GENETIC DIVERSITY IN SOME LOCAL CASSAVA CULTIVARS IN GHANA By ELIZABETH OKAI A thesis submitted in partial fulfilment of the requirements for the MASTER OF PHILOSOPHY degree in crop science at the University of Ghana, Legon CROP SCIENCE DEPARTMENT, FACULTY OF AGRICULTURE, UNIVERSITY OF GHANA LEGON APRIL, 2001. University of Ghana http://ugspace.ug.edu.gh Q 365823 s h a d ' d o i c v University of Ghana http://ugspace.ug.edu.gh DECLARATION I hereby declare that except for references to works of other researchers which have been duly cited, this work is my own original research and that neither part nor whole has been presented elsewhere for the award o f a degree. Mrs. ELIZABETH OKAI (Student) (Supervisor) Dr E. Y DANQUAH (Supervisor) I University of Ghana http://ugspace.ug.edu.gh DEDICATION Dedicated to my late husband, Isaac Teiko Okai (Nuntso), who encouraged me to start the programme and would have wished to see its successful completion, and to my two sweet children, DOCIA and ZENAS who stood with me, sacrificed, prayed and showed great concern with the progress of my work at all stages until completion. II University of Ghana http://ugspace.ug.edu.gh This thesis could not have been produced without the help of the Almighty God, many people and institutions who in diverse ways supported me prayerfully, technically, morally and financially. I wish to express my sincere thanks to my supervisors, Dr. E .Y. Danquah, an initial supervisor who went on sabbatical leave before the completion of the work and Dr. S. K. Offei for his critical suggestions and financial support at various stages of the work. I also wish to thank Dr. K. Ofori for his immense support throughout this study. I am also grateful to all the lecturers and staff of the Department of Crop Science, Legon. My special thanks go to Miss Faustina Opare and Mr. W. A. Asante for their great support and encouragement. 1 am also grateful to Drs. R Asiedu, A.G.O.Dixon, and H. D. Mignouna for their support and supervision of my work in IITA. I wish to thank all staff of Biotech. Lab.especially Amos and Josephine and TRIP staff especially Esther Mberu, Okechuku and Joseph Onyeka for their friendship and support, IITA, Ibadan, Nigeria. Sincere thanks also go to Dr. C. C. Okafor and the staff of the training unit. I wish to thank IITA for the fellowship, financial support for the laboratory work, the training and facilities made available to me. I wish to express my sincere gratitude to Dr. S. W. Alhassan, Director General of CSIR for his support, and encouragement, Dr. J. A. Otoo, Director CRI and Dr. Bafour Asafo-Agyei, Scientist IITA for their encouragement and motivation. I also wish to thank Dr. Oppong Konadu and the entire staff of CRI, Pokuase, for their support and assistance for the field work. My special thanks go to the Rev. and Rev. Mrs. Sam Korankye Ankrah, General Overseer of Royalhouse Chapel International and his family for the spiritual support and taking care of my children while studying in IITA, Nigeria. I also wish to thank all friends of RCI who assisted me in diverse ways, especially Deacon Curtis Buckman and Minister Frank Amokwando Parkes. I wish to thank my mother Mrs. Alabaster Bediako for her prayer support. Great thanks also go to my brother Mr. K. A. Bediako and my nanny, Miss Cynthia Frimpomah for supporting my children while in IITA Nigeria. Finally I wish to thank the Almighty God for His grace and preservation during the period of study in IITA, Ibadan Nigeria. To God be the glory, honour and power forever more. ACKNOWLEDGEMENTS Ill University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS Page DECLARATION 1 DEDICATION 11 ACKNOWLEDGEMENTS HI TABLE OF CONTENTS IV LIST OF ABBREVIATIONS VI LIST OF TABLES VII LIST OF FIGURES VIII ABSRACT IX CHAPTER ONE 1 INTRODUCTION 1 CHAPTER TWO 2. LITERATURE REVIEW 3 2.1 Origin and history of spread of cassava in Africa 3 2.2 Cassava plant botany 5 2.2.1 Root and shoot system 6 2.2.2 Leaves 8 2.2.3 Inflorescence 8 2 .3 Classification of cassava varieties 9 2.4 Morphological characterisation 11 2.5 Molecular characterisation 12 2.5.1 DNA based molecular techniques 14 2.5.2 PCR technique 14 2.5.3 Simple sequence repeats (SSR) technique 16 2.5.4 RAPD technique 17 2.5.5 Use of RAPD in marker assisted selection (MAS) 18 2.5 . 6 Studies using RAPD markers in cassava 19 2.5.7 Calculation of genetic distances in cassava 20 CHAPTER THREE 3. MATERIAL AND METHODS 22 3.1 Morphological characterisation 22 3.1.1 Planting Materials 22 3.1.2 Field establishment 22 3.1.3 Assessment of morphological characteristics 25 IV University of Ghana http://ugspace.ug.edu.gh 3.2 Molecular characterisation 26 3.2.1 Lvophilised leaf samples 26 3.2.2 Screen house establishment 27 3.2.3.1 DNA extraction using CTAB 27 3.2.3.2 DNA extraction using Qiagen kit 28 3.2.4 Primers 29 3.2.5 RAPD-PCR conditions 30 3.3 Data analysis 31 3.3.1 Analysis of morphological data 31 3.3.2 Analysis of RAPD data 31 CHAPTER FOUR 4. RESULTS 4.1 Variability and heritability of morphological traits 33 4.1.1 Variation in qualitative traits 33 4.1.2 Variation in quantitative traits 3 7 4.1.3 Heritability of quantitative traits 39 4.2 Associations among quantitative morphological traits 40 4.3 Clustering of cassava accessions 40 4.3.1 Morphological characterisation using quantitative traits 40 4.3.2 Morphological characterisation using qualitative traits 47 4.4 Molecular characterisation 53 CHAPTER FIVE 5. DISCUSSION 60 5.1 Morphological variation and heritability estimates 60 5.2 Genetic variation by RAPD markers 63 CHAPTER SIX 6 . CONCLUSION 65 7. LIST OF REFERENCES 67 APPENDIX Appendix I List of reagents 80 Appendix II Names of accessions and their clusters as per canonical discriminant analysis using qualitative and quantitative traits 81 Appendix III Jaccard similarity coefficient of pairwise distance matrix from RAPD data. 84 V University of Ghana http://ugspace.ug.edu.gh LIST OF ABBREVIATIONS ACMV African Cassava Mosiac Virus CIAT Centro International de Agriculture Tropical CTAB Cetyltrimethylammonium, bromide DNA Deoxyribonucleic acid DNTP’s Deoxyribonucleotides Triphosphates EDTA Ethylenediaminetetraceti acid IBPGR International Board Plant Genetic Resources IITA International Institute of Tropical Agriculture Kb Kilobases (lkb=103 base-pair) CRI Crop Research Institute CSIR Council for scientific and Industrial Research Bp bases pair MAS Marker-Assisted selection PCR polymerase Chain Reaction QTL Quantitative trait loci RAPD Random amplified polymorphic DNA RFLP Restriction fragment length polymorphism SSR Simple sequence repeat TAE Tris-acetate-EDTA buffer Tris Tris(hydroxymethyl)-aminoethane U One unit of Taq DNA polymerase enzyme that will incorporate 10 nmoles of dNTPs into insoluble material per 30 minutes at 74°C under standard analysis conditions. PVP Polyvinylpoly-pyrrolidone M Meter L Litre SAHN Sequential, agglomerative, hierachical and nested UPGMA Unweighted pair group arithmetic means algorithm MAP Months after planting WAP Weeks after Planting VI University of Ghana http://ugspace.ug.edu.gh 23 26 30 37 38 38 39 41 42 44 45 48 49 50 53 59 LIST OF TABLES Cassava accessions used for morphological and molecular characterisation Descriptor list of some morphological traits and their scales Oligonucleotide primers used for RAPD analysis of the cassava accessions Mean squares for plant establishment, number of leaf lobes, plant stand and plant height (quantitative characters) measured in 1 0 0 accessions of cassava Mean squares for size of storage root, number of tubers yield in weight and harvest index among cassava accessions Mean squares for number of nodes to first branching, percentage foliage, HCN content, ACMV and CBB severity scores at 6 MAP measured in 100 accessions of cassava Genotypic (cr2g) and phenotypic (a2p) variances and variance ratios of quantitative agronomic characters in cassava. Correlation analysis of quantitative characters in cassava Eigen values and loadings from principal component analysis of quantitative agronomic traits in cassava Distribution of accessions of cassava into the different clusters based on quantitative characters Cluster means of cassava accessions for quantitative agronomic traits Eigen values and loading from principal component analysis of qualitative agronomic features in cassava accessions Distribution of 100 accessions of cassava into different clusters based on their qualitative traits Means of clusters of cassava accessions associated according to qualitative agronomic traits in cassava Ten oligonucleotide primers used for the RAPD analysis of the cassava accessions showing the amplified and Polymorphic fragments List of core collection of cassava accessions for conservation VII University of Ghana http://ugspace.ug.edu.gh 4 7 34 34 35 35 36 36 46 51 52 54 55 57 58 LIST OF FIGURES Summary of cassava diffusion in Africa Morphology of cassava plant The distribution of petiole colour in 100 cassava accessions The distribution of stem colour in 100 cassava accessions The distribution of colour of fully expanded leaf in cassava accessions The distribution of root surface colour in cassava accessions The distribution of anthocyanin pigmentation in cassava accessions The distribution of flowering and non flowering types among 1 0 