TY  - THES
T1  - Genetic diversity in elite lines and landraces of CIMMYT spring bread wheat and hybrid performance of crosses among elite germplasm
T3  - Crop Science, Euphytica, Molecular Breeding
A1  - Dreisigacker,Susanne
Y1  - 2005/10/06
N2  - Wheat (Triticum aestivum) is one of the major cereals in the world. During the past years, the world consumption of wheat increased up to nearly 600 million tones, whereas wheat production continuously decreased. Due to land limitations, new production gains must be achieved from improved plant management systems as well as from the development of high yielding varieties. The International Maize and Wheat Improvement Center (CIMMYT) employs different strategies to enhance yield potential in wheat especially for developing countries. For instance, the wheat breeding program focuses on defined mega-environments (MEs), assuming similar growing conditions in certain countries. In the search for useful alleles, breeders often turn back to wild relatives of wheat stored in the CIMMYT gene bank. With the production of synthetic hexaploid bread wheat (SHWs), characteristics from T. durum and T. tauschii can be combined and via backcrossing incorporated into modern breeding materials. Wheat landraces (LCs) are an additional reservoir of resistances to pests and diseases as well as for environmental adaptation. The production of wheat hybrids is seen as a further option to improve yield potential. A considerable amount of genetic diversity among the materials is a prerequisite for all strategies. Due to the worldwide importance of CIMMYT wheat varieties, they represent a suitable source to examine different breeding strategies in wheat. The main objective of our research was to determine the genetic diversity in modern wheat breeding materials and genetic resources at CIMMYT. Specific research questions were: (i) Is the systematic breeding targeted for different MEs reflected in the genetic diversity among breeding lines (Experiment 1)? (ii) Does the production of SHWs (Experiment 2) and the use of LCs (Experiment 3) enhance the genetic variation in modern breeding materials? (iii) Does the development of hybrids represent an option to improve yield potential in wheat? (iv) Is it possible to predict levels of heterosis with the determination of genetic distance (GD) among hybrid parents? (v) Do genomic and EST- derived SSRs differ in the measurement of genetic diversity (Experiments 1 and 3)? (vi) Are GD values based on SSRs correlated with the coefficient of parentage (COP) (Experiments 1 to 4)? In Experiment 1, a total of 68 CIMMYT advanced breeding lines was analyzed with 99 SSRs, of which 51 were EST- and 46 genomic derived SSRs. A high level of genetic diversity (GD = 0.41) was observed among the breeding lines. The majority of variation (91%) was detected among lines targeted to one specific ME, which indicates a broad genetic base of the current CIMMYT breeding materials. Principal coordinate analysis (PCoA) could clearly separate the lines, but they clustered independently from their target MEs. Main explanations are: (i) alleles were selected that provide fitness to several MEs, (ii) adaptation depends only on a small number of genes that were not detected with the SSRs applied, or (iii) too few cycles of selection were considered to separate the germplasm. In Experiment 2, a total of 11 SHWs, 7 recurrent parent lines, and 13 families of backcross-derived lines (SBLs) were analyzed with 90 SSRs. The SHWs clustered far from the SBLs and the recurrent parents in the cluster analyses and PCoA, and formed a distinct germplasm pool with high allelic variation. Two families of SBLs were tested for a selective advantage of the SHW alleles. Six SSRs revealed non-Mendelian inheritance, indicating that the genomic region of SHWs was actively selected for. Thus, the production of SHWs provides a promising approach for the enhancement of genetic variation in modern breeding materials. In Experiment 3, gene bank accessions of 36 LCs from different countries and a total of 119 accessions from nine LCs populations collected in Turkey and Mexico were analysed with 44 and 76 SSRs, respectively. Both LC materials revealed high allelic variation (GD = 0.69 and 0.54). The 36 LC accessions could not be separated according to their continent of origin. An unexpected relationship was observed between the Chilean LC ?Trigo africano? and the Nigerian LCs ?Dikwa?. All of the nine LC populations could be discriminated except for two Turkish LCs collected from the same location. In accordance with previous studies, considerable genetic variation was observed within the LC populations. Our results contributed a lot to the characterisation of the LCs and generated important knowledge for the management of seed bank accessions.
In Experiment 4, a total of 112 wheat hybrids and their 22 parental lines were evaluated at two locations in Mexico for grain yield, plant height, days to flowering and maturity. The level of heterosis varied between -15.3% and 14.1%, but was generally too low to compensate  for the high costs of hybrid seed production. The correlations between mid-parent values and hybrid performance, as well as between parental line per se performance and general combining ability were significant (P < 0.01) for all traits, and particularly high for grain yield (r = 0.86 and 0.91). PCoA based on 113 SSR markers revealed three groups of parents. However, the correlations of GDs and COPs with the values of heterosis were negative and not significant. Thus, the prospects of large-scale cultivation of hybrid wheat in developing countries are low. The correlations between GDs and COP in Experiments 1 and 3 were generally significant but low. This can be explained by unrealistic assumptions in the calculation of COPs, which ignore the effects of selection and genetic drift. Similarly to genomic SSRs, EST-SSRs did not reflect functional diversity. The latter revealed lower degrees of polymorphism than genomic SSRs in all experiments, but the allele designation was simpler and more reliable. Across all experiments, our study demonstrates that plant breeding does not inevitably lead to a loss of genetic diversity. We confirmed that CIMMYT?s breeding strategies contributed to a successful increase in genetic variation. These results provide useful information to wheat breeders in CIMMYT and other national programs, regarding the use of wild relatives and landraces for the enhancement of the genetic base of wheat germplasm. In addition, our research provides a base of knowledge for future association studies, identification of useful alleles, and their use in marker-assisted selection.
KW  - Weizenzüchtung
KW  - Genetische Variabilität
KW  - Hybridweizen
KW  - Weizen
CY  - Hohenheim
PB  - Kommunikations-, Informations- und Medienzentrum der Universität Hohenheim
AD  - Garbenstr. 15, 70593 Stuttgart
UR  - http://opus.uni-hohenheim.de/volltexte/2005/112
ER  -