 50%. Fig 6. 1: Distribution of visual assessment score and infection rate of soyabean seeds; A (histogram of visual assessment score); B (Box plot of seed infection rate) Seed weight had a very strong positive relation with seed quality but associated weakly with PLP. Seed quality had a positive significant association with seed infection. Seed quality also related to days to maturity, plant height at maturity and length at first pod and in all cases, the associations were significant. Seed infection rate of the pathogen, Diaporthe sp (MTB-1) had a significant correlation with days to 50% flowering, days to maturity and plant height. With the exception of 100 seed weight, height at first pod had a significant positive association with all the other traits. 106 University of Ghana http://ugspace.ug.edu.gh Table 6. 1: Pair-wise relation among growth traits, seed quality and seed infection rate DTFF 0.07 DTM 0.32*** 0.72*** IR 0.1 0.15* 0.11* PHT 0.04 0.26*** 0.23** 0.17* PLP 0.09 0.27*** 0.22*** 0.07 0.32** VQ 0.30*** 0.05 0.22*** 0.13* 0.21** 0.10* %100SW DTFF DTM IR PHT PLP Note: DTF (days to 50% flowering), DTM (days to maturity), IR (seed infection rate), PHT (plant height at maturity), PLP (Plant height at first pod), and VQ (visual quality assessment), 100SW(100 seed weight) Pair-wise t-test was performed in which the fixed effects estimate for each line was compared to that of the intercept, Jenguma, a variety that is moderately resistant to the PSD infection (Table 6.2). When analyses were performed on visual quality score, only ten lines recorded significantly lower scores than the intercept. Songdaa was the only locally released variety, and the only TGX line in SARI’s germplasm holding that had a significantly lower score than the intercept. Lines that differed significantly from the intercept had lower score. Jenguma had a score of 3 which implies a fair score with infection rate between 11-30%. The ten lines presented in Table 6.2 had a mean score in the range of 1-2.5 and are the most resistant lines. Table 6: 2 : Fixed effect estimates of visual scores (VQ) for lines relative to check Std. Line Estimate Error df t value Pr(>|t|) PI476905A -3 8E-01 1E+02 -3E+00 1E-03 ** PI416873B -2 9E-01 1E+02 -2E+00 2E-02 ** PI398825 -2 9E-01 1E+02 -2E+00 2E-02 ** PI605891A -2 8E-01 1E+02 -3E+00 1E-02 ** PI594767B -2 8E-01 1E+02 -3E+00 1E-02 ** Songdaa -2 8E-01 1E+02 -3E+00 1E-02 * FT christa -2 8E-01 1E+02 -3E+00 1E-02 * PI567516C -2 8E-01 1E+02 -3E+00 1E-02 * PI423958 -2 9E-01 1E+02 -2E+00 5E-02 * PI379621 -2 7E-01 1E+02 -2E+00 3E-02 * Jenguma(check) 4 6E-01 1E+01 7E+00 5E-10 *** 107 University of Ghana http://ugspace.ug.edu.gh Pair-wise test of fixed estimates of infection rates of lines against the intercept showed 59 lines had significantly lower fixed estimates than the check (Table 6.3). Infection rates were generally higher. Only 3 lines had mean infection rate below 10%. These were the lines with fixed effect estimates of -8. Fixed estimates for lines that performed better than the intercept ranged from -8 to -4, revealing five classes of lines that performed better than intercept. Songdaa had a fixed estimate of -6 and belonged to the third class of lines that performed better than the check. 9 lines obtained from Zimbabwe had the same estimates as songdaa. 28 of the 59 lines that had performed better than the intercept were ‘TGX’ lines from Zimbabwe and IITA. Full list of the fixed effect estimates of line based on both visual assessment and infection rate are presented in Appendix 3 and Appendix 4, respectively. 108 University of Ghana http://ugspace.ug.edu.gh Table 6. 3: Fixed estimates of infection rate (IR) relative to the check Line Estimate Std. Error df t value Pr(>|t|) Sig Liu Yuemang -8 2E+01 1E+02 -4E+00 3E-04 *** PI224271 -8 2E+01 1E+02 -5E+00 1E-05 *** PI417120 -8 2E+01 1E+02 -4E+00 3E-04 *** PI423958 -7 2E+01 1E+02 -4E+00 3E-04 *** BRS 361 -7 2E+01 1E+02 -4E+00 4E-05 *** PI417013 -7 2E+01 1E+02 -4E+00 4E-05 *** PI567516C -7 2E+01 1E+02 -4E+00 4E-05 *** Liu Yuebao -7 2E+01 1E+02 -4E+00 1E-04 *** Parangoina -7 2E+01 1E+02 -3E+00 2E-03 ** PI567091A -7 2E+01 1E+02 -4E+00 1E-04 *** SARISOY 10 -7 2E+01 1E+02 -3E+00 2E-03 ** SARISOY 15 -7 2E+01 1E+02 -3E+00 2E-03 ** SARISOY 18 -7 2E+01 1E+02 -3E+00 2E-03 ** SARISOY 21 -7 2E+01 1E+02 -3E+00 2E-03 ** SARISOY 26 -7 2E+01 1E+02 -3E+00 2E-03 ** SARISOY 5 -7 2E+01 1E+02 -3E+00 2E-03 ** SARISOY 21 -7 2E+01 1E+02 -3E+00 2E-03 ** TGX 1993-4FN -7 2E+01 1E+02 -4E+00 1E-04 *** Songdaa -6 2E+01 1E+02 -4E+00 4E-04 *** BRS 313 -6 2E+01 1E+02 -3E+00 1E-03 ** PI200492 -6 2E+01 1E+02 -3E+00 7E-03 ** PI230970 -6 2E+01 1E+02 -3E+00 1E-03 ** PI416810 -6 2E+01 1E+02 -3E+00 7E-03 ** PI605791A -6 2E+01 1E+02 -3E+00 1E-03 ** PI605891A -6 2E+01 1E+02 -3E+00 1E-03 ** PI628850 -6 2E+01 1E+02 -3E+00 1E-03 ** SARISOY 27 -6 2E+01 1E+02 -3E+00 7E-03 ** SARISOY 13 -6 2E+01 1E+02 -3E+00 7E-03 ** SARISOY 18 -6 2E+01 1E+02 -3E+00 7E-03 ** TGX 1903-7F -6 2E+01 1E+02 -3E+00 1E-03 ** PI416810A -5 2E+01 1E+02 -3E+00 8E-04 *** PI 534545 -5 2E+01 1E+02 -3E+00 3E-03 ** 109 University of Ghana http://ugspace.ug.edu.gh Table 6.3: Cont’d PI417126 -5 2E+01 1E+02 -3E+00 3E-03 ** TGX 1989-75N -5 2E+01 1E+02 -3E+00 3E-03 ** TGX 2010-11F -5 2E+01 1E+02 -3E+00 3E-03 ** TGX1989-68FN -5 2E+01 1E+02 -3E+00 3E-03 ** Zamboani -5 2E+01 1E+02 -3E+00 3E-03 ** PI200526 -5 2E+01 1E+02 -2E+00 3E-02 * PI398825 -5 2E+01 1E+02 -2E+00 3E-02 * PI399096 -5 2E+01 1E+02 -3E+00 7E-03 ** PI416806A -5 2E+01 1E+02 -2E+00 3E-02 * PI416806B -5 2E+01 1E+02 -2E+00 3E-02 * PI416939 -5 2E+01 1E+02 -3E+00 7E-03 ** PI518295 -5 2E+01 1E+02 -3E+00 7E-03 ** PI584796 -5 2E+01 1E+02 -3E+00 7E-03 ** PI594538A -5 2E+01 1E+02 -3E+00 7E-03 ** PI603176A -5 2E+01 1E+02 -2E+00 3E-02 * SARISOY 23 -5 2E+01 1E+02 -2E+00 3E-02 * SARISOY 10 -5 2E+01 1E+02 -2E+00 3E-02 * SARISOY 4 -5 2E+01 1E+02 -2E+00 3E-02 * TGX 1990-95F -5 2E+01 1E+02 -2E+00 3E-02 * PI159925 -4 2E+01 1E+02 -2E+00 2E-02 * PI416778 -4 2E+01 1E+02 -2E+00 2E-02 * PI416886 -4 2E+01 1E+02 -2E+00 2E-02 * PI567058D -4 2E+01 1E+02 -2E+00 2E-02 * PI567188 -4 2E+01 1E+02 -2E+00 2E-02 * PI606405 -4 2E+01 1E+02 -2E+00 2E-02 * PI635999 -4 2E+01 1E+02 -2E+00 2E-02 * TGX 1989-42F -4 2E+01 1E+02 -2E+00 2E-02 * Jenguma(check) 8 1E+01 1E+02 6E+00 2E-09 *** 110 University of Ghana http://ugspace.ug.edu.gh 6.4 Discussions Most of the lines had very low seed quality which resulted in fewer lines recording a score between 1-2.5, the category that could be described as resistant to moderately resistant. Infection rate which scores for the presence of a specific pathogen was high but had more lines performing better than the intercept. Seed quality significantly correlated with day to maturity and so was infection rate. The association between seed infection rate and seed quality was positive but not significant. This implies that the scored pathogen was not solely responsible for the signs and symptoms that determined seed quality. The positive significant association between seed quality and infection rate, and days to maturity recorded in this trial is contrary to what has been reported. It is generally held that early maturing lines were more susceptible to PSD infection than late maturing ones (Sun et al., 2013). Thus, a reduction in infection rate with increasing days to maturity would have yielded a negative relationship. The positive association could be due to the prevailing conditions during the trial. Divilov (2014) recorded a positive non-significant correlation between PSD incidence and days maturity and concluded that plant with different maturity periods could be inoculated at the same time without any negative effect on the outcome. The incidence of PSD in this trial was very high. This may be due to high disease pressure. High disease pressure observed in this trial was due to early planting that saw the maturity period coinciding with the heavy September rainfall that created high humidity amidst rising temperature when crops were between the R6 to R8 stage. Li et al., (2015) noted that frequent rainfall during late season caused high infection by different pathogens. The early planting approach adopted in this study affected late maturing lines as the maturity occurred in late September to early October. A period where the late September rains creates 111 University of Ghana http://ugspace.ug.edu.gh considerable humidity amidst rising temperatures. This could be the reason for the positive significant correlation between PSD incidence and days to maturity as the late rains provided humid and hot conditions for PSD development. This observation agrees to report by Rupe, (1990) that warm weather during seed development increase seed infection rate. Seeds developing under warm weather record higher PSD incidence and severity than seeds developed under cooler conditions. This is because whilst latter rains provide high pod moisture for infection (Rupe and Ferriss, 1986), warm temperature enhances development of the pathogen (Tekrony, 1984). Therefore, high disease pressure due to suitable condition at seed formation stages could have led to the breakdown of the resistance in some of the lines. Li et al., (2011) noted that consistent maintenance of suitable conditions for PSD development could result in breakdown of resistance. Jenguma, the check was completely susceptible as revealed by infection rate score and so was the susceptible check. Nonetheless, some lines recorded zero PSD infection and could be considered as highly resistant. 112 University of Ghana http://ugspace.ug.edu.gh 6.5 Conclusions Prevailing weather conditions during field evaluation created high disease pressure. In late maturing lines seed development met a condition of high moisture and warm temperature. This led to high PSD incidence among late maturing line resulting in a positive significant correlation between seed infection and days to maturity. The high presence of saprophytic fungi caused a difference between visual assessment and disease infection. This makes visual assessment less reliable for identification of resistance to specific pathogen. The lines PI416806B, PI224271, PI417120 BRS 361, PI417013, TGX 1903-7F, PI594767B, PI423958, TGX 1485-1D, TGX 2006-3F and Liu Yuemang were highly resistant to PSD infection under field conditions. These lines could be used in breeding programmes to develop early maturing PSD resistant cultivars. 113 University of Ghana http://ugspace.ug.edu.gh CHAPTER SEVEN 7.0 Genotyping by sequencing, and genome-wide association studies of PSD and BP resistance, and high PLP. 7.1 Introduction Genotyping-by-sequencing (GBS) offers a more flexible and an efficient approach to genomewide level identification of genetic variation in crops (Fu and Peterson, 2011; Peterson et al., 2014). GBS yields high-density genotypic information (Poland, 2012) on samples which is suitable for genome-wide association studies, and genomic selection (Poland and Rife, 2012; He et al., 2014; Gore et al., 2009; Yu and Buckler, 2006). GBS method provides the numerous generations of ancestral recombination events required for direct mapping of variation in natural populations without generating populations through crosses (Yu and Buckler, 2006). Chang et al., (2016), successfully conducted genome-wide association studies to characterise USDA's soyabean germplasm for resistance to diseases including bacterial leaf pustule. SARI has accumulated considerable amount of germplasm in which exist wide variation for various growth and reproductive traits, and reaction to disease as revealed in chapters 3, 4, 5 and 6 of this study. A combination of GBS and GWAS presents an opportunity to simultaneously finger print lines in the germplasm assembly for conservation and identify loci associated with traits of interest including disease resistance to support marker assisted and genomic selection. This will lead to efficient utilisation of assembled germplasm for cultivar development. Bacterial leaf pustule is wide spread in most soyabean production centres worldwide including neighbouring Benin (Kime et al., 2011; Zinsou et al., 2015). In Ghana, recent disease incidence survey identified bacterial leaf pustule to be wide-spread, confirming earlier report by Offei et al. (2008). Yield losses in the range of 15-50% due to bacterial leaf pustule have been reported 114 University of Ghana http://ugspace.ug.edu.gh (Rahayu, 2007). Field screening of SARI’s germplasm for resistance to bacterial pustule presented in Chapter five showed differential response of lines to the disease. Field screening activity identified resistant lines including lines that scored 1 which describes lines with no observable lesions on 1-9 scale. Mapping of loci in identified resistant lines and their linked markers will be required to facilitate resistant cultivar development through marker assisted selection. Phomopsis seed decay (PSD) is the most important cause of poor seed quality in soyabean (Li 2011, Koenning, 2009; Wrather & Koenning, 2009). The incidence of PSD in Ghana was first reported by Asante et al. (1998) and confirmed in Chapter three of this thesis. Favourable conditions for PSD infection and development (Balducchi and McGee, 1987; Suli et al., 2013) are common to soyabean production centers in Ghana. Secondly, early maturity cultivars which are suited for the production systems of the Guinea and Sudan savannah ecologies where most of the country’s soyabean is produced are reported to be more susceptible to PSD infection (Mayhew and Caviness, 1994 and Suli et al., 2013). It is therefore imperative to incorporate PSD resistance into ongoing varietal development efforts in Ghana. To efficiently incorporate PSD resistant genes into cultivar development programmes, knowledge of resistant loci and linked markers will be very useful. High pod clearance is key to limiting yield loss when combine harvesters are deployed to harvesting soyabean (http://www.seedcogroup.com/sites, 2018). It is observed that most of the locally released cultivars have low pod clearance. Whilst low pod clearance has no consequence on yield losses in production systems where manual harvesting is practiced, it is very important in commercial production systems that use combine harvesters. As the country gradually makes effort to transit its agriculture from subsistence system into a commercial one, plant architecture that lend itself for mechanization has to be considered and incorporated into breeding 115 University of Ghana http://ugspace.ug.edu.gh programmes. To this end, Soyabean genetic resources available at SARI were evaluated for pod clearance (PLP) as reported in chapter four of this study. Wide range of height at first pod (PLP) was observed including lines that combined high pod clearance with earliness. With the aid of molecular markers, the architecture of future cultivars could be modified to have high pod clearance to facilitate the use of combine harvesters. Genome-wide association study which facilitates the identification of loci of desired traits in assembled germplasm could be used to scan the genomes of individuals within SARI’s germplasm collection for loci and markers linked to high PLP. The goal of this research work was to map loci associated with PB and PSD resistance, and high PLP through genome-wide association studies. The specific objectives of this study were to: 1. genotype a representative panel of SARI’s soyabean germplasm holding 2. identify loci associated with pod clearance 3. identify loci associated with resistance/tolerance to bacterial leaf pustule 4. identify loci associated with resistance/tolerance to Phomopsis seed decay 116 University of Ghana http://ugspace.ug.edu.gh 7.2 Materials and Methods 7.2.1 Genotyping-by-sequencing 96 sub-sample of 250 lines 7.2.2 Plant materials A 96 representative panel comprised of 49 soyabean genotypes introduced from USDA, 22 IITA lines, 15 Zimbabwean line in SARI’s germplasm holding, 5 Brazillian lines, 5 locally released varieties and 1 Chinese line 7.2.3 Sampling of plant tissue for genotyping-by-sequencing Sample collection kit comprising a collection box, a cutting mat, tubes with perforated strip caps, a cutting tool and a bag of silica gel (desiccant). For each of the 96 selected lines, leaf tissues were obtained by cutting 8 leaf discs with the cutter. These were placed in a labelled tube, covered with perforated strip cap and a desiccant containing bag fastened to it. The collected tissues were sent to LGC for genotyping-by-sequencing. 7.2.4 Genotyping-by-sequencing DNA of individuals lines was digested using the two-enzyme modification of the original genotyping-by-sequencing (GBS) protocol (Elshire et al., 2011; Poland, Brown, Sorrells, & Jannink, 2012). PstI-HF and HindIII-HF were used as rare cutters in separate digests, and MsIII was used as the common cutter in each case. The digestion yielded fragment sizes of approximately 200bp. Sequencing of fragment was performed on the Illumina NextSeq500 equipment with paired –ends of 200bp in length. A total of 144 million reads at an average of 1.5 million reads per sample were realized. SNP calling was performed using the TASSEL GBS pipeline (Glaubitz et al., 2014), using bwa (H. Li & Durbin, 2009) to align them to the standard soyabean reference genome, Glycine_max_v2.0. at a mapping rate of 98.74%. 