deforestation case study

ORIENTATION
ASSIGNMENTS
Program Home Page
LIBRARY RESOURCES
Getting Help
Engaging Course Concepts

Case Study: The Amazon Rainforest

The Amazon in context

Tropical rainforests are often considered to be the “cradles of biodiversity.” Though they cover only about 6% of the Earth’s land surface, they are home to over 50% of global biodiversity. Rainforests also take in massive amounts of carbon dioxide and release oxygen through photosynthesis, which has also given them the nickname “lungs of the planet.” They also store very large amounts of carbon, and so cutting and burning their biomass contributes to global climate change. Many modern medicines are derived from rainforest plants, and several very important food crops originated in the rainforest, including bananas, mangos, chocolate, coffee, and sugar cane.

In order to qualify as a tropical rainforest, an area must receive over 250 centimeters of rainfall each year and have an average temperature above 24 degrees centigrade, as well as never experience frosts. The Amazon rainforest in South America is the largest in the world. The second largest is the Congo in central Africa, and other important rainforests can be found in Central America, the Caribbean, and Southeast Asia. Brazil contains about 40% of the world’s remaining tropical rainforest. Its rainforest covers an area of land about 2/3 the size of the continental United States.

There are countless reasons, both anthropocentric and ecocentric, to value rainforests. But they are one of the most threatened types of ecosystems in the world today. It’s somewhat difficult to estimate how quickly rainforests are being cut down, but estimates range from between 50,000 and 170,000 square kilometers per year. Even the most conservative estimates project that if we keep cutting down rainforests as we are today, within about 100 years there will be none left.

How does a rainforest work?

Rainforests are incredibly complex ecosystems, but understanding a few basics about their ecology will help us understand why clear-cutting and fragmentation are such destructive activities for rainforest biodiversity.

High biodiversity in tropical rainforests means that the interrelationships between organisms are very complex. A single tree may house more than 40 different ant species, each of which has a different ecological function and may alter the habitat in distinct and important ways. Ecologists debate about whether systems that have high biodiversity are stable and resilient, like a spider web composed of many strong individual strands, or fragile, like a house of cards. Both metaphors are likely appropriate in some cases. One thing we can be certain of is that it is very difficult in a rainforest system, as in most other ecosystems, to affect just one type of organism. Also, clear cutting one small area may damage hundreds or thousands of established species interactions that reach beyond the cleared area.

Pollination is a challenge for rainforest trees because there are so many different species, unlike forests in the temperate regions that are often dominated by less than a dozen tree species. One solution is for individual trees to grow close together, making pollination simpler, but this can make that species vulnerable to extinction if the one area where it lives is clear cut. Another strategy is to develop a mutualistic relationship with a long-distance pollinator, like a specific bee or hummingbird species. These pollinators develop mental maps of where each tree of a particular species is located and then travel between them on a sort of “trap-line” that allows trees to pollinate each other. One problem is that if a forest is fragmented then these trap-line connections can be disrupted, and so trees can fail to be pollinated and reproduce even if they haven’t been cut.

The quality of rainforest soils is perhaps the most surprising aspect of their ecology. We might expect a lush rainforest to grow from incredibly rich, fertile soils, but actually, the opposite is true. While some rainforest soils that are derived from volcanic ash or from river deposits can be quite fertile, generally rainforest soils are very poor in nutrients and organic matter. Rainforests hold most of their nutrients in their live vegetation, not in the soil. Their soils do not maintain nutrients very well either, which means that existing nutrients quickly “leech” out, being carried away by water as it percolates through the soil. Also, soils in rainforests tend to be acidic, which means that it’s difficult for plants to access even the few existing nutrients. The section on slash and burn agriculture in the previous module describes some of the challenges that farmers face when they attempt to grow crops on tropical rainforest soils, but perhaps the most important lesson is that once a rainforest is cut down and cleared away, very little fertility is left to help a forest regrow.

What is driving deforestation in the Amazon?

Many factors contribute to tropical deforestation, but consider this typical set of circumstances and processes that result in rapid and unsustainable rates of deforestation. This story fits well with the historical experience of Brazil and other countries with territory in the Amazon Basin.

Population growth and poverty encourage poor farmers to clear new areas of rainforest, and their efforts are further exacerbated by government policies that permit landless peasants to establish legal title to land that they have cleared.

At the same time, international lending institutions like the World Bank provide money to the national government for large-scale projects like mining, construction of dams, new roads, and other infrastructure that directly reduces the forest or makes it easier for farmers to access new areas to clear.

The activities most often encouraging new road development are timber harvesting and mining. Loggers cut out the best timber for domestic use or export, and in the process knock over many other less valuable trees. Those trees are eventually cleared and used for wood pulp, or burned, and the area is converted into cattle pastures. After a few years, the vegetation is sufficiently degraded to make it not profitable to raise cattle, and the land is sold to poor farmers seeking out a subsistence living.

Regardless of how poor farmers get their land, they often are only able to gain a few years of decent crop yields before the poor quality of the soil overwhelms their efforts, and then they are forced to move on to another plot of land. Small-scale farmers also hunt for meat in the remaining fragmented forest areas, which reduces the biodiversity in those areas as well.

Another important factor not mentioned in the scenario above is the clearing of rainforest for industrial agriculture plantations of bananas, pineapples, and sugar cane. These crops are primarily grown for export, and so an additional driver to consider is consumer demand for these crops in countries like the United States.

These cycles of land use, which are driven by poverty and population growth as well as government policies, have led to the rapid loss of tropical rainforests. What is lost in many cases is not simply biodiversity, but also valuable renewable resources that could sustain many generations of humans to come. Efforts to protect rainforests and other areas of high biodiversity is the topic of the next section.

Deforestation
Going Green
Reforestation
Privacy Policy

Deforestation: Case Studies

Deforestation is putting our planet at risk, as the following case studies exemplify. It is responsible for at least 10 per cent of global greenhouse gas emissions 1 and wipes out 137 species of plants, animals and insects every day 2 . The deplorable practice degenerates soil, losing half of the world’s topsoil over the past 150 years. 3 Deforestation also leads to drought by reducing the amount of water in the atmosphere. 4

Since the 1950s, deforestation has accelerated significantly, particularly in the tropics. 5 This is primarily due to rapid population growth and a resultant increase in demand for food and resources. 6 Agriculture drives about 80 per cent of deforestation today, as land is cleared for livestock, growing animal feed or other crops. 7 The below deforestation case studies of Brazil’s Amazon rainforest and the Congo Basin provide further insights into modern deforestation.

Deforestation case study: Brazil

Nearly two-thirds of the Amazon rainforest – the largest rainforest in the world – is within Brazil’s national borders. 8 Any examination of deforestation case studies would be incomplete without considering tree felling in Brazil.

History of deforestation in Brazil

Humans first discovered the Amazon rainforest about 13,000 years ago. But, it was the arrival of Europeans in the late 15th century that spurred the conversion of the forest into farmland. Nevertheless, the sheer size of the Amazon meant that the rainforest remained largely intact until the early 20th century. It was in the latter half of the 20th century that things began to change. 9

Hoatzin bird native to amazon rainforest

Industrial activities and large-scale agriculture began to eat away the southern and eastern fringes of the Amazon, from the 1950s onwards. 10 Deforestation in Brazil received a significant boost in 1964 when a military dictatorship took power and declared the jungle a security risk. 11 By the 1970s, the government was running television ads encouraging land conversion, provoking millions to migrate north into the forest. 12 Settlements replaced trees, and infrastructure began to develop. Wealthy tycoons subsequently bought the land for cattle ranches or vast fields of soy. 13

By the turn of the 21st century, more than 75 per cent of deforestation in the Amazon was for cattle ranching. But, environmentalists and Indigenous groups drew international attention to the devastation caused and succeeded in curtailing it by 2004. Between 2004 and the early 2010s, annual forest cover loss in Brazil reduced by about 80 per cent. The decline is attributed to “increased law enforcement, satellite monitoring, pressure from environmentalists, private and public sector initiatives, new protected areas, and macroeconomic trends”. 14

Brazil’s deforestation of the Amazon rainforest since 2010

Unfortunately, however, efforts to curtail deforestation in Brazil’s Amazon have stalled since 2012. 15 Tree felling and land conversion have been trending upwards ever since. The economic incentive for chopping the rainforest down has overcome the environmental benefits of leaving it standing. 16 Political movements and lax government legislation have leveraged this to their advantage. President Jair Bolsonaro won the 2018 election with a promise to open up the Amazon to business. 17 Since his inauguration, the rate of deforestation has leapt by as much as 92 per cent. 18

However, there is still hope for the Amazon rainforest. Bolsonaro’s principal international ally was US President Trump. Now that environmentally-conscious Joe Biden has replaced him in the White House, international pressure regarding deforestation will increase heavily. 19 Biden has made this clear with a promise of USD $20 billion to protect the Amazon. 20

The impact of continued deforestation in Brazil

For its three million plant and animal species and one million Indigenous inhabitants, it is imperative that Amazonian deforestation is massively and immediately reduced. 21 As much as 17 per cent of the Amazon has been lost already. 22 If this proportion increases to over 20 per cent, a tipping point will be reached. 23 This will irreversibly break the water cycle, and at least half of the remaining forest will become savannah. 24

Impact on climate change

Losing the Amazon would also mean losing the fight against climate change. Despite the rampant deforestation in recent years, the remaining Amazon rainforest still absorbs between 5 to 10 per cent of all human CO2 emissions. 25 Cutting trees down increases anthropogenic emissions. When felled, burned or left to rot, trees release sequestered carbon. 26 A combination of reducing greenhouse gas emissions and preserving existing forests is crucial to preventing dangerous levels of global warming. 27

Deforestation case study: The Congo Basin

The Congo Basin is the second-largest rainforest in the world. 28 It has been described as the ‘second lungs’ of the Earth because of how much carbon dioxide it absorbs and how much oxygen it produces. 29 But, just as the world’s first lungs – the Amazon – is being destroyed by humans, the Congo’s rainforest is also suffering heavy casualties. 30

60 per cent of the Congo Basin is located within the Democratic Republic of the Congo (DRC). 31 The DRC is one of the world’s largest and poorest countries, though it has immense economic resources. 32 Natural resources have fuelled an ongoing war that has affected all the neighbouring countries and claimed as many as six million lives. 33 The resultant instability combined with corruption and poor governance have led to an ever-increasing rate of deforestation within the DRC’s borders. 34

Deforestation in the Democratic Republic of the Congo (DRC)

Compared to the Amazon and Southeast Asia, deforestation in the Congo Basin has been low over the past few decades. 35 Nevertheless, great swathes of primary forest have been lost. Between 2000 and 2014, an area of forest larger than Bangladesh was destroyed. 36 From 2015 until 2019, 6.37 million hectares of tree cover was razed. 37 In 2019 alone, 475,000 hectares of primary forest disappeared, placing the DRC second only to Brazil for total deforestation that year. 38 Should the current rate of deforestation continue, all primary forest in the Congo Basin will be gone by the end of the century. 39

Drivers of deforestation in the DRC’s Congo Basin

Over the past 20 years, the biggest drivers of deforestation in the DRC has been small-scale subsistence agriculture. Clearing trees for charcoal and fuelwood, urban expansion and mining have also contributed to deforestation. Industrial logging is the most common cause of forest degradation. It opens up deeper areas of the forest to commercial hunting. There has been at least a 60 per cent drop in the region’s forest elephant populations over the past decade due to hunting and poaching. 40

Between 2000 and 2014, small-scale farming contributed to about 90 per cent of the DRC’s deforestation. This trend has not changed in recent years. The majority of small-scale forest clearing is conducted with simple axes by people with no other livelihood options. The region’s political instability and ongoing conflict are therefore inciting the unsustainable rate of deforestation within the Congo Basin. 41

In future, however, industrial logging and land conversion to large-scale agriculture will pose the greatest threats to the Congo rainforest. 42 There are fears that demand for palm oil, rubber and sugar production will promote a massive increase in deforestation. 43 The DRC’s population is also predicted to grow to almost 200 million people by 2050. 44 This increase will threaten the remaining rainforest further, as they try to earn a living in a country deprived of opportunities. 45

The impact of deforestation in the Congo Basin

80 million people depend upon the Congo Basin for their existence. It provides food, charcoal, firewood, medicinal plants, and materials for building and other purposes. But, this rainforest also indirectly supports people across the whole of sub-Saharan Africa. Like all forests, it is instrumental in regulating rainfall, which can affect precipitation hundreds of miles away. The Congo Basin is a primary source of rainfall for the Sahel region, doubling the amount of rainfall in the air that passes over it. 46

The importance of the Congo Basin’s ability to increase precipitation cannot be understated. Areas such as the Horn of Africa are becoming increasingly dry. Drought in Ethiopia and Somalia has put millions of people on emergency food and water rations in recent years. Destroying the DRC’s rainforest would create the largest humanitarian crisis on Earth. 47

It would also be devastating for biodiversity. The Congo Basin shelters some 10,000 animal species and more than 600 tree species. 48 They play a hugely important role in the forest, which has consequences for the entire planet. For instance, elephants, gorillas, and other large herbivores keep the density of small trees very low through predation. 49 This results in a high density of tall trees in the Congo rainforest. 50 Larger trees store more carbon and therefore help to prevent global warming by removing this greenhouse gas from the atmosphere. 51

Preserve our forests

Preserving the Amazon and Congo Basin rainforests is vital for tackling climate change, as these deforestation case studies demonstrate. We must prioritise protecting and enhancing our existing trees if we are to limit the global temperature increase to 1.5°C, as recommended by the IPCC. 52

