Maize (Zea mays L.) is the third most important cereal crops in the world after wheat and rice. In Ethiopia, maize remains the second largest food security crop after tef [Eragrostis tef (Zucc.) Trotter.]. The mid-altitude, sub-humid agro-ecology (1000 to 1800 m above sea level) is the most important maize producing environment in Ethiopia. However, productivity of maize is low, due to several biotic and abiotic constraints. Among the biotic constraints, Turcicum leaf blight disease of maize caused by Exserohilum turcicum Pass Leonard & Suggs shows high incidence of 95-100% and inflicts significant grain losses in the country. Therefore, high yielding, Turcicum leaf blight resistant and farmers-preferred maize varieties and their production technologies should be developed and made available to growers to enhance maize production and to achieve food security. The objectives of this study were to: (1) assess farmer’s preferences, and production constraints for maize in the mid-altitude, sub-humid agro-ecology of western Ethiopia, (2) determine the genetic variability among elite maize inbred lines and select promising parents for resistance to E. turcicum, (3) determine diversity among the elite germplasm lines using SSR markers, (4) determine combining ability and heterosis among elite maize inbred lines and their hybrids, and (5) investigate genotype x environment interaction and yield stability of experimental maize hybrids developed for the midaltitude sub-humid agro-ecology of Ethiopia. A participatory rural appraisal (PRA) research was conducted involving 240 maize farmers in three representative maize growing zones of western Ethiopia; West Shoa, East Wollega and West Wollega, each represented by two districts and two subdistricts. Maize was ranked number one both as food and cash crop by 82.9% of respondents. Turcicum leaf blight was ranked as number one devastating leaf disease by 46% of respondents. Breeding for improved disease resistance and grain yield, enhancing the availability of crop input and stabilizing market price during harvest time were recommended as the most important strategies to increase maize production by small-scale farmers in western Ethiopia. Fifty inbred lines were evaluated for reaction to Turcicum leaf blight during the main cropping seasons of 2011 and 2012. Inbred lines were clustered into resistant (CML202, 144-7b, 136-a, 139-5j, 30H83-7-1, ILOO’E-1-9, SZYNA-99-F2, and 142-1-e), and susceptible (CML197, CML464, A7033 , Kuleni C1-101-1-1, CML443, SC22-430 (63), (DRB-F2-60-1-2) – B-1-B-B-B, Pool9A-4-4-1-1-1). Inbred lines (CML312, CML445, Gibe-1-158-1-1-1-1, CML395, and 124-b (113)) had intermediate response to the disease. Overall, inbred lines such as CML202, 30H83-7-1, ILOO’E-1-9-1, CML312, CML395 CML445 and 142-1-e were selected with better agronomic performance and resistance to leaf blight for breeding. Twenty selected elite parental inbred lines were genotyped with 20 polymorphic SSR markers. The genotypes used were clustered into five groups consistent with the known pedigrees. The greatest genetic distance was identified between the clusters of lines CML-202 and Gibe-1-91. Eighteen selected inbred lines were crossed using the factorial mating scheme and 81 hybrids developed to determine combining ability effects and heterosis. Inbred lines with high GCA effect (CML 202, CML395, 124-b (113), ILOO’E-1-9 and CML 197) were selected as best combiners for hybrid development. Additionally five high yielding novel single cross hybrids with grain yield of > 8 t ha-1 and high SCA effects were identified such as CML395 X CML442, DE-78-Z-126-3-2-2-1-1 X CML442, ILOO’E-1-9-1-1-1-1-1 X CML312, X1264DW-1-2-2-2-2 X CML464 and SC22 X Gibe-1-91-1-1-1-1. These experimental hybrids are recommended for direct production or as hybrid testers for hybrid development. Genotype x environment interaction (GEI) effects of 81 newly developed and three check maize hybrids were evaluated across 10 locations in the mid-altitude sub-humid agro-ecologies of Ethiopia. The AMMI-3 and GGE biplot models were used to determine stability. Hybrids such as G68, G39, G37, G77, G34 and G2 were identified as the most stable and high yielding at favorable environments such as Bako, Jima, Arsi Negelle and Pawe in Ethiopia. The genotype and genotype by environment interaction (GGE) biplot clustered the 10 environments into three unique mega-environments. Environment I included Bako, Jima, Asossa, Ambo, Finote Selam, Haramaya and Pawe while environment II represented by Arsi-Negelle and environment III Areka and Hawassa. In general, the study identified valuable maize inbred lines with high combining ability for breeding and novel single cross hybrids for large-scale production or as testers for hybrid development at the mid-altitude, sub-humid agro-ecologies of Ethiopia or similar environments in sub-Saharan Africa. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2013.