117 University of Ghana http://ugspace.ug.edu.gh A total of 158,592 SNPs were realized across all the 96 samples. TASSEL software was used to filter out genotypes with a depth of less than eight, sites with a minor allele count of less than three, and SNP position and taxa with more than 90% missing data. 7.2.5 Phenotypic data Phenotypic data for PLP, BP and PSD were generated as described in Chapter 4 (PLP), Chapter 5 (BLP) and Chapter 6 (PSD), respectively. 7.2.6 Data analysis Data were analyzed using the R statistical software (version 3.5.1) (R Core Team, 2018). A total of 32,638 single nucleotide polymorphisms (SNPs) spread across the 20 chromosomes of soyabean and scaffolds from the genotyping-by-sequencing (GBS) analysis were used. An initial genome-wide association study (GWAS) analysis was ran to identify and select the appropriate model using seven (7) models and two (2) phenotypic traits (namely: area under disease progress curve and flower colour). The selected model was deployed to analyse resistance to bacterial pustule based on AUDPC, seed decay due to Diaporthe sp. and PLP at a threshold value of to determine the existence of significant association. After which, detailed GWAS analysis was carried out on the traits with significant associations. The detailed association analysis was carried out in GAPIT (version 3.0, released: 2018.08.18) using the compressed mixed linear model (CMLM) (Z. Zhang et al., 2010) which is based on the following equation: 𝑌 = 𝑋𝑎 + 𝑄𝑏 + 𝐾𝑢 + 𝑒 Where Q is a matrix of principal component scores representing population structure, K is the kinship matrix calculated based on the VanRaden method (VanRaden, 2008; Lipka et al., 2012), X represents the genotype and Y the phenotype allowing associated values of each SNP to be calculated (X. Zhang et al., 2017). 118 University of Ghana http://ugspace.ug.edu.gh Different distance measures (mean, maximum, minimum and median) and agglomeration methods (average, complete and ward minimum variance) were used to arrive at the optimum cluster algorithm to group individuals based on their kinship. Identity of significant SNPs were graphically determined using interactive Manhattan plot. Residual variance and heritability estimates were determined at optimum cluster algorithm. 7.3 Results 7.3.1 General properties of SNP markers and individuals in the representative panel A plot of the first three principal components revealed six distinct groupings and few outliers (Fig 7.1). A heat map of the kinship matrix showed extent of correlation among subpopulations and among individuals within sub-population. Hierarchical clustering based on the kinship matrix revealed six clusters which separated into nine clusters, similar to what was inferred by cluster analysis in chapter four. Correlation within groups was high whilst correlation among sub-population was low. 119 University of Ghana http://ugspace.ug.edu.gh Fig.7.1: Population structure and kinship; A (PC1 VP C2), B(PC1VPC3); C (3D-plot of PC1, PC2 and PC3); D (Pair-wise correlation among 96 lines). Heterozygosity was generally low for both individuals and as well as SNP markers used (Fig 7.2).55 lines had heterozygosity values between 0.2 and 0.25 whilst 35 lines had heterozygosity between 0.15 and 0.2. Very few individual lines had heterozygosity above 0.25 with maximum heterozygosity value of 0.45 recorded. Though SNP markers were generally homozygous, few markers recorded heterozygosity value as high as 0.95. 120 University of Ghana http://ugspace.ug.edu.gh Fig.7.2: Extent of heterozygosity of A (individual lines) and B (SNP markers) The chance of encountering at least one SNP marker at a map distance of 1million centimogan was certain with 140000 SNP markers. This likelihood reduces to 0.4 with 50000 SNP markers and to 0.2 with 2000 SNP markers The non-random association between adjacent SNP markers, linkage disequilibrium was high at lower map distance with high accumulation of markers and generally decreased as map distance increased. Coefficient of determination (R2) was 0.6 at maximum linkage disequilibrium (LD). LD decreased to half its maximum at 280kb with variations among individual lines. 121 University of Ghana http://ugspace.ug.edu.gh Fig.7.3: A (SNP marker density), and B (decay of linkage disequilibrium with map distance) 122 University of Ghana http://ugspace.ug.edu.gh 7.3.2 Discovery of SNP markers associated with PLP Estimated genetic variance based on the compressed mixed linear model was 66.8% giving a moderately high broad sense heritability of 0.668 for PLP(Fig 7.4). Fig.7 4: Estimated variance components of PLP; genetic variance (66.8%) and Residual (33.2%) The Quantile-Quantile plot showed that most of the observed values were not in linkage disequilibrium with causal polymorphism. However, there were a few observations that were in linkage disequilibrium with causal polymorphism which produced significant P-values which deviated from that of the uniform distribution. Fig.7 5: Cross validation of significantly associated SNPs with PLP through quantile-quantile plot 123 University of Ghana http://ugspace.ug.edu.gh Further cross validation of the significance of the association between markers and PLP was carried out through a plot of the observed P-values against a range of minor allele frequencies of markers used (Fig 7.5). The most significant association occurred at a minor allele frequency of 0.15. All significant markers recorded in the QQ plot maintained their status. Fig.7 6: The effect of Minor allele frequency on the significance of association between markers and PLP Chromosome 3 had to two markers that were significantly associated with PLP as they occurred above the threshold (Fig 7.7). There were seven more associated SNP markers but they were exactly on the threshold. The most significant marker, S3_1485048, occurred at the physical map position of 1485048 in the standard soyabean reference genome (Glycine_max_v2.0). Coefficient of determination (R2) was high for S3_1485048 in the model. The second marker, S3_1991311 which occurred at the physical map position 1991311 had a lower observed Pvalue than S3_1485048 but its P-value was above a threshold computed at allele frequency 124 University of Ghana http://ugspace.ug.edu.gh of 0.01. Both markers had lower P-values than the false discovery rate. S3_1485048 and S3_1991311 both had negative allelic effect on plant length at first pod (Table 7.1). The physical map positions of the markers were probed in Phytozome12 (https://phytozome. jgi.doe.gov/phytomine/begin.do) and in Glycine max genome assembly version Glyma.Wm82 a2 (https://soybase.org/search/index.php) for the presence of candidate genes. Furthermore, the SNP sequences of the markers were aligned to deposited sequences through BLAST. The candidate gene, Glyma.03g014800, a member of the SAURlike auxin-responsive protein was identified (Fig 7.7). The candidate gene Glyma.03g014800 was found in the region between 1493449-1494342 which is about 8401 base pairs upstream of the physical position 1485048 linked to the marker S3_1485048. SAUR family of genes positively regulate cell expansion to promote hypocotyl growth (Spartz et al., 2012). The putative candidate gene was flanked by the markers, Sat_379 and BARC-042969-08479 on chromosome 3 in the soyabean reference genome. 125 University of Ghana http://ugspace.ug.edu.gh Fig.7 7: Manhattan plot displaying significant associations between markers and PLP Table 7 1: Properties of significant SNP markers and their allelic effects Rsquare.of Rsquare.of. .Model.with. FDR_Adjusted Allelic Effect P.value Model.without.SNP SNP _P-values Estimate SNP Chromosome Position 1.17E-09 0.0154 0.6511 0.00017 -26.28 S3_1485048 3 1485048 S3_1991311 3 1991311 4.32E-08 0.0154 0.5049 0.00058 -23.28 126 University of Ghana http://ugspace.ug.edu.gh Fig.7 8: Candidate gene, Glyma.0.3G.14800 associated with PLP Fig.7 9: SSR markers flanking the identified candidate gene Best linear unbiased predictor and linear best unbiased estimate were generated for the lines to support selection. Prediction of individual PLP was carried out using a combination of kinship data and phenotypic data (Table 7.1). SARISOYS27 had the highest random estimate and predicted mean PLP value. With the exception of SARISOYS27, random effects estimates were generally lower than fixed estimates. The difference between SARISOYS27 and the next, PI416873B in terms of random effect of PLP was very wide. An indication that SARISOY27 was exceptionally good. 127 University of Ghana http://ugspace.ug.edu.gh Table 7. 2:Genomic prediction of PLP for soyabean genotypes Line RefInf BLUP PEV BLUE Prediction Pred Heritable SARISOYS27 1 12.80 16.00 11.35 24.14 24.14 PI416873B 1 5.69 12.90 12.26 17.95 17.95 TGX-2011-3F 1 4.66 15.26 11.31 15.97 15.97 TGX2008-4F 1 4.32 14.68 12.72 17.04 17.04 PI567104B 1 4.30 16.13 9.49 13.79 13.79 TGX-2004-10F 1 3.26 16.26 13.11 16.37 16.37 Songdaa 1 3.20 16.98 10.53 13.74 13.74 SARISOYS30 1 1.93 17.39 10.05 11.98 11.98 PI423960A 1 1.61 15.46 11.45 13.05 13.05 TGX1988-3F 1 1.58 13.14 10.17 11.75 11.75 PI567099A 1 1.57 15.44 9.86 11.43 11.43 PI567069A 2 1.47 -39.82 12.88 14.35 14.35 PI588053B 1 1.47 16.10 11.69 13.15 13.15 PI567123A 1 1.39 17.17 10.19 11.58 11.58 PI567053 1 1.37 15.55 10.45 11.82 11.82 PI3984-77A 2 1.32 -55.29 12.76 14.09 14.09 PI567068A 1 0.91 16.37 9.96 10.87 10.87 PI398477 1 0.85 17.00 10.33 11.18 11.18 PI605824A 1 0.80 14.97 13.08 13.87 13.87 PI605891B 1 0.79 19.60 7.36 8.15 8.15 SARISOYS-14 1 0.77 16.20 10.47 11.24 11.24 PI657753 1 0.65 16.28 13.29 13.94 13.94 PI594760B 1 0.61 17.43 8.69 9.30 9.30 Sal-1 2 0.60 -89.58 12.16 12.76 12.76 BRS-326 1 0.59 15.67 9.13 9.72 9.72 TGX-1990-57F 1 0.53 14.19 13.04 13.56 13.56 TGX-1844-22E 2 0.50 -20.15 10.50 11.00 11.00 PI567076 1 0.46 14.18 11.81 12.27 12.27 PI567046A 2 0.45 -23.47 10.47 10.92 10.92 BRS-7980 1 0.35 17.32 11.13 11.49 11.49 PI606397B-2 2 0.33 -99.38 6.69 7.02 7.02 TGX-1988-3F 1 0.32 16.89 8.80 9.12 9.12 TGX1989-40F 2 0.31 -109.13 9.18 9.49 9.49 PI567129 1 0.31 19.42 7.17 7.48 7.48 SARISOY18 1 0.26 14.94 11.32 11.58 11.58 Liu yeumang 1 0.14 17.04 8.37 8.51 8.51 PI606397B 2 0.13 -96.09 6.62 6.75 6.75 PI417013 1 0.11 13.88 11.56 11.66 11.66 SARISOYS29 1 0.10 15.91 12.09 12.19 12.19 SARISOYS-1 1 0.10 14.97 11.93 12.03 12.03 PI462312-2 2 0.05 -87.86 10.41 10.46 10.46 SARISOYS-5 1 0.05 16.49 9.01 9.06 9.06 PI561381 1 0.05 16.88 10.74 10.79 10.79 128 University of Ghana http://ugspace.ug.edu.gh 7.3.3 Genome-wide scanning for loci associated with bacterial leaf pustule Estimate of heritability of lines reaction to bacterial pustule was moderate (Fig 7.8). Mahanttan plot based on the supper model revealed peaks on chromosomes 16, 8, and 10 (Fig 7.9). Two markers S16_34242886 and S16_35202484 with minor allele frequencies of 0.36 and 0.38, respectively had associations with area under bacterial pustule progress curve but the associations were not significant. The peaks found on chromosome 8 were due to the markers, S8_11831127 and S8_11842172 with minor allele frequencies of 0.30 and 0.33, respectively. On chromosome 10, the SNP marker, S10_12149388 produced the observed peak. In all cases, the associations between the markers and reaction to bacterial pustule infection were not significant. Association analyses was also repeated using the compressed mixed linear model (CMLM). The resulting manhanttan plot revealed oustanding peaks on chromosomes 1, 3, 6, 13, and 16. The markers that yielded the observed peaks were S1_38672307, S3_33047652, S6_6701252, S13_4394677, and S16_20916972 on chromosomes 1, 3, 6, 13, and 16, respectively. The peaks on chromosomes 1, 13 and 16 had higher P-values. They were however not significant as they did not significantly deviate from the expected distribution. It is noteworthy that in both models, the peak on chromosome 13 was observed. Also, the marker S8_11831127 that yielded the peak 8 on chromosome in the supper model also appeared in the compressed mixed linear model but at lower p-value. 129 University of Ghana http://ugspace.ug.edu.gh Fig.7.10: Estimate of heritability of reaction of soyabean lines to bacterial pustule infection; genetic variance (50%) and residual variance (50%) Fig.7.11: A (Manhattan plot based on the super model); B (Manhattan plot based on the compressed mixed linear model) 130 University of Ghana http://ugspace.ug.edu.gh P-values obtained from test of association based on the super model did not deviate from that of the uniform test (Fig 7.9). P-values of test of association based on the compressed mixed linear model showed deviation from that of the uniform test. The deviations were however not significant at the genome-wide association level. Peaks indicating association were observed at all frequencies of allele (Fig 7.9). Noticeable peaks were recorded at minor allele frequencies in the ranges of 0.1 to 0.5 with the highest peak occurring at 0.22. All peaks of associations did not pass the false discovery rate, irrespective of allele frequencies of markers involved. Fig 7.12: Effect of allele frequencey on false discovery rate, and test of statisitical significance of association between genomic regions and resistance to bacterial pustule Although recorded peaks were not significant associations between genomic regions and bacterial pustules resistance, the physical positions of SNPs which yielded the highest peaks were probed on the reference genome (Fig 7.13). One leucine-rich repeat receptor-like protein kinase resistant gene candidate (Glyma.01go93200) was found within the LD region of S1_386723 on chromosome 1. Two leucine-rich repeat receptor-like protein kinase resistant gene candidates (Glyma.16g187500 and Glyma16.g187400) were found within the LD regions of S16_34242886 and S16_35202484 on chromosome 16. 131 University of Ghana http://ugspace.ug.edu.gh A B Fig.7.13: Candidate genes associated with BP resistance A(candidate gene on chromosome 1), B(candidate gene on chromosome 16) 7.3.4 Genome-wide scanning for loci associated with PSD resistance Seed decay due to Diaporthe sp. was accessed using visual assessment, and assessment of pathogen infection rate through seed plating. Genomic estimate of heritability was higher for visual assessment than for pathogen infection rate based on seed plating (Fig 7.10). Inheritance of seed decay was moderate to moderately high based on method of assessment. B A Fig.7.14: Genetic and residual variance of components of soyabean seed decay due to Diaporthe sp. (variance componentof seed decay based on visual assessment), B (variance components of seed decay based on infection rate of pathogen) 132 University of Ghana http://ugspace.ug.edu.gh Three peaks were recorded on the Manhattan plot based on data from infection rate (Fig 7.16). The peaks occurred on chromosome 2 (S2_19811300), chromosome 7(S7_1562611), and on chromosome 18(S18_7792728). All peaks observed did not amount to significant associations between genomic regions and seed infection rate. Data obtained from visual assessment did not produce conspicuous peaks. Fig.7.15: Manhattan plot depicting associated with Diaporthe sp seed decay resistance on soyabean chromosomes Probing lack of significant association was performed through quantile-quantile plots (Fig 7.17). In both studies, the distribution of observed p-values clearly deviated from that of the ideal distribution. The deviations however did not point to any significant association between genomic regions and resistance to seed decay in soyabean. Fig 7. 16. Test of significance of observed p-values of association studies. A (QQ plot of p-values from) infection rate association studies), B (QQ plot of p-values from visual assessment association studies) 133 University of Ghana http://ugspace.ug.edu.gh Though not significant, physical positions of SNPs associated with conspicuous peaks were probed for identification of genes on SoyBase (https://www.soybase.org) and Phytozome2 (https://phytozome.jgi.doe.gov/pz/portal.htmland). The candidate gene Glyma.18g082100 was identified within the vicinity of the physical position of the SNP, S18_7792728 located on chromosome 18. Glyma. 18G082100 which is located at about 0.2M upstream of S18_7792728, codes for Leucine rich repeat-containing proteins and disease resistance protein. Fig 7. 