Theoretical and experimental investigations on the exploitation of heterosis in hybrid breeding

摘要

Hybrid breeding has played a key role in the improvement of the productivity of many crops. Genetic variability, an essential prerequisite in plant breeding, is expected to decrease in heterotic groups as a result of continuous breeding efforts. The consequences of the narrowing of genetic variability are a decrease in selection gain and an increase in the susceptibility of cultivars. Thus, establishment of heterotic groups and broadening the genetic base of established heterotic groups are very important research topics in hybrid breeding.Our objectives were to (1) evaluate heterosis in winter triticale and identify heterotic groups based on field and SSR marker data, (2) investigate by field evaluations and SSR markers the heterotic relationships between the Central European heterotic groups in maize and rye and exotic germplasm from the US and Eastern Europe, respectively, and (3) monitor temporal changes over 30 years in the magnitude of variances due to general combining ability (GCA) and specific combining ability (SCA) in an applied hybrid maize breeding program.Triticale has low heterosis and, therefore, no hybrid cultivars have been developed, although an effective CMS system is available. Twenty-one lines and their 210 diallel crosses were field-evaluated for grain yield at five locations in Germany. 95 SSR markers were used to analyze the parents for identifying two diverse subgroups. Hybrid performance, midparent heterosis, and estimates of variance due to GCA and SCA were determined in a diallel, a 10 × 11 factorial, and the remaining two sub-diallels with 10 and 11 parents. In addition, we applied an enumeration algorithm, which explored the entire sample space to identify diverse heterotic groups and optimize different criteria in this context. The ratio of variance due to GCA to variance due to SCA was higher for factorials between groups of parents than in diallels and subdiallels within groups. The analyses indicated a more favorable ratio in situations with genetically distinct populations compared to situations with genetically less distinct populations. Application of the enumeration algorithm improved all criteria. F1 performance followed by heterosis were the most important criteria for development and enrichment of heterotic groups. Introgression of new germplasm to broaden the genetic base of heterotic groups is required to ensure continued genetic gains in hybrid breeding. In maize, we evaluated 19 inbreds belonging to two Central European heterotic groups and US heterotic groups and their factorial crosses in F1 and F2 generations for grain yield and dry matter concentration. The parental inbreds were additionally fingerprinted with 266 SSR markers. Multi-environment evaluation was performed in three mega-environments: Central Europe, US Cornbelt, and Southeast Europe. We found higher genetic diversity in the exotic germplasm than in the Central European heterotic groups. Based on F1 performance and heterosis, we conclude that non-Stiff Stalk germplasm should be introgressed into the Flint group and the Stiff Stalk germplasm into the Dent group.In rye, we evaluated testcrosses of 610 S0 clones belonging to the two Central European heterotic groups and five East European open-pollinated varieties (OPVs). S0 clones were fingerprinted with 30 SSR markers. We found higher genetic diversity in the OPVs compared to the Central European heterotic groups. The Carsten group had a narrow genetic base and should, therefore, be the primary target for genetic broadening. Nevertheless, all five OPVs were genetically closer to Petkus than Carsten. Two OPVs were identified as good candidates for introgression into Petkus and one into Carsten. We suggest to use selected clones of these populations for introgression.Continuous selection is expected to narrow the genetic base of heterotic groups over time. We studied the nature and magnitude of genetic variability in the breeding materials of the maize program of the University of Hohenheim, which is based on two heterotic groups with continuous enrichment by other germplasm. The data generated in multilocation field trials based on inter-group factorial designs conducted from 1975 to 2004 for grain yield and dry matter concentration were analyzed. There was neither a decrease in the magnitude of genetic variance nor a change in the predominance of variance due to GCA over variance due to SCA. Consequently, for avoiding the adverse effects of selection on genetic variation and for ensuring medium and long-term selection gains, heterotic groups should not be treated as closed populations, but should be continuously enriched by introgression of new germplasm.