El Niño Tests the Amazon and Reveals Limits of One of the World’s Largest Terrestrial Carbon Sinks

Can the Amazon really release more carbon than it absorbs? Under El Niño, the heat and drought disrupt a mechanism that has operated for millions of years. A troubling fragility, at a moment when the world counts on forests to slow warming.

The delicate balance between photosynthesis and respiration at the heart of the carbon sink

On the surface, nothing seems to move beneath the canopy, save for a few insects and leaves stirred by the humid air. Yet every tree wages a microscopic battle. Through photosynthesis, its leaves capture carbon dioxide to build wood, roots, and new shoots. The forest thus becomes a vast living reservoir.

This uptake is never guaranteed. Trees respire as well, while soils, fungi and decomposers release carbon. A forest remains a sink only when its growth offsets these emissions. The IPCC already notes a weakening of the Amazonian sink, despite the fertilizing effect of atmospheric CO₂.

When heat sets in and water is scarce, trees close their stomata, tiny openings on the leaves. This reflex limits transpiration but also blocks CO₂ entry. Growth slows, some trees die and the decomposed wood then feeds a carbon debt that can last for decades.

The effects of El Niño on heat and drought in tropical Amazonia

El Niño originates thousands of kilometres away in the tropical Pacific, when surface waters become unusually warm. This oceanic upheaval shifts rainfall and temperatures across much of the globe. In the Amazon, it can bring a dangerous combination: less water, higher heat, and longer dry seasons.

The 2015-2016 episode served as a real-world stress test. In several South American tropical forests, carbon uptake collapsed under extreme temperatures. The areas already facing a dry season reacted most violently, as if their climate safety margin had vanished before that of the moister regions.

Field measurements reveal the real vulnerability of the Amazonian forests

To quantify this tipping point, researchers don’t rely on satellites alone. In hundreds of plots, teams circle trunks with meter-long tapes, tally dead trees, and sometimes return for thirty years. Behind each figure lies a colossal fieldwork effort, carried out in rain, heat and mosquitoes.

These observations let scientists distinguish a forest that is growing from one that is slowly losing biomass. They also reveal important differences between species. Some trees tolerate water shortage better, while others quickly reach their physiological limits. The Amazonian diversity could offer partial resilience, though not a foolproof shield.

Globally, forests have still absorbed more than 100 gigatonnes of CO₂ over thirty years, according to an international study reported in 2024 by the University of Leeds. But this performance masks pronounced fluctuations. A reassuring average can hide years in which certain forest blocks become temporary net carbon sources.

The repetition of climate shocks permanently weakens the role of the carbon sink

The danger lies not only in a single extreme event, but in the repetition of shocks. A tree weakened by drought may survive, only to fall in the next round. Add to this fires, fragmentation and deforestation, which locally heat landscapes and reduce the recovery capacity of ecosystems.

In July 2026, NOAA’s climate services judged a new El Niño likely to form within the year, with a strong chance of persisting through the boreal winter. The question thus goes beyond the fate of individual trees: for how long can climate policies continue to treat nature as a guarantee, rather than a balance to be protected?