Maize production in Africa is constrained by abiotic and biotic stresses. Breeders need to have information on the nature of combining ability of parents, their traits and performance in hybrid combination. This requires careful determination of genetic variability of parents, and studying associations between grain yield and adaptive traits to breed superior cultivars which are better able to withstand such stresses. Therefore, this study was aimed at selecting parental testers with best combining ability in hybrid combination with recombinant inbred lines (RILs); and studying the correlation between grain yield and its components in eastern and western South Africa. It was also aimed at determining genetic variation and associations among adaptive traits in hybrids involving RILs. The final objectives of the study were to determine cultivar superiority of testcrosses involving RILs, and to select the best cultivars within and across four different environments. The 42 RILs were crossed to 9 Zimbabwean tropical testers resulting in 1009 hybrids with sufficient seed for planting in trials. From these a sample of 87 hybrids with adequate seed were selected and planted at four sites for combining ability analysis. The hybrids were evaluated at four sites in two regions; western region (Potchefstroom research station) and eastern region (Cedara, Ukulinga and Dundee research stations), during 2011/12 season. The experiments were laid out as augmented alpha lattice design. Trials were managed in accordance with production culture for each region. All quantitative data was subjected to GenStat and SAS statistical softwares. The results from combining ability study indicated that the line general combining ability (GCA) effects played a non-significant role (p > 0.05) in determining grain yield, grain moisture and anthesis date, while they were significant (p ≤ 0.05) for the other traits such as ear prolificacy. The tester main effects were significant for all the traits except ear prolificacy and plant height. Results also revealed that all the traits were controlled by both additive and non-additive genes, where additive gene action had the most contribution to the traits. The non-additive gene action played a minor role suggesting the total GCA effects attributed to both lines and testers predominantly higher over the specific combining ability (SCA) for all traits. In general the additive effects were preponderant over the non-additive gene effects. One cross (L114 x T12) had a significant and positive SCA effect for grain yield. The correlation between grain yield and secondary traits (number of ears per plant, grain moisture content, ear height, plant height, ear position and anthesis date) suggested that indirect selection can be employed to enhance grain yield by breeding for these particular adaptive traits. Path analysis showed that plant height had the highest direct and indirect effect on grain yield indicating its importance among other secondary traits for grain yield enhancement. Phenotypic coefficient of variation (PCV) was higher than genotypic coefficient of variation (GCV) for all the studied traits across all the four environments. All the traits displayed high heritability at Potchefstroom except anthesis date which was highly heritable at Ukulinga. Cedara was the second best site for heritability of all the traits except for the number of ears per plant. The genetic advance for grain yield was the highest at Cedara followed by Potchefstroom, Dundee and Ukulinga. The hybrids exhibited different patterns of variation and distribution for all the traits. This indicated that selection strategies to exploit GCA should be emphasised. Association studies among grain yield and secondary traits such as ear length, number of ears per plant, plant height, anthesis date, silking date and ear leaf area revealed that there were significant phenotypic correlations between grain yield and secondary traits, and among the secondary traits. Ear length had the highest direct effect on grain yield at Ukulinga; number of ears per plant had the highest direct effect on grain yield at Cedara and Potchefstroom; whereas plant height had the highest direct effect on grain yield at Dundee. Grain yield was least affected by indirect factors at all the sites except Ukulinga, where anthesis date had the highest indirect effect on grain yield through silking date followed by plant height through leaf area. The study reveals that there is significant variation among the hybrids for mean performance, indicating that there is opportunity for selection. Overall the findings suggest that direct selection would be appropriate to enhance grain yield. Path analysis revealed that plant height had the highest direct and indirect effects on grain yield, indicating that plant height can be further exploited as the main trait in future breeding programmes for grain yield increment. Hybrid 10MAK10-1/N3 was the best hybrid at Ukulinga in terms of grain yield, relative yield and