Challenging modelling predictions for the impact of elevated CO2 on water limited woodlands

One of the expected effects of rising atmospheric CO2 concentration in the atmosphere is an increase in plant water use efficiency due to reduced transpiration at the leaf level. Yet, this prediction is based mostly on retrospective analyses of how the vegetation has responded to past changes in the CO2 concentration in the atmosphere for the XX century and these trends might not necessary apply to how our ecosystems will respond to the expected increases in CO2 in the atmosphere projected for the XXI century. Testing these predictions at the whole ecosystem level is very challenging, particularly for forested ecosystems and one of the few opportunities we have to do so is in the so-called FACE experiment (Free Air CO2 Enrichment). In this study, we increased the CO2 concentration in the atmosphere to mimic that of the atmosphere in the year 2050 in a water-limited native eucalyptus woodland. We expected to find a reduction in the amount of water transpired by the trees, which should have eventually led to an increase in the water available for human consumption, but this was not the case. Instead we found that there was no effective water savings by the vegetation, as the ecosystem was already strongly limited by water to start with. Our results challenge model predictions and suggest that the vegetation-climatic feedbacks might to be able to mitigate climate change as much as we expected.