Optimizing hybrid rye breeding programs using simulations

Breeding programs consist of many interacting operations including phenotypic testing and genetic evaluation of potential new varieties for yield, quality, and disease resistance, selection of crossing parents, product development etc. The consequences of changes in breeding programs can often not be evaluated until many years later. Therefore, decision support tools based on stochastic simulations can be used to improve efficiency and find the optimal<br/>breeding strategies. Using such tools, e.g. ADAM, a broad range of complex breeding programs can be simulated from the level of individual genes and genetic markers to phenotypes from different testing systems, and the use of complex statistical models for traits of interest can be evaluated. Thereby, the effects of different breeding strategies on genetic gain and genetic variance over many generations can be investigated.<br/>In this study, breeding schemes for hybrid rye were simulated to compare different genotyping, phenotyping and selection strategies, and thereby investigate the value of testing hybrids as well as parental lines from two different heterotic groups under different assumptions on variance due to additive genetic and dominance/epistatic genetic effects. Genetic gain can be increased by using genomic selection rather than phenotypic selection and by including phenotypes of parental lines in the prediction models.