Early or late? The role of genotype phenology in determining wheat response to drought under future high atmospheric CO2 levels.

Abstract

The combination of a future rise in atmospheric carbon dioxide concentration ([CO2 ]) and drought will significantly impact wheat production and quality. Genotype phenology is likely to play an essential role in such an effect. Yet, its response to elevated [CO2 ] and drought has not been studied before. Here we conducted a temperature-controlled glasshouse [CO2 ] enrichment experiment in which two wheat cultivars with differing maturity timings and life cycle lengths were grown under ambient (aCO2 approximately 400 μmol mol-1 ) and elevated (eCO2 approximately 550 μmol mol-1 ) [CO2 ]. The two cultivars, bred under dry and warm Mediterranean conditions, were well-watered or exposed to drought at 40% pot holding capacity. We aimed to explore water × [CO2 ] × genotype interaction in terms of phenology, physiology, and agronomic trait response. Our results show that eCO2 had a significant effect on plants grown under drought. eCO2 boosted the booting stage of the late-maturing genotype (cv. Ruta), thereby prolonging its booting-to-anthesis period byapproximately 3 days (p < 0.05) while unaffecting the phenological timing of the early-maturing genotype (cv. Zahir). The prolonged period resulted in a much higher carbon assimilation rate, particularly during pre-anthesis (+87% for Ruta vs. +22% for Zahir under eCO2 ). Surprisingly, there was no eCO2 effect on transpiration rate and grain protein content in both cultivars and under both water conditions. The higher photosynthesis (and transpiration efficiency) of Ruta was not translated into higher aboveground biomass or grain yield, whereas both cultivars showed a similar increase of approximately 20% in these two traits at eCO2 under drought. Overall, Zahir, the cultivar that responded the least to eCO2, had a more efficient source-to-sink balance with a lower sink limitation than Ruta. The complex water × [CO2 ] × genotype interaction found in this study implies that future projections should account for multifactor interactive effects in modeling wheat response to future climate.