Abstract
Implementation of genomic selection (GS) for the genetic improvement of forage crops, such as perennial ryegrass, requires the establishment of sufficiently large training populations ith high-quality phenotype and genotype data. This paper presents estimates of genetic and environmental variance and covariance
components, obtained in a training population of 1453 F2 families. These families were produced in 2001, 2003, 2005, and 2007, and they were tested at seven locations throughout Europe. Families were cultivated together with
commercial varieties that were used as control. Analyses focused on forage yield (green and dry matter) and six traits scored by visual inspection (i.e., rust resistance, aftermath heading, spring growth, density, winter hardiness, and heading date). Data were analyzed with linear mixed models, including fixed effects (trial and control varieties, within year and location), and random
effects (breeding values, pedigree or parents, repeated effects of family or parents within location, and within trial environmental effects, to recover interblock information). Results showed very significant genetic variances for all traits, which provide good opportunities for future GSbased breeding programs. Forage yield showed family heritabilities of up to 0.30 across locations and up to 0.60 within a location. Similar or moderately lower values were found for the
other traits. In particular, the heritabilities of rust resistance and aftermath heading were very promising. Genetic correlations between traits were generally low but positive, which increases the potential for multitrait selection
components, obtained in a training population of 1453 F2 families. These families were produced in 2001, 2003, 2005, and 2007, and they were tested at seven locations throughout Europe. Families were cultivated together with
commercial varieties that were used as control. Analyses focused on forage yield (green and dry matter) and six traits scored by visual inspection (i.e., rust resistance, aftermath heading, spring growth, density, winter hardiness, and heading date). Data were analyzed with linear mixed models, including fixed effects (trial and control varieties, within year and location), and random
effects (breeding values, pedigree or parents, repeated effects of family or parents within location, and within trial environmental effects, to recover interblock information). Results showed very significant genetic variances for all traits, which provide good opportunities for future GSbased breeding programs. Forage yield showed family heritabilities of up to 0.30 across locations and up to 0.60 within a location. Similar or moderately lower values were found for the
other traits. In particular, the heritabilities of rust resistance and aftermath heading were very promising. Genetic correlations between traits were generally low but positive, which increases the potential for multitrait selection
| Originalsprog | Engelsk |
|---|---|
| Bogserie | Crop Science Society of America. Special Publication |
| Vol/bind | 55 |
| Nummer | 2 |
| Sider (fra-til) | 631-640 |
| Antal sider | 10 |
| ISSN | 0895-9978 |
| DOI | |
| Status | Udgivet - 20 mar. 2015 |