0 cassava accessions First three variables from canonical discriminant analysis of 1 0 0 accessions of cassava based on quantitative traits First three axes from canonical discriminant analysis of 1 0 0 cassava accessions based on qualitative traits First three axes from canonical discriminant analysis of 1 0 0 accessions of cassava based on quantitative and qualitative traits Amplification products from RAPD analysis using primer OPJ20. Amplification products from RAPD analysis using primer OPR9 Dendrogram generated by UPGMA cluster analysis of pairwise distance data for 96 accessions of cassava using RAPD A map of Ghana showing sites where the cultivars were collected VIII University of Ghana http://ugspace.ug.edu.gh ABSTRACT The genetic diversity in 100 cassava accessions collected from eight regions of Ghana was assessed using morphological and molecular characterisation. Both qualitative and quantitative morphological traits were used to assess variability in the accessions. Canonical discriminant analysis of morphological data showed that quantitative traits detected more variability than qualitative traits. A set of 1 0 random sequence 1 0 mer oligonucleotide primers selected from among eighty that were screened detected polymorphisms and generated 63 amplified DNA bands. Cluster analysis based on the unweighted paired group method with arithmetic averages (UPGMA) categorised the accessions into eight groups. Some accessions with the same local name were put in the same cluster (Ankra, Bankye boodee, Bosomnsia and Steer bekum driver) and some duplicate accessions from source were clustered into different groups. RAPD markers revealed genetic variability and estimated the genetic distances, which facilitates identification of diverse parents in order to maximise the expression of heterosis. Accession Kav90004 ‘Trailasko’ showed the highest genetic distance of 8% from Eop8903 ‘Katawire’ and Dmgk05 ‘Bani bisa’ 9% Jaccard similarity coefficient. Eop9802 'Yebesi' and Eop9801 were the most closely related of 87% Jaccard similarity coefficient. Seventeen representative accessions (15%) were selected and kept as core collection for future work. There was no relationship between geographical diversity and genetic diversity. IX University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE 1. INTRODUCTION Cassava (Manihot esculenta Crantz) is a major source of energy for more than 500 million people in Africa (Jennings and Hershey, 1985). It provides over 50% of the calorie requirement of over 200 million people in sub-Saharan Africa (Osiru et al., 1996). The main nutritional component of cassava is carbohydrate, which is derived from starch accumulated in the tuberous storage roots. The storage roots also contain small amounts of proteins ranging from 1 -2 % on fresh weight basis. The leaves and tender shoots are consumed as a vegetable in many parts of Africa, and provides 7 g protein per 100 g edible portions with high lysine, minerals and vitamins (Hahn, 1989; IITA. 1990). In Ghana, cassava ranks first in both the area under cultivation and utilisation (NARP, 1994). The crop plays an important role in Ghana’s economy; it contributes 22% of the agriculture gross domestic product. Al-Hassan (1993) reported that cassava has a tremendous potential as feed for animal production. Other industrial uses are in food processing, pharmaceutical, paper, and textile. Its production, processing and utilisation offer jobs to large communities, especially women in the tropical countries (Thro et al., 1995). Although cassava is well integrated into the diverse traditional farming systems, very little genetic improvement has been achieved, because cassava planting materials have been selected and distributed by subsistence farmers (Beeching et al., 1993). Farmers have selected genotypes that best fit their needs and, thus, have generated a large number of traditional varieties. In addition, different ethnic groups have contributed to selection, thus leading to numerous vernacular names given to the same varieties according to ethnic groups (Mignouna et al., 1998). This nomenclature has led to confusion in the exact numbers and identity o f cassava l University of Ghana http://ugspace.ug.edu.gh varieties under cultivation in Ghana. There is the need to characterise the national collection o f cassava, to remove possible duplications. A number of different approaches have been applied in characterising genotypes of crop plants. These include morphological, and molecular approaches. Characterisation based on morphological traits can be highly subjective and environmentally dependent. Molecular markers based on protein and isozyme analyses show low-degree of polymorphism and are influenced by the physiological status and environment of the plant. To avert these difficulties, advances in molecular biology have introduced an alternative for genotype identification that uses DNA markers (Paterson et a i, 1991). Biodiversity studies using Polymerase Chain Reaction (PCR) based tools have offered a versatile and reliable method of generating polymorphism. The identification and characterisation o f diversity can reveal the presence of very useful genetic markers which could be introgressed into the crop. Germplasm characterisation reveal duplicates, show genetic relationships among cultivars, therefore decreasing the years for improving or developing new cultivars. Morphological methods have been used to characterise some of the cassava accessions collected and assembled in the gene banks in Ghana. Morphological characterisation is limited in the ability to reveal polymorphism. In order to generate useful and comprehensive information on the cultivars for future breeding work, morphological and molecular variation in the collection need to be studied. The objectives for this study were to: ( 1 ) investigate genetic diversity among 1 0 0 cassava cultivars using morphological markers and Random Amplified Polymorphic DNAs (RAPDs) and (2 ) identify distinct cassava genotypes to conserve as core collection for future distribution and breeding programmes. 2 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO 2. LITERATURE REVIEW 2.1 Origin and history of spread of cassava in Africa. Fregene et al. (1994) reported that all species of the genus Manihot are native of the New World tropics and occur naturally only in the Western Hemisphere between the southern USA (33°N) and Argentina (33°S). Reports indicate that cultivated and wild species were introduced in the old world tropics in the 16th Century (Jones, 1959) by Portuguese explorers when they established forts, trading stations and settlements on African coastal and nearby islands (Carter et al., 1992). The first cultivation of cassava in Africa dates from 1558 (Carter et al., 1992). The crop spread throughout Africa by various mechanisms of which the initial contacts with the Portuguese-Brazilian culture appear to be the most remarkable. Cassava spread became possible by river and overland trade by the merchandise, and by mass migration (Carter et al., 1992). In the 19th and 20th centuries, the colonial administrators promoted its spread and increased cultivation. Cassava possesses botanical characteristics that enable it to compete well with weeds and therefore can survive under various conditions o f neglect. The ability to regenerate from cuttings may have enhanced the spread of the crop to many new locations (Carter et al., 1992). Figure 2.1 presents a summary of the diffusion of cassava from historical documents and travelogues. Cassava was introduced at a number of points along the West African Coast during the 17th Century from the Gambia River by the Portuguese. Cassava has been cultivated around Accra since 1785. In Benin, cassava was at first used as a medicine for the cure of tuberculosis (Carter et, al., 1992). 3 University of Ghana http://ugspace.ug.edu.gh I 8mn. in HosstH >98/ I 9ras Cotiea. " Rossei 196? j Dapp«r in Rossel '982 4 Jow s . 1959 5. HamiS, I Ml 6 Jortcs. 1959 ? Bossti. 198-' B. Wood. 1 * 6 9 Mnuteri 1949 ?U fiarm hi Rossei.i 33? It Paseh ’ 98C 12 flesscl 1967 13 Rossei 1987 14 W j i ic a Jo re s 1359 15 jw igs to ne . 0 Aitaitso'i, ■!' Jones. ! 959 f ' i i l i . .5 Jones, U59 22 Wigtwldus, 1984 23. Daeei. m Wigboldus 1984 24 WighoWus, 1990 25 A$ootg. 1968 26 AgbooU. 1968 27 Hypoitestsed 38 Adlers, 1975 ?9. Kent, 1969 JO, Raison !972 31 Krapt m jones, 1959 32 l ivm ^ lo re , mJones, 1959 33 Liwngslone. in Jones, 1959 34 Langlands 1966 35 Cameron. n Jonas, I959 3b Langlands, 1966 37 Stanley, in Jones 5959 33 Langiands. 1966 39 Langiands. 1966 40 Kamanzi 1983 41 Meyer 1984 Fig 2.1 Summary of cassava diffusion in Africa (Carter et al., 1992) 4 University of Ghana http://ugspace.ug.edu.gh Initial attempts at improving cassava in Ghana were made by the Crops Research Institute at Kwadaso, Kumasi, where introduced cultivars, particularly accessions from East Africa were screened and elite clones such as K357, K102 and K680 were hybridised with local cultivars to generate high yielding and disease resistant clones (Doku, 1969). In collaboration with the International Institute o f Tropical Agriculture (IITA), a number o f introductions have been tested and screened to identify cultivars of high and stable yield across different agro- ecological zones. These include ‘Abasafitaa’, ‘Blemoduade’ and ‘Afisiafi’. Several genotypes are being evaluated for future release (NARP, 1994). 