Rainforest Alliance. (2018). What is the Relationship Between Deforestation And Climate Change? [online] Available at: https://www.rainforest-alliance.org/articles/relationship-between-deforestation-climate-change.
www.worldanimalfoundation.com. (n.d.). Deforestation: Clearing The Path For Wildlife Extinctions. [online] Available at: https://www.worldanimalfoundation.com/advocate/wild-earth/params/post/1278141/deforestation-clearing-the-path-for-wildlife-extinctions#:~:text=Seventy%20percent%20of%20the%20Earth.
World Wildlife Fund. (2000). Soil Erosion and Degradation | Threats | WWF. [online] Available at: https://www.worldwildlife.org/threats/soil-erosion-and-degradation.
Butler, R.A. (2001). The impact of deforestation. [online] Mongabay. Available at: https://rainforests.mongabay.com/09-consequences-of-deforestation.html.
The Classroom | Empowering Students in Their College Journey. (2009). The History of Deforestation. [online] Available at: https://www.theclassroom.com/the-history-of-deforestation-13636286.html.
Greenpeace USA. (n.d.). Agribusiness & Deforestation. [online] Available at: https://www.greenpeace.org/usa/forests/issues/agribusiness/.
Yale.edu. (2015). The Amazon Basin Forest | Global Forest Atlas. [online] Available at: https://globalforestatlas.yale.edu/region/amazon.
Time. (2019). The Amazon Rain Forest Is Nearly Gone. We Went to the Front Lines to See If It Could Be Saved. [online] Available at: https://time.com/amazon-rainforest-disappearing/.
Butler, R. (2020). Amazon Destruction. [online] Mongabay.com. Available at: https://rainforests.mongabay.com/amazon/amazon_destruction.html.
the Guardian. (2020). Amazon deforestation surges to 12-year high under Bolsonaro. [online] Available at: https://www.theguardian.com/environment/2020/dec/01/amazon-deforestation-surges-to-12-year-high-under-bolsonaro.
Earth Innovation Institute. (2020). Joe Biden offers $20 billion to protect Amazon forests. [online] Available at: https://earthinnovation.org/2020/03/joe-biden-offers-20-billion-to-protect-amazon-forests/.
Brazil’s Amazon: Deforestation “surges to 12-year high.” (2020). BBC News. [online] 30 Nov. Available at: https://www.bbc.co.uk/news/world-latin-america-55130304.
Carbon Brief. (2020). Guest post: Could climate change and deforestation spark Amazon “dieback”? [online] Available at: https://www.carbonbrief.org/guest-post-could-climate-change-and-deforestation-spark-amazon-dieback.
Union of Concerned Scientists (2012). Tropical Deforestation and Global Warming | Union of Concerned Scientists. [online] www.ucsusa.org. Available at: https://www.ucsusa.org/resources/tropical-deforestation-and-global-warming#:~:text=When%20trees%20are%20cut%20down.
Milman, O. (2018). Scientists say halting deforestation “just as urgent” as reducing emissions. [online] the Guardian. Available at: https://www.theguardian.com/environment/2018/oct/04/climate-change-deforestation-global-warming-report.
Bergen, M. (2019). Congo Basin Deforestation Threatens Food and Water Supplies Throughout Africa. [online] World Resources Institute. Available at: https://www.wri.org/blog/2019/07/congo-basin-deforestation-threatens-food-and-water-supplies-throughout-africa.
www.esa.int. (n.d.). Earth from Space: “Second lungs of the Earth.” [online] Available at: https://www.esa.int/Applications/Observing_the_Earth/Earth_from_Space_Second_lungs_of_the_Earth [Accessed 26 Feb. 2021].
Erickson-Davis, M. (2018). Congo Basin rainforest may be gone by 2100, study finds. [online] Mongabay Environmental News. Available at: https://news.mongabay.com/2018/11/congo-basin-rainforest-may-be-gone-by-2100-study-finds/.
Mongabay Environmental News. (2020). Poor governance fuels “horrible dynamic” of deforestation in DRC. [online] Available at: https://news.mongabay.com/2020/12/poor-governance-fuels-horrible-dynamic-of-deforestation-in-drc/ [Accessed 26 Feb. 2021].
DR Congo country profile. (2019). BBC News. [online] 10 Jan. Available at: https://www.bbc.co.uk/news/world-africa-13283212.
Butler, R.A. (2001). Congo Deforestation. [online] Mongabay. Available at: https://rainforests.mongabay.com/congo/deforestation.html.
Mongabay Environmental News. (2020). Poor governance fuels “horrible dynamic” of deforestation in DRC. [online] Available at: https://news.mongabay.com/2020/12/poor-governance-fuels-horrible-dynamic-of-deforestation-in-drc/.
Butler, R. (2020). The Congo Rainforest. [online] Mongabay.com. Available at: https://rainforests.mongabay.com/congo/.
Editor, B.W., Environment (n.d.). Large trees are carbon-storing giants. www.thetimes.co.uk. [online] Available at: https://www.thetimes.co.uk/article/large-trees-are-more-valuable-carbon-stores-than-was-thought-k8hnggzs8#:~:text=The%20world [Accessed 26 Feb. 2021].
IPCC (2018). Summary for Policymakers — Global Warming of 1.5 oC. [online] Ipcc.ch. Available at: https://www.ipcc.ch/sr15/chapter/spm/.

Top 10 Facts about Deforestation
Deforestation Debate
When a Tree Falls It Makes More Than a Sound
What is Deforestation?

10 Facts About Deforestation in 2021

The following facts about deforestation in 2021 show that our planet is in a precarious position. We must leave our...

Where Is Deforestation Happening in 2021?

To combat climate change, it is important to know where deforestation is happening in 2021. Deforestation currently accounts for at...

The History of Deforestation: The Past and Origins

In many ways, the history of deforestation is the same as the history of agriculture. When humans began to farm...

The Economic Impact of Deforestation

Deforestation is the permanent removal of forests. Agriculture usually replaces the woodland.Youmatter (2020). Deforestation - What Is It? What Are...

Silviculture and Timber Production

Over the past decade, millions of trees have been dying at a frightening pace across the US and Canada. Some...

How Does Planting Trees Affect Climate Change?

While climate change can be a politically controversial topic, tree planting rarely is. Planting trees and climate change are directly...

10 Causes of Deforestation: The Roots of Forest Degradation

How Much CO2 Does a Tree Absorb?

The Effect of Oil Sands Tailings Ponds on Climate Change

Privacy & Policy

Privacy Overview

Deforestation, warming flip part of Amazon forest from carbon sink to source

July 14, 2021

The study area, which represents about 20 percent of the Amazon basin, has lost 30 percent of its rainforest

New results from a nine-year research project in the eastern Amazon rainforest finds that significant deforestation in eastern and southeastern Brazil has been associated with a long-term decrease in rainfall and increase in temperature during the dry season, turning what was once a forest that absorbed carbon dioxide into a source of planet-warming carbon dioxide emissions.

The study, published in the journal Nature , explored whether these changes had altered how much carbon the Amazon stored in its vast forests.

“Using nearly 10 years of CO 2 (carbon dioxide ) measurements, we found that the more deforested and climate-stressed eastern Amazon, especially the southeast, was a net emitter of CO 2 to the atmosphere, especially as a result of fires,” said John Miller, a scientist with NOAA’s Global Monitoring Laboratory and a co-author. “On the other hand, the wetter, more intact western and central Amazon, was neither a carbon sink nor source of atmospheric CO 2 , with the absorption by healthy forests balancing the emissions from fires.”

In addition to storing vast amounts of carbon, Amazonia is also one of the wettest places on Earth, storing immense amounts of water in its soils and vegetation. Transpired by leaves, this moisture evaporates into the atmosphere, where it fuels prodigious rainfall, averaging more than seven feet per year across the basin. For comparison the average annual rainfall in the contiguous U.S. is two and half feet. Several studies have estimated that water cycling through evaporation is responsible for 25 to 35 percent of total rainfall in the basin.

But deforestation and global warming over the last 40 years have affected rainfall and temperature with potential impacts for the Amazon’s ability to store carbon. Conversion of rainforest to agriculture has caused a 17 percent decrease in forest extent in the Amazon, which stretches over an area almost as large as the continental U.S.. Replacing dense, humid forest canopies with drier pastures and cropland has increased local temperatures and decreased evaporation of water from the rainforest, which deprives downwind locations of rainfall. Regional deforestation and selective logging of adjacent forests further reduces forest cover, amplifying the cycle of drying and warming. This, in turn, can reduce the capacity of the forests to store carbon, and increase their vulnerability to fires.

The 2.8 million square miles of jungle in the Amazon basin represents more than half of the tropical rainforest remaining on the planet. The Amazon is estimated to contain about 123 billion tons of carbon above and below ground, and is one of Earth’s most important terrestrial carbon reserves. As global fossil-fuel burning has risen, the Amazon has absorbed CO 2 from the atmosphere, helping to moderate global climate. But there are indications from this study and previous ones that the Amazon’s capacity to act as a sink may be disappearing.

Over the past several decades, intense scientific interest has focused on the question of whether the combined effects of climate change and the ongoing conversion of jungle to pasture and cropland could cause the Amazon to release more carbon dioxide than it absorbs.

In 2010, lead author Luciana Gatti, who led the international team of scientists from Brazil, the United Kingdom, New Zealand and the Netherlands, set out to explore this question. During the next nine years, Gatti, a scientist with Brazil’s National Institute for Space Research and colleagues obtained airborne measurements of CO 2 and carbon monoxide concentrations above Brazilian Amazonia. Analysis of CO 2 measurements from over 600 aircraft vertical profiles, extending from the surface to around 2.8 miles above sea level at four sites, revealed that total carbon emissions in eastern Amazonia are greater than those in the west.

“The regions of southern Pará and northern Mato Grosso states represent a worst-case scenario,” said Gatti.

The southeast region, which represents about 20 percent of the Amazon basin, and has experienced 30 percent deforestation over the previous 40 years. Scientists recorded a 25 percent reduction in precipitation and a temperature increase of at least 2.7 degrees Fahrenheit during the dry months of August, September and October, when trees are already under seasonal stress. Airborne measurements over nine years revealed this region was a net emitter of carbon, mainly as a result of fires, while areas further west, where less than 20 percent of the forest had been removed, sources balanced sinks. The scientists said the increased emissions were likely due to conversion of forest to cropland by burning, and by reduced uptake of CO 2 by the trees that remained.

These findings help scientists better understand the long-term impacts of interactions between climate and human disturbances on the carbon balance of the world’s largest tropical forest.

“The big question this research raises is if the connection between climate, deforestation, and carbon that we see in the eastern Amazon could one day be the fate of the central and western Amazon, if they become subject to stronger human impact,” Miller said. Changes in the capacity of tropical forests to absorb carbon will require downward adjustments of the fossil fuel emissions compatible with limiting global mean temperature increases to less than 2.0 or 1.5 degrees Celsius, he added.

This research was supported by NOAA’s Global Monitoring Laboratory and by funding from the State of Sao Paulo Science Foundation, UK Environmental Research Council, NASA, and the European Research Council.

For more information, contact Theo Stein, NOAA Communications: [email protected]

Air sampling flasks are lined up on a manifold for intake processing in NOAA's Global Monitoring Laboratory. Credit: Lauren Lipuma, CIRES

No sign of greenhouse gases increases slowing in 2023

5 science wins from the 2023 NOAA Science Report

Red autonomous glider in the shape of a surfboard with a sail is connected to a towline for deployment

Filling A Data Gap In The Tropical Pacific To Reveal Daily Air-Sea Interactions

Satellite image showing marine stratus clouds and linear "ship-track" exhaust plumes.

Scientists detail research to assess viability and risks of marine cloud brightening

Popup call to action.

A prompt with more information on your call to action.

Our websites may use cookies to personalize and enhance your experience. By continuing without changing your cookie settings, you agree to this collection. For more information, please see our University Websites Privacy Notice .

UConn Today

School and College News
Arts & Culture
Community Impact
Entrepreneurship
Health & Well-Being
Research & Discovery
UConn Health
University Life
UConn Voices
University News

September 7, 2021 | Combined Reports - UConn Communications

Study Shows the Impacts of Deforestation and Forest Burning on Biodiversity in the Amazon

Since 2001, between 40,000 and 73,400 square miles of Amazon rainforest have been impacted by fires

Ring of fire: Smoke rises through the understory of a forest in the Amazon region. Plants and animals in the Amazonian rainforest evolved largely without fire, so they lack the adaptations necessary to cope with it. (Credit: Paulo Brando)

A new study, co-authored by a team of researchers including UConn Ecology and Evolutionary Biology researcher Cory Merow provides the first quantitative assessment of how environmental policies on deforestation, along with forest fires and drought, have impacted the diversity of plants and animals in the Amazon. The findings were published in the Sept. 1 issue of Nature .

Researchers used records of more than 14,500 plant and vertebrate species to create biodiversity maps of the Amazon region. Overlaying the maps with historical and current observations of forest fires and deforestation over the last two decades allowed the team to quantify the cumulative impacts on the region’s species.

They found that since 2001, between 40,000 and 73,400 square miles of Amazon rainforest have been impacted by fires, affecting 95% of all Amazonian species and as many as 85% of species that are listed as threatened in this region. While forest management policies enacted in Brazil during the mid-2000s slowed the rate of habitat destruction, relaxed enforcement of these policies coinciding with a change in government in 2019 has seemingly begun to reverse this trend, the authors write. With fires impacting 1,640 to 4,000 square miles of forest, 2019 stands out as one of the most extreme years for biodiversity impacts since 2009, when regulations limiting deforestation were enforced.

“Perhaps most compelling is the role that public pressure played in curbing forest loss in 2019,” Merow says. “When the Brazilian government stopped enforced forest regulations in 2019, each month between January and August 2019 was the worse month on record (e.g. comparing January 2019 to previous January’s) for forest loss in the 20-year history of available data. However, based on international pressure, forest regulation resumed in September 2019, and forest loss declined significantly for the rest of the year, resulting in 2019 looking like an average year compared to the 20-year history. This was big: active media coverage and public support for policy changes were effective at curbing biodiversity loss on a very rapid time scale.”

The findings are especially critical in light of the fact that at no point in time did the Amazon get a break from those increasing impacts, which would have allowed for some recovery, says senior study author Brian Enquist, a professor in UArizona’s Department of Ecology and Evolutionary Biology .

“Even with policies in place, which you can think of as a brake slowing the rate of deforestation, it’s like a car that keeps moving forward, just at a slower speed,” Enquist says. “But in 2019, it’s like the foot was let off the brake, causing it to accelerate again.”

Known mostly for its dense rainforests, the Amazon basin supports around 40% of the world’s remaining tropical forests. It is of global importance as a provider of ecosystem services such as scrubbing and storing carbon from the atmosphere, and it plays a vital role in regulating Earth’s climate. The area also is an enormous reservoir of the planet’s biodiversity, providing habitats for one out of every 10 of the planet’s known species. It has been estimated that in the Amazon, 1,000 tree species can populate an area smaller than a half square mile.

“Fire is not a part of the natural cycle in the rainforest,” says study co-author Crystal N. H. McMichael at the University of Amsterdam. “Native species lack the adaptations that would allow them to cope with it, unlike the forest communities in temperate areas. Repeated burning can cause massive changes in species composition and likely devastating consequences for the entire ecosystem.”

Since the 1960s, the Amazon has lost about 20% of its forest cover to deforestation and fires. While fires and deforestation often go hand in hand, that has not always been the case, Enquist says. As climate change brings more frequent and more severe drought conditions to the region, and fire is often used to clear large areas of rainforest for the agricultural industry, deforestation has spillover effects by increasing the chances of wildfires. Forest loss is predicted reach 21 to 40% by 2050, and such habitat loss will have large impacts on the region’s biodiversity, according to the authors.

