New ‘hypertropical’ climate emerging in the Amazon

Unprecedented hot drought conditions are becoming more common, exposing trees to deadly stress

News Release 10-Dec-2025

Peer-Reviewed Publication

University of California - Berkeley

The Amazon rainforest is slowly transitioning to a new, hotter climate with more frequent and intense droughts — conditions that haven’t been seen on Earth for tens of millions of years.

The conclusions come from a new study led by the University of California, Berkeley, involving a large team of national and international scientists.

The researchers predict that, if society continues to emit high levels of greenhouse gases, “hot drought” conditions could become more prevalent across the Amazon by 2100, occurring even during the wet season. This could lead to widespread tree dieoffs and impair Earth’s ability to deal with increasing levels of atmospheric carbon dioxide, since tropical forests worldwide absorb more human carbon emissions than any other biome. Recent reports found an increase in atmospheric carbon dioxide after severe droughts in the Amazon, showing that weather in the tropics has a measurable impact on the planet’s carbon budget.

The new study, to be published Dec. 10 in the journal Nature, explains why these severe tropical droughts are reducing the global uptake of carbon dioxide from the atmosphere.

The scientists refer to the new climate regime, or biome, as the hypertropics. It is emerging because of global warming, which is extending the typical July-to-September dry season as it also brings hotter-than-normal temperatures. In their study, the researchers document that hot drought conditions stress the trees and increase the normal tree mortality rate by 55%.

“When these hot droughts occur, that’s the climate that we associate with a hypertropical forest, because it's beyond the boundary of what we consider to be tropical forest now,” said study leader Jeff Chambers, a UC Berkeley professor of geography. By 2100, hot drought conditions could occur as many as 150 days each year.

Chambers and his team also discovered why the trees are dying under hypertropical conditions, which now only occur a few days to weeks during extreme droughts. Once the soil moisture content by volume decreases to about one-third, the trees either shut down carbon capture, starving to death, or develop air bubbles in their sap, akin to embolisms that cause strokes in humans.

This affects faster-growing species of trees more than slow-growing trees, the researchers found. That means that as the number of high heat-stress days increases, Amazon forests will experience a shift in tree species to those less susceptible — if that shift can take place fast enough in a rapidly changing environment.

“We showed that the fast-growing, low wood-density trees were more vulnerable, dying in greater numbers than high wood-density trees,” he said. “That implies that secondary forests might be more vulnerable to drought-induced mortality, because secondary forests have a larger fraction of these types of trees.”

Since the annual tree mortality is slightly more than 1%, an extra 0.55% may not seem like much, but it has a cumulative impact on the forest, Chambers said. Hypertropical conditions also are likely to appear outside the Amazon in rainforests in western Africa and across Southeast Asia.

Chambers emphasized that the direst outcome is predicted if society does very little to reduce carbon dioxide emissions that drive climate change.

“It all depends on what we do,” he said. “It's up to us to what extent we're actually going to create this hypertropical climate. If we're just going to emit greenhouse gasses as much as we want, without any control, then we're going to create this hypertropical climate sooner.”

Monitoring tree sap

Chambers has been conducting research in the Amazon since his graduate school days in 1993, much of that time with the Instituto Nacional de Pesquisas da Amazônia (INPA) in Manaus. His earliest research first established that the average age of rainforest trees 10 centimeters (4 inches) in diameter is about 180 years, making the region one of Earth’s longest-term storage areas for carbon. Some trees are more than 1,000 years old.

Since then, he has conducted studies to understand carbon cycling in tropical forests and forest-climate interactions. He and his international group of collaborators installed instruments on two approximately 50-meter-tall towers at two study sites north of Manaus to record temperature and humidity at different levels above the ground, as well as sunlight intensity at the top of the tree canopy and soil moisture in the forest floor. The oldest of the towers was visited in November by California Governor Gavin Newsom during his attendance at the COP30 climate summit in Belém, Brazil.

Chambers also collaborated with a team to install sensors in the trees themselves to record the flux of water from the soil up the stem and out to the atmosphere. These sensors measure sap flow, leaf temperature, water transpiration from leaves and the water potential of the soil — that is, how difficult it is for the tree to draw water from the soil to its topmost leaves through transpiration.

Using more than 30 years of data from the oldest of the two plots, which had previously been selectively logged, he and his team demonstrated a significant increase in tree die-offs the year after intense droughts. The highest tree mortality rates were among fast-growing species that are the first to sprout in logged areas and which have a low wood density.

Chambers and his colleagues also combined data from the two sites during droughts in 2015 and 2023 caused by El Niño. At both sites, they found that when the soil moisture content dropped below a threshold of about 0.32 — meaning about a third of the soil’s pores were filled with water — transpiration rates in the trees dropped rapidly, leading to increased hydraulic stress.

“The really remarkable thing is that the threshold soil moisture content in a different plot with different trees for droughts in different years — 2015 and 2023 — were essentially the same: 0.32 and 0.33,” he said. “That was really surprising to everyone.”

Eventually, when high heat continued under extended drought conditions, trees began to experience hydraulic collapse — the formation of embolisms or bubbles in the fluid-filled xylem.

“Normally, plants are pretty good at trying to compartmentalize and just say, OK, I'm willing to sacrifice that branch to keep this core piece alive,” he said. “But if there are enough embolisms, the tree just dies.”

The trees also began to starve; as the leaves closed their pores to prevent water loss, they also shut off their supply of carbon dioxide, which they need to build and repair tissue.

Finally, after exploring the changing climate conditions using published data from five different Earth system models, the researchers realized that the tropical forest was shifting to a hotter state that has no analog today, though it was found in the tropics when the Earth was much hotter between 10 and 40 million years ago. They define the hypertropics as regions that are warmer than the 99th percentile of historical tropical climates, with more frequent and intense droughts.

With more warming predicted for the future, this climate state will become more common and, depending on how quickly climate changes, may lead to broader forest die-back processes as the climate continues to warm. The hot drought conditions that drive elevated tree mortality are projected to frequently emerge during a typical dry season 20 to 40 years from now, Chambers said. But by 2100, they predicted, extreme hot drought days will no longer be confined to the peak of the dry season but will increasingly occur throughout the entire year, including during the wettest months.

“Present-day hot droughts are harbingers of this emerging climate, providing windows of opportunity to better understand tropical forest responses to increasingly extreme future conditions,” the authors wrote.

UC Berkeley co-authors of the paper include Bruno Oliva Gimenez, a former UC Berkeley postdoctoral fellow now at INPA; Anna Weber and integrative biology professor Paul Fine. Other co-authors include Adriano José Nogueira, Lima Cristina Santos da Silva, Regison Costa de Oliveira, Gustavo C. Spanner, Tatiana D. Gaui, Daisy Celestina Souza, Joaquim dos Santos and Niro Higuchi of INPA and collaborators from the U.S., United Kingdom, Brazil, Germany and Norway. The research was funded by numerous U.S. and international agencies over the past 30 years.