RT Dissertation/Thesis T1 Optimizing selection efficiency in maize for the drought prone eastern and southern African environments A1 Kebede,Aida WP 2013/10/10 AB Breeding for stress tolerance is the most cost effective way of avoiding drought-induced yield reduction in the tropics. Optimizing breeding for drought tolerance at CIMMYT could enhance the effectiveness of this multi-national breeding program and warrant fast delivery of drought tolerant materials to the farmers. Thus, the overall aim of my study was to improve the efficiency of drought tolerance breeding of maize at CIMMYT for the rapid and cost effective advancement of drought tolerant materials for the drought prone regions of the ESA (Eastern and Southern African countries). We screened a diverse source of tropical germplasm for their haploid induction rate (HIR) and the seasonal variation of this trait. We then compared various managed drought and well watered experiments conducted as line per se performance trials (LP) and testcross performance trials (TP) in Kenya and Mexico. Further, we estimated the relative selection efficiency of the principal breeding regions of CIMMYT for the tropics in ESA and Mexico with unselected and selected breeding materials. The specific objectives of my study were to (1) monitor the variation for HIR among diverse source germplasm in tropical maize, (2) determine the relative importance of general (GCA) and specific (SCA) combining abilities of the source germplasm for HIR, (3) investigate the influence of tropical summer and winter seasons and genotype × season interactions on this trait, (4) determine if LP is predictive of TP for yield under drought in sets of lines under development by the CIMMYT maize breeding program in Kenya and Mexico, (5) determine the genetic correlation between performance of lines per se under drought and testcrosses under optimal conditions and assessing its effect on yield potential, (6) examine the correlation between TP under well-watered and drought stress conditions for potential indirect selection efficiency of well-watered conditions in comparison with drought stress, (7) determine the relative importance of regional adaptation of maize hybrids to Mexico and ESA by subdividing the genotype × environment interactions and determining genotypic correlations between both regions, (8) calculate the indirect selection efficiency for selecting materials based on test results from one region on the selection gain in the other region, and (9) identify the most suitable stage for exchanging breeding materials between Mexico and ESA. Source germplasm and induction season affected HIR and MCR (mis-classification rate) considerably in tropical maize. Source germplam with high HIR and low MCR could be used in the initial stage of implementing the DH technology in the tropics. GCA effect was more important than SCA or genotype × season interaction effects for HIR in tropical maize. Thus, enhancing HIR in source germplasm can be achieved through cyclical breeding or recurrent selection. Winter season was considered the best season for induction because it provides suitable environmental conditions for higher HIR and lower MCR. Overall HIR was high enough to apply the in vivo DH technology in the routine breeding activities in tropical maize. There were moderate genotypic correlation and ISE (Indirect Selection Efficiency) values between LP and TP under drought that increased with an increase in stress level. Hence, LP trials were predictive of TP trials particularly under severe drought stress. Furthermore, screening of lines for LP under drought stress did not compromise yield potential. TP under well-watered conditions were not predictive of TP under drought stress emphasizing the need of managed drought trials to identify drought tolerant materials. With the current shift of inbred development to large scale DH line production, LP evaluations can reduce the cost of making large numbers of testcrosses and optimize breeding for drought tolerant hybrids in the tropics. The exchange of breeding materials between ESA and Mexico can be done with early and late generation materials. This is because there was negligible genotype by region interactions as compared to genotype by location interactions within each region and high genotypic correlations between the two regions. Further, ISE estimates for trials conducted in Mexico and in ESA were high. Adaptive diseases for each location might hamper the exchange of materials, however, with current molecular marker tools like marker assisted selection and genomic selection, the problem of selecting for disease resistance in the region where the disease is not prevalent seems promising. In conclusion, there are ample opportunities in the CIMMYT maize breeding program to optimize breeding for drought tolerance in the tropics through rapid and large scale production of DH lines and evaluation of these lines for LP in managed drought trials. Moreover, breeders from ESA and Mexico could benefit from each other?s materials and test results by regular exchange of breeding materials at both the early and late stages of testing. K1 Mais K1 Trockenheit K1 Afrika PP Hohenheim PB Kommunikations-, Informations- und Medienzentrum der Universität Hohenheim UL http://opus.uni-hohenheim.de/volltexte/2013/881