Abstract

Ecologists are under pressure to anticipate the ecological effects of climate change. Therefore many ecological publications (and most grant proposals) related to climate claim relevance to the projection of future climate change effects. Yet the steps leading from ecological description and understanding to reliable projection are rarely explicit. A good understanding of the factors which allow the ecological effects of climate change to be effectively anticipated is critical to both the quality of basic science and its application to public policy. We used research performed on mammals to explore scientific approaches to anticipation of climate change effects. We distinguished forecasting models based on correlations from predictive models based on cause-effect relationships, These categories represent extremes along a continuous gradient between pattern description and causal understanding. We suggest that the constraints to our capacity to anticipate fall into 6 broad categories rooted in the development and application of forecasting and predictive models. These categories help to identify the conditions that allow or prevent projection of the effects of climate change on ecosystems. This approach should also help to identify which research avenues will likely be most fruitful. © Inter-Research 2006.