Characterization of durum wheat genetic resources and evaluation under Cyprus conditions

Abstract

Breeders need to intensify their efforts to meet the challenges arising from the growing population and from climate change. Genotypic characterization, phenotypic evaluation and exploitation of high throughput technologies will be crucial in order to achieve their goals. The main objectives of this doctoral dissertation were (1) to characterize a set of 55 durum wheat accessions, including modern varieties from different breeding programs and landraces originated from Cyprus and other Mediterranean areas using microsatellites, (2) to evaluate a subset of twenty modern varieties for grain yield and other important agronomic traits under diverse climatic conditions, and (3) to evaluate the usefulness of high throughput, remote sensing techniques for plant phenotyping under a water limited and heat stressed environment. High genetic diversity was detected. Landraces formed distinct groups and depicted higher genetic variability than modern varieties reinforcing their potential use to plant breeding. The highest genetic variability was observed in Cypriot landraces suggesting the existence of unique genetic diversity within this genepool. High genetic diversity was also detected in Cypriot landrace ''Kyperounda''. The genetic diversity was higher in accessions conserved as sublines. The genetic differentiation between ''Kyprerounda'' accessions and the different clustering of landrace accessions conserved under the same name signifies that passport data is not a reliable method to eliminate duplicates in genebanks. Phenotypic evaluation revealed the existence of genetic diversity for all the traits, however, for most of them the highest percentage of variance was explained by the environmental effect. In most cases, number of seeds per m2 was significantly correlated with grain yield and the correlations were stronger in the low yielding environments. Number of seeds per spike was the trait consistently related with number of seeds per m2. Trait offs between yield components were recorded. Varieties had different yield formation strategies. The two highest yielding varieties had the highest number of fertile tillers per m2 combined with high number of seeds per spike or high kernel weight. The differences between varieties for plasticity were significant for yellow pigment content, protein content, gluten index, growing degree days to heading, thousand kernel weight and plant height. In some cases, significant correlations were found between plasticities and between plasticities and the traits per se. Statistically significant differences between varieties were observed for Spectral Vegetation Indices derived by Unmanned Aerial Vehicle imagery. Grain yield was significantly correlated with SVIs, when recorded at early reproductive stages. Both positive and negative correlations were observed indicating that the environmental conditions during grain filling can affect the sign of the correlations. The predictive ability of SVIs was affected by plant phenology. Correlations of grain yield with SVIs were stronger as the correlations of SVIs with heading were weaker or not significant. These findings highlight the potential use of SVIs derived by UAV imagery for durum wheat phenotyping at low yielding Mediterranean conditions.