economic traits. Whereas hybrid T17/L83 was the best hybrid at Cedara in terms of grain yield and relative yield; however, T11/L102 was selected as the most elite hybrid with respect to grain yield, relative yield and economic traits. Hybrid T3/L48 was identified as thebest hybrid at Dundee with respect to grain yield, relative yield and prolificacy. At Potchefstroom the standard check PAN6611 was identified as the best hybrid in terms of grain yield and relative yield followed by developmental hybrid T1/L28; however, developmental hybrid T1/L28 was the best in terms of earliness, prolificacy and ear aspect. Stability coefficients and cultivar superiority index across the sites revealed that four developmental hybrids were identified as best hybrids and they performed better than the standard check. These hybrids will be recommended for further testing in advanced trials. With respect to cultivar superiority, the desired hybrids are required to combine high grain yield with economic and adaptive traits such as high ear prolificacy, low grain moisture, and low ear aspect score (desired) for them to adapt to production environments in South Africa. There was significant variation among the top 25 yielding hybrids. At least 5 hybrids combined high grain yield with the desired complimentary adaptive traits such as quick moisture dry down, prolificacy and ear aspect. The results showed that there is variation in the performance of high yielding genotypes within all the sites, and that agronomically superior cultivars can be identified. The study shows that there is significant variation among the RILs since they interacted differently with the 9 tropical testers. Even among the top 25 selections of RILs in each environment there was still variation for combinations of the desired traits. Significant associations among grain yield and other economic and adaptive traits were observed with implications for breeding strategy. Above all the significant variation gives large score for future breeding of new unique lines. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2012.

Maize streak virus (MSV) disease, transmitted by leafhoppers (Cicadulina mbila, Naude), and maize downy mildew (DM) disease caused by Peronosclerospora sorghi (Weston and Uppal) Shaw, are major contributing factors to low maize yields in Africa. These two diseases threaten maize production in Mozambique, thus the importance of breeding Mozambican maize varieties that carry resistance to these diseases. Marker-assisted selection (MAS) was employed to pyramid MSV and DM disease resistant genes into a single genetic background through simultaneous selection. Firstly, it was essential to determine the genetic diversity of MSV disease resistance in 25 elite maize inbred lines to aid in the selection of suitable lines for the introgression of the msv1 gene; and subsequently, to introduce the msv1 resistance gene cluster from two inbred lines, CM505 and CML509, which were identified as the ideal parental lines for the introgression of MSV disease resistance into a locally adapted Mozambican inbred line LP23 that had DM background resistance. Pyramiding the resistance genes by the use of simple sequence repeat (SSR) molecular markers to track the MSV gene cluster was investigated in 118 F3 progeny derived from crosses of CML505 x LP23 and CML509 x LP23. High resolution melt (HRM) analysis using the markers umc2228 and bnlg1811 detected 29 MSV resistant lines. At the International Maize and Wheat Improvement Centre (CIMMYT) in Zimbabwe, MSV disease expression of the 118 F3 progeny lines was assessed under artificial inoculation conditions with viruliferous leafhoppers and the effect of the MSV disease on plant height was measured. Thirty-seven family lines exhibited MSV and DM (DM incidence ≤50) disease resistance. Individual plants from a total of 41 progeny lines, that exhibited MSV disease severity ratings of 2.5 or less in both locations within each of the F3 family lines, were selected based on the presence of the msv1 gene based on SSR data, or field DM disease resistance, and were then advanced to the F4 generation to be fixed for use to improve maize hybrids in Mozambique for MSV resistance. Simultaneous trials were run at Chokwe Research Station in Mozambique for MSV and DM disease assessment, under natural and artificial disease infestation, respectively. Thus the MSV and DM genes were effectively pyramided. Lines with both MSV and DM resistance were advanced to the F4 generation and will be fixed for use to improve maize hybrids in Mozambique for MSV and DM resistance, which will have positive implications on food security in Mozambique. This research discusses the results of combined selection with both artificial inoculation and the three selected SSR markers. It was concluded that a conventional maize breeder can successfully use molecular markers to improve selection intensity and maximise genetic gain. / Thesis (M.Sc.Agric)-University of KwaZulu-Natal, Pietermaritzburg, 2013.

