ELICITATION OF EXPERT JUDGMENTS OF CLIMATE CHANGE IMPACTS ON FOREST ECOSYSTEMS M. GRANGER MORGAN 1 , LOUIS F. PITELKA 2 and ELENA SHEVLIAKOVA 3 1 Department of Engineering and Public Policy, Carnegie Mellon University, Pittsburgh, PA 15213, U.S.A. E-mail: granger.morgan@andrew.cmu.edu 2 University of Maryland Center for Environmental Science, Appalachian Laboratory, 301 Braddock Road, Frostburg, MD 21532, U.S.A. E-mail: pitelka@al.umces.edu 3 Department of Ecology and Evolutionary Biology, Princeton University, Princeton, NJ 08544, U.S.A. E-mail: elena@eno.princeton.edu Abstract. Detailed interviews were conducted with 11 leading ecologists to obtain individual qual- itative and quantitative estimates of the likely impact of a 2 ×[CO 2 ] climate change on minimally disturbed forest ecosystems. Results display a much richer diversity of opinion than is apparent in qualitative consensus summaries, such as those of the IPCC. Experts attach different relative importance to key factors and processes such as soil nutrients, fire, CO 2 fertilization, competition, and plant-pest-predator interactions. Assumptions and uncertainties about future fire regimes are particularly crucial. Despite these differences, most of the experts believe that standing biomass in minimally disturbed Northern forests would increase and soil carbon would decrease. There is less agreement about impacts on carbon storage in tropical forests. Estimates of migration rates in northern forests displayed a range of more than four orders of magnitude. Estimates of extinction rates and dynamic response show significant variation between experts. A series of questions about research needs found consensus on the importance of expanding observational and experimental work on ecosystem processes and of expanding regional and larger-scale observational, monitoring and modeling studies. Results of the type reported here can be helpful in performing sensitivity analysis in integrated assessment models, as the basis for focused discussions of the state of current understanding and research needs, and, if repeated over time, as a quantitative measure of progress in this and other fields of global change research. 1. Introduction Knowledge of the likely response of forest ecosystems to possible climate change is important in the formulation of public policy, research planning, and the decision making of private organizations and individuals. Approaches such as the Delphi method (Dalkey, 1970; Linstone and Turoff, 1975) and ‘nominal group techniques’ (Gustafson et al., 1973), have been developed to obtain consensus summaries in such circumstance, but such methods have significant limitations (Seaver, 1978; Brockhoff, 1975). Without using such formal methods, scientific panels, such as those of the Intergovernmental Panel on Climate Change (IPCC, 1996) or the Climatic Change 49: 279–307, 2001. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.