| dc.description.abstract | Tobacco (Nicotiana tabaccum L.) is a crop with commercial value in most countries of the world and not much work has been done to improve its performance. Information on availability of genetic diversity expressible in the phenotype, understanding gene action influencing key traits and adaptation of tobacco varieties to different growing environments are a prerequisite for effective tobacco breeding and subsequent production. This study was conducted (a) to determine the extent of genetic diversity within the Zimbabwean flue-cured tobacco (Nicotiana tabacum L.) germplasm, (b) to determine the nature of gene action for key traits in tobacco as well as (c) to investigate genotype × environment interaction effects in tobacco production. To determine the molecular diversity of flue-cured tobacco germplasm, a sample of 93 accessions were extracted from the Tobacco Research Board germplasm collection. Genomic DNA was extracted from leaves using a Direct Zol™ Plant DNA extraction kit. Microsatellite amplifications were performed using 10 SSR markers and analysis of molecular variance (AMOVA) was used to evaluate the partitioning of diversity within and between populations and between sample groups using Arlequin. Morphological diversity among tobacco germplasm was determined by evaluating 196 accessions in a 14×14 α-Lattice design with two replicates at Kutsaga research station for two seasons. Data analysis was done using Genstat Version 17. Genetic analyses for growth rate, ripening rate and flue-cured leaf colour composition of the flue-cured tobacco was done using Griffing's Method I, Model I; and Hayman's Wr –Vr graphic analysis. A ten-parent full diallel cross including reciprocals and parentals was laid out in an α-Lattice experimental design with two replicates over a two-year period at Kutsaga research station. To investigate genotype × environment interaction effects, experiments were set out in a randomised complete block design with three replicates and twelve genotypes were evaluated across eight locations and five parameters were measured. Data was subsequently subjected to Additive Main Effects and Multiplicative Interaction (AMMI) Model and Genotype×environment interaction (GGE) biplot analysis. Results from genetic and morphological characterization showed that the studied germplasm was morphologically closely related, despite originating from diverse parts of the world since the majority of the accessions had similarity coeficiencies of above 60%. However, there were some genotypes like Coker 347, which were different from the rest of the genotypes and could contribute private alleles for future use in breeding programmes to expand the tobacco genetic base. Results from genetic analysis showed significant General Combining Ability effects for dark cured leaf colour (p<0.05) suggesting the influence of additive gene action on the trait. Reciprocal effects were not significant for all traits meaning that the parents can be interchanged in hybridization programmes. The Hayman Wr-Vr graph indicated partial dominance mode of inheritance for traits measured except dark cured leaf colour. Heritability was moderate for ripening rate and low for growth rate and dark cured leaf colour implying that slow progress can be made when selecting for the later two traits in tobacco improvement programmes. Significant genotype × interaction effects were only recorded for the traits: quality grade, grade index, and mass at untying (P<0.01) and not for saleable yield and dark cured leaf colour. The comparison biplot showed that the genotype KE1×AW3R was the most stable for quality. Generally, these results show that although the flue-cured | en_ZW |
