Sorghum [Sorghum bicolor (L.) Moench] is a major food security crop for millions of people in sub-Saharan Africa and the fourth most important crop in Africa. The potential sorghum yields are limited due to a number of abiotic, biotic and socio-economic constraints. Among the biotic stresses is the parasitic weed, Striga hermonthica, which inflicts yield losses ranging from 30-100%. Various control options have been recommended to reduce levels of Striga damage. However, these techniques need to be integrated for effective control and to boost sorghum productivity. A series of experiments was conducted to integrate host resistance improvement and the use of a biological control agent, Fusarium oxysporum f.sp. strigae to control Striga hermonthica. These studies were also focused on improving breeders‟ awareness of the traits that farmers‟ desire, on the assumption that farmers‟ variety preference traits are the missing link in technology development and adoption process for S. hermonthica management. The objectives of the study were to: 1) determine farmers‟ views on sorghum production opportunities; threats; indigenous knowledge and perceptions; breeding priorities; Striga infestation; and the coping mechanisms of farmers in the north eastern and north western Ethiopia, 2) evaluate sorghum genotypes for compatibility to F. oxysporum inoculation where grown in Striga infested soil in controlled environments, 3) determine field responses of sorghum genotypes and F. oxysporum compatibility for integrated Striga management (ISM), 4) determine the variability present among selected sorghum genotypes exhibiting S. hermonthica resistance, and compatibility with the biological control agent using phenotypic and simple sequence repeat (SSR) markers, 5) identify F. oxysporum compatible sorghum parents and hybrids with high combining ability for grain yield, yield components, and Striga resistance for ISM, and 6) undertake farmers‟ participatory assessment, and identify their preferred traits for sorghum genotypes under ISM, simultaneously with the breeders‟ evaluation. A participatory rural appraisal (PRA) research was conducted involving 315 farmers in nine districts of three administrative zones within two provinces in Ethiopia. Sorghum landraces were preferred by >85% of participants rather than previously improved released varieties. The participating farmers listed and prioritized their sorghum production constraints. In the North Shewa and North Wello zones drought was the most important constraint, followed by Striga. In the Metekel zone Striga was the number one constraint followed by a lack of genotypes with high grain quality. Controlled environment experiments were conducted involving greenhouse and laboratory tests in order to evaluate 50 sorghum genotypes for their compatibility with F. oxysporum and for possible deployment of the bio-control agent to control Striga. Striga population was reduced by 92% through the application of F. oxysporum, resulting in yield increment of 144%. Twelve sorghum genotypes were identified as promising parents for breeding and to control Striga through integration of host resistance and F. oxysporum seed treatment. During field and sick plot plot evaluations differential responses to F. oxysporum application among the sorghum genotypes were observed for various attributes including Striga plant height. Most traits showed highly significant (p<0.001) genotype X site interactions. Similarly, the main effects of F.oxysporum application were highly significant (p<0.001) across sites for most of the traits. The genotype and genotype X environment biplot identified 13 genotypes that consistently performed well following Fusarium application. The variability present among 14 selected sorghum genotypes exhibiting S. hermonthica resistance, and compatibility with a biological control agent, Fusarium oxysporum, were determined using phenotypic and 20 polymorphic simple sequence repeat (SSR) markers. Highly significant (p<0.001) differences were detected among genotypes for phenotypic traits. Principal component analysis showed three components that accounted for 73.99% of the total variability exhibited among genotypes. Cluster analysis allocated the genotypes into two major groups, one with a further two subgroups based on morphological traits, showing clear demarcations between the genotypes. The SSR markers revealed high levels of polymorphisms among genotypes, with the mean number of alleles per locus being 6.95 and the mean polymorphic information content being 0.80. The observed genetic diversity was relatively wide, with the allele sizes ranging from 203.6-334 bp. The SSR markers allocated genotypes into two distinct clusters close to the phenotypic markers. Forty sorghum hybrids were developed through a line by tester mating design involving 10 lines selected for their compatibility with F. oxysporum and high agronomic performances and four Striga resistant tester parents. The F1s and their parents were field evaluated with complementary in-vitro tests. Field evaluations were conducted at two locations: Kobo and Shewa Robit in Ethiopia, which are well known for their severe Striga infestation. Significant (p<0.05) general combining ability (GCA) effects were observed among testers and lines at both sites for days to 50% flowering and maturity, plant height, biomass, number of Striga plants and Striga plant height. Furthermore, significant (p<0.05) specific combining ability (SCA) effects were detected for days to 50% flowering, biomass, grain yield and number of Striga plants. From the complementary in-vitro experiment, highly significant variation (p<0.01) was exhibited due to line x tester interaction for maximum Striga germination distance. The study identified paternal parents with high GCA effects including SRN-39 and Birhan and maternals 235761, 2384443, IC9830, 235466, 237289,235763, and 235929 to be useful for breeding for ISM in sorghum. At Kobo, cross 235763 x N-13 and Shewa Robit IC9830 x SRN-39 had significantly negative SCA effects for the numbers of Striga plants. Progenies of these crosses will be selected in the Striga resistance breeding program. In the participatory sorghum genotypes assessment, farmers were invited to assess and select the genotypes based on their preferences at maturity and harvesting. The standard agronomic traits and Striga parameters relevant for breeding were collected by the breeders. Earliness, Striga resistance, high yield and high grain quality and threshability were the most important farmers‟-preferred traits for sorghum genotypes. Comparative analyses between farmers‟ and breeders‟ evaluations revealed highly significant correlations (p<0.01) except between Striga resistance and Striga damage and pest resistance and insect damage. Repeatability of scoring genotypes among farmers was consistent (>0.80) for all traits except Striga and pest resistance. The prioritized traits through farmers‟ participation are important for further breeding program. Overall, the study established farmers‟ preferred traits, the effectiveness of ISM to boost sorghum productivity, and identified useful parents and crosses for effective sorghum breeding to control Striga in Ethiopia. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2013.