17. Candidate genes identified in the chromosomal region linked to S18_7792728 on chromosome 18 134 University of Ghana http://ugspace.ug.edu.gh 7.4 Discussion 7.4.1 Characteristics of samples and GBS generated SNP markers Breeding for crop improvement thrives on genetic diversity. Detailed evaluation of available germplasm facilitates crop improvement through identification of the underlying allelic variation affecting crop phenotype (Zhou et al., 2015). In this study a representative sample of 96 lines from SARI’s soyabean germplasm holding were genotyped through GBS. A biplot of the first two principal components did not show clear clustering patterns among the lines but a 3D plot involving the first three principal components revealed 6 clusters with few outliers. A pair-wise correlation based on a heat map of the kinship matrix indicates low correlation among sub-groups within the representative sample. This result points to high diversity among the lines in the sample. The high genetic diversity increases the chance for capturing the allelic variation driving phenotypic variation in traits like days to maturity, seed size, and reactions to diseases like bacterial pustule. Individuals lines within the sample had low observed heterozygosity with more than 90% of them recording heterozygosity values in the range of 0.0 to 0.25. This implies that majority of lines in the sample were almost fixed across all loci and therefore little or no further segregation is expected. Thus, they are in a good state for conservation, and for studies that involve evaluating lines for multiple years as it is normally done for quantitative mapping like GWAS. 48000 (30.2%) of SNPs generated through GBS on the representative sample had heterozygosity in the range of 0.0 to 0.05. Markers with zero heterozygosity were not informative and were therefore filtered out before performing association studies. Non-random association of alleles at different loci among the lines was maximum at R2=0.6. Decay of disequilibrium was rapid reaching half its maximum value at a map distance of 2.8+e5(R2=0.3). The R2 at maximum LD was higher among lines used in this study than was 135 University of Ghana http://ugspace.ug.edu.gh reported by Liu et al. (2016) who obtained R2 of 0.48 at for 146 soyabean accessions. Strong LD exists within the 96 lines. LD is said to be stronger among cultivars than in landraces, and so the presence of advance breeding lines in the sample evaluated might have accounted for the high LD. Whilst some of the lines, especially, plant introductions from USDA were land landraces, IITA lines were advanced breeding lines. They were normally received in batches of lines from a single or few populations. Such lines with common ancestry harbour common linkage blocks of non-random associating alleles. 7.4.2. Associations studies involving PLP 66.8% of variation in PLP among the lines was due to genetic effect. This suggests that considerable gain could be made through hybridization and selection. QQ plot revealed the presence of significant deviations from uniform distribution, pointing to the existence of excess relatedness among some lines in the sample evaluated. Manhattan plot showed two markers were strongly associated with plant length at first pod on chromosome 3. The extend of the strength of the association between the two markers was pronounced in both observed p values and in R2 when each was included in the model. The most significant marker, S3_1485048 had observed p-value of 1.1e-9 and its inclusion in the model increased R2 from 0.0154 to 0.651. The second marker, S3_119311 also had p-value of 4.32E-08 and increased R2 from to 0.0154 to 0.5049. Both markers had large allelic effect on PLP. The allelic effects of both markers were negative with S3_1485048 having a higher effect. They will therefore be useful for selecting against low pod clearance but not for high clearance. Further probe at the physical map locations where these markers occurred revealed a few candidate genes including Glyma.03G014800. The candidate gene Glyma.03G014800 is SAUR like-auxin responsive protein. SAUR family of genes regulate a wide range of growth and developmental processes in plants (Ren and Gay, 2015). McClure and Guilfoyle (1987) 136 University of Ghana http://ugspace.ug.edu.gh observed that SUAR genes were associated with hypocotyl elongation in soyabean. Subsequent findings confirmed SAUR family genes positively regulate cell expansion to promote hypocotyl growth (Spartz et al., 2012; Stamm and Kumar, 2013). Cell expansion, cell division, elongation growth, tropic growth among others have been identified as the functions of genes in the SAUR family (Ren and Gray, 2015). The hypocotyl is the primary organ of extension of the young plant and develops into the stem (Das, et al., 2016). In soyabean, hypocotyl pulls the cotyledon out of the soil and determines how high above the ground the cotyledon reaches before the unifoliate leaf and its auxiliary bud are formed (Shaun, 2010). Low pod clearance (PLP) has to do with formation of pods close to the ground which suggest short hypocotyl or short distance between the cotyledon and the first trifoliate. Discovery of SNP markers linked to the SAUR family of genes responsible for elongation of the hypocotyl and extension of young plant could be useful for selection for high clearance. One source of yield loss to combined harvesting is cutting below the cut-height (Hanna, 2010). As noted in Seedco Farmer’s Guide (2018), the height of the pod off the ground determines the extent of yield loss with combined harvester. Height of soyabean crop correlates positively with days to maturity (Malik et al., 2007). This limits the choice of breeding for taller varieties to overcome low pod clearance because of the short cropping season in which most of the country’s soyabean is produced. The putative candidate gene and its linked markers could facilitate breeding for increased pod clearance without necessarily increasing plant height and its consequent effect on maturity period. A number of lines with predicted high pod clearance have been identified. Most of these lines had pod clearance of more than 10 cm and could be considered in breeding programmes to breed for increased pod clearance. 137 University of Ghana http://ugspace.ug.edu.gh 7.4.3 Loci associated with bacterial leaf pustule The compressed mixed linear model (Zhang et al., 2010) was used for evaluating the association between bacterial leaf pustule infection and genomic regions of lines. Peaks were observed on chromosomes 1, 3, 6, 13, and 16. None of the observed peaks amounted to a significant association. Quantile-quantile plot revealed a clear deviation of the observed pvalues from that of the uniform distribution but there was no sufficient statistical power to give statistical significance. A second plot was carried out based on the super model (Wang, 2014) which boast of statistical power when even small number of genetic markers are used. Manhattan plot based on the super model identified peaks on chromosomes 16, 8 and 10 with minor allele frequency in the range of 0.30 to 0.36. All peaks identified were not significant. Different peaks were recorded for the two models except the peak on the chromosome 16, which was present in both models. Although association as revealed by Manhattan plot were not significant, a cursory probe of the physical position of the peaks on chromosome 1 and 16 indicated the LD within these regions harbours resistant genes. Three putative candidate genes, Glyma.01go93200, Glyma.g187500 and Glyma.g187400 were identified on chromosomes 1 and 16, respectively. All three- candidate genes codes for leucine-rich repeat receptor-like protein kinases. Leucine rich repeat receptor like protein kinase are leucine-rich repeat receptors localised in plant cell membrane. They play a key role in plant response to pathogen infection by recognising Pathogen Associated Molecular Patterns (PAMP) and endogenous Damage Associated Molecular Patterns (DAMP) and elicit immune response (Yamaguchi et al., 2006; Huffaker et al., 2006). Chang et al. (2016) reported of the markers ss715580342, ss715609404 and ss715628133 on chromosomes 1, 11 and 17, respectively linked to genomic regions offering resistance to 138 University of Ghana http://ugspace.ug.edu.gh bacterial pustule disease. These markers were identified in a genome-wide association studies and they occurred in genomic regions codding for leucine-rich repeat receptor-like protein kinases. Before then, Kim et al 2010 fine mapped the rxp gene identified by Narvel et al. (2001) to a 33kb region and identified two candidate genes, Glyma17g09780 and Glymag09790. Glyma17g09780 is membrane protein whilst Glymag09790 was identified as a zinc finger. With regards to PSD, the candidate gene, Glyma.18G082100 was observed upstream of the physical position of the SNP S18_7792728. A number of molecules including IPR000767 (Disease resistance protein), IPR025875 (Leucine rich repeat 4) and GO:0006952 (defense response) were associated with Glyma.18G082100. Additionally, PTHR23155 (Leucine-Rich Repeat-Containing Protein), and AT3G07040.1 (NB-ARC-domain-containing disease resistance protein) were found within the physical map position of SNP S18_7792728. Recognition of invading pathogen is the important step towards plant response to diseases. Plants have thus evolved mechanisms to recognize and elicit rapid response to invading pathogen leading to basal resistance (Freeman and Beattie, 2008). Recognition of invading pathogen is achieved through resistant proteins, most of which contain central nucleotide- binding domain (NBARC domain-containing disease resistance protein) (Ooijen et al., 2008). All the molecules associated with candidate gene, Glyma.18G082100 are involve in recognition and activation of basal or immune response to pathogens (Alywin and Rammik, 2011). The presence of LD regions harbouring Leucine-rich repeat receptor like protein kinase in genomic regions confirms that deviation from the normal as depicted in QQ plot were due to presence of LD blocks. The failure to obtain significant associations for observed peaks could be as a result of lack of statistical power due to small sample size and or structure of the representative panel. Chang et al., (2016) noted that large population size does not necessarily translates into significant results despite high power. In their trial, the same significant SNP was 139 University of Ghana http://ugspace.ug.edu.gh detected with a sample size of 79 as with 9,865 samples. Furthermore, (Hart and Griffiths 2015) identified resistance to viral disease in common bean with just 84 recombinant inbred lines, and botrytis immunity in Arabidopsis with 96 accessions. Representative diversity panel has been as suggested as equally important as population for GWAS (Chang et al., 2016). Hence, failure to obtain significant association in this study could be due to the structure of 96 representative panel used. Similar pattern was observed for seed decay due to Diaporthe sp. Although, the panel was diverse, structure for traits such as maturity period existed in the panel. The gene-for-gene theory suggests that for every gene that conditions reaction in the host, there is corresponding gene that conditions pathogenicity in the parasit). The implication of this is an arm race between host and pathogen. This often leads to breakdown of resistance and the emergence or discovery of new sources of resistance (Brun et al., 2010). This warrants the continuous search of sources of resistance to control area specific races or newly emerged races of a pathogen. Twenty-eight of 250 lines screened for bacterial leaf pustule infection recorded no lesion despite high disease pressure on the field. They were classified as highly resistant. They could be very useful for bacterial pustule resistant cultivar development. 140 University of Ghana http://ugspace.ug.edu.gh 7.5. Conclusion GBS generated over 150000 markers, 32000 of which are fairly polymorphic with less than 10% missing data. This can be used for quantitative mapping including association mapping. Ninety-six diverse, highly fixed lines with strong linkage disequilibrium has also been constituted to form a representative panel of the SARI’s soyabean germplasm holding. Plant length at first pod (PLP) has been identified as a critical trait in the soyabean architecture that will influence future mechanization of soyabean harvesting. Two SNP markers, S3_14585048 and S3_1991311 with large negative effect on PLP has been identified to support marker assisted selection. The gene influencing PLP has been identified as a SAUR like-auxin responsive protein known to positively regulate cell expansion to promote hypocotyl growth. Associations between genomic locations and resistance to bacterial leaf pustule were identified but the associations were not significant as a result of low statistical power due to small sample size. Nonetheless, three SNP markers, S16_34242886, and S16_35202484 and S1_38672307 were found linked to three candidate genes, Glyma.01go93200 Glyma.g187500 and Glyma.g187400). All three candidate genes code for Leucine-rich repeat receptor like protein kinases for eliciting immune response to pathogen infection. The linked markers could be validated and used for marker assisted selection. The SNP, S18_7792728 associated with PSD resistance is linked to a genomic region that harbours genes known to play critical roles in plants response to pathogens. This confirms the existence of resistant genes for PSD among the lines in the representative panel. The identified SNP could be validated to aid with marker assisted selection. 141 University of Ghana http://ugspace.ug.edu.gh CHAPTER EIGHT 8.0 General conclusion and recommendation 8.1Conclusion Understanding the prevalence and distribution of plant diseases across production centres is required for effective prevention and control of diseases. This is even more critical when resistant genes are to be deployed to controlling diseases because of the resources and time required to identify and transfer resistant genes into cultivars. In order to ascertain the incidence, severity and distribution of soyabean diseases in Ghana, a survey was carried out covering production centres in the Northern, Upper East, Upper West, Brong Ahafo and Ashanti. The survey identified seven fungal diseases, two viral diseases and one bacterial disease. Of the ten diseases identified during the survey, bacterial leaf pustule, cercospora leaf blight, frog eye leaf spot and two viral diseases were present at varying levels of incidence and severity at almost all 15 locations surveyed. Bacterial leaf pustule and soyabean mosaic virus occurred at all 15 locations surveyed whilst frog eye leaf spot and cercospora leaf blight occurred at 13. Malzire, Karaga and Gushegu all in the northern region had higher severity of fungal and bacterial diseases encountered than other locations. Viral diseases were more important at Ejura and Wenchi than other locations surveyed. Binduri, Navrongo and Manga in the Upper east region generally recorded lower incidence and severity of identified disease than locations in the other regions. The study also identified soyabean seed decay disease, and successfully confirmed the causal agent to be Diaporthe sp. and tentatively designated the isolate as MTB-1. This is the first report of soyabean seed decay due to Diaporthe/Phomopsis complex of fungi in Ghana. SARI’s germplasm was screened for resistance to bacterial pustule which was the widest spread disease identified during the survey. Sources of resistance to bacterial leaf pustule exist within the assembled germplasm including twenty-eight lines that elicited immune response to 142 University of Ghana http://ugspace.ug.edu.gh bacterial leaf pustule. This study also identified bacterial pustule resistant plant introductions (PIs) obtained from the USDA that had their resistance broken down and will therefore not be useful for developing bacterial pustule resistant cultivars for northern Ghana. The area under chlorophyll retention curve, a trait computed in this study, had a strong association with area under disease progress curve and could therefore be further developed to indirectly screen soyabean lines for resistance to bacterial pustule. Identification of resistant loci and linked markers are required for marker assisted selection. GWAS was carried out on a 96 representative panel to identify genomic regions associated with resistance to bacterial leaf pustule. Three SNP markers, S16_34242886, and S16_35202484 and S1_38672307 linked to bacterial leaf pustule resistance were identified. Although, the association between these SNP markers and bacterial pustule resistance was not statistically significant, three candidate genes, Glyma.01go93200 Glyma.g187500 and Glyma.g187400 found at the physical positions of these markers code for Leucine-rich repeat receptor like protein kinases known for eliciting immune response to pathogen invasion. In order to support pre-emptive breeding efforts against the newly identified disease of soyabean in Ghana, Phomopsis seed decay (PSD), assembled germplasm was screened for sources of resistance. Eleven resistant lines were identified to aid with PSD resistant cultivar development. GWAS identified the SNP S18_7792728 linked to the candidate gene, Glyma.18G082100 which codes for central nucleotide-binding domain (NB-ARC domaincontaining disease resistance protein) also known for stress response including invading pathogens. In both cases of bacterial leaf pustule and PSD resistance, the association studies did not identify significant SNP due to lack of statistical power as a result of small size and structure of the representative panel. Morphological evaluation showed wide diversity among accessions in SARI’s germplasm holding and observed variation were largely under genetic control. Number of pods per plant, 143 University of Ghana http://ugspace.ug.edu.gh number of seed per plant, and days to flowering were the main drivers of variation among the assembled germplasm. Unbiased structure analysis of a 96-line representative panel revealed that accessions in the germplasm belonged to nine sub-populations with high diversity among lines in the panel which indicate that wide variations observed in morphological traits were due to genetic effect. Ninety-six diverse, nearly fixed lines with strong linkage disequilibrium has also been constituted to form a representative panel of the SARI’s soyabean germplasm holding. GBS generated over 150000 markers, 32000 of which are fairly polymorphic with less than 10% missing data. This can be used for quantitative mapping including association studies. Shaping the architecture of plant is an important objective in plant breeding. Pod clearance influences losses when combined harvesters are used for harvesting. SARI’s soyabean germplasm was evaluated for pod clearance and the result showed that clearance was highly heritable and some accessions combined high pod clearance with higher number of pods per plant. Two SNP markers, S3_14585048 and S3_1991311 with large negative allelic effects on PLP has been identified to support marker assisted selection. The gene influencing PLP has been identified as a SAUR like-auxin responsive protein known to positively regulate cell expansion to promote hypocotyl. The sequence of the candidate gene could be used to design more primers for marker assisted selection. Lines with high pod clearance have been identified through prediction based on random effects estimate. These lines could be exploited to develop productive cultivars that will be suitable for mechanisation including the use of combine harvesters. 