PhD ThesisBackground Coffee is the most important stimulant beverage in the world and a major form of cash income for millions of smallholder farmers. The world’s coffee trade is dominated by two types Arabica and Robusta. Arabica represents 60-70% whereas Robusta represents 30-40% of global exports. Coffee is one of Tanzania’s primary agricultural export commodities accounting for about 5% of total exports value, and generating export earnings averaging USD 100 million per annum over 30 years by 2012. Insufficient information of genetic diversity at molecular level of germplasms and breeding materials existing in the Institute, absence of fast, convenient and high precision screening tools/diagnostic markers for coffee berry diseases (CBD) resistance, narrow genetic base of released improved coffee varieties and little knowledge base of genetics of arabica coffee traits leads to low efficiency in development and release of improved superior arabica coffee varieties in a reduced time. The main objective of this study was to generate information, which will be used to address the above problems. The specific objectives were: to determine the genetic diversity of Tanzanian coffee germplasm and the Ethiopian Coffea arabica collection maintained in Tanzania using simple sequence repeats (SSRs) molecular markers; to incorporate coffee berry disease resistance and screen progenies of the crosses between Ethiopian accessions and Tanzanian commercial variety KP423 for Coffee Berry Disease (CBD) resistance genes using SSR markers; to determine the combining ability, heritability and relationships of growth and yield characters of progenies of commercial variety KP423 and selected accessions of Ethiopian collection. Materials and Methods Leaf samples were collected from three trees per genotype from Ethiopian collection and other germplasms maintained at Tanzania Coffee Research Institute (TaCRI) Lyamungu for genomic DNA extraction which was carried out at Sokoine University of Agriculture molecular biology laboratory. Thirty SSR primer sets were used in PCR amplification of the extracted DNA. Analysis of Molecular Variance (AMOVA), Principal Component Analysis (PCA) and Cluster Analysis were performed from SSR genetic similarity matrix data using GenStat statistical software version 15.1 (GenStat, 2012). Coffee berry disease (CBD) hypocotyl screening was applied on the progenies and parental genotypes of eleven genotypes selected from the germplasm and Ethiopian collection at TaCRI Lyamungu which were crossed to a susceptible variety KP423. Two gene specific markers Sat 235 and Sat 207 targeting a major CBD resistance gene Ck-1 were used to amplify DNA extracted from leaves of five seedlings of each parental genotypes and crosses (F1s) which survived phenotypic (hypocotyl) CBD resistance screening. For determination of the combining ability, heritability and relationships of growth and yield variables nine genotypes were selected including six from Ethiopian collection, two from the germplasm and a commercial variety KP 423 were used in a half-diallel mating design. Data were collected at Year 1 and 2 after establishment on: Stem girth, Plant height, Length of the longest primary (canopy radius), Number of primaries per plant, Number of berries (flower buds) per cluster, Internode length and Number of bearing primaries per plant. Analysis was based on the fixed effect (model 1) method II by Griffing (1956) while Path coefficients analysis was performed according to Dewey and Lu (1959) using (Sheoran et al., 1998) Statistical Package for Agricultural Research Workers. Key Findings Determination of genetic diversity revealed high observed heterozygozity (0.9993), high percent of polymorphic alleles (80 %) whereas average number of alleles per SSRs locus was 2.5 and polymorphic information content was (0.4128). Principal component analysis identified three significantly different groups with diversity percentages 56.75, 5.69 and 4.66 explained 67.11 % of the total diversity. For CBD resistance gene screening, the genotypes that were clearly amplified by SSR primer Sat 235 and Sat 207 and also showed phenotypic resistance through hypocotyl screening confirmed the presence of the CBD resistance gene Ck-1 by producing bands similar to the progenitors of CBD resistance. Both general and specific combining ability variances were significant (p = 0.01) for all variables except berries per node. Broad sense (H2bs) heritability values were higher than that for narrow sense (h2ns) for all variables. Genotypic correlations between all variables were significant and positive except the correlation between internode length (IL) and number of bearing primaries per plant (NBPP). Similarly phenotypic correlations were positive and significant for all variables except that between number of primaries per plant (NPP) and internode length (IL). Path coefficient analysis reveals that variables length of the longest primary (LLP), stem girth (SG) and internode length (IL) were found to contribute mostly in berries per node. The direct effect due to plant height was found contributing most negatively to berries per node comparing with other variables studied. Major Conclusions High heterozygozity was observed among the genotypes studied with Ethiopian accessions been most diverse. The commercial varieties Bourbon (1) and Kent (2) were grouped in the same group in the PCA (Fig. 5) and clustered in the same cluster in the dendrogram (Fig. 4) revealed low diversity among themselves. On the other hand the Ethiopian genotypes viz (4, 6, 8) and breeding lines viz (65, 66, 86) were scattered in across the groupings revealed their high heterozygozity. Marker ssrAY2449 was the most informative (PIC = 0.7390). Presence of the CBD resistance gene Ck-1 in the genotypes studied was confirmed showing phenotypic hypocotyl resistance and by SSR primer Sat 235 and Sat 207 clear amplification by producing bands similar to the progenitors of resistance. Variables length of the longest primary (canopy radius), internode length and stem girth can be used as an important criteria in selection for superior coffee arabica varieties involving Ethiopian genotypes and commercial variety KP423. The information generated will facilitate selection of the most diverse genotypes for development of superior varieties with broad genetic base for various traits at early growth stages reducing the time for selection and release of superior varieties to growers. Recommendations Accessions from the Ethiopian germplasm should be intensively used in the development of new improved arabica coffee varieties with broad genetic base and suitable agronomic characteristics including resistance to coffee berry disease. Deployment of marker assisted selection for resistance to coffee berry disease should be use microsatellites SAT 235 and SAT 207 to shorten selection cycles and new varieties release time. Variables such as length of the longest primary (canopy radius), internode length and stem girth should be used as important criteria in selection for superior Arabica coffee varieties involving Ethiopian genotypes and commercial varieties at the early growth stage. Implication of this study for Tanzania Coffee Industry The diversity of the Breeding materials at Tanzania coffee research institute (TaCRI) is high and can continue to sustain the breeding programme in development of superior coffee varieties with broad genetic base especially involving the Ethiopian Arabica coffee collection. The established use of coffee berry disease (CBD) resistance screening through the use of gene specific markers will facilitate early selection of superior genotypes and segregating populations for resistance to CBD at a shortest possible time reducing the time taken to release varieties using the phenotypic hypocotyl and field screening. Early selection of superior genotypes will also be possible through the use of yield and growth variables as revealed in the study reducing the time taken to release superior varieties. Based on these facts the industry will benefit from the fast and efficient system of releasing new superior varieties which will lead to increased coffee production at reduced cost of managing the CBD, increasing farmer’s income and national coffee export forex earnings.Tanzania Coffee Research Institute (TaCRI)