Maize grey leaf spot (GLS), caused by Cercospora zeae-maydis, and potato late blight (LB), caused by Phytophthora infestans, are foliar diseases of maize and potato, two of the most widely grown crops in KwaZulu-Natal (KZN), after sugarcane and timber. Commercial maize in KZN accounts for just on 4.3% of the national maize crop. This is worth R563 million using an average of the yellow and white maize price for the 2001/02 season (at R1 332.87 ton(-1)). In 2003 KZN produced about 5% of the national potato crop (summer crop: 7531 300 10kg pockets from 2243 hectares). This equates to a gross value of R89.4 million based on an average price of R1 188 ton(-1) in 2001. Successful commercial production of maize and potatoes depends upon control of these diseases by translaminar fungicides with highly specific modes of action. This study extends an existing model available for timing of fungicide sprays for GLS and tests and compares two LB models for two calendar-based spray programmes. The study also evaluated the use of an early blight model which is caused by Alternaria solani, and over the single season of evaluation showed potential for use in KZN. For the GLS model it was found that a number of refinements are needed, e.g., the amount of infected maize stubble at planting and not the total amount of maize residue at planting. Based on two years' data, it was found that for the LB models there are no significant differences in levels of control between using a predicted fungicide programme and a calendar-based programme. The importance of knowing initial infection sites, and hence initial inoculum, was demonstrated. This led to the creation of a KZN LB incidence map, now being used to more accurately time the start of a preventative spray programme and to time the inclusion of systemic fungicides in the preventative spray programme. This study has contributed to the further development and expansion of the Automatic Weather Station Network (AWSN) at Cedara, which now comprises 15 automatic weather stations in KZN. The AWSN is currently used to aid farmers and advisers in decision-making regarding fungicide spray timing for GLS and LB. / Thesis (M.Sc.Agric.)-University of Natal, Pietermaritzburg, 2003.

Maize (Zea mays L.) streak virus disease (MSD) is the most important virus disease in Africa but farmers are unaware of its status. A project was initiated to assess the current status of MSD and to breed for its resistance. Four populations comprised of two BC1F1 and two F2 progenies developed by backcrossing and selfing the F1 progenies of two crosses between a donor line (CMl 202) and two susceptible lines (CMl 321 and CMl 384) were developed. Conventional and molecular marker assisted selection (MAS) methods were used to screen for resistance to MSD in each of the four populations. To facilitate unbiased comparison, separate screening nurseries were established for MAS and conventional screening. The objectives of the study were five-fold; 1) to assess the status of MSD in Uganda and understand farmers' preferences and varietal selection criteria for maize using a participatory rural appraisal (PRA), 2) to screen for MSD resistance in early generations of segregating maize populations using conventional method, 3) to screen for resistance to MSD using SSR marker assisted selection , 4) to compare the effectiveness of marker assisted selection and conventional methods for selection for resistance to MSD, and 5) to compare costs associated with MAS and conventional selection methods. Results of PRA showed that unreliable rainfall and insect pests were the dominant constraints to maize productivity in Uganda. Diseases were ranked fifth among the production constraints . Maize streak virus disease was considered the most important disease constraint. Farmers showed common preference for high yielding and early maturing cultivars. However, farmers had other special preferences which were diverse and included large, white and high test density kernels for marketing, and sweet taste, particularly for home consumption. Farmers' research priorities included tolerance to drought, resistance to insect pests and diseases, sweetness, prolificacy, resistance to lodging, and drooping leaves because theyt cover the soil fast and prevent weed growth. Conventional screening for resistance to MSD showed that backcross and selfing populations segregated in 1:1 and 3:1 Mendelian ratios confirming the presence of one major gene with simple inheritance . Severity and incidence of disease were positively correlated suggesting a non-reference by the insects. In the selfing populations, the presence of complete esistance against MSD was suggested because frequency distribution patterns were highly skewed in favour of resistance. There was a decrease in disease severities with selection from BC1F1 to BC2F1 and from F2 to F3 generations indicating that high response to selection was achieved. On the other hand, one marker, umc1917, consistently polymorphic and eo-dominant was selected and used in MAS protocol. Results showed that the observed outcomes fitted the expected ratio of 1:2:1 for a F2 population and 1:1 for a BC1F1 population (X2 not significant). Evaluation of F3 and BC2F1 progeny selected using markers showed low disease severity suggesting that marker assisted selection was effective. However, the study showed that the presence of the O'Tl, was not consistent with symptom expression in the field. Evaluation of lines in three-way crosses identified ten potential lines that were high yielding, highly resistant to MSD and stable across three locations. Both MAS and conventional selection were equally effective in identifying high yielding lines although resistance was higher under MAS. Costs of MAS and conventional method varied depending on the units for comparison. The total costs of conventional method were higher than that of MAS in both first and second selection cycles. Comparing costs per row for conventional and costs per plant or data point for MAS showed that conventional selection was 2.4 times more expensive than costs per sample for MAS. However, costs per plant for MAS were 6.6 times higher than for conventional selection. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2006