144 University of Ghana http://ugspace.ug.edu.gh 8.2 Recommendation The resistance lines identified for bacterial leaf pustule and Phomopsis seed decay should be further evaluated in multi-location trials to identify stable sources of resistance for cultivar development. The information on diseases, their severity and distribution should inform extension strategies for soyabean disease management for specific locations. Also, the 96 lines genotyped and the inferred groupings should form the bases of conservation efforts at SARI. Furthermore, SNP markers generated through genotyping by sequencing could be used to genotype additional lines to increase the number of lines in the representative panel to attain the required statistically power for GWAS. SNP markers and candidate genes identified could be further probed and validated to aid with marker assisted selection. Information on pod clearance generated in this study including linked markers and predicted genotypes could be used to alter the architecture of future varieties to make them suitable for mechanisation, especially, combine harvesting. 145 University of Ghana http://ugspace.ug.edu.gh BIBLIOGRAPHY AND APPENDICES Abe, J, Xu, D, Suzuki, Y, Kanazawa, A, Shimamoto, Y. (2003). Soybean germplasm pools in Asia revealed by nuclear SSRs. Theoretical and Applied Genetics 106: 445– 453. Acuña, R. (2008). Compendio de fitopatógenos de cultivos agrícolas en Chile. 123 pp. Servicio Agrícola y Ganadero (SAG), División Protección Agrícola, Santiago, Chile. Agada, M. O. (2015). Constraints to Increasing Soybean Production and Productivity in Benue State , Nigeria, 4(4), 277–284. https://doi.org/10.9734/AJAEES/2015/13759 Aidoo, R., Mensah, J. O., Opoku, A., & Abaidoo, R. C. (2014). Factors influencing Soybean Production and Willingness to Pay for Inoculum Use in Northern Ghana, 4(3), 2014. Ali, J., Xu, J., Gao, Y., Ma, X., Meng, L., Wang, Y., … Li, Z. (2017). Harnessing the hidden genetic diversity for improving multiple abiotic stress tolerance in rice (Oryza sativa L) PLOS ONE, 12(3). https://doi.org/10.1371/journal.pone.0172515 Amagloh, K., Andrade, J., Amagloh, F. K., & Andrade, J. E. (n.d.). ECN Study-Report F Community-level acceptability and feasibility assessments of soy-containing complementary food blends by women and children in rural Ghana. Retrieved from http://soybeaninnovationlab.illinois.edu/sites/soybeaninnovationlab.illinois.edu/files/ Early Childhood Nutrition Study Report%2C 2018.pdf Anderson TR, Buzzell RI (1985). Heritability of tolerance to Diaporthe/Phomopsis seed mold of soybean. In: Proceedings of the Conference on the Diaporthe/Phomopsis Disease Complex of Soybean MM Kulik, ed. 1985, p. 59–62. Antwi-Boasiako, A., Karikari, B. and Abrokwah, A. O. (2017). Jour. Sci. Res. Allied Sci., 6 (3), 2017, 470-483 Asante, S. N. A., Wolffhechel, H., and Neergaard, E. d. (1998). First report of Diaporthe phaseolorum var. meridionalis on soybean seeds from Ghana. Plant Disease 82:1401- 1401. https://doi.org/10.1094/PDIS.1998.82.12.1401A Atakora, W.K. and Kwakye, P .K. (2016). Measurement and Modeling Nitrous Oxide Emissions from Ferric Luvisols in the Measurement and Modeling Nitrous Oxide Emissions from Ferric Luvisols in the Guinea Savanna Agro-ecological Zone of Ghana, (January). https://doi.org/10.9734/IJPSS/2016/24165 Athow, K. L. (1987). Fungal diseases. Pages 687-727 in: Soybeans Improvement, Production, and Uses. J. R. Wilcox, ed. Source Book, Soyinfo Center, Lafayette, California, USA, 2009. Bafour Asafo-Adjei. (2015). Morphology and Physiology of Soybean. Retrieved November 1, 2018, from https://es.slideshare.net/IITA-CO/morphology-and-physiology-of- soyabean 146 University of Ghana http://ugspace.ug.edu.gh Balducchi, A. J., and Mcgee, D. C. (1987). Environmental Factors Influencing Infection of Soybean Seeds by Phomopsis and Diaporthe Species During Seed Maturation. Plant Disease 17 (3). https://doi.org/10.1094/PD-71-0209. Bale, J. S., van Lenteren, J. C., & Bigler, F. (2008). Biological control and sustainable food production. Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 363(1492), 761–776. doi:10.1098/rstb.2007.2182 Baral, K., Coulman, B., Biligetu, B., & Fu, Y. B. (2018). Genotyping-by-Sequencing Enhances Genetic Diversity Analysis of Crested Wheatgrass [Agropyron cristatum (L.) Gaertn.]. International journal of molecular sciences, 19(9), 2587. Doi :10.3390/ijms19092587 Bates, D., Maechler, M., Bolker, B., & Walker, S. (2015). Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software, 67(1), 1–48. https://doi.org/doi:10.18637/jss.v067.i01 Begum, H., Spindel, J.E., Lalusin, A., Borromeo, T., Gregorio, G., Hernandez, J., Virk, P,, Collard, B., McCouch, S.R. (2008). Genome-wide association mapping for yield and other agronomic traits in an elite breeding population of tropical rice (Oryza sativa). PLoS One 10: 1–19, 2015. Berger, G. U., and Minor, H. C. (1999). An RFLP Marker Associated with Resistance to Phomopsis Seed Decay in Soybean PI 417479. Crop Sci. Vol. 39. No.3 (800- 805) oi:10.2135/cropsci1999.0011183X003900030031x Bisen, A., Khare, D., Nair, P., & Tripathi, N. (2014). SSR analysis of 38 genotypes of soybean (Glycine Max (L.) Merr.) genetic diversity in India. Physiology and Molecular Biology of Plants, 21(1), 109–115. https://doi.org/10.1007/s12298-014-0269-8 Bock, C. H., Poole, G. H., Parker, P. E. and Gottwald, T. R. (2010). Plant Disease Severity Estimated Visually, by Digital Photography and Image Analysis, and by Hyperspectral Imaging, Critical Reviews in Plant Sciences, 29:2, 59- 107, DOI: 10.1080/07352681003617285 Bommarco, R., Kleijn, D. and Potts, S.G. (2013) Ecological intensification:harnessing ecosystem services for food security. Trends in Ecology and Evolution, 28, 230– 238. Bradbury PJ, Zhang Z, Kroon DE, Casstevens TM, Ramdoss Y, Buckler ES. TASSEL: software for association mapping of complex traits in diverse samples. Bioinformatics. 2007;23:2633–2635. doi: 10.1093/bioinformatics/btm308 Brian C. Freeman and Gwyn A. Beattie. (2008). An Overview of Plant Defenses against Pathogens and Herbivores. The Plant Health Instructor. https://doi.org/10.1094/PHI-I- 2008-0226-01. Brown, E. A., Minor, H.C., and Calvert, O.(1978). A soyabean genotype resistant to phomopsis seed decay . Crop Sci.27:895-898. Brun, H., Chevre, A-M., DL Fit, B., Powers, S., Besnard, A-L., Ermel, M., Huteau, V., Marquer, B., Eber, F., Renard, M. and Andrivon, D. (2013). Quantitative resistance increases the durability of qualitative resistance to Leptosphaeria maculans in Brassica 147 University of Ghana http://ugspace.ug.edu.gh napus. New Phytologist.Vol.185, issue1: Pages 285-299. https://doi.org/10.1111/j.1469-8137.2009.03049.x Buri, M. M., Iassaka, R. N., Fujii, H., & Wakatsuki, T. (2010). Comparison of soil nutrient status of some rice growing environments in the major agro-ecological zones of Ghana. Journal of Food, Agriculture and Environment, 8(1), 384–388. Carter, S. B. (1967). Effects of Cytochalasins on Mammalian Cells Nature, volume 213, pages261–264 (1967) Chandra Sekhar Singh Bhaludra, Hari Yadla, Farhan Sachal Cyprian, Rama Rao Bethapudi, S.D. Basha and Roja Rani Anupalli, 2014. Genetic Diversity Analysis in the Genus Aloe vera (L.) Using RAPD and ISSR Markers. International Journal of Pharmacology, 10: 479-486. Chang, H.-X., Lipka, A. E., Domier, L. L., and Hartman, G. L. (2016). Characterization of disease resistance loci in the USDA soybean germplasm collection using genome-wide associations. Phytopathology 106:1139-1151. Costamilan, L., Yorinori, J., Almeida, A., Seixas, C., Binneck, E., Araújo, M., & Carbonari, J. (2008). First report of Diaporthe phaseolorum var. caulivora infecting soybean plants in Brazil. Tropical Plant Pathology, 33(5), 381–385. https://doi.org/10.1590/S1982- 56762008000500007 Council for Scientific and Industrial Research (CSIR, 2015). CSIR crop varieties released and registered in ghana. https://www.csir.org.gh/images/Doc/NewsLetter/CROP%20VARIETIES%20RELE ASED%20AND%20REGISTERED%20IN%20GHANA.pdf.Accessed in September 2017 Cregan, P.B. and Hartwig EE (1984). Characterization of Flowering Response to Photoperiod in Diverse Soybean Genotypes1. Crop Sci 24: 659. Cui , Y.‐L., Duan, C.‐X, Wang, X.‐M., Li, H.‐J. and Zhu, Z.‐D.(2009). First report of Phomopsis longicolla causing soybean stem blight in China. Plant Pathology, Vol. 58, Issue 4. https://doi.org/10.1111/j.1365-3059.2009.02057.x Cure, J. D., Acock B (1986). Crop responses to carbon dioxide doubling: a literature survey. Agric For Meteorol 38: 127–145 . Das, D., Kate, R., Onge, St., Laurentius, A. Voesenek, C. J., Pierik, R. and Sasidharan, R. (2016). Ethylene- and Shade-Induced Hypocotyl Elongation Share Transcriptome Patterns and Functional Regulators. Plant Physiology, Vol. 172, pp. 718–733DOI: https://doi.org/10.1104/pp.16.00725 Deepak, J., Singh, K and Adhikari, S. (2018). Study of Genetic Parameters in Soybean Germplasm Based on Yield and Yield Contributing Traits. Int.J.Curr.Microbiol.App.Sci. 7(01): 700-709. doi: https://doi.org/10.20546/ijcmas.2018.701.085 148 University of Ghana http://ugspace.ug.edu.gh Divilov, K. (2014). Taxonomy and Biocontrol of Diaporthe Sojae and Screening for Resistance To Phomopsis Seed Decay Caused By an Atypical Diaporthe Sojae Isolate Using Various Assays. .Thesis submitted for Master of Science in Crop Sciences in the Graduate College of the University of Illinois at Urbana-Champaign, 2014 T.Z. Dlamini, P. Tshabalala, T. Mutengwa, Soybeans production in South Africa, Oilseeds Fats Crops Lipids 21 (2014) D207 Dlamini, T.Z., Tshabalala, P., Mutengwa, T. (2014). Soybeans production in South Africa. Oilseeds Fats Crops Lipids 21 (2014) D207 Dogbe W, Etwire PM, Martey E, Etwire JC, Baba IIY, Siise A (2013). Economics of Soybean Production: Evidence from Saboba and Chereponi Districts of Northern Region of Ghana. Journal of Agric Sci 5. Desissa, D. H. (2017). Genetic Variability, Heritability and Genetic Advances of Soybean (Glycine max (L.) Merrill) Varieties Grown at Bako Tibe in Western Ethiopia. Asian Journal of Plant Science and Research, 2017, 7(5):20-26. Dong Y, Zhao L, Liu B, Wang Z, Jin Z, Sun H. 2004. The genetic diversity of cultivated soybean grown in China. Theoretical and Applied Genetics 108: 931– 936. Doyle, J., & Doyle, J. (1990). Isolation of plant DNA from fresh tissue. Focus, 12, 13–15. Domsch, K. H., Gams, W., and Anderson, T. H. (2007). Compendium of soil fungi. IHWVerlag, Eching Earl, D. A., & vonHoldt, B. M. (2012). STRUCTURE HARVESTER: A website and program for visualizing STRUCTURE output and implementing the Evanno method. Conservation Genetics Resources, 4(2), 359–361. https://doi.org/10.1007/s12686-011- 9548-7 Ellis, M. A., Ilyas, M. B., Tenne, F. D., Sinclair, J. B., and Palm, H. L. 1974. Effect of foliar applications of benomyl on internally seedborne fungi and pod and stem blight in soybean. Plant Disease Reporter 58:760-763 Elizondo, M., Rew, R.L., Jacobsen, B., Hogg, A. C. and Dyer, A.T. (2011). Distribution and prevalence of Fusarium crown rot and common root rot pathogens of wheat in Montana. Plant Disease 95:1099-1108. Elshire, R. J., Glaubitz, J. C., Sun, Q., Poland, J. A., Kawamoto, K., Buckler, E. S., & Mitchell, S. E. (2011). A Robust, Simple Genotyping-by-Sequencing (GBS) Approach for High Diversity Species. PLoS ONE, 6(5), e19379. https://doi.org/10.1371/journal.pone.0019379 Emiko Terazono. (2018). How Argentina’s drought affects the food on your plate | Financial Times. Retrieved September 21, 2018, from https://www.ft.com/content/eea8f456- 1d48-11e8-956a-43db76e69936 Evanno, G., Regnaut, S., & Goudet, J. (2005). Detecting the number of clusters of individuals using the software STRUCTURE: A simulation study. Molecular Ecology, 14(8), 2611–2620. https://doi.org/10.1111/j.1365-294X.2005.02553.x 149 University of Ghana http://ugspace.ug.edu.gh Fehr, W. (1991)."Principles of Cultivar Development: Theory and Technique” (1991). Agronomy Books. 1. https://lib.dr.iastate.edu/agron_books/1. Flor, H. H. (1971). "Current status of the gene-for-gene concept". Annu Rev Phytopathol. 9: 275–296. doi:10.1146/annurev.py.09.090171.001423. Food and Agriculture Organisation of the United Nations: FOASTAT (2018). http://www.fao.org/faostat/en/#data/QC(Retrieved in December, 2018) Forbes, G.; Pérez, W.; Andrade Piedra, J. (2014). Field assessment of resistance in potato to Phytophthora infestans. Lima (Peru). International Potato Center (CIP). 35 p. http://cipotato.org/wp-content/uploads/2014/06/006154.pdf. DOI:10.4160/9789290604402 Fu, Y.B.; Peterson, G.W. (2011). Genetic diversity analysis with 454 pyrosequencing and genomic reduction confirmed the eastern and western division in the cultivated barley gene pool. Plant Genome 4, 226–237. Garcia, M. C., Torre, M., Laborda, F. and Marina, M. L. (1997). Rapid separation of soybean globulins by reversedphase high-performance liquid chromatography. Chromatog. A 758, 75-83. Garcia, A., Calvo, E.S., Kiihl, R.A. D. and Souto, E.R. (2011). Evidence of a susceptible allele inverting the dominance of rust resistance in soybean. Crop Science 2011; 51:32–40. Gasparri, N.I., Kuemmerle, T., Meyfroidt, P., de Waroux,Y.L-P. and Kreft, H.(2015). The Emerging Soybean Production Frontier in Southern Africa: Conservation Challenges and the Role of South‐South Telecouplings. Conservation Letters: A journal of the Society for Conservation Biology. Vol. 9(1), 21–3. https://doi.org/10.1111/conl.12173 Ghose, L., Neela, F. A., Chakravort, T. C., Ali, M. R. and Alam., M. S. ( 2010). “Incidence of Photosynthetic Pigments of Infected Leaf.” Plant Pathology Journal 9 (3): 140–43. Gillen A, Mengistu A, Kremer RJ, Maxwell JE, Bellaloui N, Reddy KN, Bruns HA, Zobiole, LHS, Fisher DK, Abbas, HK, and Zablotowicz RM(2011). Soybean seed composition and quality: Interactions of environment, genotype, and management practices The Solar Corridor: New Approach in Intercropping for Efficient Crop Production View project Soil Health View project Soybean Seed Composition Glaubitz, J. C., Casstevens, T. M., Lu, F., Harriman, J., Elshire, R. J., Sun, Q., & Buckler, E. S. (2014). TASSEL-GBS: A High Capacity Genotyping by Sequencing Analysis Pipeline. PLoS ONE, 9(2), e90346. https://doi.org/10.1371/journal.pone.0090346 Goldsmith, P. (2017). The Faustian Bargain of Tropical Soybean Production. Tropical Conservation Science, 10(July), 194008291772389. https://doi.org/10.1177/1940082 917723892 Goldsmith, P.D. (2008). Executive Director’s Message. NSRL Bulletin. 15 (1): 7. Gore, M. A, Chia, J-M, Elshire, R.J, Sun, Q., Ersoz, E.S, Hurwitz, B. L, Peiffer, J.A., McMullen, M. D., Grills, G.S., Ross-Ibarra, J., Ware, D.H. and Buckler, E.S. (2009). A First-Generation Haplotype Map of Maize. Science 326: 1115–1117. 150 University of Ghana http://ugspace.ug.edu.gh Guan, R., Chang, R., Li, Y., Wang, L., Liu, Z. and Qiu, L. (2010). Genetic diversity comparison between Chinese and Japanese soybeans (Glycine max (L.) Merr.) revealed by nuclear SSRs. Genet Resour Crop Evol 57: 229-242. Guo, J., Wang Y, Song C, Zhou J, Qiu L, Huang H, Wang Y. A (2010) single origin and moderate bottleneck during domestication of soybean (Glycine max): Implications from microsatellites and nucleotide sequences. Ann Bot 106: 505–514. Hartman, G.L., West, E.D., Herman, T.K. (2011). Crops that feed the World 2. Soybean- worldwide production, use, and constraints caused by pathogens and pests, Food Secur. 