The multi-model projections by the IPCC show that the Amazon forest is the tropical forest to be most affected by higher temperatures and changing rainfall until the end of this century. Such a no-analogue climate will most certainly drive this highly diverse forest in the future to a composition, functioning and/or structure that may have no analogue in the past.
Results/Conclusions
In this presentation we first briefly review current experimental efforts to assess the impacts of climate change on the ecology and resilience of the Amazon forest at the ecosystem scale, namely rainfall exclusion, soil fertilization and CO2 enrichment experiments. Second we suggest that other novel experimental and modeling methods are necessary to deepen our incipient knowledge on these impacts. In the experimental domain for example, a novel and efficient way of testing the impacts of higher temperature at the forest ecosystem level is particularly necessary, although there is not any ongoing attempt to do that in the Amazon. For modeling assessments we argue that four approaches will allow us to better understand how the Amazon forest ecosystem will respond to climate change and, therefore, lead to more accurate projections of the no-analogue regional climate-vegetation interaction of the future: (1) the development of conceptual models that help advancing the understanding of relations between and importance of different ecological processes in a faster way compared to detailed process-based vegetation models; (2) the inclusion of relevant ecological processes not previously considered in vegetation models such as microbe-root interactions and phosphorus cycle processes; (3) the revision of the structural way of modeling vegetation, such as moving from plant functional type-based modelling to plant functional trait-based modeling. These activities should be permeated by strong model-experiment integration, in which models generate hypothesis to be tested in localized field experiments; the latter in turn provide more accurate data to parameterize and evaluate large-scale ecosystem modeling. Acquiring a better knowledge about the impacts of climate change on the Amazonian ecosystem through such model-experiment integration is perhaps the most promising path to develop effective adaptation policies for the region.