A breeding concept to improve the performance of locally cultivated bread wheat (Triticum aestivum L.) cultivars

The objective of this study was to explore the internal variability in six established F7 commercial wheat cultivars for breeding purposes. They are sown traditionally in the region of Western Macedonia, Greece. Spikes of the six cultivars were sown in 2008 in separate rows. A head to row selection scheme was applied for two growing periods in order to select lines within the cultivars, based on various traits such as: the total spike number, the spike weight per row, the 1000-kernel weight and the specific weight. The final selection was based on the specific weight and the four best rows from each cultivar were selected. All selected lines were tested in a field trial with a randomised complete block design (RCB). The original seed of the cultivars were used as controls. Statistically significant differences were found for all the studied traits. The line selections differed from the original cultivars, sometimes highly significantly. In conclusion, commercial cultivars that are sown traditionally for many years may contain exploitable variability, which reveals, that the continuous selection within cultivars is necessary to avoid cultivar deterioration and to improve the yield and other traits. The results indicate a degeneration of grain yield from 8% to 20%. Although eye-selection restricts off-types, our results mainly indicate new variability and cultivar performance deterioration under extreme biotic and abiotic stress.