3 (2011) 5–17. Index Mundi, 2017 (http://www.indexmundi.com/). Jackson, E. W., Fenn, P. and Chen, P. (2005). Inheritance of Resistance to Phomopsis Seed Decay in Soybean PI 80837 and MO/PSD-0259. Crop Sci. 45:2400–2404 (2005). doi:10.2135/cropsci2004.0525 Jackson, E. W., Feng, C., Fenn, P. and chen, P. (2009). Genetic Mapping of Resistance to Phomopsis Seed Decay in the Soybean Breeding Line MO/PSD-0259 (PI562694) and Plant Introduction 80837. Journal of Heredity 2009:100(6):777–. doi:10.1093/jhered/esp042. James, W.C. (1971). A manual of assessment keys for plant diseases. Can Dep Agric Pub, No 1458 Koch H, Hau B (1980) Ein psychologischer Aspekt beim Schätzen von Planzenkrankheiten. Z Pflanzenkr Pflanzenschutz 87: 553–587 Jannink, J. L. and Walsh, B. (200). "Association Mapping in Plant Populations", In: M. S. Kang, Ed., Quantitative Genetics, Genomics and Plant Breeding, CAB International, Oxford, 2002, pp. 59-68 Jeffers, D.L., A.F. Schmitthenner, and D.L. Reichard. (1982). Seed borne fungi, quality, and yield of soybeans treated with benomyl fungicide by various application methods. Agron. J. 74:886–888 Jiang, B., Yue Y, Gao Y, Ma L, Sun S, Wu C, et al. (2013) GmFT2aPolymorphism and Maturity Diversity in Soybeans. PLoS ONE 8(10): e77474. https://doi.org/10.1371/journal.pone.0077474 Joubert, J.C. N. and Jooste, A. A. (2013). Comparative analysis of the different regions of South African soybean industry. Pretoria: National Agricultural Marketing Council. Hall, D., Tegström, C. and Ingvarsson, P.K. (2010). Using association mapping to dissect the genetic basis of complex traits in plants. Brief Funct Genomic Proteomic 9: 157–165. Han Y, Zhao X, Liu D, Li Y, Lightfoot DA, Yang Z, Zhao L, Zhou G, Wang Z, Huang L, Zhang Z, Qiu L, Zheng H, Li W (2010). Domestication footprints anchor genomic regions of agronomic importance in soybeans. New Phytol 209: 871–884, 2016. Hanna, H. M. (2010). Combine harvest settings to reduce grain loss and improve grain quality. 151 University of Ghana http://ugspace.ug.edu.gh Proceedings of the Integrated Crop Management Conference- Iowa State University. https://lib.dr.iastate.edu/cgi/viewcontent.cgi?article=1036&context=icm Hao, D., Cheng, H., Yin, Z., Cui, S., Zhang, D., Wang. H., and Yu, D. (2012). Identification of single nucleotide polymorphisms and haplotypes associated with yield and yield components in soybean (Glycine max) landraces across multiple environments. Theor Appl Genet 124: 447–458. Hartman G, Murithi HM (1999). Field Guide to African Soybean Diseases and Pests. . Hartwig, E. E. and Lehman, S. G. (1951). Inheritance of resistance to the bacterial pustule disease in Soybean. Agronomy Journal, 1951 Vol.43 No.5 pp.226-229 pp He J, Zhao X, Laroche A, Lu Z-X, Liu H, Li Z(2014). Genotyping-by-sequencing (GBS), an ultimate marker-assisted selection (MAS) tool to accelerate plant breeding. Front Plant Sci 5: 484. Hepperly, P. R., and Sinclair, J. B. (1978). Quality losses in Phomopsis-infected soybean seed. Phytopathology, 68, 1684–1687. Holbrook, C. C., & Stalker, H. T. (2003). Peanut Breeding and Genetic Resources. (J. Janick, Ed.), Plant Breeding Reviews (Vol. 22). John Wiley & Sons, Inc. https://doi.org/10.1002/9780470650202.ch6 Hou F-F, Fwu-Fenn HOU, Thseng F-S (1992). Studies on the Screening Technique for Pregermination Flooding Tolerance in Soybean. Japanese journal of crop science 61: 447– 453. Huffaker A., Pearce G., Ryan C. A. (2006). An endogenous peptide signal in Arabidopsis activates components of the innate immune response. Proc. Natl. Acad. Sci. U.S.A. 103 10098–10103 10.1073/pnas.0603727103 Ibrahim,, S.E., Han, W.Y., Baek, I.Y., Cho and G.R.E.( 2017). Evaluating soybean germplasm for agronomic performance under irrigated cropping environment in Sudan, J. Korean Soc. Int. Agric. 29 (2017) 415–420 Ibrahim, S.E. (2011). Soybean: Crop of hope in Sudan, Symposium of Korean Society of International Agriculture, June 21, 2011, Suwon, Republic of Korea, 2011, pp. 199– 215. Ito T, Chiba T, Ozawa R, Yoshida M, Hattori M, Sakaki Y (2001). A comprehensive two hybrid analysis to explore the yeast protein interactome. Proc Natl Acad Sci U S A 98: 4569–4574. Jame, W.C. 1971. A Manual of Assessment Keys for Plant Disease. The American Phytopathological Society, St Paul, MN. Kao WY, Tsai TT, Tsai HC, Shih CN (2006) Response of three Glycine species to salt stress. Environ Exp Bot 56: 120–125 152 University of Ghana http://ugspace.ug.edu.gh Kebede, M. and Gidesa, A.(2017) . Survey and Identification of Diseases on Major Crops of Assosa and Kamashi Zones, EthiopiaGalore International Journal of Applied Sciences and Humanities (www.gkpublication.in) 21 Vol.1; Issue: 1; January-March 2017. Khojely, D. M., Ibrahim, S. E., Sapey, E., & Han, T. (2018). History, current status, and prospects of soybean production and research in sub-Saharan Africa. Crop Journal, 6(3), 226–235. https://doi.org/10.1016/j.cj.2018.03.006 Kim, M. Y., Van, K., Kang, Y. J., Kim, K. H., & Lee, S.-H. (2012). Tracing soybean domestication history: From nucleotide to genome. Breeding Science, 61(5), 445–452. https://doi.org/10.1270/jsbbs.61.445 Kim, D. H, Van, K., Kim, M. Y. and Lee, S-H. (2010). Fine mapping of a resistance gene to bacterial leaf pustule in soybean. Theor Appl Genet (2010) 120:1443–1450 DOI 10.1007/s00122-010-1266-0. Kim, K. H, Kim, M.Y., Van K, Moon, J. K., Kim, D. H. and Lee, S. H . (2008). Marker- assisted foreground and background selection of near isogenic lines for bacterial leaf pustule resistant gene in soybean. J Crop Sci Biotechnol 11:263–268 Kim, K.D., Shin, J. H., Van, K., Kim, D.H and Lee, S.H. (2009). Dynamic rearrangements determine genome organization and useful traits in soybean. Plant Physiol. 151:1066– 1076 Kmetz, K.T., C.W. Ellett, and A.F. Schmitthenner. 1979. Soybean seed decay: Sources of inoculum and nature of infection. Phytopath ology 69:798–801. Knogge, W. (1996).Fungal infection of plants. Plant Cell 8:1711–1722. doi: 10.1105/tpc.8.10.1711 Koenning, S. R. (2010). “Southern United States soybean disease loss estimate for 2009,” in Proceedings of the Southern Soybean Disease Workers 37th Annual Meeting, p. 1, Pensacola Beach, Fla, USA, 2010. Koenning, S. R. (2010). Southern United States soybean disease loss estimate for 2009. Proceedings of the Southern Soybean Disease Workers, the 37th Annual Meeting, p 1 Kolapo AL (2011). Soybean: Africa ’s Potential Cinderella Food Crop. Soyabean - Biochemistry, Chemistry and Physiology. Kunert, K.J., Vorster, B.J., Fenta, B.A., Kibido, T., Dionisio, G., Foyer, C.H. (2016). Drought Stress Responses in Soybean Roots and Nodules. Front Plant Sci 7: 1015. Kulik, M.M. and Sinclair, J.B. (1999). Phomopsis seed decay. p. 31–32. Compendium of soybean diseases 4th ed. APS Press, St. Paul, MN. Ku, Y., Yung, Y., Li, M., Wen, C., & Liu, X. (2010). Drought stress and tolerance in soybean. A Comprehensive Survey of International Soybean Research - Genetics, Physiology, Agronomy and Nitrogen Relationships, 624. https://doi.org/10.5772/45867 153 University of Ghana http://ugspace.ug.edu.gh Kuroda, Y., Kaga, A., Tomooka, N., & Vaughan, D. A. (2006). Population genetic structure of Japanese wild soybean (Glycine soja) based on microsatellite variation. Molecular Ecology, 15(4), 959–974. https://doi.org/10.1111/j.1365-294X.2006.02854.x Kawuki, R.S, Adipala, E., Lamo, J., Tukamuhabwa, P. (2003). Responding to the soybean rust epidemic in sub-Saharan Africa: A review. African Crop Science Journal 11: 301– 318,. Langewisch, T., Lenis, J., Jiang, G., Wang, D., Pantalone, V., & Bilyeu, K. (2017). The development and use of a molecular model for soybean maturity groups, 1–13. https://doi.org/10.1186/s12870-017-1040-4 Lersten, N.S., and J.B. Carlson. 2004. Vegetative morphology. In: H.R. Boerma and J.E. Specht, editors, Soybeans: Improvemment, production and uses. ASA, CSSA, and SSSA, Madison, WI. p. 15–57. Li, H., & Durbin, R. (2009). Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics, 25(14), 1754–1760. https://doi.org/10.1093/bioinformatics /btp324 Li, S.,(2015). Evaluation of Diverse Soyabean Germplasm for Resistance to Phomopsis Seed Decay, A, & Genetics, C (November), 1517–1525. https://doi.org/10.1094/PDIS-04- 14-0429-RE Li, Y. H., Li, W., Zhang, C., Yang, L., Chang, R. Z., Gaut, B. S., & Qiu, L. J. (2010). Genetic diversity in domesticated soybean (Glycine max) and its wild progenitor (Glycine soja) for simple sequence repeat and single-nucleotide polymorphism loci. New Phytologist, 188(1), 242–253. https://doi.org/10.1111/j.1469-8137.2010.03344.x Li, S. and Chen, P. (2013).Resistance to Phomopsis Seed Decay in Soybean. ISRN Agronomy Volume 2013, Article ID 738379, 8 pages http://dx.doi.org/10.1155/2013/738379 Li, S., Smith, J. and Nelson, R. (2011). Resistance to Phomopsis seed decay identified in maturity group V soybean plant introductions,” Crop Science, vol. 51, no. 6, pp. 2681– 2688, 2011 Likpa, A. E., Tian, F., Wang, Q., Peiffer, J., Li, M., Bradbury, P. J., … Zhang, Z. (2012). GAPIT: genome association and prediction integrated tool. Bioinformatics, 28(18), 2397–2399. Li, S., Hartman, G. L. and Boykin, D. L. (2010) “Aggressiveness of Phomopsis longicolla and other Phomopsis spp. on soybean,” Plant Disease, vol. 94, no. 8, pp. 1035–1040, 2010. Li, S.,Rupe, J., Chen, P., Shannon, G., Wrather, A. and Boykin, B. (2015). Evaluation of Diverse Soybean Germplasm for Resistance to Phomopsis Seed Decay. https://doi.org/10.1094/PDIS-04-14-0429-RE Liu Z., Li H., Fan X., Huang W., Yang J., Wen Z., et al. (2016). Phenotypic characterization and genetic dissection of nine agronomic traits in Tokachi nagaha and its derived cultivars in soybean (Glycine max (L.) Merr.). Plant Sci. 256 72–86. 10.1016/j.plantsci.2016.11.009 154 University of Ghana http://ugspace.ug.edu.gh Lee, G‐A,, Crawford, G.W., Liu, L., Sasaki, Y., and Chen, X. (2011). Archaeological soybean (Glycine max) in East Asia: does size matter?. PLoS ONE 6: e26720. https://doi.org/10.1371/journal.pone.0026720 Lee, C. S., Park, E. W., Kim, H. S., Kim, S. D., Hong, E. H., and Koh, M. H. (1994). Effects of benomyl application on phomopsis seed decay of early soybeans. Korean Journal of Plant Pathology 10:222-227. Lu, S., Zhao, X., Hu,Y., Liu, S., Nan, H., Li, X., Fang, C., Cao, Dong, S. , Xiny, K., Lingping, S. Tong, Z., Fengge, L., Shichen,W., Zheng, Y. and Yuan, X. (2017). Natural variation at the soybean J locus improves adaptation to the tropics and enhances yield, Nature Genetics. https://doi.org/10.1038/ng.381910.1038/ng.3819 Ma, G.Z. 1994. Review and forecast of study on frogeye leaf spot. Soybean J. 1:6–7. Malik, M. F.A., Ashraf, M., Qureshi, A. S. and Abdul Ghafoor, A. (2007). Assessment of genetic variability, correlation and path analyses for yield and its components in soybean. Pak. J. Bot., 39(2): 405-413, 2007. Mannan, M.A., Halder, E., Karim, M.A, and Ahmed, J. U.(2016). Alleviation of Adverse Effect of Drought Stress on Soybean (Glycine max. L.) by Using Poultry Litter Biochar. Bangladesh Agron. J. 2016, 19(2): 6. DOI: 10.3329/baj.v19i2.31854. Maphosa, M., Talwana, H. and Tukamuhabwa, P. (2012). Enhancing soybean rust resistance through Rpp2, Rpp3 and Rpp4 pair wise gene pyramiding. African Journal of Agricultural Research Vol. 7(30), pp. 4271-4277, 7. http://www.academicjournals.org/AJAR DOI: 10.5897/AJAR12.1123. Masudaa,T., and Goldsmith, P.D(2009). World Soybean Production: Area Harvested, Yield, and Long-Term Projections. International Food and Agribusiness Management Review. Volume 12, Issue 4. Matsuo, E., Sediyama, T., Brommonschenkel, S. H. and Damião Cruz, C. D. (2014). Crop Breeding and Applied Biotechnology 14: 209-215, 2014. ARTICLE http://dx.doi.org/10.1590/1984-70332014v14n4a33 Mayhew, W.L. , and Caviness, C.E. (1994). Seed Quality and Yield of Early-Planted, Short- Season Soybean Genotypes. Agron J 86: 16. Mbanya, W. (2011). Assessment of the Constraints in Soybean Production : A Case of Northern Region, Ghana. Journal of Developments in Sustainable Agriculture, 6, 199– 214. https://doi.org/10.11178/jdsa.6.199 McClure, B.A. and Guilfoyle, T. 1987. Characterization of a class of small auxin-inducible soybean polyadenylated RNAs. PlantMol. Biol. 9: 611-623. Mendelsohn, R., Nordhaus, W.D., Shaw, D. (1994). The Impact of Global Warming on Agriculture: A Ricardian Analysis. Am Econ Rev 84: 753–771. Mengistu A, Castlebury L, Smith R, Ray J, B. N. (2009). Seasonal progress of Phomopsis longicolla infection on soybean plant parts and its relationship to seed quality. Plant Dis, 93, 1009–1018. 155 University of Ghana http://ugspace.ug.edu.gh Mengistu, A., Arelli, P. A., Bellaloui, N., Bond, J. P., Shannon, G. J., Wrather, A. J., Rupe, J. C., Chen, P., Little, C. R., Canaday, C. H., Newman, M. A., and Pantalone, V. R. (2012). Evaluation of Soybean Genotypes for Resistance to Three Seed-borne Diseases. https://doi.org/10.1094/PHP-2012-0321-02-RS. Mian, M. A. R., Missaoui, A. M., Walker, D. R., Phillips, D. V., & Boerma, H. R. (2008). Frogeye Leaf Spot of Soybean: A Review and Proposed Race Designations for Isolates of Hara. Crop Science, 48(1), 14. doi:10.2135/cropsci2007.08.0432 Miles, M.R, Morel, W., Ray, J. D., Smith, J.R. and Frederick, R.D. (2008). Adult plant evaluation of soybean accessions for resistance to Phakopsora pachyrhizi in the field and greenhouse in Paraguay. Plant Dis. 92:96–105. https://doi.org/10.1094/PDIS-92- 1-0096 Miransari, M. (2016). Soybean production and salinity stress. Abiotic and Biotic Stresses in Soybean Production. Elsevier Inc. https://doi.org/10.1016/B978-0-12-801536- 0.00007-4 Miranda, C. (2018). Adaptation of Soybean to Tropical Environments for Small Holder Farmers. A Dissertation presented to the Faculty of the Graduate School at the University of Missouri-Columbia on May 2018 https://mospace.umsystem.edu/xmlui/bitstream/handle/10355/66176/research.pdf?se quence=1&isAllowed=y Minor, H. C., Brown, E. A. and Zimmerman, M. S.(1995). Developing soybean varieties with genetic resistance to Phomopsis spp. Journal of the American Oil Chemists' Society, vol. 72, no. 12, pp. 1431–1434, 1995. MoFA-SRID. (2017). Facts and Figures, October (2016), 12–66. MoFA-SRID. (2016). Facts and Figures, October (2016), 12–66. MoFA-SRID. (2009). Facts and Figures, October (2009), 10-15. Monteros, M. J., Missaoui, A. M., Phillips, D. V., Walker, D. R., & Boerma, H. R. (2007). Mapping and Confirmation of the “Hyuuga” Red-Brown Lesion Resistance Gene for Asian Soybean Rust. Crop Science, 47(2), 829. doi:10.2135/cropsci06.07.0462 Mullis, K. B., & Faloona, F. A. (1987). Specific Synthesis of DNA in Vitro via a Polymerase- Catalyzed Chain Reaction. Methods in Enzymology, 155(C), 335–350. https://doi.org/10.1016/0076-6879(87)55023-6 Nelson, C. G., Rosegrant, M., Koo, J., Robertson, R., Sulser,T., Zhu, T. and Msangi, S., Ringler,C., Palazzo, A., Batka, M., Magalhaes, M. and Lee, D.(2009). Climate Change: Impact on Agriculture and Costs of Adaptation. International food policy research institute. Food policy report. ISBN 10-digit: 0-89629-535-4, ISBN 13-digit: 978-0-89629-535-3. DOI: 10.2499/0896295354 Narvel, J. M., Jakkula, L. R., Phillips, D. V., Wang, T., Lee, S.-H. Boerma, H. R(2001). Molecular Mapping of RxpConditioning Reaction to Bacterial Pustule in Soybean. Journal of Heredity, Volume 92, Issue 3, May 2001, Pages 267– 270, https://doi.org/10.1093/jhered/92.3.267 156 University of Ghana http://ugspace.ug.edu.gh Ng, A., & Xavier, R. J. (2011). Leucine-rich repeat (LRR) proteins: integrators of pattern recognition and signaling in immunity. Autophagy, 7(9), 1082–1084. https://doi.org/10.4161/AUTO.7.9.16464 Nevena, M., Jelena, V. and Franić-Mihajlović, D.(1997). A comparative study of Diaporthe/Phomopsis fungi on soybean from two different regions of the world Mycopathologia, 139: 107. https://doi.org/10.1023/A:1006849020432 Mercer-Quarshie, H. and Nsowah, G. F. (1975). Soya bean in Ghana. INTSOY Series No. 6. P 200-08 D. K Whighan, ed. Soybean Production, Protection and Utilization: Proceedings of the Conference of Scientists of Africa, the Middle East, and South Asia (College of Agric., Univ. of Illinois at Urbana-Chanpaign Ooijen, G.V., Mayr, G., Albrecht, M., Cornelissen, B.J.C. and Takken, F. L. W.(2008). Transcomplementation, but not Physical Association of the CC-NB-ARC and LRR Domains of Tomato R Protein Mi-1.2 isAltered by mutations in the ARC2 Subdomain. Molecular Plant 1(3):401-10. DOI: 10.1093/mp/ssn009. Offei, S.K., Cornelius, E.W., Sakyi-Dawson, O. (2008). Crop diseases in Ghana and their management [Online]. http://ugspace.ug.edu.gh/handle/123456789/359. Perrier, X., Jacquemoud-Collet JP(2006). DARwin - Dissimilarity Analysis and Representation for Windows. Pathan, M. S., Clark, K. M., Wrather, J. A., Sciumbato, G. L., Shannon, J. G., Nguyen, H. T., and Sleper, D. A. (2009). Registration of soybean germplasm SS93-6012 and SS93- 6181 resistant to Phomopsis seed decay. Journal of Plant Registrations 3:91-93 Peterson, G., Dong, Y., Horbach, C., & Fu, Y.-B. (2014). Genotyping-By-Sequencing for Plant Genetic Diversity Analysis: A Lab Guide for SNP Genotyping. Diversity, 6(4), 665–680.doi:10.3390/d6040665 Petrović, K.L., Riccioni, M., Vidić, V., Đorđević, S., Balešević- Tubić, V., Dukić, V., & Miladinov, Z. (2016). First records of Diaporthe novem, D. foeniculina and D. rudis associated with soybean seed decay in Serbia. Plant Disease 100, 2324. Peralta, A., Wander, M. (200). Soil organic matter dynamics under soybean exposed to elevated CO2. Plant Soil. 303: Issue 1–2, pp 69–81 https://doi.org/10.1007/s11104- 007-9474-3 Pham, A. T., Harris, D. K., Buck, J., Hoskins, A., Serrano, J., Abdel-Haleem, H., Li, Z. (2015). Fine Mapping and Characterization of Candidate Genes that Control Resistance to Cercospora sojina K. Hara in Two Soybean Germplasm Accessions. PloS one, 10(5), e0126753. doi:10.1371/journal.pone.0126753 Prathuangwong, S. and Amnuaykit, K. (1987). Studies on tolerance and rate reducing bacterial pustule of soybean cultivars/lines. Kasetsart J. (Nat. Sci.) 21, 408–420. Plahar, W. A. (2006). Overview of the soybean industry in Ghana: Workshop on soybean protein for human nutrition and health-. Accra, Ghana, 28th September 2006. https://tropicalsoybean.com/sites/default/files/Soybean%20Industry%20in%20Ghana _Plahar_2006.pdf(Retrieved in September 2017). 