Predicting resistance to stripe (yellow) rust (Puccinia striiformis) in wheat genetic resources using focused identification of germplasm strategy
Impact factor: 1.082 (Year: 2014)
MetadataShow full item record
Timeless limited access
Abdallah Bari, Ahmed Amri, Kenneth Street, Michael Mackay, Eddy De Pauw, Richard Sanders, Kumarse Nazari, B. Humeid, Jan Konopka, Fida Alo. (17/4/2014). Predicting resistance to stripe (yellow) rust (Puccinia striiformis) in wheat genetic resources using focused identification of germplasm strategy. Journal of Agricultural Science, 156 (6), pp. 906-916.
Stripe rust, caused by Puccinia striiformis f. sp. tritici (Pst), is a major wheat disease that can inflict yield losses of up to 70% on susceptible varieties under favourable environmental conditions. The timely identification of plant genetic resources likely to possess novel resistance to this disease would facilitate the rapid development of resistant wheat varieties. The focused identification of germplasm strategy (FIGS) approach was used to predict stripe rust resistance in a collection of wheat landraces conserved at ICARDA genebank. Long-term climate data for the collection sites, from which these accessions originated and stripe rust evaluation scores for one group of accessions were presented to three different non-linear models to explore the trait×collection site environment interactions. Patterns in the data detected by the models were used to predict stripe rust resistance in a second and different set of accessions. The results of the prediction were then tested against actual evaluation scores of the disease in the field. The study mimics the real scenario where requests are made to plant genetic resources curators to provide accessions that are likely to possess variation for specific traits such as disease resistance. The models used were able to identify stripe rust-resistant accessions with a high degree of accuracy. Values as high as 0·75 for area under the curve and 0·45 for Kappa statistics, which quantify the agreement between the models’ predictions and the curator's disease scores, were achieved. This demonstrates a strong environmental component in the geographic distribution of resistance genes and therefore supports the theoretical basis for FIGS. It is argued that FIGS will improve the rate of gene discovery and efficiency of mining genetic resource collections for adaptive traits by reducing the number of accessions that are normally required for evaluation to identify such variation.