“Since the majority of fires in the Amazon are intentionally set by people, preventing them is largely within our control,” says study co-author Patrick Roehrdanz, senior manager of climate change and biodiversity at Conservation International. “One way is to recommit to strong antideforestation policies in Brazil, combined with incentives for a forest economy, and replicate them in other Amazonian countries.”

Policies to protect Amazonian biodiversity should include the formal recognition of Indigenous lands, which encompass more than one-third of the Amazon region, the authors write, pointing to previous research showing that lands owned, used or occupied by Indigenous peoples have less species decline, less pollution and better-managed natural resources.

The authors say their study underscores the dangers of continuing lax policy enforcement. As fires encroach on the heart of the Amazon basin, where biodiversity is greatest, their impacts will have more dire effects, even if the rate of forest burning remains unchanged.

The research was made possible by strategic investment funds allocated by the Arizona Institutes for Resilience at UArizona and the university’s Bridging Biodiversity and Conservation Science group. Additional support came from the National Science Foundation’s Harnessing the Data Revolution program . Data and computation were provided through the Botanical Information and Ecology Network , which is supported by CyVerse , the NSF’s data management platform led by UArizona.

Recent Articles

May 1, 2024

UConn Staff Raises Awareness for Humane Biomedical Research

Read the article

Researchers Explore Non-Surgical Approaches to Treat TMD Pain, Free Treatment Available

A group of college students poses for a photo in a classroom while wearing sock puppets on their hands.

Students From Public Health House Learning Community Feel Their Best Self

How to tackle the global deforestation crisis

Imagine if France, Germany, and Spain were completely blanketed in forests — and then all those trees were quickly chopped down. That’s nearly the amount of deforestation that occurred globally between 2001 and 2020, with profound consequences.

Deforestation is a major contributor to climate change, producing between 6 and 17 percent of global greenhouse gas emissions, according to a 2009 study. Meanwhile, because trees also absorb carbon dioxide, removing it from the atmosphere, they help keep the Earth cooler. And climate change aside, forests protect biodiversity.

“Climate change and biodiversity make this a global problem, not a local problem,” says MIT economist Ben Olken. “Deciding to cut down trees or not has huge implications for the world.”

But deforestation is often financially profitable, so it continues at a rapid rate. Researchers can now measure this trend closely: In the last quarter-century, satellite-based technology has led to a paradigm change in charting deforestation. New deforestation datasets, based on the Landsat satellites, for instance, track forest change since 2000 with resolution at 30 meters, while many other products now offer frequent imaging at close resolution.

“Part of this revolution in measurement is accuracy, and the other part is coverage,” says Clare Balboni, an assistant professor of economics at the London School of Economics (LSE). “On-site observation is very expensive and logistically challenging, and you’re talking about case studies. These satellite-based data sets just open up opportunities to see deforestation at scale, systematically, across the globe.”

Balboni and Olken have now helped write a new paper providing a road map for thinking about this crisis. The open-access article, “ The Economics of Tropical Deforestation ,” appears this month in the Annual Review of Economics . The co-authors are Balboni, a former MIT faculty member; Aaron Berman, a PhD candidate in MIT’s Department of Economics; Robin Burgess, an LSE professor; and Olken, MIT’s Jane Berkowitz Carlton and Dennis William Carlton Professor of Microeconomics. Balboni and Olken have also conducted primary research in this area, along with Burgess.

So, how can the world tackle deforestation? It starts with understanding the problem.

Replacing forests with farms

Several decades ago, some thinkers, including the famous MIT economist Paul Samuelson in the 1970s, built models to study forests as a renewable resource; Samuelson calculated the “maximum sustained yield” at which a forest could be cleared while being regrown. These frameworks were designed to think about tree farms or the U.S. national forest system, where a fraction of trees would be cut each year, and then new trees would be grown over time to take their place.

But deforestation today, particularly in tropical areas, often looks very different, and forest regeneration is not common.

Indeed, as Balboni and Olken emphasize, deforestation is now rampant partly because the profits from chopping down trees come not just from timber, but from replacing forests with agriculture. In Brazil, deforestation has increased along with agricultural prices; in Indonesia, clearing trees accelerated as the global price of palm oil went up, leading companies to replace forests with palm tree orchards.

All this tree-clearing creates a familiar situation: The globally shared costs of climate change from deforestation are “externalities,” as economists say, imposed on everyone else by the people removing forest land. It is akin to a company that pollutes into a river, affecting the water quality of residents.

“Economics has changed the way it thinks about this over the last 50 years, and two things are central,” Olken says. “The relevance of global externalities is very important, and the conceptualization of alternate land uses is very important.” This also means traditional forest-management guidance about regrowth is not enough. With the economic dynamics in mind, which policies might work, and why?

The search for solutions

As Balboni and Olken note, economists often recommend “Pigouvian” taxes (named after the British economist Arthur Pigou) in these cases, levied against people imposing externalities on others. And yet, it can be hard to identify who is doing the deforesting.

Instead of taxing people for clearing forests, governments can pay people to keep forests intact. The UN uses Payments for Environmental Services (PES) as part of its REDD+ (Reducing Emissions from Deforestation and forest Degradation) program. However, it is similarly tough to identify the optimal landowners to subsidize, and these payments may not match the quick cash-in of deforestation. A 2017 study in Uganda showed PES reduced deforestation somewhat; a 2022 study in Indonesia found no reduction; another 2022 study, in Brazil, showed again that some forest protection resulted.

“There’s mixed evidence from many of these [studies],” Balboni says. These policies, she notes, must reach people who would otherwise clear forests, and a key question is, “How can we assess their success compared to what would have happened anyway?”

Some places have tried cash transfer programs for larger populations. In Indonesia, a 2020 study found such subsidies reduced deforestation near villages by 30 percent. But in Mexico, a similar program meant more people could afford milk and meat, again creating demand for more agriculture and thus leading to more forest-clearing.

At this point, it might seem that laws simply banning deforestation in key areas would work best — indeed, about 16 percent of the world’s land overall is protected in some way. Yet the dynamics of protection are tricky. Even with protected areas in place, there is still “leakage” of deforestation into other regions.

Still more approaches exist, including “nonstate agreements,” such as the Amazon Soy Moratorium in Brazil, in which grain traders pledged not to buy soy from deforested lands, and reduced deforestation without “leakage.”

Also, intriguingly, a 2008 policy change in the Brazilian Amazon made agricultural credit harder to obtain by requiring recipients to comply with environmental and land registration rules. The result? Deforestation dropped by up to 60 percent over nearly a decade.

Politics and pulp

Overall, Balboni and Olken observe, beyond “externalities,” two major challenges exist. One, it is often unclear who holds property rights in forests. In these circumstances, deforestation seems to increase. Two, deforestation is subject to political battles.

For instance, as economist Bard Harstad of Stanford University has observed, environmental lobbying is asymmetric. Balboni and Olken write: “The conservationist lobby must pay the government in perpetuity … while the deforestation-oriented lobby need pay only once to deforest in the present.” And political instability leads to more deforestation because “the current administration places lower value on future conservation payments.”

Even so, national political measures can work. In the Amazon from 2001 to 2005, Brazilian deforestation rates were three to four times higher than on similar land across the border, but that imbalance vanished once the country passed conservation measures in 2006. However, deforestation ramped up again after a 2014 change in government. Looking at particular monitoring approaches, a study of Brazil’s satellite-based Real-Time System for Detection of Deforestation (DETER), launched in 2004, suggests that a 50 percent annual increase in its use in municipalities created a 25 percent reduction in deforestation from 2006 to 2016.

How precisely politics matters may depend on the context. In a 2021 paper, Balboni and Olken (with three colleagues) found that deforestation actually decreased around elections in Indonesia. Conversely, in Brazil, one study found that deforestation rates were 8 to 10 percent higher where mayors were running for re-election between 2002 and 2012, suggesting incumbents had deforestation industry support.

“The research there is aiming to understand what the political economy drivers are,” Olken says, “with the idea that if you understand those things, reform in those countries is more likely.”

Looking ahead, Balboni and Olken also suggest that new research estimating the value of intact forest land intact could influence public debates. And while many scholars have studied deforestation in Brazil and Indonesia, fewer have examined the Democratic Republic of Congo, another deforestation leader, and sub-Saharan Africa.

Deforestation is an ongoing crisis. But thanks to satellites and many recent studies, experts know vastly more about the problem than they did a decade or two ago, and with an economics toolkit, can evaluate the incentives and dynamics at play.

“To the extent that there’s ambuiguity across different contexts with different findings, part of the point of our review piece is to draw out common themes — the important considerations in determining which policy levers can [work] in different circumstances,” Balboni says. “That’s a fast-evolving area. We don’t have all the answers, but part of the process is bringing together growing evidence about [everything] that affects how successful those choices can be.”

Suggestions or feedback?

MIT News | Massachusetts Institute of Technology

Machine learning
Social justice
Black holes
Classes and programs

Departments

Aeronautics and Astronautics
Brain and Cognitive Sciences
Architecture
Political Science
Mechanical Engineering

Centers, Labs, & Programs

Abdul Latif Jameel Poverty Action Lab (J-PAL)
Picower Institute for Learning and Memory
Lincoln Laboratory
School of Architecture + Planning
School of Engineering
School of Humanities, Arts, and Social Sciences
Sloan School of Management
School of Science
MIT Schwarzman College of Computing

How to tackle the global deforestation crisis

Press contact :, media download.

Aerial view of a forest, with a large wedge of deforested, brown dirt.

*Terms of Use:

Images for download on the MIT News office website are made available to non-commercial entities, press and the general public under a Creative Commons Attribution Non-Commercial No Derivatives license . You may not alter the images provided, other than to crop them to size. A credit line must be used when reproducing images; if one is not provided below, credit the images to "MIT."

Previous image Next image

“Climate change and biodiversity make this a global problem, not a local problem,” says MIT economist Ben Olken. “Deciding to cut down trees or not has huge implications for the world.”

So, how can the world tackle deforestation? It starts with understanding the problem.

Replacing forests with farms

But deforestation today, particularly in tropical areas, often looks very different, and forest regeneration is not common.

The search for solutions

Politics and pulp

“The research there is aiming to understand what the political economy drivers are,” Olken says, “with the idea that if you understand those things, reform in those countries is more likely.”

Share this news article on:

A clean alternative to one of the world’s most common ingredients

Barefoot woman on a ladder, adjusting a solar light

Keeping humanity central to solving climate change

3 Questions: Benjamin Olken on the economic impact of climate change

The economic cost of increased temperatures

Previous item Next item

More MIT News

An elderly Francis Fan Lee wears a blue jumpsuit and an expression of pure joy while floating mid-air on a reduced-gravity aircraft.

Francis Fan Lee, former professor and interdisciplinary speech processing inventor, dies at 96

Read full story →

Fostering research, careers, and community in materials science

Three boxes demonstrate different tasks assisted by natural language. One is a rectangle showing colorful lines of code with a white speech bubble highlighting an abstraction; another is a pale 3D kitchen, and another is a robotic quadruped dropping a can into a trash bin.

Natural language boosts LLM performance in coding, planning, and robotics

Photo of Nuno Loureiro seated indoors on a white lounge chair

Nuno Loureiro named director of MIT’s Plasma Science and Fusion Center

James Simon smiles while sitting on a ping pong table, with windows in background.

Studies in empathy and analytics

Laurence Willemet stands on stage and gestures toward her research poster.

Science communication competition brings research into the real world

More news on MIT News homepage →

Massachusetts Institute of Technology 77 Massachusetts Avenue, Cambridge, MA, USA

Map (opens in new window)
Events (opens in new window)
People (opens in new window)
Careers (opens in new window)
Accessibility
Social Media Hub
MIT on Facebook
MIT on YouTube
MIT on Instagram

Drivers of Deforestation

The world loses 5 million hectares of forest to deforestation each year. what activities are driving this, cutting down forests: what are the drivers of deforestation.

Every year the world loses around 5 million hectares of forest. 95% of this occurs in the tropics. At least three-quarters of this is driven by agriculture – clearing forests to grow crops, raise livestock and produce products such as paper. 1

If we want to tackle deforestation we need to understand two key questions: where we’re losing forests, and what activities are driving it. This allows us to target our efforts towards specific industries, products, or countries where they will have the greatest impact.

In a study published in Global Environmental Change , Florence Pendrill and colleagues addressed both of these questions. 2 They quantified how much and where deforestation occurs from the expansion of croplands, pasture and tree plantations (for logging), and what products are grown on this converted land. They also combined this with global trade flows to assess how much of this deforestation was driven by international trade – we look at the role of trade specifically in a related article .

Here we’ll look at both where tropical deforestation is happening and what products are driving it .

Brazil and Indonesia account for almost half of tropical deforestation

The study by Pendrill et al. (2019) found that, between 2005 and 2013, the tropics lost an average of 5.5 million hectares of forest per year to agricultural land. That was a decade ago, but the world is still losing a similar amount today: using satellite data, researchers at Global Forest Watch estimate that global deforestation in 2019 was around 5.4 million hectares. 3 95% of this was in the tropics. But where in the tropics did we lose this forest?

In the chart we see the share of tropical deforestation by country and region. It's measured as the annual average between 2010 and 2014.

One-third of tropical deforestation happened in Brazil. That was 1.7 million hectares each year. The other single country where large forest areas are lost is Indonesia – it accounted for 14%. This means around half (47%) of tropical deforestation took place in Brazil and Indonesia. Again, if we look at more recent satellite data we find that this is still true today: in 2019, the world lost 5.4 million hectares to deforestation, with Brazil and Indonesia accounting for 52% of it. 4 As we will see later, the expansion of pasture for beef production, croplands for soy and palm oil, and increasingly conversion of primary forest to tree plantations for paper and pulp have been the key drivers of this.

The expansion of pasture lands have also had a major impact on land use in the rest of the Americas – outside of Brazil, Latin America accounted for around one-fifth of deforestation.

The expansion of agricultural land in Africa accounted for around 17.5% of deforestation. This may slightly underestimate the loss of forests in Africa, for two reasons. Much of Africa’s deforestation has been driven by subsistence agricultural activities, which are not always fully captured in national statistics. Secondly, depending on the permanence of agricultural activities such as slash-and-burn farming, some of this forest loss might be classified as temporary forest degradation rather than permanent deforestation.

Beef, soy and palm oil are responsible for 60% of tropical deforestation

If we want to tackle deforestation we also need to know what causes it. That allows us to avoid the foods that drive deforestation or innovate the ways we produce them.

In the chart here we see the breakdown of tropical deforestation by the types of agricultural production.

