Unlock the breeding value of genetic resources for improving bread wheat

Genetic resources were rarely used in elite breeding programs as they generally carry quite a lot of negative characteristics. The few times for which they were commonly used is to improve simple inherited traits. For instance, when the genetic resource carries a single resistance gene, which largely improves disease resistance. However, only this gene with some linkage drag is then introgressed into elite wheat by marker-assisted backcrossing or without cumbersome backcrossings and linkage drag by CrisprCas, where allowed.

Grain yield is the most important trait but also one of the most complex ones, that means many genes with small effects influence this trait. It is widely accepted that the use of genetic resources is important also for yield to maintain or increase the genetic variance for selection. However, the choice of the wright candidate out of the plethora of genetic resources is difficult, because the line per se performance of most resources is very poor. The limited number of studies available in this context either try to develop small but representative line collections of genetic resources for genotyping and phenotyping or choose arbitrarily few genetic resources and introgress them via MAGIC or AB-QTL-populations.

We propose a paradigm change away from investigating only small fractions of the plethora of genetic resources towards a systematic and intensive screen of whole gene banks. The full concept idea can be read here.  Roughly, this concept is based on two new high-throughput technologies available now for wheat: high density genotyping and hybrid technology. It is difficult to phenotype the breeding value of genetic resources for grain yield, because they carry many genes with large negative effects on simple adaptation traits, like tall plant height, heavy disease susceptibility, or late flowering. However, most of these traits are simply inherited by dominant major genes, which are normally fixed in latest elite lines. We therefore propose to screen genetic resources in hybrid combination with a strong elite line tester. Thus, most deleterious adaptation genes were hidden through the elite line and the breeding value of the genetic resource can rapidly been phenotyped in the field. With the new hybrid technology in wheat based on chemical hybridization agents like Crosoir, this strategy can easily be established by everybody.

These arising phenotypic information about genetic resources coupled with high density phenotyping can be used to calibrate models for genomic selection. As far as robust calibrations were established, whole gene banks should be genotyped and only genetic resources with high predicted genomic breeding value should be phenotyped enabling to largely reduce costs in the evaluation of genetic resources. To warrant widespread use of genetic resources, these data should be generated across gene banks with one standard protocol for genotyping and phenotyping and with open access to the data. First encouraging results of this concept can be found here.

Based on this concept, we are running a continous prebreeding program for wheat.