2.2 Cassava Plant Botany Cassava is a perennial crop. Most farmers usually harvest within the first two years. Abandoned stands of cassava may continue to grow for several years (Onwueme, 1978) and most often in association with other crops ( Hershey, 1993a). Cassava is propagated from stem cuttings, however, under natural conditions it can be propagated from seeds. Under favourable conditions cuttings planted in moist soil, sprout and root within one week. Seed propagation result in relatively slower plant establishment with smaller and weaker plants. The seedlings genetically segregate into different types (Ekanayake et al., 1997). A few weeks after sprouting, the shoot lengthens and the roots extend downwards and spread. Flowering may begin as early as the sixth week after planting, but this depends upon the cultivar and the environment. Storage roots formation begins about eight weeks after planting. Leaf area approaches its maximum size between 4-5 months depending on planting time. The average height of a cassava plant ranges from one to two meters; some cultivars however reach four meters (Ekanayake et al., 1997). 5 University of Ghana http://ugspace.ug.edu.gh 2.2.1 Root and Shoot Systems The cassava plant can be divided into the shoot and root system (Figure 2.2). The shoot system consists of the stem, leaves, inflorescence and the root system with feeder roots and storage roots. The cassava propagated by stem cutting (hard wood), develop adventitious roots at the base o f the cuttings during the first two to three weeks. The adventitious roots subsequently develop into a fibrous root system, which absorbs water and nutrients from the soil. Nodal roots, which are adventitious roots that develop at the base of the auxiliary buds on the cuttings, are at the nodes. The fibrous root system may reach 200 cm or more in length (HTA, 1990). Cassava plants propagated by seeds first develop a tap root system. After 30 to 60 days, some roots increase in diameter and become tuberous storage roots. As tuberisation proceeds, the tuberous roots are developed as the result o f cambium activity and starch accumulation. Tuberous roots are physiologically inactive and cannot absorb water or nutrients, however, the rest of the fibrous roots continue to function and absorb water and nutrients. The number o f roots which form storage tuberous roots depends on several factors including genotype, assimilate supply, photoperiod and temperature (Ekanayake et al, 1997). Cassava stems may grow up to four meters in height, whereas dwarf varieties may be one meter tall. The stems vary considerably in colour and usually are woody with very large pith. The older parts of the stem consist of prominent knob-like scars, which indicate nodal position where leaves were originally attached. Each nodal unit consists of a node, which subtends a leaf and an intemode. The rate of node production on each stem is about one node per day during early and active growth stages, and about one node per week in the older plants. Intemodes vary considerably depending on varieties and environmental conditions. They tend to be long under favourable conditions, and short under drought stress, and with insufficient light, they are usually abnormally long (Ekanayake et al., 1997). Two types of branching patterns exist in most 6 University of Ghana http://ugspace.ug.edu.gh in flo rescence reproductive branch ing (forking) lateral b ranch ing fib ro u s roots m ain stem s tuberous roots fruits node Fig 2.2: Cassava plant (Ekanayeke et al., 1997) 7 University of Ghana http://ugspace.ug.edu.gh varieties growing under normal conditions, namely, forking and lateral branching. The type of branching is influenced by several factors, including genotype and physical damage. 2.2.2 Leaves The number of cassava leaves, leaf longevity and whole plant products are determined by genotype and environmental conditions. Cassava leaves are arranged alternately in a spiral order on the stem. The phyllotaxis or arrangement of leaves on the stem is 2/5 spiral. Cassava leaves are simple, with lobed lamina and petiole. Each leaf is subtended by three to five stipules, each about one cm long. The number of lamina lobes varies between three and nine. Most cassava varieties grown in Africa have lobes that are elliptical or lanceolated (Onwueme, 1978). 2.2.3 Inflorescence Cassava is a monoecious plant. Flowering is frequent and regular in some cultivars, while in others it is rare or non-existent. The flowers are borne in terminal panicles, with the axis of the branch being continuous with that of the panicle inflorescence. The male flowers occur near the tip, while the female flowers occur closer to the base. Each flower, whether female or male has five yellowish or reddish perianths. The female flower opens first, while the male flower opens about a week later. Cross-pollination is usually the rule. Self pollination can occur when female and male flowers, located on different branches of the same plant open at the same time (Ekanayake et al., 1997). After pollination and fertilisation, the ovary develops into a fruit that matures in 70 to 90 days. The mature fruit is a globular capsule (diameter 1 to 1.5 cm) with six narrow longitudinal wings along which it naturally splits open when dry. The woody endocarp contains three locules, each with one seed. When the fruit is dry, the endocarp splits explosively to release the seed. The cassava seed shape is ellipsoidal and about 1.5 cm long. It has a brittle testa that is grey and mottled with dark blotches. 8 University of Ghana http://ugspace.ug.edu.gh 2.3 Classification of Cassava Varieties Numerous cassava varieties exist in each locality where the crop is grown. The cultivars have been distinguished by morphological characteristics such as leaf characteristics, colour and shape branching habit, plant height, colour o f stem, tuber shape and colour, time to maturity yield and the cyanogenic glucoside content in the roots (Dixon et al., 1994) Genetic studies on the genus Manihol has shown potential benefit to breeding for the improvement of quality traits (Beeching et al., 1993). Local landraces of crops offer a rich source of genetic diversity and these provide a valuable source of genetic material for crop improvement. However, the rapid development of high-yielding cultivars threaten such crops and efforts for their preservation should be of the highest priority (Attere, 1997). Cyanogenic glucoside potential or content as a trait has been used to place cassava cultivars into three groups: (i) cassava with high potential to generate HCN-10 mg per 100 g fresh weight or more; (ii) Intermediate types in which the levels o f HCN range between 5 and 10 mg per 100 g fresh weight (iii) cassava with low potential to generate HCN - less than 5 mg per 100 g fresh weight. The cyanogenic glucosides are often concentrated in the peel (Rao and Hahn, 1984; IITA, 1990). The breeding strategies in cassava are strongly influenced by its vegetative propagation, allowing the fixation of heterozygous genotypes at any stage of selection (Hershey, 1993b). Cassava, Manihot esculenta Crantz (Synonymous with Manihot ultissima Pohl) (Onwueme, 1978) is a dicotyledonous crop, belonging to the Euphorbiaceae family and genus Manihot. The genus includes 98 other species that are useful as gene resources in cassava improvement (Rogers and Appan, 1973; Fregene et al., 1994). The Euphorbiaceae family has members characterised by lactiferous vessels composed of secretory cells and include several commercial plants. Some of these are rubber trees (Hevea brasiliensis), oil plants (Ricinus 9 University of Ghana http://ugspace.ug.edu.gh comunis), root crops (Manihot spp) and ornamental plants (.Euphorbia spp.) (Ekanayake el al., 1997). Cassava is widely distributed in the tropical and subtropical areas. It is the only species from the genus that is widely cultivated. A few other Manihot species that have minor uses, especially as alternative sources of latex for rubber production are M. glaziovii and M. caerulescenc (Franche etal., 1991). 2.3.1 Cytology of Cassava Cytogenetics has been used as a basis for understanding the organisation o f genetic diversity in many genera including Manihot. Chromosome morphology and behaviour has shown cassava as a diploid species with 2n=2x=36 chromosomes (Hershey, 1993b). There is paucity of information on the synaptic behaviour of chromosomes in Manihot esculenta, wild Manihot species or interspecific hybrids. On the basis o f the observed number of satellite chromosomes and their behavior during the division stages, it has been postulated that cassava is a segmental allotetraploid with a basic chromosome number of X=9 (Perry, 1943; Magoon et a/., 1969a). However, analysis with ten isozyme loci revealed predominantly disomic inheritance suggesting cassava to be a diploid (Hussain et al, 1987; Lefevre and Charrier, 1993). Gomez et al. (1994) confirmed disomic inheritance in cassava based on recent studies using arbitrary PCR markers. Cassava is genetically the least understood among the major staple crops that are used for food production (Gomez et al., 1996). Apparently less than fifty percent o f the species within the genus Manihot have been examined. Bai’s (1987) account on meiotic studies, and information on the family pMphorbiaceae, suggest irregularities at cell division stage. Polyploidy manipulation in cassava has been possible using colchicine to produce artificially induced somatic tetraploids (2n=4x=72) Hahn et al., (1990). They did not show much variation and had undesirable characteristics such 10 University of Ghana http://ugspace.ug.edu.gh as low yield and stunted growth. Hahn et al. (1990) reports on occurrence o f spontaneous sexual tetraploids and triploids resulting from diploid interspecific crosses with cassava. 