Beef stands out immediately. The expansion of pasture land to raise cattle was responsible for 41% of tropical deforestation. That’s 2.1 million hectares every year – about half the size of the Netherlands. Most of this converted land came from Brazil; its expansion of beef production accounts for one-quarter (24%) of tropical deforestation. This also means that most (72%) deforestation in Brazil is driven by cattle ranching. 5 Cattle in other parts of Latin America – such as Argentina and Paraguay – also accounted for a large amount of deforestation – 11% of the total. Most deforestation for beef therefore occurs in Latin America, with another 4% happening in Africa.

Palm oil and soy often claim the headlines for their environmental impact. They are categorised as ‘oilseeds’, which also include a range of smaller commodities such as sunflower, rapeseed, and sesame. They drove 18% of deforestation.

Here we see that Indonesian palm oil was the biggest component of this. In neighbouring Malaysia the expansion of oil seeds was also a major driver of forest loss. Soybeans are the most common oilseed in Latin America. While many people immediately think of food products such as tofu or soy milk, most of global soybean production is used as feed for livestock, or biofuels. Just 6% is used for direct human food. The impact of soy production is one we look at in more detail in a related article .

Combined, beef and oilseeds account for nearly 60% of deforestation.

If we add the third largest driver – forestry products, which is dominated by paper but also includes timber – then we cover almost three-quarters. Across Europe and North America, forestry products mainly come from managed plantation forests that have been established for a long period of time, or are grown on previously unforested land. This is different from most tropical countries where forestry products also come from the logging of primary rainforests or their replacement with plantations. This destroys primary rainforests and, as shown in the chart, has been an important driver of deforestation in Indonesia and elsewhere in Asia.

We can tackle a lot of deforestation by focusing on a few key supply chains

If almost three-quarters of tropical deforestation is driven by the production of a few key products – beef, soybeans, palm oil, and paper – then we can achieve a lot by focusing our efforts on these supply chains.

There are some signs that progress is possible. Soybean production in Brazil was once also an important driver of deforestation in the Amazon region. 6 In 2006, under pressure from retailers and NGOs, the world’s major soybean traders signed Brazil’s Soy Moratorium (SoyM) – the world’s first voluntary zero-deforestation agreement. Traders agreed that they would not purchase soy that was grown on deforested lands in the Brazilian Amazon after July 2006. Overall, it was a success: in the two years before the agreement, 30% of soybean expansion in the region came at the expense of forest; afterwards, deforestation declined dramatically and by 2014 only 1% of expansion was turning forests into land for oilseed production. 7 , 8

But, as we show in our article on the impact of soy, there are also lessons to learn about how to implement these commitments more effectively. There is evidence that while the moratorium reduced deforestation rates in the Brazilian Amazon, some of this deforestation may have ‘leaked’ to neighbouring regions. Soybean production has shifted from the Amazon to the Cerrado region south of the Amazonas, often at the expense of forests there. 9 This suggests that zero-deforestation agreements can be effective but must be considered in the wider context of how they shape forest and agricultural changes elsewhere. To combat this, researchers have suggested the SoyM be expanded to not only include the Amazon but also regions such as the Cerrado. 7

If we can take similar action in the other industries – beef, palm oil and paper – then there is the potential to cut out a large share of deforestation today.

Looking to the future, a shift in focus towards Sub-Saharan Africa looks likely. The demands for increased agricultural production in Africa are going to be large, and could come at the cost of forests. 10 Solutions there will have to focus on major improvements in crop yields so African farmers can produce more food without increasing the amount of land they need to do so.

Alternative ways of making high-quality protein could also be transformative. Beef is the leading driver of deforestation, and the demand for meat across the world will continue to grow in the coming decades. Technological innovations in meat substitute and cultured meat products would allow people to continue eating meat-like products without the destruction of tropical forests that come with it.

Is our appetite for soy driving deforestation in the Amazon?

Soy has earned itself a bad reputation with many consumers. Its links to deforestation means that, alongside palm oil, soy has become a product to avoid. Is this reputation justified?

In this article we will take a look at the story of soy: how production has changed over time; where it is produced; what it is used for; and whether it really has been a key driver of deforestation. Although the research suggests that by far the largest driver of deforestation in the Brazilian Amazon has been driven by the expansion of pasture land for beef production, soy is likely to have played at least some role in the loss of forest.

More than three-quarters (77%) of global soy is fed to livestock for meat and dairy production. Most of the rest is used for biofuels, industry or vegetable oils. Just 7% of soy is used directly for human food products such as tofu, soy milk, edamame beans, and tempeh. The idea that foods often promoted as substitutes for meat and dairy – such as tofu and soy milk – are driving deforestation is a common misconception.

How has demand for soy changed over time?

Global soy production has exploded over the past 50 years. Global production today is more than 13 times higher than it was in the early 1960s. Even since the year 2000, production has more than doubled.

In the chart here we see the change in global soybean production from 1961 onwards. 11 Back in the 1960s, we were producing 20 to 30 million tonnes per year. This is now 350 million tonnes.

We can increase agricultural production in two ways: by improving yields (growing more on a given plot of land) or expanding the amount of land we use. As we will see later, although countries have seen yield gains over time, much of the increase in production has been driven by the expansion of croplands. Unfortunately, some of this has been at the expense of forests.

Which countries produce the most soy?

Global soy production has increased rapidly over the past 50 years. But to understand whether this has come at a major environmental cost, we also need to understand where this growth has come from.

Which countries are the main producers of soy?

In the map we see the distribution of soybean production across the world. Most of the world’s soy comes from only two countries: the US and Brazil. Combined, they account for more than two-thirds (69%) of global soy production. 12 In fact, they produce almost exactly the same amount: in 2018 the US produced 123 million tonnes, and Brazil 118 million tonnes. Individually, they each account for around one-third of global production. The other major producer is Argentina, which accounts for 11% (at 40 million tonnes).

To understand soy as a potential driver of deforestation, it’s also useful to understand how each country’s production has changed over time. It is the change in production, and how this was achieved (either increased yields or cropland expansion), that is the potential driver of deforestation.

In the other chart we see the change in soy production in Brazil and the US. The US was already producing a lot of soy throughout the 1960s, 70s and 80s and so its growth in recent decades has been much slower than Brazil. The US was growing 20 million tonnes per year as early as 1960. Brazil did not reach this level of output until 1990 – three decades later. This means that its rate of growth in the last 30 years has been much faster.

More than three-quarters of global soy is fed to animals

Before we look at the evidence for whether Brazilian soy is responsible for the clearing of the Amazon, we should first understand what products have been driving this growth.

When someone mentions soy we often think about foods such as tofu, soy milk, tempeh or edamame beans. This feeds into the argument that meat and dairy substitutes – such as switching from meat to high-protein tofu, or from dairy to soy milk – is in fact worse for the environment. But, only a small percentage of global soy is used for these products. More than three-quarters (77%) of soy is used as feed for livestock.

In the chart here we see the breakdown of what the world’s soy was used for in 2018. On the left we have total global soy production; in the middle, the three categories of uses (direct human food, animal feed, and industrial processes); and on the right we have the end use products. This data is sourced from an analysis published by the University of Oxford’s Food Climate Research Network (FCRN), which relies on the USDA’s PSD database. 13 Over one-third (37%) of global soy is fed to chickens and other poultry; one-fifth to pigs; and 6% for aquaculture. Very little soy is used for beef and dairy production – only 2%.

One-fifth of the world’s soy is used for direct (i.e. not from meat and dairy) human consumption. Most of this is first processed into soybean oil. Typical soy products such as tofu, soy milk, tempeh and edamame beans account for just 7% of global demand.

Soy can also be used for industrial purposes. Around 4% is used for biofuels, lubricants and other industrial processes. Biodiesel alone accounts for 2.8%.

We might therefore conclude that the increased demand for soy has been driven by a growing appetite for meat, dairy and soybean oil. But to double-check we should look beyond this static single-year view and see how demand has changed over time. Maybe demand for these products has always been high, and instead the growth in demand has come from the increased popularity of products such as soy milk and tofu.

In the chart we see the allocation of soybeans to three categories. Processed products include all animal feed from soybean cake (i.e. 70% of global demand); soybean oil; and industrial products such as biofuels. Direct human food includes all non-animal-sourced foods from soy excluding oils . Direct animal feed is soybeans fed directly to livestock (rather than been processed into soybean cake first).

We see that most of this growth has come from the increased demand for processed soy – animal feed, biofuels and vegetable oil. This rise has been particularly steep since 1990. By 2013, it had increased from 88 million to 227 million tonnes. Over this period, demand for human food products such as tofu and soy milk increased by only 3 million tonnes (from 7.4 to 10.7 million).

This should not surprise us. Global meat production has more than tripled over the last 50 years. This increase has been most marked for poultry – the largest consumer of soy feed.

Using the ‘change country’ toggle you can see how this allocation varies by country. For example, we see that this distribution is similar for Brazil.

Is soy production driving deforestation?

So far we’ve established that demand for soy has increased rapidly over the past 50 years; most soy is produced in the US, Brazil (and to a lesser extent, Argentina); and most of this has been driven by increased demand for animal feed, biofuels and vegetable oils.

The big question is whether this has been a key driver of deforestation.

The first step to answering this is to understand whether the amount of land we use to grow soy has increased. If improvements in crop yields were able to keep up with this rise in production then we wouldn’t need any additional land, and there would be no need to cut down forests.

In the chart we see the change in soy production, yields, and area harvested. This tells us the percentage change in each of these variables relative to the first year shown. Here the starting year is 1961 but you can change this by adjusting the time slider at the bottom of the chart.

Crop yields have not been able to keep up with production. Since 1961 global yields increased by 150%. But production increased by 1200%. This means the area used to grow soy has more than quadrupled.

This is also true for Brazil [you can see this using the “Change country” toggle]. Since Brazilian soy production was very low in 1961, it makes sense to adjust the time slider to see the change since 1980. Soy yields have doubled since 1980. This is impressive but not enough to keep up with demand: soy production increased by 680%. Instead, Brazil has had to devote more and more land to soy production: land use has tripled since 1980.

Brazil has devoted increasing amounts of land to soy production. But has this come at the expense of forests? Has the Amazon been cleared to make room for it?

This is a question many researchers have tried to answer. What makes it complex is that agricultural and land use systems are intricately linked: it’s possible to assess whether soy production has a direct impact on deforestation, but much harder to understand whether it is indirectly causing harm elsewhere.

In a study published in Science, Alexandra Tyukavina et al. (2017) look at the drivers of forest loss in the Brazilian Legal Amazon. 14 The change in these drivers from 2000 to 2013 is shown in the chart. Note that this overall trend suggests a major decline in forest loss since 2000 – unfortunately in the years since 2013, rates have increased again.

From this, we would conclude that the dominant driver of deforestation in the Brazilian Amazon was the expansion of pasture for beef production. If we look at forest loss from commercial crops – which is mainly soybeans – we see a significant decline, especially following the introduction of ‘Brazil’s Soy Moratorium’. The ‘Soy Moratorium’ was a campaign involving civil agencies and soybean companies, which stipulated that farmers who grew soy on illegal or legal deforestation areas would not be able to sell them to suppliers. Since 2009, satellite imagery has been used to help to identify soybean crops being grown on deforested areas. 15

Numerous studies have reached a very similar conclusion: it is pasture, not soybean production that is driving most deforestation in the Brazilian Amazon. 16 But, this only looks at the direct drivers of deforestation. In other words, the cutting down of forest today to make space for cropland for soy production.

The indirect impacts of soy production

Soy may no longer be a direct driver of deforestation in the Brazilian Amazon. But we should also think about the indirect impacts of increased production. We know that the area being used to grow soy is still increasing, and that land has to come from somewhere. In particular regions, such as Mato Grosso, studies have found that instead of replacing forest, these croplands are replacing pasture. 17 If this pasture land is simply shifting into forested areas, we could argue that soy is still a major underlying cause of deforestation.

A recent study by researchers Nikolas Kuschnig, Jesús Crespo Cuaresma, and Tamás Krisztin reached this conclusion. 18 They combined high-resolution satellite imagery from the Mato Grosso region in Brazil with socioeconomic panel data to not only quantify the direct impacts of soy production, but also the spillover effects. The results suggest that by only looking at direct impacts we are underestimating the role that soy continues to play in deforestation.

It’s also the case that the Amazonas region in Brazil gets most of the attention. But most soy is now grown in other regions of the country. In 2015, only 13% came from the Amazon, while 48% came from the Cerrado region. 19 Some researchers therefore make a strong case that interventions such as the ‘Soy Moratorium’ need to be extended to cover regions beyond the Amazon if they are to be effective. Without wider implementation of these policies, we will continue to see deforestation simply shift elsewhere.

Whilst the expansion of pasture for beef production is the leading driver of deforestation in Brazil, soy still plays a significant role when we take its indirect impacts into account. To end deforestation, there are a couple of key actions we can take. For consumers, since most deforestation is driven by expanding pastures for beef, or soy to feed poultry and pigs, reducing meat consumption is an effective way to make a difference. For companies and regulators, zero-deforestation policies must be more widely implemented (i.e. not only focused on the Amazon) and must be more carefully designed to take spillover effects into account.

Why boycotting palm oil could do more harm than good

In a large-scale consumer survey across the UK population on perceptions of vegetable oils, palm oil was deemed to be the least environmentally-friendly. 20 It wasn’t even close. 41% of people thought palm oil was ‘environmentally unfriendly’, compared to 15% for soybean oil; 9% for rapeseed; 5% for sunflower; and 2% for olive oil. 43% also answered ‘Don’t know’, meaning that almost no one thought it was good.

Retailers know that this is becoming an important driver of consumer choices. From shampoos, to detergents, from chocolate to cookies, companies are trying to eliminate palm oil from their products. There are now long lists of companies that have done so [Google ‘palm-oil free’ and you will find an endless supply] . Many online grocery stores now offer the option to apply a ‘palm-oil free’ filter when browsing their products. 21

Why are consumers turning their back on palm oil? And is this reputation justified?

In this article I address some key questions about palm oil production: how has it changed; where is it grown; and how this has affected deforestation and biodiversity. The story of palm oil is not as simple as it is often portrayed. Global demand for vegetable oils has increased rapidly over the last 50 years. Being the most productive oilcrop, palm has taken up a lot of this production. This has had a negative impact on the environment, particularly in Indonesia and Malaysia. But it’s not clear that the alternatives would have fared any better. In fact, because we can produce up to 20 times as much oil per hectare from palm versus the alternatives, it has probably spared a lot of environmental impacts from elsewhere.

