Temperate and sub-tropical maize germplasm: Heterotic patterns, combining abilities and genetic similarities

This study emphasizes: (1) The narrow germplasm base of commercial US maize (Zea mays L.) and the potential of South African germplasm in broadening this base. (2) The use of US germplasm to enhance the productivity of South African germplasm.; The main objective was to rationalize the heterotic patterns prevalent in Southern African germplasm by conducting two studies. The first was evaluating a ten parent diallel cross between elite sub-tropical and tropical inbreds from South Africa, Zimbabwe, Argentina and the US. In the second study, nine South African public inbreds were evaluated in testcrosses with Reid Yellow Dent (RYD) and Lancaster Sure Crop (LAN). The second objective, was evaluating the entries for resistance to the foliar disease, Gray Leaf Spot ( Cercospora zeae-maydis). The final objective was using Simple Sequence Repeats (SSR) and genetic similarities among entries to assign the inbreds to either the RYD or LAN heterotic groups.; From a US perspective we conclude: (1) An introgression approach is advocated. (2) RYD (in one form or another) remains the female and tester of choice. (3) The inbreds A272, E2558W, NPP, SWZ and POS (except for white maize) have no immediate value. (4) Breeding crosses of ITY (I137TN), M37W and FTY (F2834T) with either Iodent (IDT) or LAN and selection for yield, dry-down and Gray Leaf Spot (GLS) resistance. (5) D940Y should be considered as an alternative to M37W.; From a South African perspective: (1) Immediate gains are possible with using Corn Belt germplasm. (2) Maintain ITY (I137TN), M37W and FTY (F2834T) as separate heterotic groupings. RYD and LAN, which have poor per se adaptation to South Africa, could be used on the male side. (3) IDT has potential for use in breeding crosses with South African genotypes.; There are alleles for GLS resistance present in the South African germplasm not found in Corn Belt germplasm. A suggested breeding strategy would be using the South African genotypes in breeding crosses with LAN and testcross with RYD.; Using 70 SSR markers, we were able to make assignments that are accurate to either the RYD or LAN heterotic groups.