2.4 Morphological Characterisation Morphological traits have been used as a powerful tool in the classification of cultivars as well as study their taxonomic status (Roger and Appan,1973). Certification of new cultivars is usually based on the genetic purity of a particular crop. However, traditionally those assessments depend on the botanical traits. Breeders and geneticists have used morphological characteristics such as leaf and flower attributes to follow segregation of genes and hybrids, but most agronomic traits are not associated with easily observable phenotypic markers (Kochert, 1994). Most of the descriptors are ambiguous, and have limited use for cultivar identification (Stegemann, 1984). Such characteristics are often controlled by multiple genes and subject to varying degrees of environmental modifications and interactions. Many o f these traits are also difficult to analyse because they do not have the simple genetic control assumed by many populations in genetic models and they are of very little use (Tanksley et al., 1989). Morphological characterisation has been used to identify duplicates, study genetic variation patterns and correlation with characteristics of agronomic potential. The use o f morphological traits involve a lengthy survey of plant growth over time that is costly, labour intensive and vulnerable to environmental conditions. Variation in traits do not reflect only the genetic constitution of the cultivar, but also the interaction of the genotype with the environment (G x E) within which it is expressed (Dixon and Nukenine, 2000). In cassava breeding programs most emphasis has been on the collection and conservation o f gene pools (CIAT, 1993) and characterisation of the collections. Morphological data on cassava have been useful in correcting ambiguities of cultivars. The IBPGR (International Board for Plant Genetic Resources, 1983) descriptors have been used to 11 University of Ghana http://ugspace.ug.edu.gh characterize cultivars. The IBPGR has identified a set o f relatively stable morphological traits useful for characterisation of cassava genotypes. The descriptors include qualitative and quantitative measurements for cassava root and shoot characters. Since the past decades, a number of alternative laboratory methods have been successfully developed, such as isozyme analysis (Nienhuis et al., 1995), seed storage protein electrophoresis and high performance liquid chromatography (Buehler et al., 1989). 2.5 Molecular characterisation The main constraint to morphological characterisation is the limited amount of polymorphism they are able to detect among closely related genotypes (Hu and Quiros, 1991). Isozymes often show low levels o f polymorphism and problems with reproducibility arise due to tissue type and conditions. A large number o f polymorphic markers are required to measure genetic relationships and genetic diversity in a reliable manner (Soller and Beckmann, 1983; Mignouna, 1994). This limits the use of morphological characters and isozymes. Advances in molecular biology have introduced DNA based procedures for cultivar identification. DNA sequences show greater variation than amino acid changes in isozymes. The composition of DNA is also consistent between tissues and is not affected by environmental changes (Beeching et al., 1993). Target genes in a segregating population can be identified with the assistance of DNA markers so as to accelerate traditional breeding. (Thottappilly et al., 2000). One extensive use of these molecular markers is for development of detailed genetic and physical chromosome maps in a variety of organisms among animal and human systems and among plant systems. Molecular markers in plant systems have been found very useful in conventional breeding by carrying out indirect selection through molecular markers linked to the traits of interest. Molecular markers have been used for both simple and quantitative trait loci (QTL) because the environment does not influence these markers and can be scored at all stages 12 University of Ghana http://ugspace.ug.edu.gh of plant growth. In addition to these two major applications, DNA markers can also be used in plant system for germplasm characterisation, genetic diagnostics, characterisation of transformants, study of genome organisation and phylogenetic analysis, (Rafalski et al., 1993). Although each marker system has some advantages and disadvantages, the choice o f any marker system is dictated to a large extent by the intended application, convenience, cost and time consideration, number of samples, how quickly the data is needed and the technique that will best yield the maximum data (Gupta et al., 1999;Thottappilly et al., 2000). Molecular markers can be broadly grouped into: 1) Hybridization-based DNA markers such as restriction fragment length polymorphisms (RFLPs) and oligonucleotide fingerprinting. 2) PCR-based DNA markers such as random amplified polymorphic DNAs (RAPDs), which can also be converted into sequence characterised amplified regions (SCARs), simple sequence repeats (SSRs) or microsatellites, sequence-tagged sites (STS), amplified fragment length polymorphisms (AFLPs), inter-simple sequence repeat amplification (ISA), cleaved amplified polymorphic sequences (CAPs) and amplicon length polymorphisms (ALPs). 3) DNA chip and sequencing-based DNA markers such as single nucleotide polymorphisms (SNPs) 4) In addition to the above three groups of markers, microsatellite-primed polymerase chain reaction (MP-PCR), arbitrarily- primed PCR (AP-PCR), allele-specific PCR (AS-PCR) and DNA amplification fingerprinting (DAF) have also proved useful in the detection of polymorphism (Gupta e ta l , 1999; Thottappilly etal., 2000). 13 University of Ghana http://ugspace.ug.edu.gh 2.5.1 DNA-based molecular techniques DNA fingerprinting is a technique, which has been widely adopted to differentiate among organisms at the species and subspecies levels (McClean et al., 1994). The techniques used for cultivar identification are designed to detect the presence o f specific DNA sequences or combination of sequences that uniquely identify the plant. Cultivar identification can be achieved more accurately using DNA fingerprinting data, especially in materials characterised by high genetic variation between cultivars. The most closely related cultivars are usually distinguished with the DNA fingerprinting methods (Beckmann and Soller, 1986). Another advantage of DNA fingerprinting over morphological markers is the dominance and the absence of environmental effects that are shared. The application of DNA fingerprinting could be very valuable in the identification of cultivars and species and could help to create more efficient breeding programs through the detection of genetic linkages between DNA fingerprinting bands and agriculturally important quantitative trait loci (QTL). The high variability o f DNA fingerprinting described in humans, animals and plants allows the identification of different individuals, genotypes, and species (Lin et al., 1993). 2.5.2 PCR Technique The polymerase chain reaction (PCR) Saiki et al. (1988) has been the basis o f a growing range of new techniques for genome analysis based on the selective amplification of genomic DNA fragments. Williams et al. (1990) reported on the use o f PCR with short oligonucleotide primers of arbitrary (random) sequence to generate markers, the basis of the random amplified polymorphic DNA (RAPD). Welsh and McClelland (1990) also reported on arbitrarily primed polymerase chain reaction (AP-PCR) while Caetano-Anolles et a l.( 1991) report on DNA amplification fingerprinting (DAF). 