Simulating Mediterranean forest landscape dynamics in the context of climate change

Climate change has the potential to directly and indirectly affect forest ecosystems through changes in natural disturbance regimes, tree physiology, forest productivity, and tree species distribution. In Mediterranean environmental conditions and where forest ecosystems have been shaped by centuries of human activities these changes can be much more remarkable. Furthermore, the amount of uncertainty associated with climate prediction, forest ecosystem complexity, the broad temporal and spatial scales of forest management, and the response of the forest sector to climate change pose tremendous challenges. The goal of this study was to investigate to what and extent tree species distribution and productivity are likely to be affected by predicted regional climate change and to assess the combined influence of climate change, wild fires and forest management.  We used, for the first time in Europe, the forest landscape dynamics simulator LANDIS-II (Scheller et al., 2007) to model two climate (climate change Vs. current climate) and two disturbance/management scenarios (wildfires and no-forest management Vs. wildfires and forest management) in a typical Mediterranean landscape in Southern Italy. This is a micro-heterogeneous landscape dominated by semi-natural grasslands and croplands and comprising residual coppice woodlands (Quercus pubescens Willd. and Q. calliprinos Webb), scrubs, coniferous plantations (Pinus halepensis Miller, Cupressus spp.). Our results show that climate change affects tree species diversity and that different species show individual responses in terms of productivity. The combined effect of disturbances, either wildfires alone or wildfires and current forest management leads to a further decline in productivity mainly for coniferous species and deciduous oak whilst sclerophyllous oak species appear to be favored. We conclude that LANDIS II even when applied to micro-heterogeneous landscapes seems to be capable of representing forest biomass dynamics and of providing information on their response to climate change. In addition it is clear that the current main negative effect on forest productivity is mainly driven by the combined effect of climate change and current wildfire regime. Therefore further iterations of the simulator considering different wildfire management options might provide useful information to reduce the impact on both biodiversity and carbon stocks.