Palm oil production has grown to meet rising demands for vegetable oils

Palm oil production has increased rapidly over the past 50 years. In 1970, the world was producing only 2 million tonnes. This is now 35 times higher: in 2018 the world produced 71 million tonnes. The change in global production is shown in the chart. 22

The rise of palm oil follows the rapid increase in demand for vegetable oils more broadly. The breakdown of global vegetable oil production by crop is shown in the stacked area chart. Global production increased ten-fold since the 1960s – from 17 to 170 million tonnes in 2014. As we will see later in this article, more recent data for 2018 comes to 218 million tonnes.

The story of palm oil is less about it as an isolated commodity, but more about the story of the rising demand for vegetable oils. Palm oil is a very productive crop; as we will see later, it produces 36% of the world’s oil, but uses less than 9% of croplands devoted to oil production. It has therefore been a natural choice to meet this demand.

Who uses palm oil and what is it used for?

Why has the market for palm oil – and vegetable oils more broadly – increased so rapidly? What is it used for?

Palm oil is a versatile product which is used in a range of products across the world:

Foods : over two-thirds (68%) is used in foods ranging from margarine to chocolate, pizzas, breads and cooking oils;
Industrial applications : 27% is used in industrial applications and consumer products such as soaps, detergents, cosmetics and cleaning agents;
Bioenergy : 5% is used as biofuels for transport, electricity or heat.

While food products dominate globally, this breakdown varies from country-to-country. Some countries use much more palm oil for biofuels than others. In Germany, for example, bioenergy is the largest use , accounting for 41% (more than food at 40%). A push towards increased biofuel consumption in the transport sector has been driving this, despite it being worse for the environment than normal diesel (more on this later).

In the next section we will look at what countries produce palm oil, but here we see a map of palm oil imports. Although production is focused in only a few countries across the tropical belt, we see that palm oil is an important product across the world.

Where is palm oil grown?

Oil palm is a tropical plant species. It thrives on high rainfall, adequate sunlight and humid conditions – this means the best growing areas are along a narrow band around the equator. 23 Palm oil is therefore grown in many countries across Africa, South America, and Southeast Asia. In the map we see the distribution of production across the world.

Small amounts of palm oil are grown in many countries, but the global market is dominated by only two: Indonesia and Malaysia. In 2018, the world produced 72 million tonnes of oil palm. Indonesia accounted for 57% of this (41 million tonnes), and Malaysia produced 27% (20 million tonnes).84% of global palm oil production comes from Indonesia and Malaysia.

In the chart we see the production of the palm oil plant across a number of countries. Other producers include Thailand, Colombia, Nigeria, Guatemala, and Ecuador. As we’d expect, all of these countries lie along the zone of ‘optimal conditions’ around the equator.

How has land use for palm oil changed over time?

How has the world achieved such a rapid expansion of palm oil production? There are only two ways in which we can produce more of a given crop: increase yields (growing more on a given amount of land) or expand the amount of land we use to grow it.

Global palm yields have increased over time, but far short of the increase in demand. This means that over the last 50 years the amount of land devoted to growing palm oil has increased a lot. In the chart here we see the change in land use. Since 1980 the amount of land the world uses to grow palm has more than quadrupled, from 4 million to 19 million hectares in 2018. Indonesia and Malaysia account for 63% of global land use for palm. This is low when we consider that it accounts for 84% of production . This is because both countries achieve high yields .

19 million hectares might sound like a lot of land. But we should consider this in the context of all land used to grow oilcrops. The world devotes more than 300 million hectares for oilcrop production. Palm oil accounts for 6% of this land use, which is small when we consider that it produces 36% of the oil.

Is palm oil responsible for deforestation?

Land use to grow palm has more than quadrupled since 1980. But has this expansion came at the expense of tropical forest?

This seems like a simple question, but is not as straightforward as we might expect. The IUCN (International Union for Conservation of Nature) Oil Palm Task Force conducted an in-depth review of the literature to understand the impact of palm oil on deforestation. 24 There was a lot of variability in the results, depending on how a forest was defined, the geographic focus of the analysis, and timeframe that was considered.

Having done my own review of the literature I conclude that palm oil has been a significant driver of tropical deforestation, especially in Southeast Asia.

Some studies suggest that palm oil has played a very small role in global forest loss. One study suggests that palm oil is responsible for around 2% of global tree loss. 25 Another suggests it is responsible for only 0.2% of intact forest loss. 26 This seems very low. But there are a couple of caveats to these figures. Firstly, they measure forest loss , which combines both permanent deforestation (where trees do not regrow) and forest degradation (which is a temporary thinning of forests with subsequent regrowth). As we discuss in our related article on this, deforestation accounts for just one-quarter of global tree loss. Palm oil’s contribution to deforestation – which has greater environmental impacts than degradation – would be higher.

Secondly, the 0.2% figure is based on intact forest loss. Intact forests are a specific subset of primary forests which are very rich sites of biodiversity, and are largely undisturbed by human activity. Only 6% of global intact forest is in Southeast Asia, the hotspot of palm oil expansion. Only 2.8% was in Indonesia, and 0.2% in Malaysia. So, even if all of these countries’ forests were wiped out by palm oil plantations, it would still make very little difference to this metric of global intact forest loss. The authors of the original study make it clear that intact forest loss should not be confused with primary forest loss.

So, how much of palm oil’s expansion has really come at the expense of forests? Let’s focus on the two key countries driving production: Indonesia and Malaysia. In the chart here we see the drivers of deforestation in Indonesia from 2001 to 2016. 27 Oil palm plantations were the largest driver of deforestation over this period, accounting for 23%. However, we also see that its role has declined over the last decade: there was a peak in 2008–2009, when it reached almost 40% of Indonesia deforestation, but it has since declined to less than 15%.

Part of what makes this question so challenging to answer is that it depends on whether you only consider palm plantations which immediately and directly replaced existing forest, or whether you include plantations which very quickly replaced forests which had been logged for wood, paper and pulp. In a paper published in Nature , David Gaveau and colleagues (2016) used satellite imagery to assess what types of land industrial plantations replaced in the Borneo region. 28 Industrial plantations include palm oil and pulpwood tree plantations, but are dominated by the former.

This is shown for the Indonesian and Malaysian Borneo in the chart.

75% of these plantations were grown on land that was previously forested in 1973. Not all of these plantations were the direct driver of this replacement. In fact, only one-quarter of post-1973 plantations in the Borneo was driven by rapid conversion of this land to plantations (20–21% oil-palm; 4.3–4.8% pulpwood). This is shown in green in the chart. Direct deforestation for palm oil played a larger role in Malaysia; 60% was driven by plantations, whilst in Indonesia it was only 16%.

This study only looks at the Borneo region. But this is reasonably consistent with studies that have looked at the expansion of palm plantations more broadly. A global study of palm-driven deforestation found that in Southeast Asia, 45% of oil palm plantations came from areas that were forests in 1989. 29 In Indonesia this was 54% and in Malaysia, 40%.

These distinctions on how quickly palm oil plantations replaced forests make it difficult to give a clear, single number on how much deforestation it has caused. But, most of the research concludes that, particularly in tropical forests in Southeast Asia, palm expansion has played a significant role.

Palm oil versus the alternatives

Palm oil has been an important driver of deforestation. But would the alternatives have fared any better?

There are a couple of reasons why palm oil has been the favored crop to meet growing demand for vegetable oils. Firstly, it has lowest production costs. 30 Secondly, its composition means it’s versatile and can be used for food and non-food purposes alike: some oils are not suited for cosmetic uses such as shampoos and detergents. Third, it gets incredibly high yields.

If we weren’t meeting global oil demand through palm oil, another oilcrop would have to take its place. Would the alternatives be any better for the environment?

We can compare crops in terms of their yields – how much oil we can produce from one hectare of land. This comparison is shown in the chart. 31 Palm oil stands out immediately. It achieves a much higher yield than the alternatives. From each hectare of land, you can produce about 2.9 tonnes of palm oil. That’s around four times higher than alternatives such as sunflower or rapeseed oil (where you get about 0.7 tonnes per hectare); and 10 to 15 times higher than popular alternatives such as coconut or groundnut oil . 32

Let’s take a look at how this comparison affects the global landscape of oilcrops in terms of production and land use. In the chart we see the breakdown of global vegetable oil production in 2018. On the left we have each crop’s share of global land use for vegetable oils; on the right we have its share of production.

We know from our yield comparison that palm oil achieves a much higher yield. What this means is that it accounts for a very high share of oil production without taking up much land. In 2017 it produced 36% of our vegetable oil, but took up only 8.6% of the land.

Sunflower oil was almost exactly proportional in terms of how much oil it produced relative to how much land it took up: it produced 9% of oil, and required 8.3% of land. Rape and mustardseed oil were also in proportion. The rest – soybean, olive, coconut, groundnut, and sesameseed – used more land than they gave back in oil production. Coconut oil, for example, provided only 1.4% of global oil but required 3.6% of the land.

It’s of course true that some crops provide co-products in the process. The non-oil fraction of soybeans, for example, can be allocated to other uses such as high-protein animal feed. Therefore using this land to grow the crop is meeting other food demand at the same time. But this doesn’t change the fact that if the world requires a given amount of vegetable oil, it is the oil yield per hectare of each crop that we care about – regardless of whether it provides co-products in the process.

What would be the impact of substituting palm oil?

Palm oil achieves very high yields relative to other oilcrops. Why does this matter?

If we want to limit our environmental impact, reducing the amount of land we devote to agriculture is key. To make space for more croplands and pastures, we have been displacing forests, grasslands and peatlands – areas of rich biodiversity. The less land we need for farming, the better.

The high yields from palm oil means that it, in some sense, ‘spares’ the world additional farmland we would need if we want to meet global oil demand from the alternatives. We can look at this in terms of the amount of land we would need if total global demand for vegetable oil was met from a single crop alone. In other words, how much land would be needed if the 218 million tonnes of oil was only produced from palm oil, or only produced from rapeseed? This comparison is shown in the chart. As we’d expect, we see vast differences.

Let’s give these numbers some context. Currently the world devotes around 322 million hectares to oilseed crops. That’s an area similar to the size of India. If global oil was supplied solely from palm, we’d need 77 million hectares, around four times less. If we got it from rapeseed we’d need an area similar to the size we use today; from coconuts, an area the size of Canada; and in the most extreme case, we’d have to devote 2 billion hectares to sesame seeds – a bit more than Canada, the USA, and India combined.

In this sense, palm oil has been a ‘land sparing’ crop. Switching to alternatives would mean the world would need to use more farmland, and face the environmental costs that come with it. A global boycott on palm oil would not fix the problem: it would simply shift it elsewhere, and at a greater scale because the world would need more land to meet demand.

This is true for other tropical oilcrops such as coconut, groundnut, or soy. However, we might argue that this oversimplifies the comparison to more temperate crops: it assumes the environmental impact of devoting one hectare of land for sunflower seeds in Europe is the same as cutting down tropical rainforests to grow palm or coconut plantations. We know that tropical forests are incredibly rich in biodiversity , and store a lot of carbon from the atmosphere. In some cases, especially for European domestic markets, some substitution for rapeseed or sunflower seed oils could have a positive environmental impact even if it required using a bit more land.

How can we use palm oil without destroying tropical forests?

There are a number of steps we can take to ensure we meet global demand for oils, without destroying our tropical forests.

Substitutes for palm oil do not always exist . As we’ve discussed, substituting palm oil with alternatives can do more harm than good. But it’s also true that alternative oils are not always suitable for the products we need. Palm oil is unique in its versatility, meaning it is suitable for a range of foods, cosmetics, industrial applications and biofuels. Substitution would be feasible for most food products. Substitution in industrial processes would be more difficult, especially if we want to replace it with oils grown in temperate countries: sunflower or rapeseed oil is not suited to products such as soaps, detergents or cosmetics. One sector where alternatives do exist is bioenergy, which brings us to our next point.

European countries should stop using palm oil for biofuels . The EU – after China and India – is the third largest importer of palm oil . There are some products where using palm is our best option. This is not the case for biofuels, yet two-thirds of the EU’s imported palm oil goes to bioenergy production. Using palm oil as a biofuel is worse for the environment than petrol. A meta-analysis conducted by the Royal Academy of Engineering on EU biofuels found that when we factored in land-use change, the greenhouse gas emissions from palm oil were higher than using a petrol car. 33 Other studies have shown that when we take the additional environmental impacts into account, these biofuels are much worse than conventional fuels. 34 By using palm oil, EU countries are not only increasing emissions, they’re passing the responsibility and accountability of these emissions on to other countries. A ban on using palm oil for biofuels would reduce this impact while allowing global palm oil to be used for purposes where there are few better alternatives – food and cosmetic products.

Increase companies sourcing from suppliers with sustainability certification . There are now a number of certification schemes which help to verify whether palm oil is being produced in a sustainable way. The most well-known is the Roundtable on Sustainable Palm Oil (RSPO). The RSPO was launched in 2004, and provides certification for suppliers who produce their crop in a more sustainable way by conducting impact assessments, managing high-value areas of biodiversity, not clearing primary forest and avoiding land clearance through fires. 35 For example, suppliers can only be certified if their plantations since 2005 have not replaced primary forest or areas rich in biodiversity. Consumer demand for sustainable palm oil puts pressure on food and cosmetic companies to source from certified suppliers, and ultimately rewards the most sustainable growers.

Only 19% of palm oil production is covered by the RSPO. 36 Research into the impact of the RSPO found that it was successful in reducing deforestation. 37 However, it found that this most avoided deforestation from older plantations, which is not where most tropical forests remain. To have a real, permanent impact, certification needs to cover a much larger number of growers.

Increased crop yields. If we want to reduce agricultural expansion, we want to maximise crop yields through effective management practices, improved varieties and choosing the most productive areas of land.

This combination of interventions involves actors across the full production line, from agricultural scientists improving crop varieties, to palm oil growers, governments, food and cosmetic producers, retailers and consumers. Pressure from consumers can filter through to growers. To do this effectively, understanding for consumers has to be clearer. Many believe that boycotting palm oil is how we make a difference. But as we’ve seen, the alternatives are not necessarily better; more sustainable palm oil (used for food, not fuel) rather than no palm oil is what we should be pushing for.

There are a number of estimates on what share of deforestation is driven by agriculture. The differences in these estimates are often explained by the different treatment of deforestation versus forest degradation – we cover the differences between these types of forest loss in a related post . Agriculture accounts for 70-80% of tropical deforestation – the permanent conversion of forested land to another land use. It accounts for a smaller percentage when degradation – the temporary loss of forest prior to regrowth – is included.

