Using crop modelling to determine the meteorological conditions to be implemented in an Ecotron facility - Prerequisites to improve the experimental design?

An Ecotron is a facility where ecosystems are confined in experimental chambers, allowing the simultaneous control of environmental conditions and the on-line monitoring of processes. Under the threats of climate change and the pressure of a world growing population, such facilities will be of major importance to study the relations between climate change and agro-ecosystems.As it can quickly become time- and money-consuming, conducting experiments in an Ecotron will force researchers to cautiously select the climate of interest to be generated. They will thus need reliable tools to help them support the decision making process.Here, we present an innovative methodology, supported by the use of crop model, to assist researchers in finding the climatic conditions under which crop services will be impacted.The meteorological datasets among which the choice can be done were generated by the ALARO-0 model (RMI, Belgium) for current and future climatic conditions. Runs were conducted for the historical period 1981-2010, and for two time frames - 2041-70 and 2071-2100 - under two emissions scenarios - RCP 4.5 and 8.5.A crop model (STICS, INRA, FR) was run over the entire database. Crop model outputs were synthesized for the main crop phenological phases, i.e. the juvenile, vegetative and reproductive phases. A particular emphasis was put on agronomical outputs (biomass and grain yield) and crop growth stresses (deficit and excess of water, thermal and nutrient stresses).Using these outputs as selection criteria, a novel multi-criteria approach was designed to retro-select the specific climatic conditions allowing to reach certain outcomes (e.g. yield target) while simultaneously exhibiting given thresholds of stresses for any considered crop stages.