14 University of Ghana http://ugspace.ug.edu.gh The PCR reaction requires deoxynucleotides to provide both energy and nucleotides for synthesis of DNA, DNA polymerase primer, template and buffer containing magnesium (Taylor, 1991). Typical PCR amplification utilises oligonucleotide primers that hybridise to opposite strands. The product of DNA synthesis of one primer serves as template for another primer. The PCR process requires repeated cycles of DNA denaturation, annealing and extension by DNA polymerase leading to amplification of the target sequence. The result is an exponential increase in the number of copies o f the region bounded by the primer (Saiki et al., 1988; Mullis,1990). The technique can be applied to detect polymorphism in various plants, animals, bacterial species and fungi (Williams etal., 1990). The introduction o f the PCR technique has revolutionized standard molecular techniques and has allowed for the proliferation of new tools serving to detect DNA polymorphism (Hu and Quiros, 1991). The electrophoresis pattern of fragments generated by each primer for one isolate can be used as DNA fingerprints for assaying diversity (Tommerup et al., 1995) Polymorphism between two individuals is generally scored as a presence or absence (non-amplification) of a particular DNA fragment. The absence may result from deletion of a priming site or insertion rendering site too distant for successful amplification. Insertion can change the size of a DNA fragment without preventing its amplification (Williams et al., 1990). PCR is simple, fast, specific, sensitive and relatively low cost. The main advantages of this technique over other techniques are its inherent simplistic analysis (a single reaction can contain all reagents) and the ability to conduct PCR test with extremely, small quantities of tissue for DNA extraction (Welsch et al., 1991). On the other hand PCR is limited in its usefulness because of the time and cost required to obtain the DNA sequence information required for primer design (Samec and Nasinec, 1995;Thottappilly et al., 2000). 15 University of Ghana http://ugspace.ug.edu.gh 2.5.3 Simple Sequence Repeats (SSR) technique Simple sequence repeats are usually 1 - 6 base pair repeat motifs (example TA, CA, GTG, TAA GAT A) repeats. They are ubiquitous in eukaryotic genomes and their study has been greatly facilitated by recent advances in PCR technology. The first report o f microsatellites in plants was made by Condit and Hubbel (1991), who suggested their abundance in the plant system, and subsequently confirmed by Gupta et al. (1996). Later Akkaya et al. (1992) reported on the length of sequence and polymorphisms o f SSRs in soybean. Simple or short sequence repeats also known as short tandem repeats STRs or ‘microsatellites’ or incomplete amplification using primers to the sequences flanking these repeats can be used to generate polymorphism because of frequent variation in the length of the repeat regions. The simple sequence repeats has been reported as useful markers in many plant species including cassava (Maroof et al., 1994; Roder el al., 1995). Chavarriaga-Aguirre et al. (1999); Agyare-Tabbi et al.(\991) reported the presence of some useful repetitive and microsatellite DNA in cassava. The main limitation with the application of this technique in cultivar identification is the difficulty in cloning and sequencing the regions flanking the SSR. This must be done for each species since the flanking regions are relatively species specific and are not usually useful for application to even closely related species (Agyare-Tabbi etal.,1997). In some instances SSR primers may not reveal any or detect low levels of polymorphism (Taylor et al., 1992; Agyare-Tabbi et al.,1997). Notwithstanding, the reliability and reproducibility o f the markers, especially between laboratories, makes them attractive alternatives to other techniques such as RAPD and PCR. The SSR sequences are mostly found in introns and the 5’ flanking regions of plant genes (Roder et al., 1995). 16 University of Ghana http://ugspace.ug.edu.gh 2.5.4 RAPD technique Another PCR-based technique in use is the random amplified polymorphic DNA RAPD markers are generated by the use of short ( 1 0 -mer) synthetic oligonucleotides in single strand primer (Williams et al., 1990). In this technique a decamer primer o f arbitrary sequence is allowed to anneal at a relatively low temperature priming the amplification of DNA fragments distributed at random in the genome (Wiliams et al., 1990). Amplification products are visualised by separation on agarose gel and stained with ethidium bromide. They usually result in DNA fragment patterns that are polymorphic between genotypes, therefore detecting diversity within them (Tommercup et al., 1995). The main issue associated with its use is ensuring reproducible amplification profiles. The nature of amplification process with short primers is that many sites in the genome are potential templates and the profile may be influenced by any variation in the method used to prepare DNA template and the exact reaction composition (Muralidharan and Wakeland, 1993). However, a key requirement for reliable and reproducible RAPD results is a consistent approach to sample preparation and DNA isolation. There are several advantages of RAPDs compared to other DNA based techniques. It is simple, rapid and does not involve radioactivity and costs less (Varghese et al., 1997). Moreover it does not require target DNA sequence information and can provide markers in genomic regions not accessible by other analysis (Williams et al, 1990). The technique uses a very small amount of genomic DNA (few hundred nanograms) which allows the analysis of single seeds or young seedlings (Hu and Quiros, 1991). Another advantage of the RAPD method is that a universal set of random primers can be used for genomic analysis o f any organism (Welsch and McClelland, 1990) Short random primers have been used to reproducibly amplify segments o f genomic DNA from a wide variety of species including plants (Williams et al., 1990; Quiros et al., 1991, Fregene et al., 1997). Polymorphism detected by RAPDs are inherited in a mendelian fashion as 17 University of Ghana http://ugspace.ug.edu.gh dominant markers (William et al., 1990; Welsch et al., 1991). This polymorphism has been proved to be useful for identifying variation at different levels. RAPD analysis enables differentiation between very closely related organisms due to the high resolution o f the technique (Tommerup et al., 1998). The polymorphic fragments generated by RAPDs are useful as genetic markers to identify organisms (Williams et al., 1990) and the relative degree of similarity between individual populations and species (Yang and Quiros, 1993; Tonukari et al., 1997). In cassava, RAPD technology has been used to assess genetic diversity in collections, relationship between Manihot species, develop linkage maps and for studies of evolutionary relationships (Marmey et al., 1994; Fregene etal., 1995) In studies comparing molecular markers there have been similarity in results using other DNA based techniques and RAPDs in Manihot esculenta (Beeching et al., 1993; Marmey et al, 1994), thus concluding that any of the techniques can be used for the evaluation o f genetic diversity. 2.5.5 Use of RAPD in marker assisted selection (MAS) Molecular markers used as a tag for various traits have been proposed and used in breeding programs. This approach relies upon the establishment of a linkage between a molecular marker and the characteristic to be selected. Once this has been achieved, breeding selection can be conducted in the laboratory and does not require the expression of the associated phenotype (e.g. disease resistance). In tomato (Lycopersicum esculenta) RAPD markers for genes conferring resistance to mosaic virus were identified from the wild relative L. peruvianum located on the long arm of chromosome 9 (Ohmori et al., 1995). RAPDs has been used effectively for cultivar identification and screening for resistance in a variety of plant species including cassava and yam (Tonukari et a/., 1997; Mignouna and Dixon, 1997). RAPDs have been used on crops such as apple (Malus domestica Borki,) Landry 18 University of Ghana http://ugspace.ug.edu.gh et a l , 1994), asparagus {Asparagus officinalis L.), (Khand ka et al., 1996), lemon (Citrus lemon) (Deng et al., 1995), celery (Aipum agraveolens L.) (Yang and Quiros, 1993), pepper (Capsicum cmnuum), (Prince et a l , 1994), rasberry (Rubus spp), (Parent et al., 1993), sweet potato (Ipomea batatas L.), (Commoly et al., 1994), yam (Dioscorea spp), (Asemota el al., 1996), peas (Pisum sativum L ), (Samec and Masinec, 1995), wheat (Triticum aestivum), (Myburg et a l, 1996) and cocoa (Theobromin cacao L.) (Russel et al., 1993). 