One of the most-widely cited studies on this comes from Noriko Hosonuma et al. (2012) who estimate that 73% of tropical deforestation is driven by agriculture. This is similar to estimates by Geist and Lambin (2002) who estimated that around 80% of deforestation in the 1980s and 1990s was driven by agriculture. In fact, over the longer analysis of deforestation from 1840 to 1990, they estimate 96% of deforestation was driven by agriculture. Gibbs et al. (2010) estimate that during the 1980s and 1990s, 83% of agricultural land expansion replaced forest.

Geist, H. J., & Lambin, E. F. (2002). Proximate Causes and Underlying Driving Forces of Tropical Deforestation . BioScience , 52 (2), 143-150.

Gibbs, H. K., Ruesch, A. S., Achard, F., Clayton, M. K., Holmgren, P., Ramankutty, N., & Foley, J. A. (2010). Tropical forests were the primary sources of new agricultural land in the 1980s and 1990s . Proceedings of the National Academy of Sciences , 107 (38), 16732-16737.

Hosonuma, N., Herold, M., De Sy, V., De Fries, R. S., Brockhaus, M., Verchot, L., ... & Romijn, E. (2012). An assessment of deforestation and forest degradation drivers in developing countries . Environmental Research Letters , 7 (4), 044009.

Pendrill, F., Persson, U. M., Godar, J., Kastner, T., Moran, D., Schmidt, S., & Wood, R. (2019). Agricultural and forestry trade drives large share of tropical deforestation emissions . Global Environmental Change , 56 , 1-10.

The Global Forest Watch programme categorizes forest loss drivers based on permanent deforestation – the conversion of forest to another land use – and degradation (which includes logging of tree plantations and wildfires). ‘Commodity-driven deforestation’ – which includes some activities such as mining but is predominantly agricultural commodities – totalled 5.4 million hectares in 2019.

A paper by Philip Curtis et al. (2018) discusses this classification in detail. We also look at these categories in more detail in a related article .

Curtis, P. G., Slay, C. M., Harris, N. L., Tyukavina, A., & Hansen, M. C. (2018). Classifying drivers of global forest loss . Science , 361 (6407), 1108-1111.

One-third (1.8 million hectares) came from Brazil, and 19% (1 million hectares) from Indonesia.

Since Brazil accounts for 33% of tropical deforestation, and Brazilian cattle account for 24%, cattle accounts for 72% of Brazil’s total [24 / 33 * 100 = 72%].

Tyukavina, A., Hansen, M. C., Potapov, P. V., Stehman, S. V., Smith-Rodriguez, K., Okpa, C., & Aguilar, R. (2017). Types and rates of forest disturbance in Brazilian Legal Amazon, 2000–2013 . Science Advances , 3 (4), e1601047.

Gibbs, H. K., Rausch, L., Munger, J., Schelly, I., Morton, D. C., Noojipady, P., ... & Walker, N. F. (2015). Brazil's soy moratorium . Science , 347 (6220), 377-378.

Boucher, D., Roquemore, S., & Fitzhugh, E. (2013). Brazil's success in reducing deforestation . Tropical Conservation Science , 6 (3), 426-445.

Kuschnig, N., Crespo Cuaresma, J., & Krisztin, T. (2019). Unveiling Drivers of Deforestation: Evidence from the Brazilian Amazon .

Clark, M. A., Williams, D. R., Buchanan, G. M., Ficetola, G. F., Rondinini, C., & Tilman, D. (2020). Proactive conservation to prevent habitat losses to agricultural expansion . Nature Sustainability .

This data is sourced from the UN Food and Agriculture Organization (FAO).

In 2018 global soy production was 349 million tonnes. The US produced 123 million tonnes [123M / 349M * 100 = 35%] and Brazil produced 118 million tonnes [118M / 349M * 100 = 34%]. Combined, they accounted for 69% of global production.

United States Department of Agriculture. PSD Online. Available at: https://apps.fas.usda.gov/psdonline/app/index.html#/app/advQuery.{/ref }

The majority (77%) of the world’s soy is fed to livestock for meat and dairy production. 7% is fed directly to animals as soybeans, but the remainder is first processed into soybean ‘cake’.{ref}Soybean cake (sometimes referred to as soybean meal) is a high-protein feed made from the pressurisation, heat-treatment and extraction processing of soybeans. The oil is extracted from the soybeans to leave a protein-rich product.

Rudorff, B. F. T., Adami, M., Aguiar, D. A., Moreira, M. A., Mello, M. P., Fabiani, L., ... & Pires, B. M. (2011). The soy moratorium in the Amazon biome monitored by remote sensing images . Remote Sensing , 3 (1), 185-202.

Brandão, ASP, de Rezende, GC, Costa Marques, RW, & de Aplicada, IPE (2005). Agricultural growth in the period 1999-2004, explosion of the area planted with soybeans and the environment in Brazil .

Müller, C. (2003). Expansion and modernization of agriculture in the Cerrado–the case of soybeans in Brazil’s center-West. Brasília: Departamento de Economia, Universidade de Brasília .

Barona, E., Ramankutty, N., Hyman, G., & Coomes, O. T. (2010). The role of pasture and soybean in deforestation of the Brazilian Amazon . Environmental Research Letters , 5 (2), 024002.

Kuschnig, Nikolas & Crespo Cuaresma, Jesus & Krisztin, Tamás, 2019. " Unveiling Drivers of Deforestation: Evidence from the Brazilian Amazon ," Ecological Economic Papers 32, WU Vienna University of Economics and Business.

Soterroni, A. C., Ramos, F. M., Mosnier, A., Fargione, J., Andrade, P. R., Baumgarten, L., ... & Carvalho, A. X. (2019). Expanding the soy moratorium to Brazil’s Cerrado . Science Advances , 5 (7), eaav7336.

Ostfeld, R., Howarth, D., Reiner, D., & Krasny, P. (2019). Peeling back the label—exploring sustainable palm oil ecolabelling and consumption in the United Kingdom . Environmental Research Letters , 14 (1), 014001.

Ocado, an online UK retailer is just one example .

Wahid, M. B., Abdullah, S. N. A., & IE, H. (2005). Oil palm . Plant Production Science , 8 (3), 288-297.

Meijaard, E., Garcia-Ulloa, J., Sheil, D., Wich, S.A., Carlson, K.M., Juffe-Bignoli, D., and Brooks, T.M. (eds.) (2018). Oil palm and biodiversity. A situation analysis by the IUCN Oil Palm Task Force . IUCN Oil Palm Task Force Gland, Switzerland: IUCN. xiii + 116pp.

Pendrill, F., Persson, U. M., Godar, J., & Kastner, T. (2019). Deforestation displaced: trade in forest-risk commodities and the prospects for a global forest transition . Environmental Research Letters , 14 (5), 055003.

Potapov, P., Hansen, M. C., Laestadius, L., Turubanova, S., Yaroshenko, A., Thies, C., ... & Esipova, E. (2017). The last frontiers of wilderness: Tracking loss of intact forest landscapes from 2000 to 2013 . Science advances , 3 (1), e1600821.

Austin, K. G., Schwantes, A., Gu, Y., & Kasibhatla, P. S. (2019). What causes deforestation in Indonesia? . Environmental Research Letters , 14(2), 024007.

Seymour, F., & Harris, N. L. (2019). Reducing tropical deforestation . Science , 365 (6455), 756-757.

Gaveau, D. L., Sheil, D., Salim, M. A., Arjasakusuma, S., Ancrenaz, M., Pacheco, P., & Meijaard, E. (2016). Rapid conversions and avoided deforestation: examining four decades of industrial plantation expansion in Borneo . Scientific reports , 6 (1), 1-13.

Vijay, V., Pimm, S. L., Jenkins, C. N., & Smith, S. J. (2016). The impacts of oil palm on recent deforestation and biodiversity loss . PloS one , 11 (7), e0159668.

Carter, C., Finley, W., Fry, J., Jackson, D., & Willis, L. (2007). Palm oil markets and future supply . European Journal of Lipid Science and Technology , 109 (4), 307-314.

Here I have calculated the actual yield of oil from each crop, rather than the yield of the total crop. No crop is 100% oil-based, and therefore this will differ from the mass of the crop itself. Ultimately this is what matters if our question is how to meet global oil demand.

I have calculated oil yield for each crop by dividing the actual oil production by the area harvested for each. Both of these metrics are sourced from the UN Food and Agriculture Organization (FAO).

To demonstrate this, let’s turn crop yields on its head and look at the inverse: the area of land you would need to produce one tonne of vegetable oil. We can simply calculate this as 1 / Yield (in tonnes per hectare).

This comparison by crop is shown here . To produce one tonne of oil we need only 0.3 hectares of land from palm oil; 1.4 hectares from sunflower or rapeseed; 3.7 hectares from coconut; or 7 hectares from groundnut. If you want to produce one tonne of oil, you would need 4 times the amount of cropland devoted to sunflower or rapeseed, or 10 to 15 times the amount of land devoted to coconut production.

Royal Academy of Engineering (2017). Sustainability of liquid biofuels .

R. Zah et al., Ökobilanz Von Energieprodukten: Ökologische Bewertung Von Biotreibstoffen (EMPA, Abteilung Technologie und Gesellschaft, St. Gallen, Switzerland, 2007).

This is based on reported figures as of November 2020.

Carlson, K. M., Heilmayr, R., Gibbs, H. K., Noojipady, P., Burns, D. N., Morton, D. C., ... & Kremen, C. (2018). Effect of oil palm sustainability certification on deforestation and fire in Indonesia . Proceedings of the National Academy of Sciences , 115 (1), 121-126.

Cite this work

Our articles and data visualizations rely on work from many different people and organizations. When citing this article, please also cite the underlying data sources. This article can be cited as:

BibTeX citation

Reuse this work freely

All visualizations, data, and code produced by Our World in Data are completely open access under the Creative Commons BY license . You have the permission to use, distribute, and reproduce these in any medium, provided the source and authors are credited.

The data produced by third parties and made available by Our World in Data is subject to the license terms from the original third-party authors. We will always indicate the original source of the data in our documentation, so you should always check the license of any such third-party data before use and redistribution.

All of our charts can be embedded in any site.

Our World in Data is free and accessible for everyone.

Help us do this work by making a donation.

First-of-its-kind study definitively shows that conservation actions are effective at halting and reversing biodiversity loss

Download images

A new study published online today, April 25, in the scientific journal Science provides the strongest evidence to date that not only is nature conservation successful, but that scaling conservation interventions up would be transformational for halting and reversing biodiversity loss—a crisis that can lead to ecosystem collapses and a planet less able to support life—and reducing the effects of climate change.

The findings of this first-ever comprehensive meta-analysis of the impact of conservation action are crucial as more than 44,000 species are documented as being at risk of extinction , with tremendous consequences for the ecosystems that stabilize the climate and that provide billions of people around the world with clean water, livelihoods, homes, and cultural preservation, among other ecosystem services. Governments recently adopted new global targets to halt and reverse biodiversity loss, making it even more critical to understand whether conservation interventions are working.

“If you look only at the trend of species declines, it would be easy to think that we’re failing to protect biodiversity, but you would not be looking at the full picture,” said Penny Langhammer, lead author of the study and executive vice president of Re:wild. “What we show with this paper is that conservation is, in fact, working to halt and reverse biodiversity loss. It is clear that conservation must be prioritized and receive significant additional resources and political support globally, while we simultaneously address the systemic drivers of biodiversity loss, such as unsustainable consumption and production.”

Although many studies look at individual conservation projects and interventions and their impact compared with no action taken, these papers have never been pulled into a single analysis to see how and whether conservation action is working overall. The co-authors conducted the first-ever meta-analysis of 186 studies, including 665 trials, that looked at the impact of a wide range of conservation interventions globally, and over time, compared to what would have happened without those interventions. The studies covered over a century of conservation action and evaluated actions targeting different levels of biodiversity—species, ecosystems, and genetic diversity.

The meta-analysis found that conservation actions—including the establishment and management of protected areas, the eradication and control of invasive species, the sustainable management of ecosystems, habitat loss reduction, and restoration—improved the state of biodiversity or slowed its decline in the majority of cases (66%) compared with no action taken at all. And when conservation interventions work, the paper’s co-authors found that they are highly effective .

For example:

Management of invasive and problematic native predators on two of Florida’s barrier islands, Cayo Costa and North Captiva, resulted in an immediate and substantial improvement in nesting success by loggerhead turtles and least terns, especially compared with other barrier islands where no predator management was applied.
In the Congo Basin, deforestation was 74% lower in logging concessions under a Forest Management Plan (FMP) compared with concessions without an FMP.
Protected areas and Indigenous lands were shown to significantly reduce both deforestation rate and fire density in the Brazilian Amazon. Deforestation was 1.7 to 20 times higher and human-caused fires occurred four to nine times more frequently outside the reserve perimeters compared with inside.
Captive breeding and release boosted the natural population of Chinook salmon in the Salmon River basin of central Idaho with minimal negative impacts on the wild population. On average, fish taken into the hatchery produced 4.7 times more adult offspring and 1.3 times more adult second generation offspring than naturally reproducing fish.

“Our study shows that when conservation actions work, they really work. In other words, they often lead to outcomes for biodiversity that are not just a little bit better than doing nothing at all, but many times greater,” said Jake Bicknell, co-author of the paper and a conservation scientist at DICE, University of Kent. “For instance, putting measures in place to boost the population size of an endangered species has often seen their numbers increase substantially. This effect has been mirrored across a large proportion of the case studies we looked at.”

Even in the minority of cases where conservation actions did not succeed in recovering or slowing the decline of the species or ecosystems that they were targeting compared with taking no action, conservationists benefited from the knowledge gained and were able to refine their methods. For example, in India the physical removal of invasive algae caused the spread of the algae elsewhere because the process broke the algae into many pieces, enabling their dispersal. Conservationists could now implement a different strategy to remove the algae that is more likely to be successful.

This might also explain why the co-authors found a correlation between more recent conservation interventions and positive outcomes for biodiversity— conservation is likely getting more effective over time . Other potential reasons for this correlation include an increase in funding and more targeted interventions.

In some other cases where the conservation action did not succeed in benefiting the target biodiversity compared with no action at all, other native species benefitted unintentionally instead. For example, seahorse abundance was lower in protected sites because marine protected areas increase the abundance of seahorse predators, including octopus.

“It would be too easy to lose any sense of optimism in the face of ongoing biodiversity declines,” said study co-author and Associate Professor Joseph Bull , from the University of Oxford’s Department of Biology. “However, our results clearly show that there is room for hope. Conservation interventions seemed to be an improvement on inaction most of the time; and when they were not, the losses were comparatively limited."

