Coupling agricultural plant growth and hydrological models for climate change projections

Abstract

The continuous increase of the greenhouse gas carbon dioxide (CO2) is expected to impact a wide range of processes within the soil-plant system, including biomass production and transpiration. In C3 and C4 plants, elevated CO2 (eCO2) is reported to decrease stomatal conductance which in turn reduces water loss through transpiration at the leaf level. However, eCO2 is observed to increase biomass production of C3 plants, which might counteract the water saving effect at the canopy level by increased leaf area. The direct CO2-fertilizating effect is not observed for C4 plants, but a combination of eCO2 and drought stress has been observed to distinctly increase C4 biomass. Free-air carbon dioxide enrichment (FACE) experiments have been developed to investigate the effect of eCO2 on the soil-plant system under field conditions providing a number of parameters valuable for crop modelling. Process-based models, which are used to project climate change effects on agricultural systems, need to be capable of simulating the effects observed in the field. However, recent crop model ensemble studies revealed strong limitations, for instance in simulating the distinct biomass increase of the C4 crop maize under eCO2 and water stress. To improve the representation of the dynamic behavior of the soil-plant system, two independent process-based models with a high degree of process representation, i.e. a plant growth and a soil hydrological model, were coupled in this work, and straightforward CO2 response functions regulating stomatal conductance and biomass accumulation were implemented. A comprehensive parameter uncertainty analysis based on Latin Hypercube sampling has been undertaken for the established model. The coupled model was applied to long-term data of a FACE experiment on a temperate C3 grassland. Results imply that temperate, mown, wet-dry C3 grasslands could benefit from biomass increase while maintaining water consumption, already with a modest increase of CO2 concentration of 20%. Further, the expected water saving effect at the leaf level could be offset at a stand level as a result of increased transpiration, caused by a biomass gain under eCO2. For simulating the combined effect of eCO2 and water stress on C4 crops, the coupled model was applied to a two-year long FACE experiment where maize was grown under combined eCO2 and water limited conditions. The clear benefit of maize biomass from eCO2 under water-limited conditions was well simulated. Results indicate that the coupled hydrological-plant growth model is capable of simulating the relevant climate change feedback mechanisms on plant growth of C4 plants. The obtained values of calibrated response parameters could be used in other crop models to project maize yields under climate change conditions. Based on the results of this work, the importance of plant‐specific CO2 response factors obtained by using comprehensive FACE data is emphasized. Further, for the rigorous assessment of crop models and their applicability to project yields and water fluxes under climate change, datasets that go beyond single criteria (only biomass response) and single effects (only eCO2) are needed.