2.5.6 Studies using RAPD markers in cassava Tools for traditional linkage analysis and morphological markers are few in cassava. There exist no classical genetic map for cassava (Ocampo et al., 1995). Linkage analysis using isozymes has been carried out in cassava (Lefervre and Cherrier, 1993). Seventeen isozymes loci from enzyme systems produced 59 alleles organised into three linkage groups. Fregene et al. (1995) have used 200 RFLP and RAPD markers in a FI population derived from intraspecific cross between two elite cassava clones to construct a preliminary linkage map. Eight hundred primers were screened with single dose fragments in a RAPD analysis, and 42 per cent polymorphisms identified. Five restriction enzymes were employed in the parental survey, 30 per cent polymorphisms were identified. Results o f this study revealed about 5 per cent of all RFLP markers and 3 percent RAPD markers which did not segregate. About 30% were found to be lined in repulsion for both RFLP and RAPD markers. Cassava ancient polyploid, but not behaving as a diploid linkage map of cassava constructs from RFLPs microsatellites, isozymes and RAPD markers, 30% of the markers were found to be limited in repulsion, suggesting preferential paring among some chromosomes and random pairing among others (Gomez et al., 1996). RAPD markers have been used to measure genetic relationship within the genus Manihot. Reports on studies on the origin of cassava, the diversity and evolution of Manihot genus have 19 University of Ghana http://ugspace.ug.edu.gh been carried out using different species including cassava germplasm. Genetic improvement of cassava is limited by poor knowledge o f genetic diversity within the species (Marmey et al., 1994). From 98 Manihot species reported by Rogers and Appan (1973), 80 occured in Brazil, hence an important genetic resource centre for the genus. There are two hypotheses on the origin of cassava. Rogers (1963) considers cassava as a cultigen, that is a species selected by humans and not occurring naturally in the wild form. Therefore, cassava is thought to have originated through a series of hybridisation and introgression events involving several Manihot species. Allem (1987) however, supports the alternative hypothesis of wild progenitor species, having reported that wild cassava has been found in Brazil. The use of RAPDs to detect phylogenic relationship is controversial. However, RAPDs have detected fine-scale variation in Manihot, and can be used to distinguish Manihot esculenta cultivars which can help resolve ambiguous relationships such as was observed in some Brazilian Mamhot species andM esculenta (Schall etal., 1995). Tonukari et al. (1997) studied genetic polymorphism o f cassava within the Republic of Benin, Laminski et al. ( 1997) also did a similar study for cassava elite lines in South Africa and RAPD markers were found useful in both studies. 2.5.7 Calculation of genetic distances Data obtained from DNA fingerprinting can be used in two ways. The first approach in the parsimony analysis, where phenogram representing phylogenetic relationship are constructed on the basis of the lowest number of characteristic state transformation that yields a particular phenogram. Dendrogram is the second which is most widely used. Diagrams o f genetic relationships are constructed using the cluster analysis based on the pairwise genetic distance. An input data matrix containing absence (0) and presence (1) value of all RAPD markers is directly used to calculate pairwise genetic distance. 20 University of Ghana http://ugspace.ug.edu.gh Generally the fraction o f bands shared between any two observed taxonomic units is used to calculate a similarity coefficient (S) from which genetic distance (D) is derived. The genetic distance, D=l-S or D= - In (S). All possible pairwise grouping of individuals have their pairwise distance values calculated and grouped in a table of pairwise distance matrix. The index of genetic similarity (F) of Nei and Li (1979), among other formula has been used in most of the studies and is suitable to calculate the pairwise distance matrix from RAPD data and SSR. This formula F=2 (nx.y)/nx+ny was developed for RFLP data where F (similarity) is the ratio of shared bands between individuals X and Y, 2nxy is the number o f shared bands, and nx and % are the numbers of bands observed in cultivars x and y respectively. The F ratio is used to calculate the distance value (D=I-F) for a pairwise combination. Jaccard’s similarity coefficient (Jaccard, 1908), simple matching and Rogers’ distance (Rogers, 1972) are other coefficients which have frequently been used to calculate genetic distance. Computer software programs include Numerical Taxonomy and Multivariate Analysis program package (Rohlf, 1993), RAPD distance, phylogeny inference package and phylogenetic analysis using parsimony (Swafford, 1991). These programs can be used to calculate distance matrixes using the formula F=2 (nxy)/nx+ny. The resultant data are processed with cluster analysis using methods such as unweighted pair group mean arithmetic analysis (Sneath and Sokal, 1973) and then plotted in dendrograms representing the genetic relationship among the genotypes in the pairwise genetic matrix. 21 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE 3. MATERIALS AND METHODS An experiment was conducted to characterise 100 cassava accessions using morphological and molecular traits. There were two phases of the study, the first part being morphological characterisation at Pokuase, Ghana and the second phase, molecular characterisation at the Biotechnology Unit, o f the International Institute of Tropical Agriculture (IITA), Ibadan, Nigeria. 3.1 Morphological characterisation 3.1.1 Planting materials One hundred cassava accessions used for the study were obtained from four institutions, namely, Plant Genetic Resources Centre (PGRC) Bunso, University of Cape Coast (UCC), Cape Coast, Savanna Agriculture Research Institute (SARI), Nyankpala and Crop Research Institute (CRI) out station, Pokuase. The name of each accession, region of collection, and source/institution are presented in Table 3.1. 3.1.2 Field establishment The experimental area was prepared by ploughing and harrowing. Cassava cuttings were planted in the field in October 1998. A 10 x 10 simple lattice design with two replications was used. Each of the accessions was represented by two plots, with eight plants per plot. A spacing of lm by lm was used. Supplementary irrigation was provided by watering during the initial four weeks. Weeding was done at 1 MAP (month after planting), 4 MAP, and 9 MAP. No fertilisers were applied, and no pest or disease control was practiced.. 22 University of Ghana http://ugspace.ug.edu.gh Table 3.1. Cassava accessions used for morphological and molecular characterisation NO. ACCESSION 1 Jk90016 D1 2 Jk90013 D2 3 Acw86 4 Jk90006 5 Jk90022 6 Acw157 7 Acw91 8 Acw82 9 Acw19 10 Acw11 11 Jk90021 12 Jk90009 13 Jk90026 14 Jk90001 15 Jk90012 16 Jk90005 17 Acw6 18 Jk90008 D3 19 Acw5 20 Acw23 21 Acw125 22 Acw126 23 Acw10 24 Acw9 25 Jka90013 D2 26 Acw2 27 Acw37 28 Acw36 29 Acw75 30 Jka90016 D1 31 Acw107 32 Jka90008* D3 33 Acw111 34 Acw16 35 Acw29 36 Acw71 37 Acw24 38 Acw62 39 Acw61 40 Jk90033 41 Jk90031 42 Mc90028 43 Mc90033 44 Mc90015 45 Mc90013 46 Mc90012 47 Mc90019 48 Mc90021 LOCAL .NAME. STEER BEKUM DRIVER* 1 ABISH KWESI TUTU KOFI ASEM ETIBIRE ANKRA * 2 STEER BEKUM DRIVER* 1 ADZOCONGO AWONA CONGO DUAFRA ESI PANYIN KASA FANTI OKWADWOFO APPIAH ADUANE ASA AKWADA PE KWESI TU MANOA BESREBEMA* 5 NIMME ESI. PAN YA KWESI ITU TABU ADUO BILE ABISH CONGO ADWOFUFULE ANUONLABEDE BESERESBEMA* 5 STEER BEKUM DRIVER* 1 ESSABAYEM MANOA BANKYE SANTUM AGRIC TABOO NEME ADROBLE BESERABEMA* 5 NDA-AYA ADAKOLA NSAWAM HALF ASS IN BANKYE MPAPRO EFUA BEYAW ADJOA BEESIWA BANKYE SANTUM AWONA BANKYE SANTUM KETEKE WUSIE SOURCE. REGION PGRC Western PGRC Western UCC Western PGRC Western PGRC Western UCC Western UCC Western UCC Western UCC Western UCC Western PGRC Western PGRC Western PGRC Western PGRC Western PGRC Western PGRC Western UCC Western UCC Western UCC Western UCC Western UCC Western UCC Western UCC Western UCC Western PGRC Western PGRC Western UCC Western UCC Western UCC Western PGRC Western UCC Wes fern PGRC Western UCC Western UCC Western UCC Western UCC Western UCC Western UCC Western UCC Western PGRC Central PGRC Central PGRC Central PGRC Central PGRC Central PGRC Central PGRC Central UCC Central UCC Central 23 University of Ghana http://ugspace.ug.edu.gh Table 3.1 continues 49 Mc90029 50 Mc90022 51 Mc90031 52 Mc90034 53 Mc90020 54 Mc90030 55 MC90036 56 Mc90010 D4 57 Mca90010 D4 58 Acc497 59 Mc90006 60 Kaa90060 61 Kaa90830 62 Kaa90067 63 Kaa90087 64 Kaa90062 65 Mc90005 66 Kae90050 67 Kae90033 68 Btl97006 69 Mc90004 70 B89003 71 B89037 72 Kae90034 73 B89031 74 Kae90046 75 B89013 76 Kaa90063 77 B89029 78 B89015 79 B89001 80 B89032 81 Eop984 82 Eop9801 83 Eop9802 84 Eop9803 85 Kav90004 86 Kav90002 87 Kav90007 88 Kav90028 89 Kav90013 90 Kav90008 91 Kav90006 92 Kav90022 93 Kav90010 94 Kav90005 95 Eop9805 96 Dmgk05 97 Dmgk12 98 Dmgk11 BANKYE BODEE* 3 GYANOA SAAKWA STEER BEKUM DRIVER* 1 SAAKWANYA GARIBANKYE PAAFIO KWAKU SAM KWAKU SAM OHYEWKAW ADIMADIN ASONA NO LOCAL NAME BETEA BASARE BANKYE BANKYE BODEE* 3 ANKRA TUAKA * 2 BANKYE SONO BOSOM NSIA BTL97006 TUAKA* 4 B89003 B89037 AGEGE B89031 BANKYE TUNTUM B89 013 AKOSUATUMTUM B89029 B89 015 IPGRU B89032 QUARANTINE BOSOM NSIA YEBESI KATAWIRE TRAILASKO AKPANYA KATAWIA SAKPA SOKLA TUAKA* 4 HWIAKPO ANKRA* 2 KPENYIVIA ADUSE AFISIAFI BIAN BASI KPLASO AGRIC PGRC Central PGRC Central PGRC Central PGRC Central PGRC Central PGRC Central PGRC Central PGRC Central PGRC Central UCC Central PGRC Ashanti PGRC Ashanti PGRC Ashanti UCC Ashanti PGRC Ashanti PGRC Ashanti PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern PGRC Eastern CRI Greater Accra CRI Greater Accra CRI Greater Accra CRI Greater Accra PGRC Greater Accra PGRC Greater Accra PGRC Volta PGRC Volta PGRC Volta PGRC Volta PGRC Volta PGRC Volta PGRC Volta PGRC Volta IITA/CRI Volta SARI Northern SARI Northern SARI Northern 24 University of Ghana http://ugspace.ug.edu.gh Table 3.1 continues 99 Dmgk02 "MADE IN DOZEN SARI Northern 10 0_______So96002 BANKE PGRC Upper East Source ;UCC University of Capecoast,CRI;Crop Research Institute,SARI; Savanna Agriculture Research Institute,PGRC Plant Genetic Resource Centre (Dl,2,3,4 : Duplicate accessions from source *1,2,3,4,5: share common local names ) Accessions in italics 5,31,39 and 77 were not used for the molecular (DNA) work. 3.1.3 Assessment of morphological characteristics Morphological data were collected from all plants of each cultivar. Data