Popcorn is increasingly becoming popular as a snack and is consumed widely all over the world. It is a high value crop, with possible multiplier effects like income generation for the under-resourced communities in the second economy. Despite its popularity, developing countries are battling to meet the demand and rely on importing popcorn grain due to challenges which include poor agronomic traits and slow breeding progress. Most of the imported varieties are not adapted to stress-prone local environments, which are prevalent in tropical sub-Saharan Africa. The objective of the study was to evaluate newly developed hybrids and inbred lines for agronomic and popping quality traits with the possibility for commercialization in future. The study aimed at determining variability for popping ability in inbred lines and hybrids, grain yield and its secondary traits, the nature of gene action, relationships among agronomic and popping quality traits, effect of genotype x environment interaction on agronomic traits and popping method x genotype interaction effects. To determine popping ability, 128 inbred lines were evaluated at the University of KwaZulu-Natal, South Africa, in June 2011 using two popping methods, the microwave method and the hot-air method. The popping quality attributes measured were flake volume, popping fold, number of unpopped kernels, kernel size and quality score. Variability among inbred lines was significant (P<0.05) for all traits. Flake volume ranged from 63 cm3 to 850 cm3, popping fold ranged from 2.5 to 34 times the original volume. Kernel size had a significant positive correlation (r= 0.49) with the number of unpopped kernels. There was a significant strong and negative correlation between flake volume and the number of unpopped kernels (r= -0.62), indicating that either of the two traits would be effective for measuring popping ability. Experimental hybrids were then developed from 87 out of the possible 128 inbred lines. Only the inbred lines with sufficient seed were crossed to develop hybrids. Random crosses were generated at Makhathini Research Station during the winter season of 2011. Crosses were made at random among parents that managed to synchronize their flowering dates, resulting in 119 hybrids with sufficient seed for planting in trials. To determine agronomic superiority, the 119 experimental hybrids and the standard check P618 were evaluated at the Cedara Research Station and Ukulinga Research Farm in the Midlands of KwaZulu-Natal during the summer of 2011/2012. The experiments were laid out as 10 x 12 alpha lattice design, with two replications at each site. Standard cultural practices for maize were followed. The data were subjected to analysis of variance and line x tester analysis in Genstat and SAS statistical programmes. Results indicated that hybrids were significantly different for all agronomic traits. Means for grain yield ranged from 1.0 t/ha to 5.2 t/ ha. General combining ability effects were significant for all agronomic traits, suggesting that additive gene effects were governing these traits. Specific combining ability effects were significant for ear length, number of ears per plant and yield indicating, that non-additive gene effects were influential for these traits. Generally, agronomic traits were highly heritable. Grain yield showed significant and positive correlation with ear length, plant height, ear position, shelling percentage and number of ears per plant, indicating that these were the major yield-determining secondary traits which should be enhanced in popcorn. Although site main effects were highly significant for secondary traits, the hybrid x site interaction was not significant. The results therefore indicate that the hybrids were ranked similarly at both sites. The 119 experimental hybrids and the standard check P618 were evaluated for popping quality, using the microwave and the hot-air popping method. There was a significant variability observed among hybrids for popping quality traits. Flake volume across sites and across popping methods ranged from 734 cm3 to 1288 cm3. Popping fold ranged from 14.69 to 25.75 times the original volume. Additive gene action was more prominent than non-additive action for all popping quality traits. The SCA effects were significant for flake volume, popping fold and number of kernels per 10 g. All popping quality traits had high heritability, indicating that selection would be effective to improve popping. Flake volume was negatively correlated to quality score, indicating that popping expansion is reflected on the quality score and a significant negative correlation between flake volume and number of unpopped kernels. There was significant and strong positive correlation between kernel size and number of unpopped kernels. Hybrid x site interaction was only significant for quality score and kernel size. Hybrid x method interaction was not significant, indicating that popping ability was not dependent on the method. Inbred lines showed significant variation for popping quality and therefore have utility for hybrid development. Significant genotypic variation was also observed among hybrids for agronomic and popping quality traits. Additive gene action was predominantly responsible for both agronomic and popping quality traits. Both agronomic and popping quality traits were highly heritable and positive relationships were identified among traits. Overall, the study indicates opportunities for further breeding progress through selection. / Thesis (M.Sc.)-Unversity of KwaZulu-Natal, Pietermaritzburg, 2012

Maize (Zea mays L.) is one of the most important food crops in sub-Saharan Africa (SSA); however its production is constrained by many factors. Grain yield is compromised by poor genetic performance and poor agronomic management. This calls for need to develop hybrids and exploiting heterosis of single crosses which are adapted to challenging environments. Currently, there is no popcorn hybrids developed in South Africa which is adapted to local conditions. As such, there is need to develop hybrids that cater for smallscale farmers in marginal environments. The objectives of the study were to determine i) standard heterosis, levels of variation and heritability for phenotypic traits in dent and flint maize hybrids; ii) the association between genetic distances and phenotypic traits in dent and flint maize hybrids; iii) mid-parent heterosis in popcorn hybrids, iv) the effect of secondary traits on grain yield in dent, flint and popcorn hybrids; v) genetic diversity and the relationship between traits in widely grown selected hybrids in Southern Africa; and vii) to compare effectiveness of phenotypic analysis models for determining genetic distances between hybrids. Popcorn, dent and flint hybrids were evaluated at two sites. The data was analysed using SAS, Genstat and Power marker statistical packages. The results revealed that the relationship between genetic distance and heterosis is dependent on the environment. Hybrids in top 10 at both sites were different indicating that there was a significant genotype x environment interaction. 13 new heterotic patterns that performed better than the controls can be utilized in heterosis breeding; however there is need to test them in different environments to check on their stability. Grain texture cannot be used to discriminate hybrids for yield because all patterns of dent x dent, dent x flint and flint x flint were present in the top 10 hybrids. Lines DXL124 and DXL158 dominated parentage of the top 10 hybrid rank for yield qualifying them as potential testers for specific combining ability in future studies. Heterosis in popcorn hybrids that performed better than the mid-parent can be utilized in heterosis breeding to exploit vigour, though there is need to test the hybrids in a number of different environments. The main direct factors contributing to yield were ear prolificacy, ear aspect, number of plants and shelling percentages qualifying them to be selected to boost grain yield. Phenotypic data and 91 SNP markers were used to estimate the genetic distance between the hybrids. The results indicated that hybrids that were in the same cluster belong to the same brand and were related in origin and pedigree. Both molecular and phenotypic data were effective in discriminating the hybrids into different clusters according to genetic background. SNP markers revealed nine clusters of hybrids, 12-trait model revealed eight clusters and five-trait model revealed six clusters at 85% genetic distance. The study indicates strategies that can be adopted to boost grain yield in dent, flint and popcorn hybrids. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2013.