The maize weevil (Sitophilus zeamais Motschulsky) is one of the most destructive storage insect pest of maize (Zea mays L.) in tropical Africa and worldwide, especially when susceptible varieties are grown. Therefore, grain resistance against the maize weevil should be part of a major component of an integrated maize weevil management strategy. The specific objectives of this study were to: i) determine farmers’ perceptions about weevil resistance in maize cultivars; ii) determine the genotypic variation for maize weevil resistance in eastern and southern Africa maize germplasm lines; iii) study the gene action conditioning weevil resistance in the inbred line populations from eastern and southern Africa maize germplasm and to measure their combining ability for yield and weevil resistance; iv) determine the effectiveness of two cycles of modified S1 recurrent selection in improving a tropical maize population “Longe5” for weevil resistance and agronomic superiority and v) evaluate the effectiveness of the “weevil warehouse techniques” compared to the “laboratory bioassay technique” as methods of maize screening against the maize weevil. A participatory rural appraisal (PRA) was conducted in three districts between December 2010 and January 2011, to gather information on the maize weevil pest status in Uganda and farmers’ perceptions about improved maize varieties and the major attributes desired in new maize varieties. Over 95% of farmers knew the maize weevil and its pest status, and were reportedly controlling the maize weevil using wood ashes, red pepper and Cupressus sempervirens. The estimated postharvest weight losses attributed to weevil damage was over 20% within a storage period of four months. The most highly ranked attributes desired in the new maize varieties included high grain yield, tolerance to drought and low nitrogen stresses, resistance to field pests and diseases, good storability and resistance to storage pests. In the search for new sources of weevil resistance, a total of 180 inbred lines from three different geographical areas were screened for weevil resistance using the laboratory bioassay technique. Eight inbred lines (MV21, MV23, MV75, MV102, MV142, MV154, MV157, and MV170) were consistently grouped in the resistant class, and therefore selected as potential donors for weevil resistance in the maize improvement programs. Large significant genetic variations for weevil resistance, and high levels of heritability (89 – 96%) were observed. The results revealed that there was no significant association between maize weevil resistance andgrain yield; suggesting that breeding for maize weevil resistance can be achieved without compromising grain yield. Eight weevil resistant and two susceptible inbred line parents were crossed in a 10 x 10 full diallel mating design and the resulting 45 experimental hybrids and their reciprocal crosses evaluated for grain yield and secondary traits under four environments, and also to determine the gene action regulating their expression. The F1 hybrid seed, F2 full-sib and F2 half-sib grain generated from the 45 experimental hybrids and their reciprocals under two environments in Namulonge, were evaluated for weevil resistance using F1 weevil progeny emergence, median development period (MDP), Dobie’s index of susceptibility (DIS), and parental weevil mortality as susceptibility parameters. The general combing ability (GCA), specific combining ability (SCA), and reciprocal effects were all significant for grain yield, with SCA accounting for over 80% of the hybrid sum of squares. Inbred line parent MV44 exhibited positive significant GCA for grain yield and thus can be utilized in the development of synthetics and hybrids. Hybrids MV21 x MV13, MV154 x MV44, and MV154 x MV102 and all hybrids between parent MV142 and the rest of the parental lines exhibited positive and significant SCA effects. For the weevil resistance parameters, the general combining ability (GCA), specific combining ability (SCA) and reciprocal effects were all significant for F1 weevil progeny emergence, MDP, and DIS in the three seed categories. The results revealed that weevil resistance was governed by additive gene action, non-additive, and maternal effects. Parents MV170 and MV142 were consistently exhibiting weevil resistance in the three seed categories and thus recommended for future breeding strategies. Furthermore, most of the hybrids generated from parental line M142 were noted to exhibit outstanding performance in terms of grain yield and weevil resistance. Another study was conducted to determine the effectiveness of two cycles of modified S1 recurrent selection towards the improvement of weevil resistance in a maize population Longe5. Over 540 selfed ears were selected from the source population (C0) and screened for weevil resistance in the laboratory at Namulonge. Based on weevil resistance characteristics, 162 genotypes were selected from C0 and recombined in an isolated field to generate cycle C1. The same procedure was used for generating cycle C2 from cycle C1, but instead 190 weevil resistant C1 genotypes were selected and recombined to form C2. Seed from cycles C1 and C2, together with that from the source population (C0), was used to plant an evaluation trial in three locations, to compare the performance of the three cycles in terms of grain yield and reaction to the major foliar diseases, and