were recorded using the International Board for Plant Genetic Resources (IBPGR) descriptor list for cassava as a guide. The descriptors used for morphological characterisation are listed in Table 3.2. Data were taken from the first month after planting (MAP). Data recorded included qualitative and quantitative measurements. Qualitative traits evaluated using different scales included, colour of unexpanded leaf, pubescence of young leaves, number of leaf lobes, petiole colour, distribution of anthocyanin pigmentation, growth habit o f young stem, pubescence on young stem, stem colour, and flowering characteristics. At harvest, other traits measured included, storage root peduncle, storage root form, root constriction, position of roots, storage root surface colour, root surface texture, storage root length, storage diameter, ease of root periderm removal (outer-skin) ease of root cortex removal (inner skin), colour of outer surface of storage root cortex (inner skin) and storage root pulp colour. Quantitative traits were measured at different stages of growth of the cassava plants. Data on the severity of African cassava mosaic virus (ACMV) and cassava bacterial blight CBB were taken at 3MAP and 6MAP. Incidence was scored as the number of infected plants over the total number of plants, severity was assessed based on a scale o f 1 to 5 in increasing severity ( IBPGR). Other quantitative traits measured included; length of petiole, height at first apical branch, height of plant, number of levels of branching, number of nodes to first apical branching and number of leaf lobes. At harvest time, root yield characteristics measured were recorded as 25 University of Ghana http://ugspace.ug.edu.gh average number of roots, average fresh root weight per plant, and top weight. Hydrogen cyanide potential or content was determined using picrate acid test ( Almazan,1988). Table 3.2 Descriptor list of some morphologic traits and their scales Trait code Description Scale T i l Germination of stakes Actual proportion Tl_2 Initial vigour 3-Low 5-Intermediate 7-High Colour of unexpanded 3-Light green 5-Dark green 7-Green purple 9-Purple Tl_3 Apical leaf Colour of first fully Tl_4 expanded apical leaf 3-Light green 5-Dark green 7-Green purple 9Purple Tl_6 Leaf vein colour 3-Light green 5-Dark green 7-Green purple 9Purple Pubescence of young Tl_7 leaves 0-Absent 3-Little pubescence 5-Moderate pubescence 7-High Pubescence Position of young T i l l leaves 1-Erect 2-Horizontal 3-Deflexed 4-Retorse Angle of petiole T115 insertion 0-Not branching 3-(15-30°) 5-(45-60°) 7-(75-90°) T 117 Petiole length 0-Absent 3-Short 5-Medium 7-Long T118 Petiole colour 3-Light green 5-Dark green 7-Green purple 8-Red 9-Purple T122 Stem colour 1-Silver green 2-Light brown 3-Dark brown 4-Dark green Number of levels T127 of branching Actual levels (at harvest) T128 Angle of branching 0-Not branching 3-(15-30°) 5-(45-60°) 7-(75-90°)° Height of first apical T129 branching In cm ((months after planting) Storage root surface T5_6 texture 3-Smooth 5-Medium- 7-Rough T5 9 Number of storage root Actual number Abridged from (International Board for Plant Genetic Resources) (IBPGR) descriptor list for cassava, 1983. 3.2 Molecular characterisation 3.2.1 Lyophilised leaf samples Leaf samples from the 100 cassava accessions were sent from Ghana as lyophilised leaf samples. The DNA extracted from the lyophilised leaf samples was of poor quality and degraded and hence was not suitable for DNA analysis and hence was discarded. 26 University of Ghana http://ugspace.ug.edu.gh 3.2.2 Screen house establishment Stem cuttings were made from all the 100 accessions and packaged in labelled polythene and sent to IITA, Ibadan Nigeria. The stem cuttings were raised in nursery bags with topsoil in the screen house. Out of the 100 cassava accessions raised in the screen house, 96 survived and these were used for the genetic characterisation. Accessions 5,31,61 and 77 (Table 3.1) were the entries that did not survive in the screen house. Young fresh leaves were harvested and used for the laboratory work. 3.2.3.1 DNA Extraction using Cetyltrimethylammonium bromide (CTAB) DNA extractions from fresh leaf tissues were prepared using the modified DNA isolation method described by Rogers and Bendich (1985). The DNA extraction buffer contained 2% CTAB, w/v, 100 mM of Tris-HCl (pH 8.0), 20 mM of EDTA (pH 8.0) 1.4 mM NaCl, 1% PVP (Poly Vinyl Pirrolidone) and 0.3 B-mercaptoethanol added just before use. The leaf samples were ground with mortar and pestle in liquid nitrogen into fine powder. The powdered tissue was transferred into 50 ml falcon tubes and 15 ml o f preheated (at 65°C) extraction buffer added. The extract was mixed thoroughly by gentle shaking, and incubated in a water bath at (60°C-65°C) for lhr with occasional mixing. The mixture was extracted with phenol/chloroform (chlorofronrisoamyl 24:1 v/v) and centrifuged for 10 minutes at 10,000 rpm in a Sorval RC-5C automatic super speed refrigerated centrifuge (Beckman, USA). The supernatant was transferred into 30 ml corex tubes and an equal volume of ice cold isopropanol, and a tenth volume o f 3M sodium acetate (pH 7.0) added. The DNA was then precipitated at -20°C for 2 hours and centrifuged at 10,000 rpm for 10 min to pellet the DNA. DNA was air dried and dissolved in IxTE (10 mM Tris pH 8.0, 0.1 mM EDTA) and treated with 10 ng of RNAse at 37°C for 30 mins. The DNA was extracted with phenol chloroform and precipitated with ice cold isopropanol and 3 M sodium acetate at -20°C for 2 hrs. The mixture was centrifuged for 5 mins 27 University of Ghana http://ugspace.ug.edu.gh and the pellet washed with 500 (il cold 70% ethanol by centrifugation at 10,000 rpm for 5 mins. The DNA was air-dried and the pellet resuspended in 1 ml T.E. The extracted DNA was examined on 1% agarose in lxTAE buffer (45 mM Tris -HC1, 1 mM EDTA, 3% Glacial acetic acid, pH 8.0). The DNA concentration was measured photospectrometrically using DU-65 UV spectrophotometer (Beckman Instruments USA) at 260 nm. The DNA samples were diluted to a final concentration of 10 ng/|il and stored at -20°C for subsequent experiments. The CTAB protocol was used to extract DNA from the 96 accessions that survived at the screen house in IITA. Seventy-two (72) of the samples gave good quality DNA that could be amplified with RAPD primers. 3.2.3.2 DNA extraction using Qiagen kit DNA was extracted with the Qiagen Kit according to the manufacturer’s protocol. The constitutions of the accompanying buffers were not disclosed. Each fresh leaf sample (about 1 g) was ground in 1.5 ml eppendorf tube with plastic pestle in liquid nitrogen into fine powder. 400 |il of preheated (65°C) extraction buffer API and 4 |il of Rnase was added to the powdered tissue and vortexed vigorously for cells to be lysed. The mixture was incubated at 65°C in water bath for 10 min , with occasional mixing by inverting 2 or 3 times. Buffer AP2 (130 ul) was added to the lysate, mixed and incubated for 5 min on ice to precipitate detergent, proteins, and polysaccharides. The lysate was introduced into the QIAshredder spin column sitting in a 2 ml-collection tube and centrifuged for 2 min at maximum speed to remove precipitates or cell debris. The flow-through lysate (usually 450 jj.1) was introduced into a new eppendorf tube and 0.5 volume of Buffer AP3 and 1 volume o f ethanol added and mixed by pippetting. The mixture was then introduced into Dneasy mini spin column in a 2 ml collection tube, centrifuged for lmin at (8,000 rpm). Flow through was discarded and the step repeated for the rest of the sample. The Dneasy column membrane with the DNA was washed with 500 |il of 28 University of Ghana http://ugspace.ug.edu.gh Buffer AW by centrifugation for 1 min (8,000 rpm). The step was repeated at maximum speed (11,000 rpm) for 2 min to dry the column membrane. Since residual ethanol could interfere with subsequent reactions, the column was dried at 50°C for 15 min. The mini column is supported in a 1.5 ml eppendorf tube and 100 (nl of preheated Buffer AE (65°C) directly dispensed into the Dneasy column membrane, incubated for 5 min at room temperature and then centrifuged for 1 min at (8,000 rpm) to elute the DNA. This was repeated with another 100 |il of Buffer AE and collected into a second tube. RNase was added to eliminate RNA contamination. DNA yield and purity was determined by measuring the absorbance at 260 nm (A260) and (A280) in a spectrophotometer. The DNA was checked on 1% agarose in Ix TAE buffer (40 mM Tris - acetate and 1 mM EDTA, pH 8.0). The DNA samples were diluted to a final concentration of 10 ng/^1 and stored at -5°C for the PCR experiments. 3.2.4 Primers A total o f eighty random 10-mer oligonucleotides primers from the A, O, P, Q, R, S, J, K, and E set kits ( Operon Technologies Inc. Alameda CA, USA) were screened ( see appendix 3). Ten of the primers, OPR-9, OPR-2, OPK-1, OPK-11, OPK-16, OPQ-18, OPS-18, OPJ-20, OPJ- 5, and OPJ-14 were selected to amplify the DNAs. The primers were selected on the basis of their ability to generate highly unambiguous scorable bands. 