157 University of Ghana http://ugspace.ug.edu.gh Pioli, R. N., Morandi, E. N., Martínez, M. C., Lucca, F., Tozzini, A., Bisaro, V., and Hopp, H. E. (2003). Morphologic, Molecular, and Pathogenic Characterization of Diaporthe phaseolorum Variability in the Core Soybean-Producing Area of Argentina. Phytopathology 93:136-146. Poland, J. A., Brown, P. J., Sorrells, M. E., & Jannink, J.-L. (2012). Development of High- Density Genetic Maps for Barley and Wheat Using a Novel Two-Enzyme Genotyping- by-Sequencing Approach. PLoS ONE, 7(2), e32253. https://doi.org/10.1371/journal.pone.0032253 Poland, J. and Rife, T.W.(2012). Genotyping-by-sequencing for plant breeding and genetics. Plant Genome 2012, 5, 92–102. Pritchard, J. K., Stephens, M., & Donnelly, P. (2000). Inference of population structure using multilocus genotype data. Genetics, 155(2), 945–959. https://doi.org/10.1111/j.1471- 8286.2007.01758.x Racedo, J., Salazar, S. M., Castagnaro, A. P., and Díaz Ricci, J. C. (2013). A strawberry Rahayu, M. (2007). Tanggapan Varietas Kedelai terhadap Penyakit Pustul Xanthomonas axonopodis pv. glycines dan Potensi Ekstrak Nabati untuk Pengendaliannya. Di dalam Inovasi teknologi kacang-kacangan dan umbi-umbian mendukung kemandirian pangan dan kecukupan energi. Prosiding Seminar Nasional Balai Penelitian Tanaman Kacangkacangan dan Umbi-umbian Malang,19 November2007 Raja, H. A., Miller, A. N., Pearce, C. J., & Oberlies, N. H. (2017). Fungal Identification Using Molecular disease caused by Acremonium strictum. Eur J Plant Pathol 137:649-654. Ray, J.D, Morel W., Smith, J.R., Frederick, R. D., Milles, M.R. (2009). Genetics and mapping of adult plant rust resistance in soybean PI 587886 and PI 587880A. Theor Appl Genet 119:271–280. DOI:10.1007/s00122-009-1036-z R Core Team. (2018, July). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Reinhardt, D., and Kuhlemeier, C. (2002).Plant Architecture. MBO reports vol. 3: no. 9, pp 846–851. DOI: 10.1093/embo-reports/kvf177. Ren, H., & Gray, W. M. (2015). SAUR Proteins as Effectors of Hormonal and Environmental Signals in Plant Growth. Molecular plant, 8(8), 1153–1164. doi:10.1016/j.molp.2015.05.003 Rimbaud, L., Papaïx, J., Rey, J-F., Barrett, L.G., Thrall, P.H. (2018) Assessing the durability and efficiency of landscape-based strategies to deploy plant resistance to pathogens. PLoS Comput Biol 14(4): e1006067. https://doi.org/10.1371/journal.pcbi.1006067 Riccioni, L., Conca, G., Pucci, N., (2003). Identification by PCR-RFLP of Rivera-Varas, V. V., Freeman, T. A., Gudmestad, N. C., and Secor, G. A. (2007). Phomopsis/Diaporthe species on Italian soybean. Proceedings of the 8th International Congress of Plant Pathology, 2003. Christchurch, New Zealand: ISPP, 322. 158 University of Ghana http://ugspace.ug.edu.gh Mycoparasitism of Helminthosporium solani by Acremonium strictum. Phytopathology 97:13311337. Roberts EH, Qi A, Ellis RH, Summerfield RJ, Lawn RJ, Shanmugasundaram S. Use of field observations to characterise genotypic flowering responses to photoperiod and temperature: A soyabean exemplar. Theor Appl Genet 93: 519–533. Ross, J. P. (1986). Registration of eight soybean germplasm lines resistant to seed infection by Phomopsis spp. Crop Sci, 1986 vol. 26, pg. 210 Rossman, A.Y., Farr, D. F. and Castlebury, L. A. (2007). A review of the phylogeny and biology of the Diaporthales. Mycoscience,Volume 48, Issue 3, June 2007, Pages 135- 144. https://doi.org/10.1007/S10267-007-0347-7 Roy, K. W., Keith, B. C., and Andrews, C. H. (1994). Resistance of hard seeded soybean lines to seed infection by Phomopsis, other fungi and soybean mosaic virus. Can. J. Plant Pathol. 16:122-128. Rupe, J. C. (1990). Effects of temperature on the rate of infection of soybean seedlings by Phomopsis longicolla,” Canadian Journal of Plant Pathology, vol. 12, no. 1, pp. 43– 47. https://doi.org/10.1080/07060669009501041 Rupe, J. C. and Ferriss, R. S. (1986). Effects of pod moisture on soybean seed infection by Phomopsis sp., Phytopathology, vol. 76, pp. 273–277 Santos, J. M., Vrandečić, K., Cosić, J., Duvnjak, T., & Phillips, A. J. (2011). Resolving the Diaporthe species occurring on soybean in Croatia. Persoonia, 27, 9–19. doi:10.3767/003158511X603719 Sedivy, E. J, Wu, F. and Hanzawa, Y.(2017). Soybean domestication: the origin, genetic architecture and molecular bases. New Phytologist, Volume214, Issue2. https://doi.org/10.1111/nph.14418 Seedco Farmers' guide (2018). http://www.seedcogroup.com/sites/default/files/Agronomy%20Manual.pdf.page 34 Shannon, D.A. and Kalala, M.M.(1994). Adoption of soybean in Sub-Saharan Africa: a comparative analysis of production and utilization in Zaire and Nigeria, Agric. Syst. 46 (1994) 369–384. Shaun, C. (2010). Soybean Physiology: How Well Do You Know Soybeans?. Purdue University Soybean Extension Specialist . ©2010-11, Purdue University - (https://www.agry.purdue.edu/ext/soybean/arrivals/10soydevt.pdf. Accessed on September, 2018). Shukla A K. 1994. Pilot estimation studies of soybean (Glycine max), yield losses by various levels of bacterial pustule (Xanthomonas campestris pv. glycines) infection.International Journal of Pest Management 40 (3): 249-251. Shurtleff, W and Aoyagi, A. (2009). History of Soybeans and Soyfoods in Africa (1857– 2009): Extensively Annotated Bibliography and Source Book, Soyinfo Center, Lafayette, California, USA, 2009 159 University of Ghana http://ugspace.ug.edu.gh Sinclair, J. B. (1982). Seed testing: Detection of pathogens. INTSOY Ser. University of Illinois, Urbana- Champaign, No. 22, 112–115. Sinclair, J. B. and Hartman, G. L, (1999). Diaporthe-Phomposis Complex. In Compendium of soybean diseases, 4th edition (p. P.13). APS Press, The American Phytopathological Society, St. Paul MN, USA. Sinclair, T. R., Marrou, H., Soltani, A., Vadez, V., and Chandolu, K. C. (2014). Soybean production potential in Africa. Global Food Security, 3(1), 31– 40.doi:10.1016/j.gfs.2013.12.001 Singh, R. B. and Jain, J. P.(1988). Chemical control of bacterial pustule in soybean. Journal of Turkish Phytopathology 1988 Vol.17 No.1 pp.31-36 ref.12. Smith, S., Fenn, P., Chen, P, & Jackson, E. (2008). Inheritance of resistance to Phomopsis seed decay in PI360841 soybean. J. Heredity 99: 588-592. Silva, A. F., Sediyama, T., Silva, F. C. S., Bezerra, A. R. G., Ferreira, L.V. (2015). Correlation and path analysis of soybean yield components. International Journal of Plant, Animal and Environmental Sciences Volume-5, Issue-1, Page: 177-179 Song, Q., Hyten, D. L., Jia, G., Quigley, C. V., Fickus, E. W., Nelson, R. L., & Cregan, P. B. (2013). Development and Evaluation of SoySNP50K, a High-Density Genotyping Array for Soybean. PLoS ONE, 8(1). https://doi.org/10.1371/journal.pone.0054985 Spartz, A.K, Lee, S.H, Wenger, J. P., Gonzalez, N., Itoh, H., Inzé, D., Peer, W.A, Murphy, A.S., Overvoorde, P. J. and Gray, W.M(2012). The SAUR19 subfamily of Small Auxin UP RNA genes promote cell expansion. Plant J. 2012; 70:978–990. doi: 10.1111/j.1365-313X.2012.04946.x. Stamm, P. and Kumar, P.P. (2013). Auxin and gibberellin responsive Arabidopsis SMALL AUXIN UP RNA36 regulates hypocotyl elongation in the light. Plant Cell Rep. 2013;32:759–769 Statista (2018).Area planted with soybean in Brazil from crop year 2010/11 to 2017/2018 (in million hectares). //www.statista.com/statistics/740030/area-planted-soybean-brazil/ (accessed in December, 2018) Subrahmanyam, P., McDonald, D., Gibbons, R. W., Nigam, S. N., & Nevill, D. J. (1982). Resistance to Rust and Late Leafspot Diseases in Some Genotypes ofArachis hypogaea1. Peanut Science, 9(1), 6–10.doi:10.3146/i0095-3679-9-1-2 Summerfield, R.J., Wilcox, J.R. (1978). Effects of photoperiod and air temperature on growth and yield of economic legumes. In: Summerfield, R.J., Bunting, A.H. (Eds.), Advances in Legume Sciences. Royal Botanic Gardens, Kew, England, pp. 17–36. Sun, S., Kim, M. Y., Chaisan, T., Lee, Y.-W., Van, K., & Lee, S.-H. (2013). Phomopsis(Diaporthe) Species as the Cause of Soybean Seed Decay in Korea. Journal of Phytopathology, 161(2), 131–134.doi:10.1111/jph.12034 Sonah, H., Bastien, M., Iquira, E., Tardivel, A., Le´gare´, G. (2013) An Improved Genotyping by Sequencing (GBS) Approach Offering Increased Versatility and Efficiency of SNP 160 University of Ghana http://ugspace.ug.edu.gh Discovery and Genotyping. PLoS ONE 8(1): e54603. doi:10.1371/journal.pone. 0054603 TeKrony, D.M., Egli, D.B., Balles, J., Tomes, L. and Stucke, R.E. (984). Effect of date of harvest maturity on soybean seed quality and Phomopsis sp. seed infection. Crop Sci. 24:189–193 Thseng, F.S., Tsai, S. J., Abe, J., Wu, S.T. (1999). Glycine formosana Hosokawa in Taiwan: pod morphology, allozyme, and DNA polymorphism. Bot Bull Acad Sin 40:251–257 Untergasser, A., Cutcutache, I., Koressaar, T., Ye, J., Faircloth, B. C., Remm, M., & Rozen, S. G. (2012). Primer3--new capabilities and interfaces. Nucleic acids research, 40(15), e115. doi:10.1093/nar/gks596 United States of America/ Foreign Agricutlure Services (USDA/FAS/ 2018). World Agricultural production, page 23. //apps.fas.usda.gov/psdonline/circulars/production.pdf(accessed in December, 2018) VanRaden, P. M. (2008). Efficient methods to compute genomic predictions. J. Dairy Sci., 91, 4414–4423. van Rensburg, J. C., Lamprecht, S. C., Groenewald, J. Z., Castlebury, L. A., & Crous, P. W. (2006). Characterisation of Phomopsis spp. associated with die-back of rooibos (Aspalathus linearis) in South Africa. Studies in mycology, 55, 65–74. Van Ee JH(2009). Soy constituents: Modes of action in low-density lipoprotein management. Nutr Rev 67: 222–234. Walters, H. J. and Caviness, C. E. (1973). Breeding for improved soybean seed quality. Arkansas Farm Research, vol. 22, article 5, 1973. Wang, Q., Tian, F., Pan, Y., Buckler, E. S., & Zhang, Z. (2014). A SUPER powerful method for genome wide association study. PloS one, 9(9), e107684. doi:10.1371/journal.pone.0107684 Wei, W., Mesquita, A., Figueiró, A. A., Wu, X., Manjunatha, S., Wickland, D. P., … Clough, S. J. (2017). Genome-wide association mapping of resistance to a Brazilian isolate of Sclerotinia sclerotiorum in soybean genotypes mostly from Brazil. BMC genomics, 18(1), 849. doi:10.1186/s12864-017-4160-1 Wrather, J. A. and Koenning, S. R. (2009). “Effects of diseases on soybean yields in the United States 1996 to 2007,” Plant Health Progress, 2009. Wrather, A., Sleper, D. A., Stevens, W. E., Shannon, J. G., and Wilson, R. F.(2003) “Planting date and cultivar effects on soybean yield, seed quality, and Phomopsis sp. seed infection,” Plant Disease, vol. 87, no. 5, pp. 529–532, 2003. Wrather, J.A., Anderson, T.R., Arsyad, D.M., Tan,Y., Ploper, L.D., Porta-Puglia, A., Ram, H.H.and Yorinori, J.T. (2001. Soybean disease lossestimates for the top ten soyabean-producingcountries in 1998.Canadian Journal of Plant Pathology 23:115- 121 161 University of Ghana http://ugspace.ug.edu.gh Woomer, P.L., Baijukya, F., and Turner, A. (2012). Progress Towards Achieving the Vision of Success of N2Africa, http://www.n2africa.org/sites/n2africa.org/files/N2Africa_Progress%20towards%20a chieving%20the%20vision%20of%20success%20in%20N2Africa_0.pdf, 2012. Wu, W.S. and Lee, M. C. (1985). Occurrence, pathogenicity and control of Phomopsis sojae on soyabean. MeM. Coll. Agric. Taiwan University, 1985, 24, 16-24. Xu, D., Abe, J., Gai, J. and Shimamoto, Y. (200). Diversity of chloroplast DNA SSRs in wild and cultivated soybeans: evidence for multiple origins of cultivated soybean. Theor Appl Genet (2002) 105: 645. https://doi.org/10.1007/s00122-002-0972-7. Yamaguchi Y., Pearce G., Ryan C. A. (2006). The cell surface leucine-rich repeat receptor for AtPep1, an endogenous peptide elicitor in Arabidopsis, is functional in transgenic tobacco cells.Proc. Natl. Acad. Sci. U.S.A. 103 10104–10109 10.1073/pnas.0603729103 Yamanaka, N., Lemos, N. G., Uno, M., Akamatsu, H., Yamaoka, Y., Abdelnoor, R. V, … Suenaga, K. (2013). Resistance to Asian soybean rust in soybean lines with the pyramided three Rpp genes. Crop Breeding and Applied Biotechnology, 13(1), 75–82. https://doi.org/10.1590/S1984-70332013000100009 Yamanaka, N., Morishita, M.,Mori, T., Lemos, N. G., Hossain, Md. M., Akamatsu, H., Kato, M., Yamaoka, Y. (2015) Trop. plant pathol. (2015) 40: 283. https://doi.org/10.1007/s40858-015-0038-4. Yang, H. A., Li, C. D., Lam, H. M., Clements, J., Yan, G. J., and Zhao, S. C. (2015). Sequencing consolidates molecular markers with plant breeding practice. Theor. Appl. Genet. 128:779-795 Yang, K.W., Lee, Y.H., Ko, J.M., Jeon, M.G. (2011). Development of Molecular Markers Conferring Bacterial Leaf Pustule Resistance Gene, rxp, using Resistant and Susceptible Cultivars in Soybean. AGRIS since: 2013, Volume : 43,Issue : 4(282- 287). http://lib.rda.go.kr/newlib/adlib_en/index.html Yan, W. (2014). Crop Variety Trials: Data Management and Analysis. John Wiley & Sons, Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK. Yu J, Buckler ES(2006). Genetic association mapping and genome organization of maize. Curr Opin Biotechnol 17: 155–160. Zhang, X., Zhang, J., He, X., Wang, Y., Ma, X., & Yin, D. (2017). Genome-Wide Association Study of Major Agronomic Traits Related to Domestication in Peanut. Frontiers in Plant Science, 8. https://doi.org/10.3389/fpls.2017.01611 Zhang, Z., Ersoz, E., Lai, C.-Q., Todhunter, R. J., Tiwari, H. K., Gore, M. A., Ordovas, J. M. (2010). Mixed linear model approach adapted for genome-wide association studies. Nature Genetics, 42(4), 355. Zhou, Z., Jiang, Y., Wang, Z., Gou, Z., Lyu, J., Li, W., Yu, Y., Shu, L., Zhao, Y., Ma, .Y. (2015). Resequencing 302 wild and cultivated accessions identifies genes related to domestication and improvement in soybean. Nat Biotechnol 33: 408–414. 162 University of Ghana http://ugspace.ug.edu.gh Zinsou, V. A., Afouda, L. A. C., Zoumarou-Wallis, N., Pate-Bata, T., Dossou, L., Götz, M., & Winter, S. (2015). Occurrence and characterisation of xanthomonas axonopodis pv. glycines, causing bacterial pustules on soybean in Guinea savanna of Benin. African Crop Science Journal (Vol. 23). Retrieved from http://www.bioline.org.br/pdf?cs 15017. Zimmerman, M.S., and Minor, H.C. (1993). Inheritance of Phomopsis Seed Decay Resistance in Soybean PI 417479. In AGRIS since : 1994. Volume : 33 (1), 96-100. 163 University of Ghana http://ugspace.ug.edu.gh APPENDICES Appendix.1: severity scores of 250 lines in reaction to bacterial leaf pustule infection Line DS Line DS Line DS Line DS BRS 8660A 1 BRS 313 1.5 PI567069A 2 TGX 19110-6F 3 BRS-7980B 1 BRS TIANA 1.5 PI605824A-2 2 TGX 1989-20FA 3 ISS-8203 1 PI200492 1.5 SAL-II 2 TGX 1990-80F 3 P587886 1 PI379621 1.5 SARISOY 12 2 PI398477 3.5 Paragoina 1 PI416873B 1.5 SARISOY 2 2 PI605823 3.5 PI274954B 1 PI59025A 1.5 TGX 1799-8FA 2 TGX 1844-22E 3.5 PI398512 1 SARISOY 14 1.5 TGX 1990-52F 2 TGX 1988-3F 3.5 PI398874 1 SARISOY 19 1.5 TGX 1990-95F 2 TGX 1989-19FA 3.5 PI416778A 1 SARISOY 27 1.5 TGX 1993-4FN 2 TGX 1989-42F 3.5 PI416810A 1 SARISOY 4 1.5 TGX 2008-4F 2 BRS 361 4 PI416939 1 TGX 1799-8F 1.5 PI588053B 2.3 HUANGZADO 4 PI417126A 1 TGX 1903-7F 1.5 PI200488 2.5 PI416810 4 PI506677 1 TGX 1989-20F 1.5 PI416810-1 2.5 PI459025A 4 PI506939 1 TGX 1990-57F 1.5 PI417503 2.5 PI468919A 4 Sal-1 1 TGX 1990-78F 1.5 PI423957 2.5 PI507004 4 SARISOY 13A 1 TGX 2006-37F 1.5 PI606440A 2.5 PI507009 4 SARISOY 15 1 TGX 2011-3F 1.5 TGX 1989-11F 2.5 PI561356 4 SARISOY 20B 1 TGX1845-10E 1.5 TGX 1989-45F 2.5 PI603170 4 SARISOY 2A 1 1.5 WAMINI 2.5 SARISOY 23 4 TGX-198948FNB SARISOY 5 1 ZAMBOANI 1.5 ISS-8203A 3 SARISOY 24 4 TGX 1989-20FC 1 BRS 90-8360 2 JENGUMA 3 TGX 1834-5E 4 TGX 1990-80FA 1 BRS PEROLA 2 PI468919 3 TGX 1844-19F 4 TGX-1835-10E 1 BRS-360 2 PI594538A 3 TGX 2007-11F 4 TGX-1989-40FB 1 BRS-7980 2 SARISOY 1 3 TGX-1989-40F 4 TGX-1989-48FN 1 FT christaline 2 SARISOY 10 3 TGX-198968FNA 4 164 University of Ghana http://ugspace.ug.edu.gh TGX-1989-48FNA 1 PI224270 2 SARISOY 26 3 TGX-1990-46F 4 TGX1989-68FN 1 PI379621A 2 TGX 1445-3E 3 TGX-1990-55F 4 TGX-1990-21F 1 PI3984 77A 2 TGX 1485-1D 3 BRS- 7525 4.5 Afayak 1.5 PI567020A 2 TGX 1740-2F 3 BRS 8660 4.5 PI506491 4.5 SARISOY 7 5.5 PI567099A 6.5 PI594767A 7.5 SARISOY 17 4.5 SARISOY 9 5.5 PI567104B 6.5 PI594767B-1 7.5 TGX 1805-8F 4.5 TGX 1988-5E 5.5 PI567190 6.5 PI605865B 7.5 TGX 1989-19F 4.5 TGX 1989-41F 5.5 SARISOY 28 6.5 PI605891B-1 7.5 TGX 1989-20FB 4.5 TGX 2010-11F 5.5 Songdaa 6.5 PI60591B 7.5 TGX 1989-75N 4.5 SARISOY 8 5.75 TGX 1990110FN 6.5 PI81042A 7.5 TGX 1990-93F 4.5 CROTON 6 TGX-1989-68FN 6.5 SARISOY 20 7.5 Appendix 1 Cont’d: Severity scores of 250 lines in reaction to bacterial leaf pustule Line DS Line DS Line DS Line DS TGX 2004-10F 4.5 PI 534545 6 PI594760B 7 PI594767B 8 TGX1987-10F 4.5 PI203398 6 PI628850 7 PI605824A 8 BRS-326 5 PI224271 6 PI628932 7 PI605829 8 LG 04 -6000 5 PI398594 6 PI635999 7 PI605854B 8 PI230970 5 PI399096 6 SARISOY 16 7 PI606397B 8 PI274954 5 PI462312 6 SARISOY 18 7 TGX 184419FB 8 PI398387 5 PI506491A 6 SARISOY 30 7 PI398825 8.5 PI398836C 5 PI5670466 6 TGX 1990 -114 FN 7 PI417120A 8.5 PI41219B 5 PI567076 6 TGX 2004-13FB 7 PI506938 8.5 PI423963 5 PI567104 6 OZARK 7.5 PI567056A 8.5 PI506491C 5 PI567123A 6 PI398836 7.5 PI567188 8.5 PI603176A 5 PI567191A 6 PI417085 7.5 PI567516C 8.5 PI605824A-1 5 PI584796 6 PI417126 7.5 PI587886 8.5 PI605865B-2 5 PI594538A-1 6 PI423960A 7.5 PI605865B-1 8.5 PI657753 5 Quarshie 6 PI476905A 7.5 PI605891A 8.5 SARISOY 21 5 SARISOY 11 6 PI567053 7.5 PI605891B 8.5 165 University of Ghana http://ugspace.ug.edu.gh TGX 1987-62F 5 SARISOY 13 6 PI567129 7.5 PI81042 8.5 TGX 2004-13FA 5 SARISOY 22 6 PI567191 7.5 PI427120 8.5 TGX-1989-40FA 5 TGX 1988-5EB 6 TGX 2006-3F 7.5 TGX1989-48FN 5 TGX 1990-93FB 6 PI274454 8 TGX 2004-13F 5.3 TGX-1990-40F 6 PI398836A 8 PI398836B 5.5 159925 6.5 PI416806 8 PI417119 5.5 TGX-1990-49F 6.75 PI416810B-1 8 PI518295 5.5 BRS-7980A 7 PI417013 8 SARISOY 25 5.5 PI416778 7 PI417134 8 SARISOY 29 5.5 PI423972 7 PI423958 8 PI200526 6.5 PI567058D 7 PI561381 8 PI340898A 6.5 PI567090 7 PI567034 8 PI567054C 6.5 PI578457A 7 PI567046A 8 PI567068A 6.5 PI588053A 6.5 PI587905 8 Appendix. 