Conservation news

Adapt to a changing Amazon now, or pay far higher price later, experts say

  • A new study estimates the costs of delaying adaptation to a hotter, dryer Amazon would be orders of magnitude higher than acting now, despite uncertainties.
  • The study is the first comprehensive impact analysis of the Amazon Forest Dieback hypothesis, which posits that there exists a definitive climate-driven deforestation tipping point beyond which large swaths of the rainforest would be rapidly replaced with savanna.
  • The study’s authors estimate the costs of such a catastrophic loss of forest could be as high as $3.6 trillion over a 30-year period.
  • They also estimate that the cost of a series of adaptation measures taken now would be $122 billion, a fraction of the economic losses estimated if no actions were taken.

The costs of acting now to adapt to a predicted, dramatic loss of Amazon rainforest would be at least one order of magnitude lower than the economic fallout if we waited and did nothing, a new study says.

It’s the first comprehensive impact analysis of a contentious prediction: that there exists a definitive climate-driven deforestation tipping point beyond which large swaths of the Amazon rainforest would be rapidly replaced with savanna.

In their study, published in the Proceedings of the National Academy of Sciences, David Lapola from the University of Campinas in São Paulo, Brazil, and colleagues estimated the costs of such a catastrophic loss of forest would be between $957 billion and $3.6 trillion over a 30-year period. The cost of adapting now, on the other hand, could be as little as $122 billion.

Born out of modeling studies in the late 1990s, the Amazon Forest Dieback (AFD) hypothesis suggests that with severe increases in temperature and reductions in rainfall projected under some climate models, the Amazon biome would experience sudden and catastrophic loss of biomass, transforming lush rainforest into a more drought-tolerant savanna.

Savanna in Serra da Canastra National Park in Minas Gerais, Brazil. Image by Klaus Balzano on (CC BY-NC).

The Amazon generates up to 50 percent of its own rainfall by recycling water through plants in a process known as transpiration, though climate change is resulting in deepening drought. Rapid and large-scale deforestation, reducing the number of trees transpiring, could push the biome closer to the tipping point for massive forest dieback. That tipping point was initially estimated at 40 percent deforestation, but was more recently revised down to the 20-25 percent range. The Amazon is presently estimated to be roughly 17-19 percent deforested.

In their study, the team estimated the economic costs to different sectors based on the costs reported for past droughts or from published estimates generated by computer models. They found that reduced crop yields, fishing stocks and hydroelectric output, as well as disrupted transport networks and large-scale climate-driven migration caused by even moderate levels of AFD, would cost $44 billion over a period of 30 years – an annual cost of 1.9 percent of the total GDP of the Brazilian Amazon in 2015. AFD could also lead to negative social impacts, such as reduced food security and increased risk of mosquito-borne diseases like malaria.

The researchers also considered other ecosystem services that are not valued based on current economic indicators, such as carbon stocks, pollination, provision of clean water resources, local climate regulation, and cultural and recreational benefits. They estimated that the true total bill could be many times higher: between $1 trillion and $3.6 trillion over 30 years, or an annual cost of between 42 percent and 107 percent of the Amazon’s GDP in 2015.

Delphine Clara Zemp, an ecologist and biogeographer at the University of Göttingen in Germany, who was not involved in the present study, said the findings represented “a great step forward to show policymakers that AFD should be considered seriously despite the large uncertainties.”

A single tree stands in a soy field next to rainforest, south of Santarem and along the BR 163 highway. Roads like this one, which cuts through the Brazilian Amazon for 1,700 kilometers (1,056 miles), allow access to the rainforest, and are a major cause of new deforestation. Image by Daniel Beltrà/Greenpeace.

Even small climate-driven changes in forest composition, such as an increase in drought-tolerant tree species that transpire less, could have far-reaching socioeconomic impacts. “A reduction in the transference of humidity from the [Amazon] forest to the atmosphere would have serious implications for the volume of rain falling in that region and in the Plata basin farther south,” Lapola said. Recently published data from a 30-year study of plots across the Amazon basin shows this pattern already starting to emerge, with more drought-tolerant species replacing aging rainforest trees.

