|dc.description.abstract||Development of testers from new inbred lines that are high yielding and high discriminating abilities in diverse environments, and in stress conditions is very important in maize breeding in southern and eastern Africa. Genetic gain decreases when new lines are combined with old testers. It is, therefore, important to develop new testers that perform better in harsh environmental conditions to replace the old single-cross testers; CML 312 x CML 442 (heterotic group A) and CML 444 x CML 395 (heterotic group B). These are currently dominating the maize breeding programmes at the Agricultural Research Council (ARC) and most of the National Agricultural Research stations (NARS) in eastern and southern Africa. The objectives of this study were therefore; i) to identify elite single-cross hybrids suitable for further evaluation as potential testers in development of three-way cross hybrids, ii) to determine the correlations between grain yield and secondary traits in single-cross maize hybrids, and iii) to estimate phenotypic and genetic variance components, heritability and genetic advances for yield and its related components in single-cross maize hybrids. The trials were established under three different environments which are: random drought (RD), optimum environment (OPT) and low nitrogen (Low N) at ARC Potchefstroom and Cedara using alpha lattice 0, 1 design (32 x 5). The single-cross hybrids used in this study were obtained from CIMMYT Zimbabwe and they belong to two different heterotic groups as follows;, heterotic group A with 160 entries among which 155 were experimental single-cross hybrids and five were check entries and heterotic group B with 160 entries among which 157 were experimental single-cross hybrids and three were hybrid checks.
High significant differences (P< 0.01) among single-cross hybrids were observed on days to 50% anthesis (AD), grain yield (GY), ear height (EH), and ears per plant (EPP) in heterotic group B under optimum environment. Hybrid 139 had a mean yield of 8.65 t/ha, which was higher than the average grain yield of 6.22 t/ha. Days to 50% anthesis (AD), anthesis-silk interval (ASI), plant height (PH), and EH varied significantly (p<0.01) among single-cross hybrids in the low nitrogen environment. Hybrid 65 had a mean yield of 4.04 t/ha, which was higher than the average yield of 2.29 t/ha. Days to 50% anthesis were slightly higher in low nitrogen environment than in optimal environment. Hybrid 92 had a mean yield of 7.53 t/ha, which was higher than the average yield of 4.77 t/ha in random drought environment. As in heterotic group B, significant variations were observed in heterotic group A, in random drought and optimum environments. Grain yields of 9.44 t/ha and 6.42 t/ha for maize hybrids 134 and 52 were higher than average mean yields of 6.75 t/ha and 3.43 t/ha from optimum and random drought environments, respectively. Hybrids with higher trait values than the average may be advanced for further use in breeding in their respective environments. Maize single-cross hybrids 1, 23, 127, 15, 122, 8, 134, 109, 34, and 31 from heterotic group A and 69, 81, 65, 97, 92, 40, 117, 58, 101, and 44 from heterotic group B were selected for further use in breeding. The hybrids had consistently higher mean yields across the environments.
Significant, positive and negative correlations were observed among secondary traits in all environments. Yield (t/ha) was positively correlated with cob length (CBL), EH, field weight (FW), grain weight (GW) and shelling percentage (SP). It was however, negatively correlated with days to 50% silking (SD) and days to 50% anthesis (AD). A breeding programme aimed at improving CBL, EH, FW, GW and SP and reducing traits SD and AD may indirectly result in improvement of maize yields in random and optimum environments. High genotypic coefficient of variation (GCV) and phenotypic coefficient of variation (PCV) were observed for ASI, yield t/ha, PH and ear height therefore, and selection can be done. Improvement of maize yield based on AD and ASI selection would be successful due to their high broad sense heritability estimates. The genetic advance observed in this study were high and therefore some trait values can be increased in the next generation through selection.||en_US