2: Fixed effects estimates of AUDPC for soyabean lines relative to the check Std. Error Line Estimate df t value Pr(>|t|) SARISOY 5 -28.8 31.7 211 -0.91 0.364 SARISOY 20B -28.6 31.7 211 -0.90 0.368 BRS 8660A -28.2 31.7 211 -0.89 0.374 BRS-7980B -28.2 31.7 211 -0.89 0.374 PI274954B -28.2 31.7 211 -0.89 0.374 TGX-198948FNA -28.2 31.7 211 -0.89 0.374 TGX1989-68FN -27.0 25.8 211 -1.05 0.296 TGX-1990-21F -26.9 25.8 211 -1.04 0.298 TGX-1989-48FN -26.9 25.8 211 -1.04 0.299 PI506677 -26.8 25.8 211 -1.04 0.301 PARANGOINA -26.7 25.8 211 -1.03 0.303 PI416810A -26.5 25.8 211 -1.03 0.306 PI416939 -26.5 25.8 211 -1.03 0.306 ISS-8203 -26.5 25.8 211 -1.03 0.306 TGX-1835-10E -26.5 25.8 211 -1.02 0.307 P587886 -26.4 25.8 211 -1.02 0.308 166 University of Ghana http://ugspace.ug.edu.gh PI398512 -26.4 25.8 211 -1.02 0.309 PI506939 -26.1 25.8 211 -1.01 0.313 PI398874 -25.9 25.8 211 -1.00 0.318 SARISOY 2A -25.4 31.7 211 -0.80 0.423 BRS 313 -25.1 25.8 211 -0.97 0.333 SARISOY 13A -25.0 31.7 211 -0.79 0.430 SARISOY 15 -24.9 31.7 211 -0.79 0.432 TGX 1903-7F -24.9 25.8 211 -0.96 0.337 PI59025A -24.8 25.8 211 -0.96 0.337 TGX 1989-20FC -24.8 31.7 211 -0.78 0.434 Afayak -24.8 25.8 211 -0.96 0.338 PI379621 -24.8 25.8 211 -0.96 0.338 PI417126A -24.7 31.7 211 -0.78 0.435 Sal-1 -24.7 31.7 211 -0.78 0.435 SARISOY 14 -24.6 25.8 211 -0.96 0.341 ZAMBOANI -24.6 25.8 211 -0.95 0.341 TGX1845-10E -24.6 25.8 200 -0.95 0.342 BRS TIANA -24.6 25.8 200 -0.95 0.342 Appendix 2 Cont’d. Std. Error Line Estimate df t value Pr(>|t|) PI416778A -24.6 31.7 211 -0.78 0.439 TGX-1989-40FB -24.6 31.7 211 -0.78 0.439 TGX 1990-80FA -24.6 31.7 211 -0.78 0.439 TGX 1799-8F -24.5 25.8 211 -0.95 0.343 PI200492 -24.5 25.8 211 -0.95 0.345 BRS-360 -22.8 25.8 211 -0.88 0.378 PI416873B -21.8 25.8 211 -0.85 0.399 TGX-198948FNB -21.5 25.8 211 -0.83 0.407 TGX 2011-3F -21.4 25.8 211 -0.83 0.408 PI379621A -21.2 31.7 211 -0.67 0.503 SARISOY 4 -21.2 25.8 211 -0.82 0.413 TGX 2006-37F -21.1 25.8 211 -0.82 0.415 SARISOY 19 -21.0 25.8 211 -0.81 0.418 PI567020A -20.9 31.7 211 -0.66 0.510 BRS-7980 -19.7 25.8 211 -0.76 0.447 167 University of Ghana http://ugspace.ug.edu.gh TGX 2008-4F -19.4 25.8 211 -0.75 0.454 PI567069A -19.2 25.8 211 -0.75 0.457 SAL-I -17.8 31.7 211 -0.56 0.574 TGX 1990-57F -15.9 25.8 211 -0.61 0.540 TGX 1990-78F -15.9 25.8 211 -0.61 0.540 PI605824A-2 -14.3 31.7 211 -0.45 0.651 SARISOY 2 -14.3 25.8 211 -0.55 0.580 PI3984 77A -14.3 25.8 211 -0.55 0.581 TGX 1990-52F -14.2 25.8 211 -0.55 0.584 SARISOY 12 -14.0 25.8 211 -0.54 0.587 TGX 1993-4FN -13.9 25.8 211 -0.54 0.590 TGX 1799-8FA -13.8 31.7 211 -0.44 0.664 PI416810-1 -12.6 25.8 211 -0.49 0.626 TGX 1989-11F -12.4 25.8 211 -0.48 0.632 FT christa -11.1 25.8 211 -0.43 0.669 PI588053B -9.9 23.6 208 -0.42 0.674 TGX 1989-20F -9.2 25.8 211 -0.36 0.723 PI417503 -8.9 25.8 211 -0.35 0.730 PI200488 -8.8 25.8 211 -0.34 0.734 TGX 1990-95F -7.2 31.7 211 -0.23 0.820 Appendix 2 Cont’d. Std. Error Line Estimate df t value Pr(>|t|) SARISOY 27 -7.1 25.8 211 -0.28 0.784 BRS 90-8360 -6.0 25.8 211 -0.23 0.816 PI224270 -5.4 25.8 211 -0.21 0.836 BRS PEROLA -5.3 25.8 211 -0.20 0.839 TGX 19110-6F -3.9 31.7 211 -0.12 0.903 JENGUMA -3.7 25.8 211 -0.14 0.886 TGX 1989-20FA -2.0 25.8 211 -0.08 0.938 ISS-8203A -1.9 25.8 211 -0.07 0.942 TGX 1989-45F -0.3 25.8 200 -0.01 0.992 PI423957 0.0 25.8 211 0.00 1.000 PI468919 1.1 25.8 211 0.04 0.966 SARISOY 10 1.6 25.8 211 0.06 0.951 TGX 1805-8F 3.3 31.7 211 0.10 0.918 TGX 1990-80F 4.7 25.8 211 0.18 0.856 TGX 1988-3F 5.1 25.8 200 0.20 0.845 TGX 1485-1D 5.4 25.8 211 0.21 0.834 168 University of Ghana http://ugspace.ug.edu.gh PI468919A 5.5 25.8 211 0.21 0.833 TGX 1740-2F 6.8 25.8 211 0.26 0.792 TGX 1445-3E 6.8 25.8 211 0.26 0.792 TGX 1844-22E 8.6 25.8 211 0.33 0.738 PI398477 8.9 25.8 211 0.34 0.732 TGX 1989-42F 12.1 25.8 211 0.47 0.641 WAMINI 12.2 25.8 211 0.47 0.639 PI561356 13.3 31.7 211 0.42 0.675 SARISOY 24 13.4 31.7 211 0.42 0.672 TGX-1989-40F 13.5 25.8 211 0.52 0.601 SARISOY 1 13.6 25.8 211 0.53 0.599 TGX-1990-55F 13.8 25.8 211 0.53 0.594 TGX 1989-19FA 15.4 25.8 211 0.60 0.552 BRS 361 15.5 25.8 211 0.60 0.549 SARISOY 26 16.9 25.8 211 0.65 0.514 GIN HUANGZADOU 16.9 25.8 211 0.66 0.513 TGX 1834-5E 20.7 25.8 211 0.80 0.425 TGX- 198968FNA 21.1 31.6 191 0.67 0.505 TGX-1990-46F 21.2 25.8 211 0.82 0.413 PI605823 24.1 25.8 211 0.93 0.351 Appendix .2: Cont’d. Std. Error Line Estimate df t value Pr(>|t|) Sig. PI507004 24.8 25.8 200 0.96 0.337 TGX-1989-40FA 27.4 31.7 211 0.87 0.387 PI459025A 27.8 25.8 211 1.07 0.284 TGX 1844-19F 27.9 25.8 211 1.08 0.281 BRS 8660 29.5 25.8 211 1.14 0.255 TGX1987-10F 30.9 25.8 211 1.20 0.233 SARISOY 23 31.0 25.8 211 1.20 0.231 PI398836C 31.2 31.7 211 0.99 0.326 TGX 1989-19F 31.7 25.8 211 1.23 0.222 BRS- 7525 31.7 25.8 211 1.23 0.222 TGX 2007-11F 32.7 25.8 211 1.27 0.207 PI416810 32.9 25.8 211 1.27 0.204 TGX 1990-93F 36.4 25.8 211 1.41 0.160 PI603170 36.5 22.4 210 1.63 0.104 169 University of Ghana http://ugspace.ug.edu.gh TGX 1987-62F 40.4 25.8 200 1.57 0.119 SARISOY 22 41.7 25.8 211 1.61 0.108 TGX 1990-93FB 41.7 31.7 211 1.32 0.189 PI567104B 41.7 25.8 211 1.62 0.108 PI594538A 43.1 25.8 211 1.67 0.097 BRS-326 43.8 25.8 200 1.70 0.091 PI567191A 47.0 25.8 211 1.82 0.071 SARISOY 9 47.2 25.8 211 1.83 0.069 PI584796 48.7 31.7 211 1.54 0.126 SARISOY 17 48.7 25.8 200 1.89 0.060 PI567123A 48.8 25.8 211 1.89 0.060 TGX 1989-41F 49.0 25.8 200 1.90 0.059 PI657753 50.6 25.8 211 1.96 0.052 PI398836B 50.7 25.8 211 1.96 0.051 PI603176A 52.0 25.8 211 2.01 0.045 * PI507009 52.2 31.7 211 1.65 0.101 PI567090 52.2 25.8 211 2.02 0.045 * TGX 1989-20FB 52.2 25.8 211 2.02 0.044 * PI605865B-2 52.8 31.7 211 1.67 0.097 . TGX 2004-13F 55.1 25.8 211 2.13 0.034 * SARISOY 21 55.6 25.8 211 2.15 0.032 * TGX 1988-5EB 55.8 25.8 211 2.16 0.032 * PI567104 57.6 25.8 211 2.23 0.027 * Note: *, **, *** represents significance at 5, 1 and 0.1 %, respectively. Appendix.2: Cont’d Std. Error Line Estimate df t value Pr(>|t|) Sig. TGX 2004-10F 57.8 25.8 211 2.24 0.026 * PI423963 58.9 25.8 211 2.28 0.024 * PI398387 59.1 25.8 211 2.29 0.023 * TGX1989-48FN 59.1 25.8 211 2.29 0.023 * SARISOY 7 59.3 25.8 211 2.30 0.023 * PI203398 59.4 25.8 211 2.30 0.022 * SARISOY 29 59.5 25.8 211 2.30 0.022 * PI506491C 59.6 31.7 211 1.88 0.061 . PI417119 60.4 25.8 211 2.34 0.020 * PI399096 61.1 25.8 211 2.37 0.019 * (Intercept) 61.5 18.4 206 3.35 0.001 *** PI594538A-1 62.8 25.8 211 2.43 0.016 * PI41219B 63.1 31.7 211 1.99 0.048 * 170 University of Ghana http://ugspace.ug.edu.gh PI506491 64.5 25.8 211 2.50 0.013 * PI230970 66.4 31.7 211 2.10 0.037 * SARISOY 25 67.8 25.8 211 2.62 0.009 ** PI274954 68.1 25.8 211 2.63 0.009 ** PI506491A 68.2 25.8 211 2.64 0.009 ** PI628850 69.4 25.8 211 2.69 0.008 ** TGX-1990-40F 69.6 25.8 211 2.69 0.008 ** PI567076 71.5 25.8 200 2.77 0.006 ** PI567190 73.2 25.8 211 2.84 0.005 ** PI224271 73.2 25.8 211 2.84 0.005 ** PI605824A-1 73.3 31.7 211 2.32 0.022 * CROTON 74.3 25.8 211 2.88 0.004 ** LG 04 -6000 74.8 25.8 211 2.90 0.004 ** TGX 1989-75N 76.4 25.8 211 2.96 0.003 ** TGX 2004-13FA 76.5 31.7 211 2.42 0.016 * SARISOY 20 76.7 25.8 211 2.97 0.003 ** PI567058D 77.1 31.7 211 2.43 0.016 * Quarshie 78.5 25.8 211 3.04 0.003 ** PI567099A 80.2 25.8 211 3.11 0.002 ** PI567054C 80.2 25.8 211 3.11 0.002 ** PI340898A 80.3 25.8 200 3.11 0.002 ** PI578457A 80.4 31.7 211 2.54 0.012 * Note: *, **, *** represents significance at 5, 1 and 0.1 %, respectively Appendix. 2 :Cont’d. Std. Error Line Estimate df t value Pr(>|t|) Sig TGX 1988-5E 80.5 25.8 211 3.12 0.002 ** PI518295 80.5 25.8 211 3.12 0.002 ** SARISOY 8 81.3 25.8 211 3.15 0.002 ** PI416778 81.8 25.8 211 3.17 0.002 ** PI628932 82.0 25.8 211 3.17 0.002 ** TGX 1990 -114 FN 82.5 25.8 211 3.19 0.002 ** PI398836A 83.8 31.7 211 2.65 0.009 ** PI398594 83.8 31.7 211 2.65 0.009 ** 159925 83.8 25.8 211 3.25 0.001 ** SARISOY 11 84.2 31.7 211 2.66 0.008 ** PI462312 85.6 25.8 211 3.31 0.001 ** SARISOY 16 88.7 25.8 211 3.43 0.001 *** PI423972 88.9 25.8 211 3.44 0.001 *** BRS-7980A 90.3 31.7 211 2.85 0.005 ** 171 University of Ghana http://ugspace.ug.edu.gh PI587905 90.6 31.7 211 2.86 0.005 ** PI588053A 90.7 25.8 211 3.51 0.001 *** SARISOY 18 90.9 25.8 211 3.52 0.001 *** PI200526 92.2 25.8 211 3.57 0.000 *** TGX 2010-11F 92.4 25.8 211 3.58 0.000 *** TGX 2004-13FB 93.9 31.7 211 2.97 0.003 ** SARISOY 30 93.9 25.8 211 3.64 0.000 *** PI 534545 94.3 25.8 211 3.65 0.000 *** PI567053 96.0 25.8 211 3.71 0.000 *** SARISOY 28 TGX 96.0 25.8 211 3.72 0.000 *** 1990- 110FN 96.2 25.8 211 3.72 0.000 *** SARISOY 13 97.8 25.8 211 3.78 0.000 *** Songdaa 97.8 25.8 211 3.79 0.000 *** PI567068A 99.5 25.8 211 3.85 0.000 *** PI417085 99.6 25.8 211 3.85 0.000 *** PI605824A 99.9 25.8 211 3.87 0.000 *** TGX-1990-49F 100.3 25.8 211 3.88 0.000 *** PI594767B 100.9 25.8 200 3.91 0.000 *** PI635999 101.0 25.8 211 3.91 0.000 *** TGX 2006-3F 101.3 25.8 211 3.92 0.000 *** PI605829 101.5 25.8 211 3.93 0.000 *** Note: *, **, *** represents significance at 5, 1 and 0.1 %, respectively. 172 University of Ghana http://ugspace.ug.edu.gh Appendix 2 Cont’d. Line Estimate Std. Error df t value Pr(>|t|) Sig PI567191 102.6 25.8 211 3.97 0.000 *** TGX-1989-68FN 102.8 25.8 211 3.98 0.000 *** OZARK 106.1 25.8 211 4.11 0.000 *** PI594767A 106.5 25.8 211 4.12 0.000 *** PI605865B 106.7 25.8 211 4.13 0.000 *** PI5670466 111.5 31.7 211 3.52 0.001 *** PI81042A 111.5 25.8 211 4.32 0.000 *** PI567129 111.9 25.8 211 4.33 0.000 *** PI606397B 113.3 25.8 211 4.38 0.000 *** TGX 1844-19FB 114.9 31.7 211 3.63 0.000 *** PI398836 115.3 25.8 211 4.46 0.000 *** PI476905A 115.6 25.8 211 4.47 0.000 *** PI416810B-1 115.6 31.7 211 3.65 0.000 *** PI417126 118.8 25.8 211 4.60 0.000 *** PI594760B 120.7 25.8 200 4.68 0.000 *** PI594767B-1 121.9 25.8 211 4.72 0.000 *** PI60591B 122.1 25.8 211 4.72 0.000 *** PI605865B-1 122.3 25.8 211 4.73 0.000 *** PI567034 122.5 31.7 211 3.87 0.000 *** PI423958 124.1 25.8 211 4.80 0.000 *** PI567046A 129.0 25.8 211 4.99 0.000 *** PI423960A 131.1 25.8 211 5.07 0.000 *** PI417013 132.8 25.8 211 5.14 0.000 *** PI567188 134.5 25.8 211 5.21 0.000 *** PI417134 138.2 25.8 211 5.35 0.000 *** PI416810B 139.4 25.8 211 5.40 0.000 *** PI274454 141.3 25.8 211 5.47 0.000 *** PI567091A 143.2 31.7 211 4.52 0.000 *** PI605891A 143.3 25.8 211 5.55 0.000 *** PI567180 144.6 25.8 211 5.60 0.000 *** LIU YUEMANG 146.6 25.8 211 5.67 0.000 *** PI416806 148.3 25.8 211 5.74 0.000 *** PI605891B 150.3 25.8 211 5.82 0.000 *** PI605891B-1 150.5 25.8 211 5.83 0.000 *** PI567025A 151.9 25.8 200 5.89 0.000 *** PI567056A 152.1 25.8 211 5.89 0.000 *** PI561381 153.7 25.8 211 5.95 0.000 *** PI587886 154.1 25.8 211 5.97 0.000 *** PI605791A 159.0 25.8 211 6.15 0.000 *** PI417120A 162.6 25.8 211 6.29 0.000 *** PI81042 166.2 25.8 211 6.43 0.000 *** 173 University of Ghana http://ugspace.ug.edu.gh LIU YUEBAO 176.4 25.8 211 6.83 0.000 *** PI417120 176.5 25.8 200 6.84 0.000 *** PI605854B 177.7 31.7 211 5.61 0.000 *** PI606405 183.3 25.8 211 7.09 0.000 *** PI398825 190.5 25.8 211 7.37 0.000 *** PI506938 197.5 25.8 211 7.64 0.000 *** PI567516C 209.6 25.8 211 8.11 0.000 *** Note: *, **, *** represents significance at 5, 1 and 0.1 %, respectively. Appendix. 3: Fixed effects estimates of lines’ reaction to PSD based visual assessment Line Estimate Std. Error df t value Pr(>|t|) (Intercept) 4.01E+00 5.65E-01 8.02E+01 7.101 4.50E-10 Croton -1.58E-02 7.71E-01 1.31E+02 -0.02 0.98368 Wamini - 8.71E-01 7.72E-01 1.31E+02 -1.128 0.26146 Afayak 8.44E-01 9.45E-01 1.31E+02 0.893 0.37352 BRS- 7525 1.13E+00 9.45E-01 1.31E+02 1.194 0.23458 BRS- 7525B -1.15E+00 9.45E-01 1.31E+02 -1.22 0.22464 BRS-326 4.90E-01 7.71E-01 1.31E+02 0.636 0.52582 BRS-7980 9.80E-01 7.71E-01 1.31E+02 1.271 0.20595 BRS 8660 9.86E-01 7.71E-01 1.31E+02 1.28 0.20298 BRS 8660A 8.39E-01 9.45E-01 1.31E+02 0.888 0.3764 BRS Perola -1.15E+00 9.45E-01 1.31E+02 -1.219 0.22497 BRS TIANA -1.01E+00 7.71E-01 1.31E+02 -1.309 0.19292 BRS 313 -2.10E-02 7.71E-01 1.31E+02 -0.027 0.97836 BRS 361 -1.01E+00 7.71E-01 1.31E+02 -1.303 0.19472 BRS 90-8360 -1.32E-02 7.71E-01 1.31E+02 -0.017 0.98633 FT christa -2.01E+00 7.71E-01 1.31E+02 -2.608 0.01017 174 University of Ghana http://ugspace.ug.edu.gh ISS-8203 -1.15E+00 9.45E-01 1.31E+02 -1.22 0.22464 Jenguma 9.78E-01 7.71E-01 1.31E+02 1.269 0.2066 LG 04 -6000 -2.54E-02 7.71E-01 1.31E+02 -0.033 0.97375 Liu Yuebao -1.02E+00 7.71E-01 1.31E+02 -1.319 0.18953 Liu Yuemang -1.16E+00 9.45E-01 1.31E+02 -1.227 0.22202 NE 3400 -1.01E+00 7.71E-01 1.31E+02 -1.31 0.19233 Ozark 1.12E+00 9.45E-01 1.31E+02 1.185 0.23823 OZARK 8.59E-01 9.45E-01 1.31E+02 0.908 0.36528 Parangoina -8.68E-01 9.45E-01 1.31E+02 -0.918 0.36009 PARANGOINA 8.33E-01 9.45E-01 1.31E+02 0.881 0.38001 PI 534545 4.72E-01 7.71E-01 1.31E+02 0.612 0.54158 PI159925 -5.45E-03 7.71E-01 1.31E+02 -0.007 0.99437 PI200492 1.15E+00 9.45E-01 1.31E+02 1.218 0.2256 PI200526 -1.15E+00 9.45E-01 1.31E+02 -1.219 0.225 PI224271 -5.15E-01 7.71E-01 1.31E+02 -0.668 0.50557 PI229358 -9.96E-01 7.71E-01 1.31E+02 -1.292 0.19849 PI230970 -9.98E-01 7.69E-01 1.21E+02 -1.297 0.197 PI274454 -2.45E-03 7.71E-01 1.31E+02 -0.003 0.99747 PI274954A 8.54E-01 9.42E-01 1.14E+02 0.907 0.36634 PI340898A -5.21E-01 7.71E-01 1.31E+02 -0.675 0.50064 PI379621 -1.51E+00 6.68E-01 1.31E+02 -2.258 0.0256 PI398387 9.80E-01 7.71E-01 1.31E+02 1.271 0.20587 PI398477 -8.68E-01 9.45E-01 1.31E+02 -0.918 0.36027 PI398512 -1.01E+00 7.71E-01 1.31E+02 -1.313 0.19154 PI398543 -1.15E+00 9.45E-01 1.31E+02 -1.219 0.2251 PI398594 -1.01E+00 7.71E-01 1.31E+02 -1.31 0.1925 PI398825 -2.15E+00 9.45E-01 1.31E+02 -2.277 0.0244 PI398836 -1.01E+00 7.71E-01 1.31E+02 -1.304 0.19455 PI398874 -1.02E+00 7.71E-01 1.31E+02 -1.327 0.18681 PI399096 -5.12E-01 7.71E-01 1.31E+02 -0.664 0.50784 175 University of Ghana http://ugspace.ug.edu.gh PI416778 -9.97E-01 7.71E-01 1.31E+02 -1.294 0.19805 PI416806A -1.15E+00 9.45E-01 1.31E+02 -1.219 0.225 PI416806B 1.01E+00 7.71E-01 1.31E+02 1.303 0.19475 PI416810 -1.15E+00 9.45E-01 1.31E+02 -1.219 0.22516 PI416810A -9.75E-01 7.04E-01 1.31E+02 -1.386 0.16821 PI416810B -8.64E-01 9.45E-01 1.31E+02 -0.915 0.36211 PI416810B-1 -1.15E+00 9.45E-01 1.31E+02 -1.218 0.22526 PI416810C -1.53E-01 9.45E-01 1.31E+02 -0.162 0.87155 PI416810E 8.63E-01 9.45E-01 1.31E+02 0.913 0.36279 PI416873B -2.17E+00 9.45E-01 1.31E+02 -2.292 0.02351 PI416886 4.97E-01 7.71E-01 1.31E+02 0.645 0.52015 PI416939 -5.07E-01 7.69E-01 1.21E+02 -0.659 0.51118 PI417013 -1.52E+00 7.69E-01 1.22E+02 -1.977 0.05035 PI417119 -1.00E+00 7.71E-01 1.31E+02 -1.299 0.19624 PI417120 -8.54E-01 9.42E-01 1.14E+02 -0.907 0.36634 PI417120A -8.70E-01 9.45E-01 1.31E+02 -0.92 0.35931 PI417126 -1.02E+00 7.71E-01 1.31E+02 -1.326 0.18721 PI417134 -1.50E+00 7.71E-01 1.31E+02 -1.947 0.05373 PI423958 -1.88E+00 9.45E-01 1.31E+02 -1.984 0.04938 PI423960A 4.98E-01 7.71E-01 1.31E+02 0.646 0.51956 PI423972 -2.43E-02 7.71E-01 1.31E+02 -0.031 0.97493 PI459025A -5.21E-01 7.71E-01 1.31E+02 -0.675 0.50068 PI462312 8.31E-01 9.45E-01 1.31E+02 0.879 0.38082 PI468919A -1.56E-01 9.45E-01 1.31E+02 -0.165 0.86918 PI476905A -2.51E+00 7.71E-01 1.31E+02 -3.255 0.00144 PI506677 -1.37E-01 9.45E-01 1.31E+02 -0.145 0.88519 PI506938 -1.00E-02 7.71E-01 1.31E+02 -0.013 0.98963 PI506939 -1.37E-01 9.45E-01 1.31E+02 -0.145 0.88519 PI507004 4.83E-01 7.71E-01 1.31E+02 0.627 0.53199 PI507004A -1.03E+00 7.71E-01 1.31E+02 -1.333 0.18498 176 University of Ghana http://ugspace.ug.edu.gh PI518295 -1.01E+00 7.69E-01 1.21E+02 -1.319 0.18975 PI561381 -1.01E+00 7.71E-01 1.31E+02 -1.315 0.19069 PI567025A 4.76E-01 7.71E-01 1.31E+02 0.617 0.53818 PI567053 -1.35E-02 7.71E-01 1.31E+02 -0.018 0.98606 PI567054C -1.02E+00 7.71E-01 1.31E+02 -1.32 0.1893 PI567056A -1.13E-02 7.71E-01 1.31E+02 -0.015 0.98829 PI567058D -5.09E-01 7.71E-01 1.31E+02 -0.66 0.51016 PI567068A -1.51E+00 7.71E-01 1.31E+02 -1.961 0.05204 PI567076 -1.50E+00 7.71E-01 1.31E+02 -1.947 0.05371 PI567090 -1.02E+00 7.71E-01 1.31E+02 -1.324 0.18781 PI567091A -1.51E+00 7.71E-01 1.31E+02 -1.952 0.05307 PI567099A -1.01E+00 7.71E-01 1.31E+02 -1.316 0.19052 PI567104B -1.89E-02 7.71E-01 1.31E+02 -0.025 0.98044 PI567123A -2.19E-02 7.71E-01 1.31E+02 -0.028 0.97738 PI567129 -1.02E+00 7.71E-01 1.31E+02 -1.319 0.18939 PI567188 -5.27E-01 7.71E-01 1.31E+02 -0.683 0.49555 PI567190 -1.52E-01 9.45E-01 1.31E+02 -0.161 0.8722 PI567191 -1.87E+00 9.45E-01 1.31E+02 -1.977 0.05018 PI567516C -2.01E+00 7.71E-01 1.31E+02 -2.604 0.01028 PI578457A 1.00E+00 7.71E-01 1.31E+02 1.297 0.19681 PI584796 -9.48E-03 7.71E-01 1.31E+02 -0.012 0.9902 PI587886 -1.16E+00 9.45E-01 1.31E+02 -1.223 0.22352 PI587905 -1.01E+00 7.71E-01 1.31E+02 -1.313 0.19152 PI588053A -1.87E+00 9.45E-01 1.31E+02 -1.976 0.05024 PI588053B -1.01E+00 7.71E-01 1.31E+02 -1.315 0.19071 PI594538A -1.00E+00 7.71E-01 1.31E+02 -1.3 0.19577 PI594760B -1.50E+00 7.71E-01 1.31E+02 -1.949 0.05347 PI594767A -1.51E+00 7.71E-01 1.31E+02 -1.964 0.05164 PI594767B -2.02E+00 7.71E-01 1.31E+02 -2.625 0.00969 PI603170 -7.67E-03 7.71E-01 1.31E+02 -0.01 0.99207 177 University of Ghana http://ugspace.ug.edu.gh PI603176A 1.51E-01 9.45E-01 1.31E+02 0.159 0.87353 PI605791A -1.02E+00 7.71E-01 1.31E+02 -1.321 0.18874 PI605823 1.00E+00 7.71E-01 1.31E+02 1.297 0.19706 PI605824A -1.69E-01 9.45E-01 1.31E+02 -0.179 0.85846 PI605829 -1.01E+00 