|dc.description.abstract||Maize (Zea mays L.), is one of the staple crops in Rwanda that contributes to national economic growth. Furthermore, the genetic plasticity of maize permits its adaptation to a wide spectrum of environments ranging from 900 m to over 2400 m above sea level (asl). However, grain yield is compromised by various limiting factors, among these, the lack of appropriate varieties, especially hybrids and scarcity of maize seed of varieties that can withstand various production constraints. Among other factors, productivity can be enhanced by developing a range of hybrids, which are higher yielding than open pollinated varieties. However, to lay a strong foundation for a viable hybrid-breeding programme, knowledge on genetic diversity, genetic effects governing yield and other traits in inbred lines and effective germplasm management requiring heterotic groups and patterns establishment is needed. The objectives of this study were, therefore; i) to determine the genetic distances and clusters among potential lines selected for the mid-altitudes and highlands of Rwanda; ii) to estimate the general and specific combining ability, heterosis and gene action for grain yield; iii) to determine heterotic groups and heterotic patterns among Rwandan newly developed lines and introduced lines based on line x tester mating scheme and diallel analysis, respectively; and iv) to investigate the magnitude of genotype–by-environment (G x E) interaction and stability of new hybrids for grain yield in the target environments.
To determine genetic diversity; 71 maize inbred lines selected for the mid-altitudes and highlands of Rwanda were genotyped with ninety two SNP markers. The unweighted pair group method with arithmetic mean (UPGMA) revealed a random allocation of the inbred lines into two major clusters regardless of their origin. Genetic clustering information acquired from the current study would be suitable information not only for maize hybrid programme establishment in Rwanda, but also for other collaborative tropical maize breeding programmes.
Estimation of the general and specific combining ability, heterosis and gene action for grain yield was done using forty-five single cross hybrids from a 10 x 10 half- diallel mating design. Among these parents, three of them were adopted as testers. The hybrids were evaluated in a 6 x 8 (forty-five crosses plus three checks) alpha-lattice design across twelve environments in Rwanda. General combining ability (GCA) and specific combining ability (SCA) effects were both highly significant (P<0.001-0.01) suggesting the presence of both additive and non-additive effects, but with higher magnitude of GCA for grain yield effects when all environments were combined.
The highest heterotic patterns were realized between groups S4 and S6/S7 (S4/S5) and within S4 group (S4/S8) and would be potentially useful for maize hybrid production in Rwanda.
Furthermore, nineteen maize inbred lines were crossed with four testers (20(T1), 21(T2), 22(T3), 23(T4)), following a line x tester mating scheme and generating 76 test crosses. These were evaluated together with two checks in 6 x 13 α-lattice design at four locations in 2015B and 2016A seasons, along with their 23 parental lines in adjacent trials. Generally, most of the lines exhibited positive heterosis with all testers. However, there was more inclination firstly towards tester T2 and then T3. The highest heterosis was displayed by line 8 with T3. Regardless of heterotic grouping method applied, the lines were discriminated in different heterotic groups different from the four heterotic groups of the testers. Two and nine heterotic groups were identified based on standard heterosis and SCA effects, respectively. Genetic distance was correlated to heterosis, SCA effects and test cross performance however, this was specific to some testers.
To investigate the magnitude of G x E interactions and stability of new hybrids for grain yield in the target environments; 126 experimental hybrids were evaluated in four environments representing the major agro-ecologies of Rwanda. One set of 78 hybrids was evaluated over two seasons (8 environments in total), while the other set of 48 hybrids was evaluated over three seasons (12 environments in total). Genotype and genotype by environment interaction (GGE) biplot method was applied for graphical display of the data. Hybrid 26 (ACR29 x 21) and 31(ECA1 x 22) from test crosses and diallel hybrid 3 (R10164 x ET4) and 25 (ET4 x ECA13) were identified as the best performers and then qualified as desirable hybrids. The GGE biplot revealed three mega-environments for test crosses and two mega-environments for diallel hybrids. Environments Rwerere first season (RWA), Rwerere second season (RWB) and Rubona second season (RBB) for test crosses and Rubona first season (RB1), Rubona second season (RB2) and Nyagatare second season (NY2) for the single crosses were the most powerful in discriminating genotypes.
Overall, the acquired information from genetic diversity and heterotic groupings is useful in designing the hybrid maize programme in Rwanda. This will guide the programme towards identifying suitable heterotic patterns as well as combining ability of the inbred lines selected from this study. Furthermore, the study revealed valuable maize inbred lines with desirable combining ability and new single cross hybrids. Consequently, the maize breeding programme will consider development of hybrids, such as single crosses and three way crosses using the inbred lines and F1 hybrids identified.||en_US