Conservation news

Amazon forest to savannah tipping point could be far closer than thought (commentary)

The Amazon cloud factory. Photo by Rhett A. Butler / Mongabay

In Brazil during the 1970s, when the first deforestation was spreading along the route of the Belém-Brasília highway, the Amazon forest seemed endless and eternal. It was mostly a place of resource extraction – rubber, Brazil nuts and more – and a place for science.

In the middle of that decade, Brazilian Scientist Eneas Salati published some extraordinary results. By analyzing isotopic ratios of oxygen in rainwater collected from the estuary all the way to the Peruvian border, he was able to demonstrate unequivocally that the Amazon makes half of its own rainfall. The moisture recycles five to six times as the air mass moves from the Atlantic until reaching the Andes. There the uplift caused major rainfall, creating the greatest river system on Earth, holding 20 percent of all river water globally.

These were paradigm shattering results. Hitherto the unquestionable dogma was that vegetation is simply the consequence of climate and that it had no influence on climate whatsoever. But that influence is actually visible when plumes of moisture rise from the forest after a rainstorm. It is the consequence of evaporation off the complex surfaces of the forest as well as the transpiration of the trees themselves. Moisture rising from the forest contributes more importantly in the central and eastern Amazon because large-scale factors for formation of rainfall are weaker there.

Those results almost immediately raised the question of how much deforestation could cause this hydrological cycle to degrade to the point – a tipping point – where there would be dieback of the forest in the south and southeast and replacement by a somewhat degraded savannah vegetation. It is something we have talked about over the ensuing years and which was addressed by modeling from Nobre’s group in 2007. The conclusion was that the tipping point would be at approximately 40 percent deforestation.

A newly cleared section of Amazon forest. Photo by Rhett A. Butler / Mongabay

Moisture from the Amazon actually contributes in important ways to rainfall, ecology and human wellbeing south of the Amazon itself (contributing winter rainfall in the La Plata basin, even south to southern Brazil, Paraguay, Uruguay and central-eastern Argentina).

The importance for Brazilian agriculture (extant and aspired to) is complex but still significant. Evapotranspiration from pastures is relatively trivial compared to that produced by forests. That being true, a longer dry season seems to be in the offing from deforestation.

The above would be important in itself, but today Amazon deforestation also interacts with climate change and widespread use of fire. The latter is known to desiccate adjacent forest, making it tinder for major wildfires the following year. So it becomes sensible to reevaluate the tipping point to include those two other factors.

We believe that with the addition of those two factors the tipping point is much closer – in the vicinity of 20-25 percent deforestation. Past that point, the east, south and central Amazon could flip from forest to non-forest ecosystems. Nobre’s modeling group made some calculations in 2016 that considered the synergistic effect of deforestation, climate change, increased forest fires and also the so-called ‘CO2-fertilization’ effect of increased atmospheric concentrations of CO2, assumed to be positive for vegetation. Their results fully support this conclusion.

Evapotranspiration from Amazon cattle pastures is relatively trivial compared to that produced by forests. Photo by Rhett A. Butler / Mongabay

There is a good reason to think that the tipping point is close at hand. Historically unprecedented droughts (2005, 2010 and 2015) would seem to be first flickers of such change. Indeed, there is a suite of changes such as warmer temperatures over the tropical North Atlantic associated with changes in the land. And severe floods in 2009 and 2012 (and in Southwest Amazonia in 2014) suggest the Amazon system is oscillating.

So what would be the sensible way forward? Clearly there is no sense in the least in discovering the tipping point by tipping it. We believe it is critical to build a margin of safety by reducing the deforested area to less than 20 percent. The current official figure for Brazil is 17 percent but some of the remaining forest is degraded and thus contributing less moisture. So strongly controlling deforestation, and combining that with reforestation, is the sensible course.

Brazil committed in Paris 2015 to 12 million hectares of reforestation by 2030 and to significantly curbing deforestation. That commitment should be re-examined to make sure that the nation can also contribute to avoiding the tipping point for the benefit of Brazil and adjacent South America.

Scientifically Brazil has contributed centrally to our understanding of this environmental challenge. It should also contribute with concomitant action.

The tipping point conversion of a large portion of the Amazon from forest to savannah would have a devastating impact on the region’s biodiversity. Photo by Rhett A. Butler
The loss of large swathes of Amazon forest would also heavily impact the lives and livelihoods of indigenous and traditional people. Photo by Rhett A. Butler . Mongabay


Nobre et al., 2016. The Fate of the Amazon Forests: Land-use and climate change risks and the need of a novel sustainable development paradigm. Proceedings of the National Academy of Sciences,

FEEDBACK: Use this form to send a message to the author of this post. If you want to post a public comment, you can do that at the bottom of the page.

Thomas Lovejoy is an ecologist who has worked in the Amazon since 1965. He is University Professor of Environmental Science and Policy at George Mason University.
Carlos Nobre is an Earth System scientist from Brazil, with expertise on biosphere-atmosphere interaction in the Amazon. He is a member of the Brazilian Academy of Sciences and a Senior Fellow of World Resources Institute-Brazil.