“The finding that avoidance is one or more orders of magnitude cheaper than accepting damages is quite robust across many types of ecosystem stress,” said Bob Scholes, systems ecologist at the University of Witwatersrand in Johannesburg, South Africa, who was not involved in the study. However, he noted that the high proportion of the final tab contributed by these ecosystem services made the study sensitive to large uncertainties when estimating both their socioeconomic value and the impact of different degrees of AFD on the provision of those services.

Many questions remain about how forests will respond to changes in atmospheric carbon, temperature and rainfall patterns, making it difficult to evaluate the chances of a catastrophic AFD event. For example, higher atmospheric carbon dioxide levels are predicted to increase forest productivity by providing a key ingredient for photosynthesis. If true, this “CO2 fertilization” effect could counteract the effects of increasing drought, averting forest dieback. But experts disagree on the strength of such an effect, and there is evidence that it might be limited by other factors, including a lack of micronutrients to support rapid growth.

In 2014, the Fourth Assessment Report of the United Nations’ Intergovernmental Panel on Climate Change, co-authored by Scholes, ascribed low confidence to the hypothesis of an Amazon tipping point. “I think it is an ‘even chance,’” Scholes told Mongabay, adding that it was “about as likely as not.”

In contrast, earlier this year, noted Amazon scientists Thomas Lovejoy and Carlos Nobre, the latter a co-author of the present study, expressed concern that exceeding 25 percent deforestation in the Amazon region could be enough to send the ecosystem into a cascade of degradation that leaves a new habitat in its wake. Some estimates suggest we have already passed 20 percent deforestation in the biome. “The tipping point is very close at hand,” Lovejoy said.

A jaguar lounges by an Amazon river. The Amazon’s rich biodiversity is threatened by large-scale deforestation that could turn vast swaths of the lush rainforest into savanna. Image by Rhett A. Butler/Mongabay.

The likelihood that AFD will ever occur continues to be debated. But in light of such enormous projected losses, the present study recommends that governments start investing in a response to it now. “We should not await the impacts of climate change to realize how dependent we — even urban dwellers — are on the Amazon forest,” Lapola said.

There are two obvious ways to try to avoid hitting the tipping point in the Amazon: reduce global greenhouse gas emissions, and/or curb Amazon deforestation. While Lapola says the prospect of a significant reduction of emissions in the coming years seems unlikely, the researchers estimated that restoring a large fraction of deforested area and preventing further deforestation would cost about $64 billion.

“Reforestation … and minimiz[ing] further deforestation is central to maintaining the integrity of the [water] cycle,” Lovejoy said. However, Lapola said he favored strategies to adapt to AFD rather than prevent it, because the success of the latter was uncertain, whereas adaptive approaches would be beneficial no matter what happened.

For example, decentralizing energy generation and providing local electricity-generating capacity to more remote regions could offer greater availability and reliability of electricity across the Amazon. New technology, like in-stream turbines, which allow small-scale energy generation with minimal ecological impact, could be an important part of such a decentralized energy portfolio.

Other proposed measures include changes in agricultural practices in favor of more drought-tolerant crops, discouraging the deliberate lighting of fires, and diversifying agroforestry to make farms more resilient to climatic extremes. “Strategies to prevent fire … and restore biodiversity in degraded forest” are needed, said Zemp, citing their utility in helping maintain Amazon rainfall while also preserving vital ecosystem services.

The study estimated that the cost of 20 such “no-regrets” adaptation measures would be $122 billion, a fraction of the economic losses estimated if no actions were taken.

The study’s “calculation of [the] economic costs makes AFD more tangible for policymakers,” Zemp said, and means that uncertainties about the processes powering AFD are no longer an excuse for politicians’ inaction.


Lapola, D. M., Pinho, P., Quesada, C. A., Strassburg, B. B., Rammig, A., Kruijt, B., … & Vergara, W. (2018). Limiting the high impacts of Amazon forest dieback with no-regrets science and policy action. Proceedings of the National Academy of Sciences115(46), 11671-11679.