More than half of the world’s GDP, almost $44 trillion , is moderately or highly dependent on nature. According to previous studies, a comprehensive global conservation program would require an investment of between US$178 billion and US$524 billion , focused primarily in countries with particularly high levels of biodiversity. To put this in perspective, in 2022, global fossil fuel handouts--which are destructive to nature—were US$7 trillion . This is 13 times the highest amount needed annually to protect and restore the planet. Today more than US$121 billion is invested annually into conservation worldwide , and previous studies have found the cost-benefit ratio of an effective global program for the conservation of the wild is at least 1:100 .

“Conservation action works—this is what the science clearly shows us,” said Claude Gascon, co-author and director of strategy and operations at the Global Environment Facility. “It is also evident that to ensure that positive effects last, we need to invest more in nature and continue doing so in a sustained way. This study comes at a critical time where the world has agreed on ambitious and needed global biodiversity targets that will require conservation action at an entirely new scale. Achieving this is not only possible, it is well within our grasp as long as it is appropriately prioritized.”

The paper also argues that there must be more investment specifically in the effective management of protected areas, which remain the cornerstone for many conservation actions. Consistent with other studies, this study finds that protected areas work very well on the whole . And what other studies have shown is that when protected areas are not working, it is typically the result of a lack of effective management and adequate resourcing. Protected areas will be even more effective at reducing biodiversity loss if they are well-resourced and well-managed.

Moving forward, the study’s co-authors call for more and rigorous studies that look at the impact of conservation action versus inaction for a wider range of conservation interventions, such as those that look at the effectiveness of pollution control, climate change adaptation, and the sustainable use of species, and in more countries.

“For more than 75 years, IUCN has advanced the importance of sharing conservation practice globally,” said Grethel Aguilar, IUCN director general. “This paper has analyzed conservation outcomes at a level as rigorous as in applied disciplines like medicine and engineering—showing genuine impact and thus guiding the transformative change needed to safeguard nature at scale around the world. It shows that nature conservation truly works, from the species to the ecosystem levels across all continents. This analysis, led by Re:wild in collaboration with many IUCN Members, Commission experts, and staff, stands to usher in a new era in conservation practice.”

This work was conceived and funded through the International Union for Conservation of Nature (IUCN) by the Global Environment Facility.

Lindsay Renick Mayer

[email protected]

+1 512-686-6225

Devin Murphy

+1 512-686-6188

The paper ‘The positive impact of conservation action’ has been published in Science: https://www.science.org/doi/full/10.1126/science.adj6598

Additional quotes

Thomas Brooks, co-author and chief scientist, IUCN

“This paper is not only extremely important in providing robust evidence of the impact of

conservation actions. It is also extremely timely in informing crucial international policy processes, including the establishment of a 20-year vision for IUCN, the development of an IPBES assessment of biodiversity monitoring, and the delivery of the action targets toward the outcome goals of the new Kunming-Montreal Global Biodiversity Framework.”

Stuart Butchart, co-author and chief scientist, BirdLife International

“Recognising that the loss and degradation of nature is having consequences for societies worldwide, governments recently adopted a suite of goals and targets for biodiversity conservation. This new analysis is the best evidence to date that conservation interventions make a difference, slowing the loss of species’ populations and habitats and enabling them to recover. It provides strong support for scaling up investments in nature in order to meet the commitments that countries have signed up to.”

Jamie Carr, co-author and researcher in climate change and biodiversity governance, Leverhulme Centre for Anthropocene Biodiversity, University of York, UK “This work represents a huge effort on the part of many conservation professionals, all of whom are committed to reversing the loss of the world's biodiversity. It is encouraging to find that the past work of other conservationists has had a positive impact on nature, and I sincerely hope that our findings inspire those working now and in the future to ramp up their efforts."

Piero Genovesi, ISPRA, co-author and chair, IUCN SSC Invasive Species Specialist Group

“Species and ecosystems are facing a dramatic crisis, and the Biodiversity Plan of the United Nations is an urgent global call to action. This paper shows that eradication, control and management of invasive alien species have the largest impact in terms of conservation, and can help reverse the current trends of biodiversity loss, potentially saving hundreds of species from extinction. It is essential that governments and donors support the struggle against invasive alien species if we want to meet the agreed biodiversity targets by 2030.”

Mike Hoffmann, co-author and head of wildlife recovery, Zoological Society of London

“The major advance of this study is its sheer weight of evidence. We can point to specific examples, such as how captive breeding and reintroductions have facilitated the return of scimitar-horned oryx to the wild in Chad, but these can feel a bit exceptional. This study draws on more than 650 published cases to show that conservation wins are not rare. Conservation mostly works—unfortunately, it is also mostly significantly under-resourced.”

Madhu Rao, chair, IUCN World Commission on Protected Areas “With less than six years remaining to achieve ambitious biodiversity targets by 2030, there is a great sense of urgency for effective conservation action. We can take proven methods to conserve nature, such as protected areas, and scale them up for real conservation impact. This research clearly demonstrates that conservation actions are successful. We just need to take them to scale.”

Jon Paul Rodriguez, chair of the IUCN Species Survival Commission

“Anyone involved in the field of conservation will have witnessed the power of nature to regenerate and grow, given a chance to do so. From fishery exclusion zones, to ecological restoration on land, and animal, fungi and plant recovery efforts, there are numerous examples of halting and reversing biodiversity declines. Langhammer and colleagues synthesize knowledge on the impact of conservation action, and demonstrate that evidence-based conservation efforts indeed work in the majority of cases, not just in a few hand-picked examples. Much more money is spent on destroying nature than on protection and recovery. The authors show that tipping the balance in favor of nature is likely to help us deliver the world's ambitious biodiversity conservation targets.”

Gernot Segelbacher, co-author, professor and co-chair of Conservation Genetic Specialist Group, University Freiburg

“Conservation matters! While we so often hear about species declining or going extinct, this study shows that we can make a difference.”

Stephen Woodley, co-author, ecologist and vice chair for science and biodiversity, IUCN World Commission on Protected Areas

“The world needs hope that conservation action can work to halt and reverse biodiversity loss. This paper demonstrates that a range of conservation actions are highly effective. We just need to do more of them.”

Re:wild protects and restores the wild. We have a singular and powerful focus: the wild as the most effective solution to the interconnected climate, biodiversity and human wellbeing crises. Founded by a group of renowned conservation scientists together with Leonardo DiCaprio, Re:wild is a force multiplier that brings together Indigenous peoples, local communities, influential leaders, nongovernmental organizations, governments, companies and the public to protect and rewild at the scale and speed we need. Learn more at rewild.org .

University of Oxford

Oxford University has been placed number 1 in the Times Higher Education World University Rankings for the eighth year running, and number 3 in the QS World Rankings 2024. At the heart of this success are the twin-pillars of our ground-breaking research and innovation and our distinctive educational offer. Oxford is world-famous for research and teaching excellence and home to some of the most talented people from across the globe. Our work helps the lives of millions, solving real-world problems through a huge network of partnerships and collaborations. The breadth and interdisciplinary nature of our research alongside our personalised approach to teaching sparks imaginative and inventive insights and solutions.

Durrell Institute of Conservation and Ecology (DICE)

The Durrell Institute of Conservation and Ecology (DICE) is a research centre at the University of Kent. Its teaching and research is designed to break down the barriers between the natural and social sciences and produce real-world impact. Its mission is to conserve biodiversity and the ecological processes that support ecosystems and people, by developing capacity and improving conservation management and policy through high-impact research.

University of Kent

The University of Kent in England is renowned internationally for the quality of its teaching and research, with many of its academic schools and centres being among the best in their disciplines across the arts and humanities, sciences, and social sciences. Its campuses at Canterbury and Medway welcome more than 17,000 students from over 150 countries. The University of Kent is individually and collectively in the pursuit of progress, with a student-focused approach which is supportive, challenging and rewarding, and interdisciplinary research driven by collaboration to create positive impact. We are proud to be a values-driven university and work hard to ensure that our students are at the heart of all we do. We are committed to offering one of the best education and student experiences in the UK, undertaking research and innovation of the highest standard, and being a civic university that serves and contributes to our communities.

International Union for Conservation of Nature (IUCN)

IUCN is a membership Union composed of both government and civil society organisations. It harnesses the experience, resources and reach of its more than 1,400 Member organisations and the input of more than 16,000 experts. IUCN is the global authority on the status of the natural world and the measures needed to safeguard it.

IUCN World Commission on Protected Areas (WCPA)

The World Commission on Protected Areas (WCPA) is the world's premier network of protected and conserved areas expertise. The Commission has over 2500 members spanning 140 countries who provide strategic advice to policymakers and work to strengthen capacity and investment for protected areas establishment and management.

Arizona State University

Arizona State University has developed a new model for the American Research University, creating an institution that is committed to access, excellence and impact. ASU measures itself by those it includes, not by those it excludes. As the prototype for a New American University, ASU pursues research that contributes to the public good, and ASU assumes major responsibility for the economic, social and cultural vitality of the communities that surround it. www.asu.edu

BirdLife International

BirdLife International is the world's largest nature conservation Partnership: a global family of 122 national NGOs covering all continents, landscapes and seascapes. BirdLife is driven by its belief that local people, working for nature in their own places but connected nationally and internationally through the global Partnership, are the key to sustaining all life on this planet. This unique local-to-global approach delivers high impact and long-term conservation for the benefit of nature and people.

Global Environment Facility (GEF)

The Global Environment Facility (GEF) is a multilateral family of funds dedicated to confronting biodiversity loss, climate change, and pollution, and supporting land and ocean health. Its financing enables developing countries to address complex challenges and work towards international environmental goals. The partnership includes 186 member governments as well as civil society, Indigenous Peoples, women, and youth, with a focus on integration and inclusivity. Over the past three decades, the GEF has provided nearly $25 billion in financing and mobilized another $138 billion for thousands of priority projects and programs. The family of funds includes the Global Environment Facility Trust Fund, Global Biodiversity Framework Fund (GBFF), Least Developed Countries Fund (LDCF), Special Climate Change Fund (SCCF), Nagoya Protocol Implementation Fund (NPIF), and Capacity-building Initiative for Transparency Trust Fund (CBIT).

Zoological Society of London (ZSL)

Founded in 1826, ZSL is an international conservation charity, driven by science, working to restore wildlife in the UK and around the world; by protecting critical species, restoring ecosystems, helping people and wildlife live together and inspiring support for nature. Through our leading conservation zoos, London and Whipsnade, we bring people closer to nature and use our expertise to protect wildlife today, while inspiring a lifelong love of animals in the conservationists of tomorrow. Visit www.zsl.org for more information.

DISCOVER MORE

Support Oxford's research
Partner with Oxford on research
Study at Oxford
Research jobs at Oxford

You can view all news or browse by category

Sustainable Landscapes

Forest landscape restoration, protected & conserved areas, forest sector transformation & valuation, forests & climate, deforestation- and conversion-free supply chains & governance.

Case Studies
#Connect2Forests
Terms & Conditions

Case studies

Lessons from practitioners on conserving forests for nature and people.

Insights from WWF experts on solutions to safeguard forests.

Stories of people and places behind WWF's forest conservation work around the world.

Explore our work through the eyes of our people and partners.

wwf approach

Livestock farmers lead the way in implementing sustainable land use practices and reducing deforestation in peru, how argentina could emerge as a leader in mainstreaming beef free from deforestation, using wood forensic science to deter corruption and illegality in the timber trade, community-based forest monitoring in colombia, brazil's amazon soy moratorium, share your experience.

Replication of successful approaches and learning lessons from other forest practitioners can make conservation work more impactful. Take a few minutes to capture and share your experience and tips.

Work & Careers
Life & Arts
Currently reading: Deforestation opens paths for disease spread from animals to humans
Climate change highlights risk of malaria resurgence
Climate change risks fuelling antibiotic-resistant ‘superbugs’
Doubled estimate of fatal fungal infections sharpens scientific focus
Lyme disease bites as ticks gain ground in US
Climate change increases chances of zoonotic disease ‘spillover’
Climate change a ‘significant’ risk to health, insurers warn

Deforestation opens paths for disease spread from animals to humans

Deforestation opens paths for disease spread from animals to humans on x (opens in a new window)
Deforestation opens paths for disease spread from animals to humans on facebook (opens in a new window)
Deforestation opens paths for disease spread from animals to humans on linkedin (opens in a new window)
Deforestation opens paths for disease spread from animals to humans on whatsapp (opens in a new window)

Attracta Mooney

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

Six years after Malaysia first announced it had halted malaria spread by mosquitoes from person to person, the country is grappling with another big problem: a rapid rise in cases of the disease that stem from animals. And scientists say deforestation is a driving factor in this.

Around the world, deforestation — often to clear land for agriculture — has been linked to outbreaks of diseases that have jumped from animal to human. HIV, Zika, Sars, mpox and Ebola are some of the diseases that have emerged from tropical forests.

However, 20 years ago, Malaysia’s malaria problem was caused by parasites spread by mosquitoes from human to human. Then, between 2008 and 2017, there was a 861 per cent increase in cases of a type of malaria — zoonotic Plasmodium knowlesi — initially found in animals, according to academic research .

Zoonotic diseases are illnesses that can be transmitted directly or indirectly between animals and humans, and they include many vector-borne diseases, where an arthropod, such as a mosquito, plays a role in transmission.

Thousands of people a year are now infected with malaria carried by mosquitoes that first bit macaque monkeys, notes Kimberly Fornace, who heads the Climate, Environment and Health Programme at the National University of Singapore’s Saw Swee Hock School of Public Health and has spent years studying malaria in Malaysia.

“The human risk to this type of malaria is really quite closely linked with deforestation , ” she says. Malaysia lost about a third of its total tree cover in the first two decades of this millennium.

Deforestation “is a really big problem” for human health worldwide, Fornace stresses. “It has been identified as a major driver of human infections and diseases.”

Three-quarters of new or emerging diseases that infect humans originate in animals, according to the US Centers for Disease Control and Prevention. So, when humans destroy natural habitats and infringe on biodiversity-rich forests, diseases are able to spill over to them from wildlife more easily.

In Malaysia, for example, deforestation means some monkeys are living in much closer proximity to people. Similarly, the outbreak of Ebola has been linked to deforestation in Guinea. And, earlier this month in Uganda, scientists looking at virus spillovers that can cause pandemics found the disappearance of an important food source due to deforestation had left chimpanzees, monkeys and antelope eating bat excrement containing a range of viruses.

Neil Ward, vice-president of PacBio Emea, a biotechnology company, says biodiversity loss linked to deforestation increases the likelihood of pandemics. “Disease escape is a major risk, considering that many pandemics have zoonotic origins,” he explains.