29 University of Ghana http://ugspace.ug.edu.gh Table 3.3 Oligonucleotide primers used for the RAPDs analysis of the Cassava accessions Serial no Operon code Nucleotide sequence 1 OPJ-14 CACCCGGATG 2 OPJ-20 AAGCGGCCTC 3 OPJ-5 CTCCATGGGG 4 OPK-1 CATTCGAGCC 5 OPK-11 AATGCCCCAG 6 OPK-16 GAGCGTCGAA 7 OPQ-18 AGGCT GGGT G 8 OPR-2 CACAGCTGCC 9 OPR-9 TGAGCACGAG 1 0 OPS-18 CTGGCGAACT 3.2.5 RAPD-PCR conditions Polymerase Chain Reaction amplification reactions were performed in 25 |il reaction volumes. The reaction mixture composition was 5 ng template DNA, 2.5 pi o f 5% Tween 20, 2.5 jj.1 of 10 x Taq buffer, 1 jj.1 o f 2.5 mM of each deoxyribonucleotide (dNTPs), l(j.l o f Primer and 2 units o f Taq DNA polymerase, (Promega Corporation Madison,WI,USA) and 10.6^1 o f sterile distilled The reaction mixture was overlaid with a drop o f mineral oil to avoid evaporation, PCR amplification was carried out in Perkin Elmer 9600 thermocycler. The amplification profiles programmed involved an initial denaturation step o f 94°C for 3 mins followed by 45 cycles at 94°C for 1 min, 36°C for 1 min for primer to anneal, and extension at 72°C for 2 min. A final elongation step at 72°C for 7 mins was included. Amplification products were maintained at 4°C until electrophoresed. The PCR amplification products were separated by electrophoresis on 1.5% agarose gel in lxTAE buffer (40 mM Tris-acetate and ImM EDTA, pH 8.0), with a 1-kb DNA ladder (Promega Corporation, Madison,WI,USA) used as molecular size marker. The gels were stained with ethidium bromide (0.5pg/ml) and viewed under ultra-violet transiluminator and photograph taken using the computer soft ware for gel documentation called Grab-it. 30 University of Ghana http://ugspace.ug.edu.gh 3.3 Data analysis 3.3.1 Analysis of morphological data Morphological and genetic markers were used to detect genetic diversity in the cassava cultivars from Ghana. Morphological data was subjected to both univariate and multivariate analyses. Univariate analysis was performed where each trait was tested using the analysis o f variance by general linear model procedure. Multivariate analysis included principal component analysis and canonical variate analysis were performed on all the traits measured. A non-hierarchical clustering was done using results from the canonical discriminant analysis. Phenotypic correlation analysis for the morphological characters was carried out. Heritability estimates were calculated on only quantitative traits. Broadsense heritabilities were calculated using the formula: hb 2 = Vg/Vp =Vg/ (Vg + Ve) ( Wricke et al., 1986) 3.3.2 Analysis of RAPD data RAPD markers that were consistently reproduced were used for the analysis. The position of scorable RAPD bands and sizes were estimated from the gel photograph by a comparison with the lOObp and 1-KB ladder marker. Scoring was done using the gel photograph system. The entire set of reproducible RAPD bands generated in the 96 cassava accessions by the 10 primers was used for the RAPDs analysis. The bands were scored as present (1) or absent (0). A binary character matrix was generated with fragments in column and cultivar in rows. Pairwise distance matrices were compiled by the NTSYS-PC 2.02j software package (Rohlf, 1989) using the Jaccard similarity coefficient (Jaccard, 1908).The similarity index of Jaccard between cultivars (i and j ) was given by: Sy = a / (a + b + c) 31 University of Ghana http://ugspace.ug.edu.gh Where, a is the number of bands present in both i and j, b is the number of bands present in i but not in j, c the number of bands present in j but not in i Clustering was done using the unweighted pair group method with arithmetic means (UPGMA) algorithm (Sneath and Sokal, 1973). The resultant data was used to generate a dendrogram showing the phylogenetic relationships. The RAPD results analysed was subjected to similarity analysis using Jaccard (1908) coefficient. Hierarchical clustering was done using the sequential, agglomerative, hierachical and nested (SAHN) method with UPGMA values. A phylogenetic tree or dendrogram was then generated. 32 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR 4.0 RESULTS 4.1 Variability and heritability of morphological traits 4.1.1 Variation in qualitative traits Figures 4.1 shows the distribution of leaf petiole colour among the cassava accessions. More than half of the accessions (53%) had dark purple petiole; 22% had red, 14% had green and 11% light green colour. Stem colour distribution showed that 84% of the accessions were dark brown, 14% brown and 2% silver green (Figure 4.2). Thirty percent of the accessions had light green purple leaf colour, 24% dark green, 15% green and 14% light green (Figure 4.3). Variation in storage root surface colour is shown in Figure 4.4. Fifty-nine percent of the accessions had dark brown tuber colour, 26% light brown and 15% cream colour. Anthocyanin pigmentation showed considerable variation among the accessions with 49% of the accessions showing anthocyanin pigmentation on the entire petiole, 1 1% ventral portions, 26% patches on petiole and 14% did not show any anthocyanin pigmentation (Figure 4.5). Fifty-six percent of the cassava accessions flowered and 44% of the accessions did not flower (Figure 4.6). 33 University of Ghana http://ugspace.ug.edu.gh Petiole colour Figure 4.1: The distribution of petiole colour in 100 accessions of cassava. % frequency Dark brown Brown Stem colour Silver green Figure 4.2: The distribution of stem colour in 100 accessions of cassava. 34 University of Ghana http://ugspace.ug.edu.gh 30 25 20 % frequency 15 10 5 0 Figure 4.3: The distribution of colour of fully expanded leaf in 100 cassava accessions. d e Purple Green purple Dark green Light green purple Leaf colour Green Light green Cream 15% Light brown 26% Dark brown 59% Figure 4.4: The distribution of root surface colour in 100 cassava accessions. 35 University of Ghana http://ugspace.ug.edu.gh 50 45 40 35 30 % frequency 25 20 15- ID 5 0 Entire Ventral Patches Absence Anthocyanin pigmentation Figure 4.5: The distribution of anthocyanin pigmentation in 100 cassava accessions. Non-Flowering 44% Figure 4.6: The distribution of flowering and non-flowering types among 100 cassava accessions. 36 University of Ghana http://ugspace.ug.edu.gh 4.1.2 Variation in quantitative traits Plant establishment at four weeks after planting, number of leaf lobes, plant stand at harvest, plant height at harvest and height at first branching showed significant (p<0 .0 1 ) variation among the accessions (Table 4.1). All the yield traits measured showed significant (p<0.01) variation (Table 4.2). There were significant variation in the accessions with regard to the diseases evaluated. The severity of African cassava mosaic virus (ACMV) disease and cassava bacterial blight (CBB) disease both showed significant (p<0.01) variation among the accessions (Table 4.3). Hydrogen cyanide content and percentage foliage retention (Table 4.3) also showed significant (p<0 .0 1 ) variation among the 1 0 0 accessions. Table 4.1: Mean squares for plant establishment, number of leaf lobes, plant stand and plant height (quantitative characters) measured in 100 accessions of cassava Sources of variation Degrees of Freedom Plant establishment at 4 weeks Number of leaf lobes Plant stand at harvest Plant height at harvest Height at first branching (cm) Replication 1 271 0 . 2 8 8 40 10503 Block (Replication) 18 250 0 . 2 1 198 986 3219 Accession 99 5 9 9 ** 0.61** 586** 3472** 8199** Error 76 127 0.18 204 612 3165 ** Significant at 5% and 1% respectively 37 University of Ghana http://ugspace.ug.edu.gh Table 4.2: Mean squares for size of storage root, number of tubers per plot, yield and harvest index among 100 cassava accessions Sources of variation Degrees of Freedom Storage root length (cm) Storage root diameter (cm) Number of fresh tuber per plot Weight of fresh tuber yield tons/ha Harvest index Replications 1 6.51 4.13 139 119 0 . 0 1 Block (Replication) 18 1.35 1.14 124 42 0 . 0 1 Accession 99 2 .6 8 ** 1.58** 8 6 6 ** 321** 0.03** Error 76 0.77 0.56 97 41 0 . 0 1 ** Significant at 5% and 1% respectively Table 4.3: Mean squares for number of nodes to first branching, foliage retention, HCN content, ACMV and CBB severity scores at 6 MAP among 100 accessions of cassava Sources of variation Degrees of freedom Number of nodes to first branching % foliage retention Hydrogen cyanide content (mg) ACMV severity' scores at 6 MAP CBB severity scores at 6 MAP Replications 1 5008 52 15.6 3.25 0 . 1 2 Block (Replication) 18 1923 2 2 1.96 0.29 1.69 Accession 99 4207** 60** 2.99* 0.75** 2 .1 0 ** Error 76 1683 2 2 1.81 0.17 0.58 ** Significant at 5% and 1% respectively 38 University of Ghana http://ugspace.ug.edu.gh 4.1.3 Heritability of quantitative traits Genotypic and phenotypic variances and ratios of genotypic and error variances to the phenotypic variance are shown in Table 4.4. A large proportion of the phenotypic variance of the length of stipules (84%), plant height at harvest (82%), number of fresh tubers (89%) and fresh tuber yield (87%) were attributable to genotypic differences among the accessions. Hydrogen cyanide (HCN) content had a low heritability estimate of 0.39, indicating that differences in HCN content were mostly conditioned by the environment. Table 4.4 Genotypic (o2g), phenotypic (o2p) variances and variance ratios of quantitative agronomic characters in cassava Variance Character ( ° \ ) (o2p) (crg/(crp) (