Abstract
The Mediterranean forests are critical to preserve the high biodiversity which characterizes them and to provide essential ecosystem services, such as soil protection, water resources conservation and climate regulation. The climatic models simulate, for the 21th century in that region, a significant warming and a reduction of rainfall, with a significant increase of extreme events, which can considerably reduce the forest productivity. Quantification of these impacts to help the forest manager to take appropriate decisions was the main objective of REFORME. To reach it, we adopted a strategy based on measurements in forest experimental station (Puéchabon, Lamanon), of tree-ring data at broad spatio-temporal scale and a hierachy of vegetation models (MAIDEN, SIERRA, MODIS/GPP, BILHY). The vulnerability studies have been based on the B2 scenario (comparatively moderate) by the model ARPEGE of Météo-France. Main acquired results are the following: The climate variability from April to June drives the variability of the radial tree-ring increment of the evergreen oak and of the fluxes of carbon. There is a very good convergence between flux measurements and dendrochronology. At a broader scale, we showed that remote sensing images such as they are provided by MODIS is a good tool to estimate the productivity of the Mediterranean forests, but it must be corrected for some systematic biases. Puéchabon was subjected in 2005 to a caterpillar attack with more effects than successive droughts on the vulnerability of the oak. This phenomenon must be taken into account in vulnerability studies to climatic change, because the increase of the spring temperature risks to induce a decrease of the return period in future. For the Aleppo pine, the heat wave of 2003 is also a good model for the impact of climatic warming. There was a reduction from 30 to 60 % of the size of needles, of the number of the needles formed, of the length of the annual shoot on branches and fructification. The polycyclism disappeared almost entirely after 3 years of drought. The model BILHY showed a loss, in terms of wood production, about 28 % for Font-Blanche, and for Lamanon a 30 % loss. The comparison of MAIDEN simulations between the oak and the pine showed that both species arrive at a maximum of growth during the first decade of the 21th century, with a three times stronger productivity for the Alep pine. Then, drought becoming more important, the species see their productivity diminishing till the end of the 21th century, of 28 % for the oak and 8 % for the pine (smaller values than with BILHY). The latter therefore seems to resist better to water stress. If we take into account the fertilisation effect by CO2, both species seem to resist much better (with a productivity slightly increased). A statistical approach conducted in parallel showed the importance of the delayed effect of the extremes of the previous years. The latter are able, by degrading the health state of the tree and by subsequent defoliation, to exponentially cumulate with several successive events. This delayed effect has therefore the potentiality to attenuate the fertilisation effect. Finally Aleppo pine, likely by its capacity to early close its stomatae, seems to better resist than evergreen oak. Our results have limitations linked together to data and to models. Our measurements still do not cover optimum periods of time. The vegetation models have some more lacunae which must be filled up progressively in future. Finally, our vulnerability studies are based on a single scenario (IPCC-B2) of a single climatic model. They therefore do not have value of prediction but simply of indication. It will be necessary, to complete this approach, to use simulation ensembles from several climate models to deal with the climate evolution under probabilistic forms.