PREDICTING THE RELATIVE SENSITIVITY OF FOREST PRODUCTION IN IRELAND TO SITE QUALITY AND CLIMATE-CHANGE

Most model-based predictions of climate change effects on forest ecosy stems have used either potential or static descriptions of vegetation and site, removing the effects of direct management or land use. In th is paper we use a previously developed and validated model of carbon a nd water balances in forest ecosystems (PnET-II) to assess the relativ e sensitivity of forest production in Ireland to predicted climate cha nge and to ambient variability in site quality. After validating the m odel against measured productivity for 2 sets of stands, we ran the mo del using existing variation in site quality, represented as differenc es in foliar N concentration, and also for predicted changes in climat e and atmospheric CO2. Resulting variations in productivity were compa red with those due to potential errors in the specification of input p arameters and to variation in current ambient climate across the regio n. The effects on net primary production (NPP) and wood production of either ambient variation in climate or predicted changes in temperatur e, precipitation and CO2 are quite small (0 to 30%) relative to the ef fects of ambient variability in site quality (up to 400%). The range o f possible variation in other user-specified physiological parameters resulted in changes of less than 10% in model predictions. We conclude that site-specific conditions and management practices result in a ra nge of forest productivity that is much greater than any likely to be induced by climate change or CO2 enrichment. We also suggest that it i s essential to understand and map spatial variability in site quality, as well as to understand how the productive capacity of landscapes wi ll change in response to management and pollution loading, if we are t o predict the actual role that climate change will play in altering fo rest productivity and global biogeochemistry.