Time: Wednesday, 14.05.25 - 11:15 AM
Location: Main building
Room: Lecture hall XII (HG)
Topic:
Gene banks preserve the crop diversity developed by farmers and breeders over generations for possible uses in the future. The International Rice Genebank housed by the International Rice Research Institute (IRRI) holds seeds of more than 130,000 accessions which are predominantly traditional varieties and wild relatives. These possess traits and genes that can improve crop productivity and sustainability by increasing resistance to environmental stresses and pests or diseases when bred into modern varieties. This resource is invaluable to adopt crops to the challenges brought by a changing climate. Unfortunately, very little information exists on these accessions apart from the date of collection and country of origin. How to utilize this resource in a targeted way therefore remains a challenge since detecting the useful traits and genes resembles the proverbial search for a needle in a haystack. Traditionally the only entry point into their utilization is to phenotype a set of accessions and clear limitations exist regarding the number of accessions that can be handled in experiments, especially if traits are complex and highly dependent on environmental variability.
An entry point into more targeted exploration of rice gene bank resources was provided by an international effort to sequence 3000 rice accessions (3K set) and to make the genetic diversity detected publicly available (SNP-Seek.irri.org). We have utilized this resource to improve the adaptation of rice varieties to abiotic stresses and to increase their nutritional value. Using novel genomic and bioinformatic tools, our aim was to identify donors from among the 3K set and to utilize these in conventional rice breeding to improve existing modern varieties with high yield potential. To phenotype the entire 3K set in field trials at multiple locations for multiple traits would require excessively large resources and is therefore not feasible. To overcome this phenotyping bottleneck, we employed Genomic Prediction (GP), where only 10% of accessions are phenotyped and their performance simulated based on the available genetic and phenotypic data. This simulation is then extended to predict the phenotypic values of the remaining 90% of 3K accessions. This process has let to the identification of new donors that are now utilized in applied breeding in Madagascar. In my lecture I will further describe these breeding activities in more detail and highlight recent successes that cumulated in the release of three new rice varieties with improved adaptation to the stressful environments encountered in smallholder farms in Madagascar.