Complexity, cascades, and compensation in ecosystems
Some, but not all, population dynamics feed back to alter ecosystem processes. Thus species interactions have important implications for the nature and predictability of ecosystem responses to environmental change, pollutions, and exploitation. Ecological predictions will always be probabilistic, and are now possible for only a few potential stressors in a few of Earth’s ecosystems. This paper shows how large-scale experiments can improve the predictability of responses to ecosystem stress. We also argue that such experiments must be combined with long-term and comparative studies to prevent the need to study each new stressor and each ecosystem as if they were unique.
Global Environmental Forum