PARENT SESSION
Session #22: Observation, manipulation, experimentation and extrapolation: Limitations and opportunities in ecosystem-level studies. Organized by: R.H. Gardner and J.E. Petersen.
Friday, August 10, 2001. Lecture Hall
Time will tell: Paleoecology and the hidden impacts of variability on ecosystem experiments.
LEAVITT, PETER1, 1
ABSTRACT- Experiments elucidate ecosystem structure and function best when manipulation scale and magnitude are realistic and when results can be generalized to other sites. However, scientists do not often consider the impacts of past conditions on ecosystem organization, nor whether experiments themselves alter fundamental ecosystem properties. Paleoecology provides essential insights in both cases. For example, fossil analysis of whole-lake fertilization experiments demonstrates that aquatic ecosystems can be destabilized by eutrophication, thereby reducing the sensitivity of experiments and obscuring causes of variability. Paleoecological analyses also show that the impacts of historical disturbance (e.g., land-use practices, fisheries management, prior experiments) can last for decades after removal of stressors, but that ecosystems may remain subject to rapid changes between stable states. Similarly, such retrospective analyses are often the sole means to quantify the impacts of cyclic (climate, migrations) and discontinuous phenomena (catastrophe, invasion) on ecosystem production and composition. Because paleoecology is also a rapid, cost-effective method for quantifying the spatial variability of ecosystems, this approach is effective for evaluating the relevance of ecosystem experiments on a landscape scale. In conclusion, I argue that national funding initiatives are needed to exploit natural linkages among long-term (LTER), experimental and paleoecological research programs and that these represent the best means to improve our understanding of ecosystems.
KEY WORDS: paleoecology, ecosystem, lake, experiment