Genetic Diversity among CIMMYT Maize Inbred Lines Investigated with SSR Markers: I. Lowland Tropical Maize miscellaneous origin but a small proportion of flint lines. De

Abstract

sal et al., 1992a). However, the mixed genetic constitution of the populations and pools made the task of Detailed knowledge about the genetic diversity among germplasm assigning them to genetically diverse and complemen-is important for hybrid maize (Zea mays L.) breeding. The objectives of our study were to (i) investigate the genetic diversity among tropical tary heterotic groups difficult. Nevertheless, the germ-lowland inbred lines and (ii) delimit heterotic groups in this germ-plasm was categorized based on their yield performance plasm. Simple sequence repeat (SSR) analysis of 79 markers distrib-into different heterotic groups, and some promising het-uted uniformly throughout the maize genome was performed for 155 erotic patterns are under development in RRS programs tropical lowland inbred lines from the International Maize and Wheat (Vasal et al., 1999). Improvement Center (CIMMYT). Inbred lines were extracted from Since its inception in 1984, the hybrid maize program 60 broad-based populations and pools with mixed origin. We observed of CIMMYT has developed and released 497 CIMMYT an average of 7.4 alleles per marker with a range from 2 to 18. The maize inbred lines (CMLs) derived from the above men-polymorphic information content (PIC) of the 79 SSRs ranged from tioned broad-based pools and populations. These CMLs, 0.13 to 0.87, with an average of 0.60. Cluster analysis of the tropical and CIMMYT germplasm in general, have played an yellow and white lines revealed a lack of structure within this germ-plasm, which can be explained by the mixed origin of the populations important role in hybrid maize production in developing used to extract these lines and the specific choice of testers for recipro-countries (Morris, 2001). The lowland tropical CMLs cal recurrent selection (RRS). An SSR analysis can support the choice were selected from approximately 60 populations or of representative testers for evaluating inbred lines in such broad-pools based on their per se performance and combining based populations. Thus, a more complete exploitation of the out-ability when crossed to testers. Little information is standing flint-dent heterotic pattern is possible. available on the relationships among these populations and pools, and consequently on the CMLs derived from them. F ew agronomic improvements during the 20th cen-Detailed knowledge regarding genetic diversity and tury rival the development of hybrid maize (Brum-the relationship among breeding materials is indispens-mer, 1999). Yields in maize increased dramatically as able for the development of new maize inbred lines, the breeders moved away from open-pollinated cultivars assignment of maize inbred lines to heterotic groups, and began developing doublecross and later single-cross and the choice of testers for trials of hybrid combina-hybrids (Duvick, 2001). This yield advance can be attrib-tions in maize breeding. Molecular genetic markers are uted to the successful harnessing of heterosis. Clearly a powerful tool to delimit heterotic groups and to assign defined heterotic groups and patterns improved by RRS inbred lines into existing heterotic groups (Melchinger, programs are of fundamental importance for a system-1999). The SSR markers offer advantages in reliability, atic exploitation of heterosis. In temperate maize, heter-reproducibility, discrimination, standardization, and cost otic patterns were established empirically by relating the effectiveness over other marker types (Smith et al., 1997). heterosis observed in crosses with the origin of the parents The objectives of our study were to (i) investigate the included in the crosses (Hallauer et al., 1988). genetic diversity among 155 CIMMYT tropical lowland Breeding efforts at CIMMYT in the early 1960s and inbred lines with 79 SSR markers and (ii) delimit heter-1970s were focused on intrapopulation improvement otic groups in this germplasm. via recurrent selection based on the formation of 100 populations and 30 genetically broad-based backup MATERIALS AND METHODS pools. The formation of these populations disregarded known racial complexes (Vasal et al., 1999). With the Plant Materials decision to embark on a hybrid breeding program, sev-In the present study, 86 white and 69 yellow CMLs repre-eral diallel studies were performed to identify suitable senting the lowland tropical gene pool were analyzed with 79 germplasm for hybrid breeding (Crossa et al., 1990; Va-SSR markers (Table 1). The lines included represent nearly all the lowland tropical maize inbreds created by CIMMYT breeders, excluding many of the very closely related sister X.C. Xia, Inst. of Crop Breeding and Cultivation, Chinese Academy lines. These lines were extracted from populations, which were of Agricultural Sciences, Zhongguancun South Street No. 12, 100081, established mostly by intermixing germplasm from different Beijing, China; D.A. Hoisington and M.L. Warburton, CIMMYT, Int. racial complexes resulting in a huge number of lines with