Flag leaf senescence, as evaluated by numerical image analysis, and its relationship with yield under drought in durum wheat

Senescence was evaluated at different stages of the grain-filling period in eight durum wheat varieties using numerical image analysis (NIA). The varieties were grown under early, severe drought conditions on the high plains of Setif in Algeria. After flowering, three different irrigation treatments were applied. Treatment effect was small, while a genotypic effect was noted for most of the senescence parameters. Senescence correlated to biomass, while the maximal rate of senescence, V-smax, correlated to thousand-kernel weight. The potential of the method of numerical image analysis for monitoring flag leaf senescence, detecting genotypic variability and selecting genotypes with delayed senescence is discussed.