We have been investigating these questions from the cellular to the level of the whole plant by using the model plant Arabidopsis thaliana. We have used mutants in which particular genes are disrupted, to investigate their stomatal behaviour and thus understand the role played by these specific genes in the traits of interest. We gathered a unique collection of 50 mutants affected in stomatal development or function. We set up original phenotyping tools to allow a tight control of the CO2 scenarios experienced by the plants and by setting up a platform made of connected balances for continuous measurement of plants transpiration dynamics. This was coupled with RGB imaging, with automated image analysis pipelines to measure plant growth, and with thermal imaging to track changes in plants temperature, a proxy for stomatal conductance. Building on these biological resources and phenotyping facilities, we examined the role of the signalling pathway of a stress hormone (abscisic acid, ABA) in the plant response to elevated CO2, depending on the timing of a drought event during its development cycle. We used mutants deficient in ABA biosynthesis or in the receptors for the hormone. Plants were subjected to different scenarios of drought while growing in elevated CO2 and we measured their growth, water use, stomatal opening and photosynthesis to understand the role of these genes in maintaining stomatal function upon these complex combinations of environmental factors, and how this impacts on whole plant performance. We further addressed how the elevated CO2 signal is sensed by the plants on the long term, by investigating the role of a carbonic anhydrase newly identified by the group. In order to extend our study to many mutants and many combinations of CO2 x drought scenarios, we used the PHENOPSIS high-throughput phenotyping platform (Montpellier, EPPN2020 funding). We further extended our study to the natural variability within a major European crop, grapevine, by screening large populations for stomatal responses to CO2, and linking their variability to allelic variations. In collaboration with INRAE (secondment and funding The Company of Biologists), we deployed an experiment in which grapevine leaves from 279 varieties were measured for their transpiration in response to elevated CO2, together with leaf morphological structures. Genome wide analysis studies have allowed the identification of genomic regions controlling the variations in the traits of interest.