Improving models and plant phenotyping pipelines for a smart agriculture under abiotic stress combination and elevated CO2 – MODCARBOSTRESS

current crop models currently have large uncertainties in particular under combinations of drought and high temperatures and are little armed to predict the behavior of plants under elevated atmospheric CO2. We hypothesize that this could be due to poor model parametrization under stress combination and questionable modelling hypotheses. In particular, crop models can be broadly split into two distinct categories depending on whether growth is essentially limited by photosynthesis and C capture and its sensitivity to stresses (classically called ‘source’ limitation) or by the inability of organs to grow when challenged by stresses (called ‘sink’ limitation). Drought as well as elevated CO2 are suspected to shift growth limitation from source to sink while elevated temperature could on the contrary promote source limitation by stimulating respiration and reserve depression.
In this context, the MODCARBOSTRESS project will use two model archetype that are distinct in terms of source-sink formalisms and will benefit from various European phenotyping platforms in order to:
- deliver simple, low cost, principles and solutions for manipulating combined stresses, including elevated CO2, in experimental set-ups.
- improve model parameterisation thanks to experiments performed in the frame of the project, revisit questionable model hypotheses and evaluate model performance using available field data.

in progress

A final outcome of the project will be to propose model improvements and to run them against climate model projections for Europe.

Several papers are in preparation

Climate change accelerates the need for a smarter, more efficient, more secure agriculture. Because climate change is predicted to increase spatial and temporal variability, crop models able to predict the best local allele/phene combinations within a species, in addition to the best management systems (such as, for instance, species choice, rotations, sowing dates…) will be of great value for farmers and breeders worldwide. However, current crop models have large uncertainties in particular under drought and high temperatures that often occur in combination and while their occurrences are likely to increase in several regions of the world. Accounting for the impact of elevated atmospheric CO2 in the picture will add another level of difficulty with possible positive or negative infleunces depending on complex interactions We thus raise the double hypothesis that important reasons for crop model uncertainties are (i) Lack of accurate dataset under combined stresses hampering proper parameterisation. (ii) Inappropriate modelling hypotheses. Because CO2 control in experimental facilities is the exception rather than the rule, our project will aim at delivering to simple, low cost, principles and solutions for manipulating combined stresses, including elevated CO2, in experimental set-ups. Crop models can be broadly split into 2 distinct categories depending on whether growth is essentially source or sink limited. However, drought and CO2 are likely to shift growth limitation from source to sink while elevated temperature could shift growth limitation towards the source. A possibility is thus that both types of models find their limits under stress combinations. Our project will thus assess models of these two types in front of stress combinations. We will both improve model parameterisation thanks to the experiments performed in the frame of this project and evaluate model performance using field data obtained from other consortia (in particular FACE experiments). A final outcome of the project will be to propose model improvements and to run them against climate model projections for Europe. Two crop species, bread wheat (Triticum aestivum L.) and oilseed rape (Brassica napus L.), will used but the project intends to revisit crop model rationales in a species independent manner. In both species, a set of genotypes contrasted for stress sensitivity and for which field data are available will be selected.