also to produce seed for subsequent screening against weevil ii infestation. A total of 54 seed samples were screened for weevil resistance in a laboratory at Namulonge, in an experiment laid out in a randomized complete block design. A reduction in grain weight loss of 65% was registered in the C2 seed, whereas in C1 seed it was 15%. A similar trend was observed for F1 weevil progeny emergence and grain damage. Grain yield results indicated a yield gain of 19% realized from cycle C2 while a yield gain of 7% was realized from cycle C1. Furthermore, reductions in disease severity of 27%, 10% and 13% were exhibited for Turcicum leaf blight (TLB), grey leaf spot (GLS) and rust disease, respectively in cycle C2. The results indicated that Longe5 can be improved for maize weevil resistance, grain yield, and resistance to foliar diseases through selection. Further recurrent selection cycles would be recommended. The last study was aimed at evaluating the potential of shelled grain and suspended ear options of the weevil warehouse technique in discriminating maize genotypes into different susceptibility classes, based on genotype response to weevil attack. It involved comparing the effectiveness of the two options under the weevil warehouse technique with the laboratory bioassay technique using grain damage and grain weight loss as the maize grain susceptibility parameters. Fourteen maize genotypes were screened using the weevil warehouse and the laboratory bioassay techniques at Namulonge. On grouping the 14 genotypes into different response classes, high levels of consistency were observed in the three screening techniques. Therefore, the two weevil warehouse screening options being faster and effective in discriminating maize cultivars towards weevil attack, they were found to be better than the laboratory bioassay technique. The minimum evaluation period required to discriminate genotypes by the two weevil warehouse options was two months from the onset of the experiment. The maize weevil was noted to be an important storage pest constraining maize production in Uganda. The major weevil control measures included proper postharvest handling procedures and use of indigenous technical knowledge. The results also revealed that host plant resistance could significantly reduce grain damage. It was further revealed that grain resistance against the maize weevil could be enhanced through hybridization and recurrent selection; thus the germplasm identified in the study can provide new sources of maize weevil resistance for commercial deployment and further breeding. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2013.

Maize (Zea mays L.) supports millions of livelihoods in sub-Saharan Africa (SSA) in terms of food and feed. Production of the crop is however limited by several factors, among these, maize streak virus (MSV) disease. Although extensively studied, MSV remains a serious problem in SSA due to several challenges in breeding MSV resistant maize varieties. These include integration of MSV resistant germplasm from different backgrounds, reliance on a few resistant sources, and genotype x environment interactions. This study was designed to assess the breeding potential of several MSV resistant lines in hybrid combinations. Understanding architecture of genetic divergence and background of these genotypes would greatly aid in breeding high yielding and stable MSV resistant hybrids. Experiments were conducted during 2010 to 2012 seasons in Kenya. Diallel crosses and SSR markers were used to characterize MSV resistant maize inbred lines from three programs of CIMMYT, KARI and IITA. In general, this study revealed that MSV is still an important problem in Kenya with high incidence and severity levels in the farmers’ fields. The levels of MSV resistance in locally grown hybrids needs to be improved. Farmers challenged breeders to develop new hybrids that combine early maturing, high yield potential and MSV resistance. The study was successful in identifying the best eight inbred lines for use in breeding new maize hybrids with MSV resistance. The nature of gene effects was established for the first time, in particular the role of epistasis and G x E in conditioning MSV resistance in hybrids. Results indicate serious implications for previous models that ignored epistasis in studying MSV resistance in maize. The inbreds Z419, S558, CML509 and Osu23i, displayed high levels of epistasis for MSV resistance. Unless strong sources of MSV resistance, such as MUL114 and CML509, are used, breeding resistant hybrids will require parents that carry dominant resistance genes. The additive-dominance model was adequate to explain northern leaf blight (NLB) resistance in hybrids, indicating fewer complications in breeding NLB resistant hybrids. The study also reveals that SSR genetic distance data can be used to predict hybrid performance, especially when the correct set of markers is used. Many previous studies have not found any significant relationship between genetic distance and heterosis, due to large G x E and use of a wrong set of markers. The diallel analysis and SSR data established the important heterotic groups, which will be exploited for efficient development of MSV resistant maize hybrids. These strategies will be recommended to programs that emphasize MSV resistance in maize hybrids. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2013.