Plant community feedbacks and long-term ecosystem responses to multi-factored global change

Published by Ecoss on

While short-term plant responses to global change are driven by physiological mechanisms, which are represented relatively well by models, long-term ecosystem responses to global change may be determined by shifts in plant community structure resulting from other ecological phenomena such as interspecific interactions, which are represented poorly by models. In single-factor scenarios, plant communities often adjust to increase ecosystem response to that factor. For instance, some early global change experiments showed that elevated CO2 favours plants that respond strongly to elevated CO2, generally amplifying the response of ecosystem productivity to elevated CO2, a positive community feedback. However, most ecosystems are subject to multiple drivers of change, which can complicate the community feedback effect in ways that are more difficult to generalize. Recent studies have shown that (i) shifts in plant community structure cannot be reliably predicted from short-term plant physiological response to global change and (ii) that the ecosystem response to multi-factored change is commonly less than the sum of its parts. Here, we survey results from long-term field manipulations to examine the role community shifts may play in explaining these common findings. We use a simple model to examine the potential importance of community shifts in governing ecosystem response. Empirical evidence and the model demonstrate that with multi-factored change, the ecosystem response depends on community feedbacks, and that the magnitude of ecosystem response will depend on the relationship between plant response to one factor and plant response to another factor. Tradeoffs in the ability of plants to respond positively to, or to tolerate, different global change drivers may underlie generalizable patterns of covariance in responses to different drivers of change across plant taxa. Mechanistic understanding of these patterns will help predict the community feedbacks that determine long-term ecosystem responses.