Future climate scenarios forecast a 10-50% decline in rainfall in the north-eastern coasts of South America and in central Amazonia. Yet, little is known of the mechanisms which ensure food web re-assembly after severe drought in “naïve” tropical ecosystems, and of the extent to which food-web responses mediate post-drought resilience of ecosystem functions beyond currently observed stress intensities. Taking advantage of a natural system that is small, contained, and widely distributed in neotropical forests (tank-forming bromeliads and their aquatic microbial-faunal food webs), we will emulate complete drying and then rewetting of this ecosystem in a field experiment. The effects of treatments (droughts) will range from current average conditions to extreme events, and will be evaluated in time against control bromeliads. We will follow-up core ecosystem functions (detrital mass loss, microbial respiration, photosynthetic activity of algae) and their potential drivers (fungal:bacterial biomass, predator:prey biomass ratio, autotrophic:heterotrophic microorganism biomass ratio), as well as ecosystem multifunctionality (the average of normalized functions). The relative importance of in situ resistance vs recolonisation by immigration will be evaluated by quantifying resistance forms in rehydration experiments. First, we expect ecosystem functions to show optimal functioning near current climatic conditions, so multifunctionality will decline or collapse under prolonged stress due to sublethal effects on influential species and/or mortality. Second, we predict that most invertebrate species will not withstand dehydration in the active larval or adult stage (weak ‘internal resilience’), so biotic resilience will mostly rely on immigration from source patches. Ecologists have a limited timeframe in which studies on consequences of climate change will be informative/useful to society, so need to seek experimental shortcuts by which generic mechanisms can be brought out. While providing high-quality measures of ecosystem-level physiological response and resilience capacity to drought, this project will provide a fresh approach on how to predict the ecosystem consequences of climate change.