7.71E-01 1.31E+02 -1.315 0.19081 PI605854B -1.50E+00 7.71E-01 1.31E+02 -1.944 0.05402 PI605891A -2.03E+00 7.71E-01 1.31E+02 -2.626 0.00966 PI605891B -1.02E+00 7.71E-01 1.31E+02 -1.32 0.18925 PI605891B-1 -1.52E-01 9.45E-01 1.31E+02 -0.161 0.87213 PI606405 -1.50E+00 7.71E-01 1.31E+02 -1.945 0.05386 PI628850 -4.47E-03 7.71E-01 1.31E+02 -0.006 0.99538 PI628932 -5.18E-01 7.69E-01 1.22E+02 -0.674 0.50179 PI635999 -9.99E-01 7.71E-01 1.31E+02 -1.296 0.19717 PI657753 4.85E-01 7.71E-01 1.31E+02 0.628 0.5308 PI81042 -1.13E+00 9.45E-01 1.31E+02 -1.196 0.23397 Quarshie 1.32E-01 9.45E-01 1.31E+02 0.14 0.88884 SARISOY 27 1.33E-01 9.45E-01 1.31E+02 0.141 0.88823 SARISOY 1 1.33E-01 9.45E-01 1.31E+02 0.141 0.88828 SARISOY 10 1.51E-01 9.45E-01 1.31E+02 0.159 0.87353 SARISOY 12 -8.94E-01 9.45E-01 1.31E+02 -0.946 0.34602 SARISOY 13 -8.68E-01 9.45E-01 1.31E+02 -0.918 0.36008 SARISOY 14 1.13E+00 9.45E-01 1.31E+02 1.197 0.23329 SARISOY 15 1.13E+00 9.45E-01 1.31E+02 1.197 0.2333 SARISOY 16 -3.57E-01 7.72E-01 1.31E+02 -0.462 0.64482 SARISOY 17 1.10E-01 7.72E-01 1.31E+02 0.143 0.88669 SARISOY 18 1.32E-01 9.45E-01 1.31E+02 0.139 0.88945 SARISOY 19 1.12E+00 9.45E-01 1.31E+02 1.185 0.23823 SARISOY 21 -8.66E-01 9.45E-01 1.31E+02 -0.916 0.36131 SARISOY 23 1.33E-01 9.45E-01 1.31E+02 0.141 0.88823 SARISOY 25 -1.87E+00 9.45E-01 1.31E+02 -1.975 0.05036 178 University of Ghana http://ugspace.ug.edu.gh SARISOY 26 -8.67E-01 9.45E-01 1.31E+02 -0.917 0.36074 SARISOY 28 1.06E-01 9.45E-01 1.31E+02 0.112 0.91082 SARISOY 30 1.13E+00 9.45E-01 1.31E+02 1.194 0.23458 SARISOY 4 1.52E-01 9.45E-01 1.31E+02 0.161 0.87239 SARISOY 5 1.13E+00 9.45E-01 1.31E+02 1.196 0.2339 SARISOY 7 -8.64E-01 9.45E-01 1.31E+02 -0.914 0.36219 SARISOY 9 1.13E+00 9.45E-01 1.31E+02 1.197 0.2333 SEEDCO 27 -1.66E-01 9.45E-01 1.31E+02 -0.176 0.86065 SEEDCO 8 -1.57E-01 9.45E-01 1.31E+02 -0.166 0.86841 SEEDCO 1 8.51E-01 9.45E-01 1.31E+02 0.9 0.36969 SEEDCO 10 -1.15E+00 9.45E-01 1.31E+02 -1.216 0.22633 SEEDCO 11 8.48E-01 9.45E-01 1.31E+02 0.897 0.37126 SEEDCO 12 8.33E-01 9.45E-01 1.31E+02 0.881 0.38001 SEEDCO 13 -1.17E+00 9.45E-01 1.31E+02 -1.242 0.21657 SEEDCO 14 8.34E-01 9.45E-01 1.31E+02 0.882 0.37936 SEEDCO 15 8.51E-01 9.45E-01 1.31E+02 0.9 0.36969 SEEDCO 16 3.29E-01 7.72E-01 1.31E+02 0.426 0.67047 SEEDCO 17 -1.15E+00 9.42E-01 1.14E+02 -1.216 0.2264 SEEDCO 18 8.51E-01 9.45E-01 1.31E+02 0.9 0.36969 SEEDCO 19 -1.13E+00 9.45E-01 1.31E+02 -1.199 0.23275 SEEDCO 21 -1.15E+00 9.42E-01 1.14E+02 -1.216 0.2264 SEEDCO 25 -1.15E+00 9.45E-01 1.31E+02 -1.218 0.22526 SEEDCO 26 -1.57E-01 9.45E-01 1.31E+02 -0.166 0.86841 SEEDCO 28 8.40E-01 9.45E-01 1.31E+02 0.889 0.37561 SEEDCO 29 -1.37E-01 9.45E-01 1.31E+02 -0.145 0.88519 SEEDCO 30 8.49E-01 7.72E-01 1.31E+02 1.099 0.27368 SEEDCO 4 -1.16E+00 9.45E-01 1.31E+02 -1.227 0.22202 SEEDCO 5 8.48E-01 9.45E-01 1.31E+02 0.897 0.37155 SEEDCO 7 -1.46E-01 9.45E-01 1.31E+02 -0.155 0.87719 SEEDCO 9 8.54E-01 9.45E-01 1.31E+02 0.903 0.36811 179 University of Ghana http://ugspace.ug.edu.gh Songdaa -2.01E+00 7.71E-01 1.31E+02 -2.609 0.01013 TGX-1835-10E -5.10E-01 7.71E-01 1.31E+02 -0.662 0.50906 TGX-1989-40FA -1.87E+00 9.45E-01 1.31E+02 -1.978 0.05005 TGX-1989-48FN -1.61E-01 9.45E-01 1.31E+02 -0.17 0.86497 TGX-1990-21F -1.02E+00 7.71E-01 1.31E+02 -1.32 0.18899 TGX-1990-40F 1.01E+00 7.69E-01 1.21E+02 1.311 0.19243 TGX-1990-46F -5.25E-01 7.71E-01 1.31E+02 -0.68 0.49746 TGX-1990-49F -1.50E+00 7.71E-01 1.31E+02 -1.946 0.05375 TGX-1990-55F -1.53E-02 7.71E-01 1.31E+02 -0.02 0.9842 TGX 1445-3E -1.01E+00 7.71E-01 1.31E+02 -1.311 0.19203 TGX 1485-1D 1.52E-01 9.45E-01 1.31E+02 0.161 0.87239 TGX 1740-2F 4.86E-01 7.71E-01 1.31E+02 0.63 0.52961 TGX 1799-8F -7.08E-03 7.71E-01 1.31E+02 -0.009 0.99269 TGX 1805-8F -1.01E+00 7.71E-01 1.31E+02 -1.313 0.19162 TGX 1805-8F-1 -1.56E-01 9.45E-01 1.31E+02 -0.165 0.86918 TGX 1834-5E 4.86E-01 7.71E-01 1.31E+02 0.631 0.52922 TGX 1844-19F -5.10E-01 7.71E-01 1.31E+02 -0.661 0.50967 TGX 1903-7F -9.98E-01 7.69E-01 1.21E+02 -1.298 0.19682 TGX 19110-6F -1.00E+00 7.71E-01 1.31E+02 -1.298 0.19666 TGX 1987-62F 9.89E-01 7.71E-01 1.31E+02 1.283 0.20161 TGX 1988-3F -1.06E+00 7.04E-01 1.31E+02 -1.499 0.13622 TGX 1988-3FA -1.37E-01 9.45E-01 1.31E+02 -0.145 0.88519 TGX 1989-11F -5.08E-01 7.71E-01 1.31E+02 -0.659 0.51096 TGX 1989-19F -5.18E-01 7.71E-01 1.31E+02 -0.671 0.50316 TGX 1989-20FA -1.14E+00 9.45E-01 1.31E+02 -1.203 0.2313 TGX 1989-41F -1.15E+00 9.45E-01 1.31E+02 -1.219 0.22516 TGX 1989-42F -5.08E-01 7.71E-01 1.31E+02 -0.66 0.51072 TGX 1989-75N -5.31E-01 7.71E-01 1.31E+02 -0.688 0.49244 TGX 1990-110FN-1.61E-01 9.45E-01 1.31E+02 -0.17 0.86497 TGX 1990-52F 4.75E-01 7.71E-01 1.31E+02 0.617 0.53853 180 University of Ghana http://ugspace.ug.edu.gh TGX 1990-57F 2.68E-01 7.04E-01 1.31E+02 0.381 0.7037 TGX 1990-78F 4.98E-01 7.71E-01 1.31E+02 0.646 0.51958 TGX 1990-80F -3.78E-02 7.71E-01 1.31E+02 -0.049 0.96097 TGX 1990-93F -1.13E+00 9.45E-01 1.31E+02 -1.199 0.23275 TGX 1990-95F 1.11E+00 9.45E-01 1.31E+02 1.172 0.24351 TGX 1990 -114 FN-5.05E-01 7.71E-01 1.31E+02 -0.655 0.51347 TGX 1993-4FN -2.91E-03 7.69E-01 1.21E+02 -0.004 0.99699 TGX 2004-10F 9.78E-01 7.71E-01 1.31E+02 1.268 0.20706 TGX 2004-13F -5.20E-01 7.71E-01 1.31E+02 -0.674 0.5016 TGX 2006-3F -7.63E-01 6.68E-01 1.31E+02 -1.143 0.25497 TGX 2007-11F -1.50E+00 7.71E-01 1.31E+02 -1.949 0.05348 TGX 2008-4F -5.10E-01 7.71E-01 1.31E+02 -0.662 0.50935 TGX 2010-11F -1.01E+00 7.69E-01 1.22E+02 -1.31 0.19278 TGX 2011-3F -1.61E-02 7.71E-01 1.31E+02 -0.021 0.98342 TGX1845-10E -1.02E+00 7.71E-01 1.31E+02 -1.322 0.1886 TGX1987-10F -1.27E-04 7.71E-01 1.31E+02 0 0.99987 TGX1989-48FN -4.14E-03 7.71E-01 1.31E+02 -0.005 0.99572 TGX1989-68FN -1.03E-02 7.71E-01 1.31E+02 -0.013 0.98938 Zamboani -2.48E-02 7.71E-01 1.31E+02 -0.032 0.97441 Appendix. 4: Fixed effect estimate of lines based PSD infection rate Line Estimate Std. Error df t value Pr(>|t|) Sign. (Intercept) 7.50E+01 1.17E+01 1.31E+02 6.419 2.31E-09 *** Croton 2.50E+00 1.65E+01 1.31E+02 0.151 0.879966 Wamini -2.00E+01 1.65E+01 1.31E+02 -1.21 0.228283 Afayak -3.00E+01 2.02E+01 1.31E+02 -1.483 0.140608 BRS- 7525 -1.50E+01 2.02E+01 1.31E+02 -0.741 0.459868 BRS-326 -3.00E+01 1.65E+01 1.31E+02 -1.816 0.071705 BRS-7980 -2.50E+01 1.65E+01 1.31E+02 -1.513 0.13267 181 University of Ghana http://ugspace.ug.edu.gh BRS 8660 -3.00E+01 1.65E+01 1.31E+02 -1.816 0.071705 BRS 8660A -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 BRS Perola 2.50E+01 2.02E+01 1.31E+02 1.235 0.218886 BRS TIANA 2.00E+01 1.65E+01 1.31E+02 1.21 0.228283 BRS 313 -5.50E+01 1.65E+01 1.31E+02 -3.329 0.001133 ** BRS 361 -7.00E+01 1.65E+01 1.31E+02 -4.237 4.25E-05 *** BRS 90-8360 1.50E+01 1.65E+01 1.31E+02 0.908 0.365628 FT christa -1.00E+01 1.65E+01 1.31E+02 -0.605 0.546074 Jenguma 2.50E+01 1.65E+01 1.31E+02 1.513 0.13267 LG 04 -6000 -3.50E+01 1.65E+01 1.31E+02 -2.118 0.036038 * Liu Yuebao -6.50E+01 1.65E+01 1.31E+02 -3.934 0.000135 *** Liu Yuemang -7.50E+01 2.02E+01 1.31E+02 -3.706 0.000309 *** NE 3400 -2.50E+01 1.65E+01 1.31E+02 -1.513 0.13267 Ozark 1.50E+01 2.02E+01 1.31E+02 0.741 0.459868 OZARK 5.00E+00 2.02E+01 1.31E+02 0.247 0.80523 Parangoina -6.50E+01 2.02E+01 1.31E+02 -3.212 0.001658 ** PARANGOINA -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 PI 534545 -5.00E+01 1.65E+01 1.31E+02 -3.026 0.002983 ** PI159925 -4.00E+01 1.65E+01 1.31E+02 -2.421 0.016852 * PI200492 -5.50E+01 2.02E+01 1.31E+02 -2.718 0.007459 ** PI200526 -4.50E+01 2.02E+01 1.31E+02 -2.224 0.027878 * PI224271 -7.50E+01 1.65E+01 1.31E+02 -4.539 1.26E-05 *** PI229358 2.50E+01 1.65E+01 1.31E+02 1.513 0.13267 PI230970 -5.50E+01 1.65E+01 1.31E+02 -3.329 0.001133 ** PI274454 -1.50E+01 1.65E+01 1.31E+02 -0.908 0.365628 PI274954A -3.50E+01 2.02E+01 1.31E+02 -1.73 0.08606 . PI340898A -3.00E+01 1.65E+01 1.31E+02 -1.816 0.071705 . PI379621 -3.75E+01 1.43E+01 1.31E+02 -2.621 0.009812 ** PI398387 -2.50E+01 1.65E+01 1.31E+02 -1.513 0.13267 PI398477 1.50E+01 2.02E+01 1.31E+02 0.741 0.459868 182 University of Ghana http://ugspace.ug.edu.gh PI398512 -1.00E+01 1.65E+01 1.31E+02 -0.605 0.546074 PI398543 -2.50E+01 2.02E+01 1.31E+02 -1.235 0.218886 PI398594 -6.90E-13 1.65E+01 1.31E+02 0 1 PI398825 -4.50E+01 2.02E+01 1.31E+02 -2.224 0.027878 * PI398836 -2.50E+01 1.65E+01 1.31E+02 -1.513 0.13267 PI398874 1.50E+01 1.65E+01 1.31E+02 0.908 0.365628 PI399096 -4.50E+01 1.65E+01 1.31E+02 -2.724 0.00734 ** PI416778 -4.00E+01 1.65E+01 1.31E+02 -2.421 0.016852 * PI416806A -4.50E+01 2.02E+01 1.31E+02 -2.224 0.027878 * PI416806B -4.50E+01 2.02E+01 1.31E+02 -2.224 0.027878 * PI416810 -5.50E+01 2.02E+01 1.31E+02 -2.718 0.007459 ** PI416810A -5.17E+01 1.51E+01 1.31E+02 -3.425 0.00082 *** PI416810B 5.00E+00 2.02E+01 1.31E+02 0.247 0.80523 PI416810B-1 5.00E+00 2.02E+01 1.31E+02 0.247 0.80523 PI416873B -2.50E+01 2.02E+01 1.31E+02 -1.235 0.218886 PI416886 -4.00E+01 1.65E+01 1.31E+02 -2.421 0.016852 * PI416939 -4.50E+01 1.65E+01 1.31E+02 -2.724 0.00734 ** PI417013 -7.00E+01 1.65E+01 1.31E+02 -4.237 4.25E-05 *** PI417119 2.00E+01 1.65E+01 1.31E+02 1.21 0.228283 PI417120 -7.50E+01 2.02E+01 1.31E+02 -3.706 0.000309 *** PI417120A -3.50E+01 2.02E+01 1.31E+02 -1.73 0.08606 . PI417126 -5.00E+01 1.65E+01 1.31E+02 -3.026 0.002983 ** PI417134 -1.50E+01 1.65E+01 1.31E+02 -0.908 0.365628 PI423958 -7.47E+01 2.02E+01 1.31E+02 -3.69 0.000328 *** PI423960A -2.50E+01 1.65E+01 1.31E+02 -1.513 0.13267 PI423972 2.50E+01 1.65E+01 1.31E+02 1.513 0.13267 PI459025A -2.00E+01 1.65E+01 1.31E+02 -1.21 0.228283 PI462312 5.00E+00 2.02E+01 1.31E+02 0.247 0.80523 PI468919A -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 PI476905A 5.00E+00 1.65E+01 1.31E+02 0.303 0.762664 183 University of Ghana http://ugspace.ug.edu.gh PI506938 -7.01E-13 1.65E+01 1.31E+02 0 1 PI507004 -5.00E+00 1.65E+01 1.31E+02 -0.303 0.762664 PI507004A -3.00E+01 1.65E+01 1.31E+02 -1.816 0.071705 . PI518295 -4.50E+01 1.65E+01 1.31E+02 -2.724 0.00734 ** PI561381 -3.00E+01 1.65E+01 1.31E+02 -1.816 0.071705 . PI567025A -3.50E+01 1.65E+01 1.31E+02 -2.118 0.036038 * PI567053 -2.50E+01 1.65E+01 1.31E+02 -1.513 0.13267 PI567054C -3.50E+01 1.65E+01 1.31E+02 -2.118 0.036038 * PI567056A -1.50E+01 1.65E+01 1.31E+02 -0.908 0.365628 PI567058D -4.00E+01 1.65E+01 1.31E+02 -2.421 0.016852 * PI567068A 5.00E+00 1.65E+01 1.31E+02 0.303 0.762664 PI567076 2.00E+01 1.65E+01 1.31E+02 1.21 0.228283 PI567090 -1.00E+01 1.65E+01 1.31E+02 -0.605 0.546074 PI567091A -6.50E+01 1.65E+01 1.31E+02 -3.934 0.000135 *** PI567099A -7.13E-13 1.65E+01 1.31E+02 0 1 PI567104B -2.00E+01 1.65E+01 1.31E+02 -1.21 0.228283 PI567123A -1.50E+01 1.65E+01 1.31E+02 -0.908 0.365628 PI567129 -3.00E+01 1.65E+01 1.31E+02 -1.816 0.071705 . PI567188 -4.00E+01 1.65E+01 1.31E+02 -2.421 0.016852 * PI567190 5.00E+00 2.02E+01 1.31E+02 0.247 0.80523 PI567191 -3.50E+01 2.02E+01 1.31E+02 -1.73 0.08606 . PI567516C -7.00E+01 1.65E+01 1.31E+02 -4.237 4.25E-05 *** PI578457A -5.00E+00 1.65E+01 1.31E+02 -0.303 0.762664 PI584796 -4.50E+01 1.65E+01 1.31E+02 -2.724 0.00734 ** PI587886 -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 PI587905 2.50E+01 1.65E+01 1.31E+02 1.513 0.13267 PI588053A -1.50E+01 2.02E+01 1.31E+02 -0.741 0.459868 PI588053B -6.92E-13 1.65E+01 1.31E+02 0 1 PI594538A -4.50E+01 1.65E+01 1.31E+02 -2.724 0.00734 ** PI594760B -2.50E+01 1.65E+01 1.31E+02 -1.513 0.13267 184 University of Ghana http://ugspace.ug.edu.gh PI594767A -3.00E+01 1.65E+01 1.31E+02 -1.816 0.071705 . PI594767B -6.00E+01 1.65E+01 1.31E+02 -3.631 0.000403 *** PI603170 -3.00E+01 1.65E+01 1.31E+02 -1.816 0.071705 . PI603176A -4.50E+01 2.02E+01 1.31E+02 -2.224 0.027878 * PI605791A -5.50E+01 1.65E+01 1.31E+02 -3.329 0.001133 ** PI605823 -1.00E+01 1.65E+01 1.31E+02 -0.605 0.546074 PI605824A -3.00E+01 2.02E+01 1.31E+02 -1.483 0.140608 PI605829 5.00E+00 1.65E+01 1.31E+02 0.303 0.762664 PI605854B -2.50E+01 1.65E+01 1.31E+02 -1.513 0.13267 PI605891A -5.50E+01 1.65E+01 1.31E+02 -3.329 0.001133 ** PI605891B -1.50E+01 1.65E+01 1.31E+02 -0.908 0.365628 PI605891B-1 -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 PI606405 -4.00E+01 1.65E+01 1.31E+02 -2.421 0.016852 * PI628850 -5.50E+01 1.65E+01 1.31E+02 -3.329 0.001133 ** PI628932 -1.50E+01 1.65E+01 1.31E+02 -0.908 0.365628 PI635999 -4.00E+01 1.65E+01 1.31E+02 -2.421 0.016852 * PI657753 1.00E+01 1.65E+01 1.31E+02 0.605 0.546074 PI81042 -3.50E+01 2.02E+01 1.31E+02 -1.73 0.08606 . Quarshie 5.00E+00 2.02E+01 1.31E+02 0.247 0.80523 SARISOY 27 -3.50E+01 2.02E+01 1.31E+02 -1.73 0.08606 . SARISOY 1 -1.50E+01 2.02E+01 1.31E+02 -0.741 0.459868 SARISOY 10 -6.50E+01 2.02E+01 1.31E+02 -3.212 0.001658 ** SARISOY 12 2.50E+01 2.02E+01 1.31E+02 1.235 0.218886 SARISOY 13 -3.50E+01 2.02E+01 1.31E+02 -1.73 0.08606 . SARISOY 14 -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 SARISOY 15 -6.50E+01 2.02E+01 1.31E+02 -3.212 0.001658 ** SARISOY 16 -2.00E+01 1.65E+01 1.31E+02 -1.21 0.228283 SARISOY 17 -3.50E+01 1.65E+01 1.31E+02 -2.118 0.036038 * SARISOY 18 -6.50E+01 2.02E+01 1.31E+02 -3.212 0.001658 ** SARISOY 19 -2.50E+01 2.02E+01 1.31E+02 -1.235 0.218886 185 University of Ghana http://ugspace.ug.edu.gh SARISOY 21 -6.50E+01 2.02E+01 1.31E+02 -3.212 0.001658 ** SARISOY 23 -4.50E+01 2.02E+01 1.31E+02 -2.224 0.027878 * SARISOY 25 -1.50E+01 2.02E+01 1.31E+02 -0.741 0.459868 SARISOY 26 -6.50E+01 2.02E+01 1.31E+02 -3.212 0.001658 ** SARISOY 28 2.50E+01 2.02E+01 1.31E+02 1.235 0.218886 SARISOY 30 -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 SARISOY 4 -3.50E+01 2.02E+01 1.31E+02 -1.73 0.08606 . SARISOY 5 -6.50E+01 2.02E+01 1.31E+02 -3.212 0.001658 ** SARISOY 7 -2.50E+01 2.02E+01 1.31E+02 -1.235 0.218886 SARISOY 9 1.50E+01 2.02E+01 1.31E+02 0.741 0.459868 SEEDCO 27 -5.50E+01 2.02E+01 1.31E+02 -2.718 0.007459 ** SEEDCO 8 -2.50E+01 2.02E+01 1.31E+02 -1.235 0.218886 SEEDCO 1 -1.50E+01 2.02E+01 1.31E+02 -0.741 0.459868 SEEDCO 10 -4.50E+01 2.02E+01 1.31E+02 -2.224 0.027878 * SEEDCO 11 -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 SEEDCO 12 1.50E+01 2.02E+01 1.31E+02 0.741 0.459868 SEEDCO 13 -5.50E+01 2.02E+01 1.31E+02 -2.718 0.007459 ** SEEDCO 14 -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 SEEDCO 15 -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 SEEDCO 16 -6.89E-13 1.65E+01 1.31E+02 0 1 SEEDCO 17 -3.50E+01 2.02E+01 1.31E+02 -1.73 0.08606 . SEEDCO 18 -5.50E+01 2.02E+01 1.31E+02 -2.718 0.007459 ** SEEDCO 19 -3.50E+01 2.02E+01 1.31E+02 -1.73 0.08606 . SEEDCO 21 -6.50E+01 2.02E+01 1.31E+02 -3.212 0.001658 ** SEEDCO 25 -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 SEEDCO 26 2.50E+01 2.02E+01 1.31E+02 1.235 0.218886 SEEDCO 28 2.50E+01 2.02E+01 1.31E+02 1.235 0.218886 SEEDCO 29 5.00E+00 2.02E+01 1.31E+02 0.247 0.80523 SEEDCO 30 -2.50E+01 1.65E+01 1.31E+02 -1.513 0.13267 SEEDCO 4 -4.50E+01 2.02E+01 1.31E+02 -2.224 0.027878 * 186 University of Ghana http://ugspace.ug.edu.gh SEEDCO 5 -3.50E+01 2.02E+01 1.31E+02 -1.73 0.08606 . SEEDCO 7 5.00E+00 2.02E+01 1.31E+02 0.247 0.80523 SEEDCO 9 1.50E+01 2.02E+01 1.31E+02 0.741 0.459868 Songdaa 1.50E+01 1.65E+01 1.31E+02 0.908 0.365628 TGX-1835 -10E1.50E+01 1.65E+01 1.31E+02 0.908 0.365628 TGX-1989-40FA 5.00E+00 2.02E+01 1.31E+02 0.247 0.80523 TGX-1989-48FN -2.50E+01 2.02E+01 1.31E+02 -1.235 0.218886 TGX-1990-21F -1.00E+01 1.65E+01 1.31E+02 -0.605 0.546074 TGX-1990-40F 2.00E+01 1.65E+01 1.31E+02 1.21 0.228283 TGX-1990-46F -6.92E-13 1.65E+01 1.31E+02 0 1 TGX-1990-49F -1.50E+01 1.65E+01 1.31E+02 -0.908 0.365628 TGX-1990-55F -1.00E+01 1.65E+01 1.31E+02 -0.605 0.546074 TGX 1445-3E -2.50E+01 1.65E+01 1.31E+02 -1.513 0.13267 TGX 1740-2F 2.50E+01 1.65E+01 1.31E+02 1.513 0.13267 TGX 1799-8F 5.00E+00 1.65E+01 1.31E+02 0.303 0.762664 TGX 1805-8F -2.50E+01 1.65E+01 1.31E+02 -1.513 0.13267 TGX 1805-8F-1 -1.50E+01 2.02E+01 1.31E+02 -0.741 0.459868 TGX 1834-5E -6.99E-13 1.65E+01 1.31E+02 0 1 TGX 1844-19F -1.00E+01 1.65E+01 1.31E+02 -0.605 0.546074 TGX 1903-7F -5.50E+01 1.65E+01 1.31E+02 -3.329 0.001133 ** TGX 19110-6F -1.00E+01 1.65E+01 1.31E+02 -0.605 0.546074 TGX 1987-62F 2.50E+01 1.65E+01 1.31E+02 1.513 0.13267 TGX 1988-3F -5.00E+00 1.51E+01 1.31E+02 -0.331 0.740798 TGX 1989-11F -1.50E+01 1.65E+01 1.31E+02 -0.908 0.365628 TGX 1989-19F -1.50E+01 1.65E+01 1.31E+02 -0.908 0.365628 TGX 1989-20FA -2.50E+01 2.02E+01 1.31E+02 -1.235 0.218886 TGX 1989-41F 1.50E+01 2.02E+01 1.31E+02 0.741 0.459868 TGX 1989-42F -4.00E+01 1.65E+01 1.31E+02 -2.421 0.016852 * TGX 1989-75N -5.00E+01 1.65E+01 1.31E+02 -3.026 0.002983 ** TGX 1990-110FN -5.00E+00 2.02E+01 1.31E+02 -0.247 0.80523 187 University of Ghana http://ugspace.ug.edu.gh TGX 1990-52F -3.00E+01 1.65E+01 1.31E+02 -1.816 0.071705 TGX 1990-57F -2.17E+01 1.51E+01 1.31E+02 -1.436 0.153247 TGX 1990-78F -6.88E-13 1.65E+01 1.31E+02 0 1 TGX 1990-80F -3.50E+01 1.65E+01 1.31E+02 -2.118 0.036038 * TGX 1990-93F -1.50E+01 2.02E+01 1.31E+02 -0.741 0.459868 TGX 1990-95F -4.50E+01 2.02E+01 1.31E+02 -2.224 0.027878 * TGX 1990 -114 FN -3.00E+01 1.65E+01 1.31E+02 -1.816 0.071705 TGX 1993-4FN -6.50E+01 1.65E+01 1.31E+02 -3.934 0.000135 **** TGX 2004-10F -1.00E+01 1.65E+01 1.31E+02 -0.605 0.546074 TGX 2004-13F -2.00E+01 1.65E+01 1.31E+02 -1.21 0.228283 TGX 2006-3F -9.00E+00 1.43E+01 1.31E+02 -0.629 0.530462 TGX 2007-11F -2.00E+01 1.65E+01 1.31E+02 -1.21 0.228283 TGX 2008-4F 2.50E+01 1.65E+01 1.31E+02 1.513 0.13267 TGX 2010-11F -5.00E+01 1.65E+01 1.31E+02 -3.026 0.002983 ** TGX 2011-3F -1.00E+01 1.65E+01 1.31E+02 -0.605 0.546074 TGX1845-10E -5.00E+00 1.65E+01 1.31E+02 -0.303 0.762664 TGX1987-10F -3.00E+01 1.65E+01 1.31E+02 -1.816 0.071705 . TGX1989-48FN 5.00E+00 1.65E+01 1.31E+02 0.303 0.762664 TGX1989-68FN -5.00E+01 1.65E+01 1.31E+02 -3.026 0.002983 ** Zamboani -5.00E+01 1.65E+01 1.31E+02 -3.026 0.002983 ** 188