Deforestation has long prompted concerns about the loss of biodiversity and vital carbon dioxide storage, but many people are only now waking up to the risks to human health, reckons Serge Morand, senior scientist at the French National Centre for Scientific Research, the state research organisation.

Disease escape is a major risk, considering that many pandemics have zoonotic origins Neil Ward, PacBio Emea

Morand was the lead author of a 2020 study of the links between deforestation and human health, taking into account population growth. The research found that increases in outbreaks of zoonotic and vector-borne diseases from 1990 to 2016 were heavily associated with deforestation. This was especially the case in the southern hemisphere, where deforestation was “really clearly linked in zoonotic and vector-borne disease”, says Morand.

Companies and investors are also realising the risks of deforestation, says Tim Steinweg, head of stewardship for nature at the PRI, a group focused on responsible investment — with many considering it as more than just a climate issue in which trees act as carbon sinks.

Mercedes Bustamante, a biologist and a professor at the University of Brasília, says that after Covid-19 and other outbreaks, people are “more aware of these risks” posed by deforestation. “But, globally, I don’t think we are trying to do a good job to change the situation.”

She adds that, while every effort must be made to prevent deforestation and to ensure existing forests are healthy, more work also needs to be done to share information and resources within different organisations and departments, both nationally and globally.

“We live in a much more connected world — people move around the world in aeroplanes,” says Bustamante. “There are now ways where disease can spread much faster than in the past. That makes it much harder to control. That is why we need to increase international co-operation.”

Morand believes a global treaty on forests is needed, akin to those agreed for desertification and the seas. “We need governments to step up,” he says, and calls for a proper framework that takes into account the regulation of disease transmission.

But he warns that simply planting trees is not enough to halt the risk of disease spreading from animals to humans. Morand’s study found that plantations focused heavily on one type of tree, as well as badly executed reforestation or afforestation (new forests), often came with their own risks to human health, as well. In temperate countries, reforestation was linked to disease, as were oil palm plantations.

In Malaysia, Fornace says a rise in so-called forest edges — the transition zones between woodland and other open spaces, which can be increased by deforestation — as well as patchy forest cover have both been linked to the rise in malaria cases.

“It’s a very important public health issue,” she says. “It’s important not just for Malaysia, but actually for the region.” She adds that it shows the challenges countries face in eradicating malaria and other illnesses.

Whether the world can halt the spread of further vector-borne and zoonotic disease now, after years of deforestation, remains to be seen, she says. “Hopefully, [the world is] not messed up too much, but there’s a need for more research. [We] need to think long term about how to develop sustainable landscapes.”

Climate Capital

Where climate change meets business, markets and politics. Explore the FT’s coverage here .

Are you curious about the FT’s environmental sustainability commitments? Find out more about our science-based targets here

Promoted Content

Explore the series.

A close-up of someone’s hands in green gloves scrubbing a white wall covered in patches of mold. There is a spray bottle likely containing cleaner on the left, and the person is holding a sponge in the right hand, actively cleaning

Follow the topics in this article

Climate change Add to myFT
Disease control and prevention Add to myFT
Science Add to myFT
Health sector Add to myFT
Deforestation Add to myFT

International Edition

Publications
Conferences & Events
Professional Learning
Science Standards
Awards & Competitions
Instructional Materials
Free Resources
American Rescue Plan
For Preservice Teachers
NCCSTS Case Collection
Partner Jobs in Education
Interactive eBooks+
Digital Catalog
Regional Product Representatives
e-Newsletters
Bestselling Books
Latest Books
Popular Book Series
Prospective Authors
Web Seminars
Exhibits & Sponsorship
Conference Reviewers
National Conference • Denver 24
Leaders Institute 2024
National Conference • New Orleans 24
Submit a Proposal
Latest Resources
Professional Learning Units & Courses
For Districts
Online Course Providers
Schools & Districts
College Professors & Students
The Standards
Teachers and Admin
eCYBERMISSION
Toshiba/NSTA ExploraVision
Junior Science & Humanities Symposium
Teaching Awards
Climate Change
Earth & Space Science
New Science Teachers
Early Childhood
Middle School
High School
Postsecondary
Informal Education
Journal Articles
Lesson Plans
e-newsletters
Science & Children
Science Scope
The Science Teacher
Journal of College Sci. Teaching
Connected Science Learning
NSTA Reports
Next-Gen Navigator
Science Update
Teacher Tip Tuesday
Trans. Sci. Learning

MyNSTA Community

My Collections

The Deforestation of the Amazon

A Case Study in Understanding Ecosystems and Their Value

By Philip Camill

Share Start a Discussion

In this case study, students examine tropical deforestation in the Amazon from the perspective of three dominant stakeholders in the region: a peasant farmer, logger, and environmentalist. As part of the exercise, students perform a cost-benefit analysis of clearing a plot of tropical forest in the Amazon from the perspective of one of these stakeholder groups. Developed for a course in global change biology, this case could also be used in courses in general ecology, environmental science, environmental ethics, environmental policy, and environmental/ecological economics.

Download Case

Date Posted

Understand the political, cultural, and economic history leading to tropical deforestation in Amazonia. Understand issues facing the major stakeholders in the Amazon.
Understand the concern for such a large loss in biodiversity.
Understand the concepts of market and non-market valuation of ecosystems, benefit-cost analysis, and opportunity cost.
Perform a cost-benefit analysis of clearing a plot of tropical forest in the Amazon, from the perspective of a peasant farmer, logger, and environmentalist.
Critically evaluate economic vs. ethical valuation of ecosystems.
Appreciate the political, social, economic, and ecological complexity of tropical deforestation.
Appreciate how difficult decisions must me made in the face of limited or nonexistent data.

Deforestation; Amazon; tropical forest; rainforest; ecosystem; biodiversity; bioprospecting; ecotourism; ecological economics; cost-benefit analysis; tropics; developing world; South America

Subject Headings

EDUCATIONAL LEVEL

High school, Undergraduate lower division, Undergraduate upper division

TOPICAL AREAS

Ethics, Policy issues, Social issues, Social justice issues

TYPE/METHODS

Teaching Notes & Answer Key

Teaching notes.

Case teaching notes are protected and access to them is limited to paid subscribed instructors. To become a paid subscriber, purchase a subscription here .

Teaching notes are intended to help teachers select and adopt a case. They typically include a summary of the case, teaching objectives, information about the intended audience, details about how the case may be taught, and a list of references and resources.

Download Notes

Answer Keys are protected and access to them is limited to paid subscribed instructors. To become a paid subscriber, purchase a subscription here .

Download Answer Key

Materials & Media

Supplemental materials, you may also like.

Web Seminar

Join us on Thursday, June 13, 2024, from 7:00 PM to 8:00 PM ET, to learn about the science and technology of firefighting. Wildfires have become an e...

Join us on Thursday, October 10, 2024, from 7:00 to 8:00 PM ET, for a Science Update web seminar presented by NOAA about climate science and marine sa...

share this!

April 25, 2024

This article has been reviewed according to Science X's editorial process and policies . Editors have highlighted the following attributes while ensuring the content's credibility:

fact-checked

peer-reviewed publication

First-of-its-kind study shows that conservation actions are effective at halting and reversing biodiversity loss

First-of-its-kind study definitively shows that conservation actions are effective at halting and reversing biodiversity loss

A study published April 25, in the journal Science provides the strongest evidence to date that not only is nature conservation successful, but that scaling conservation interventions up would be transformational for halting and reversing biodiversity loss—a crisis that can lead to ecosystem collapses and a planet less able to support life—and reducing the effects of climate change.

Governments recently adopted new global targets to halt and reverse biodiversity loss, making it even more critical to understand whether conservation interventions are working.

"If you look only at the trend of species declines, it would be easy to think that we're failing to protect biodiversity, but you would not be looking at the full picture," said Penny Langhammer, lead author of the study and executive vice president of Re:wild.

"What we show with this paper is that conservation is, in fact, working to halt and reverse biodiversity loss. It is clear that conservation must be prioritized and receive significant additional resources and political support globally, while we simultaneously address the systemic drivers of biodiversity loss, such as unsustainable consumption and production."

The co-authors conducted the first-ever meta-analysis of 186 studies, including 665 trials, that looked at the impact of a wide range of conservation interventions globally, and over time, compared to what would have happened without those interventions. The studies covered over a century of conservation action and evaluated actions targeting different levels of biodiversity—species, ecosystems and genetic diversity.

The meta-analysis found that conservation actions—including the establishment and management of protected areas, the eradication and control of invasive species, the sustainable management of ecosystems, habitat loss reduction and restoration—improved the state of biodiversity or slowed its decline in the majority of cases (66%) compared with no action taken at all. And when conservation interventions work, the paper's co-authors found that they are highly effective.

For example:

Management of invasive and problematic native predators on two of Florida's barrier islands, Cayo Costa and North Captiva, resulted in an immediate and substantial improvement in nesting success by loggerhead turtles and least terns, especially compared with other barrier islands where no predator management was applied.
In the Congo Basin, deforestation was 74% lower in logging concessions under a Forest Management Plan (FMP) compared with concessions without an FMP.
Protected areas and Indigenous lands were shown to significantly reduce both deforestation rate and fire density in the Brazilian Amazon. Deforestation was 1.7 to 20 times higher and human-caused fires occurred four to nine times more frequently outside the reserve perimeters compared with inside.
Captive breeding and release boosted the natural population of Chinook salmon in the Salmon River basin of central Idaho with minimal negative impacts on the wild population. On average, fish taken into the hatchery produced 4.7 times more adult offspring and 1.3 times more adult second generation offspring than naturally reproducing fish.

"Our study shows that when conservation actions work, they really work. In other words, they often lead to outcomes for biodiversity that are not just a little bit better than doing nothing at all, but many times greater," said Jake Bicknell, co-author of the paper and a conservation scientist at DICE, University of Kent.

"For instance, putting measures in place to boost the population size of an endangered species has often seen their numbers increase substantially. This effect has been mirrored across a large proportion of the case studies we looked at."

This might also explain why the co-authors found a correlation between more recent conservation interventions and positive outcomes for biodiversity—conservation is likely getting more effective over time. Other potential reasons for this correlation include an increase in funding and more targeted interventions.

In some other cases where the conservation action did not succeed in benefiting the target biodiversity compared with no action at all, other native species benefited unintentionally instead. For example, seahorse abundance was lower in protected sites because marine protected areas increase the abundance of seahorse predators, including octopus.

"It would be too easy to lose any sense of optimism in the face of ongoing biodiversity declines," said study co-author and Associate Professor Joseph Bull, from the University of Oxford's department of biology. "However, our results clearly show that there is room for hope. Conservation interventions seemed to be an improvement on inaction most of the time; and when they were not, the losses were comparatively limited."

More than half of the world's GDP, almost $44 trillion , is moderately or highly dependent on nature.

According to previous studies, a comprehensive global conservation program would require an investment of between US$178 billion and US$524 billion , focused primarily in countries with particularly high levels of biodiversity. To put this in perspective, in 2022, global fossil fuel handouts—which are destructive to nature—were US$7 trillion .

This is 13 times the highest amount needed annually to protect and restore the planet. Today more than US$121 billion is invested annually into conservation worldwide , and previous studies have found the cost-benefit ratio of an effective global program for the conservation of the wild is at least 1:100 .

"Conservation action works—this is what the science clearly shows us," said Claude Gascon, co-author and director of strategy and operations at the Global Environment Facility.

"It is also evident that to ensure that positive effects last, we need to invest more in nature and continue doing so in a sustained way. This study comes at a critical time where the world has agreed on ambitious and needed global biodiversity targets that will require conservation action at an entirely new scale. Achieving this is not only possible, it is well within our grasp as long as it is appropriately prioritized."

The paper also argues that there must be more investment specifically in the effective management of protected areas, which remain the cornerstone for many conservation actions. Consistent with other studies, this study finds that protected areas work very well on the whole. And what other studies have shown is that when protected areas are not working, it is typically the result of a lack of effective management and adequate resourcing. Protected areas will be even more effective at reducing biodiversity loss if they are well-resourced and well-managed.

Moving forward, the study's co-authors call for more and rigorous studies that look at the impact of conservation action versus inaction for a wider range of conservation interventions, such as those that look at the effectiveness of pollution control, climate change adaptation, and the sustainable use of species, and in more countries.

"For more than 75 years, IUCN has advanced the importance of sharing conservation practice globally," said Grethel Aguilar, IUCN director general.

"This paper has analyzed conservation outcomes at a level as rigorous as in applied disciplines like medicine and engineering—showing genuine impact and thus guiding the transformative change needed to safeguard nature at scale around the world. It shows that nature conservation truly works, from the species to the ecosystem levels across all continents. This analysis, led by Re:wild in collaboration with many IUCN Members, Commission experts, and staff, stands to usher in a new era in conservation practice."

Journal information: Science

Provided by Re:wild

Explore further

Feedback to editors

Archaea can be 'picky eaters': Study shows a group of parasitic microbes can change host metabolism

EPA underestimates methane emissions from landfills and urban areas, researchers find

2 hours ago

This Texas veterinarian helped crack the mystery of bird flu in cows

Researchers discover key functions of therapeutically promising jumbo viruses

Marine sharks and rays 'use' urea to delay reproduction, finds study

Researchers unlock potential of 2D magnetic devices for future computing

Researchers build new device that is a foundation for quantum computing

3 hours ago

Satellite images of plants' fluorescence can predict crop yields

New work reveals the 'quantumness' of gravity

Mystery behind huge opening in Antarctic sea ice solved

4 hours ago

Relevant PhysicsForums posts

The cass report (uk).

6 hours ago

Is 5 milliamps at 240 volts dangerous?

Apr 29, 2024

Major Evolution in Action

Apr 22, 2024

If theres a 15% probability each month of getting a woman pregnant...

Apr 19, 2024

Can four legged animals drink from beneath their feet?

Apr 15, 2024

Mold in Plastic Water Bottles? What does it eat?

Apr 14, 2024

Scientist explores sufficiency as an overlooked strategy for protecting biodiversity

Apr 2, 2024

More for less: A smarter way to protect biodiversity

Mar 6, 2024

Over half of threatened species require targeted recovery actions

Jul 18, 2022

New study offers hope that conservation can help nature adapt to climate change

May 19, 2022

Q&A: How can Canada best meet its commitment to protecting 30% of its land by 2030?

Nov 24, 2023

Conservationists propose 'global conservation basic income' to safeguard biodiversity

May 18, 2023

Recommended for you

Mimicry allows lesser necklaced laughingthrush birds to benefit from living among larger related species

Apr 30, 2024

How polyps of the moon jellyfish repel viral attacks on their microbiome

Missing link in species conservation: Pharmacists, chemists could turn tide on plant, animal extinction

Researchers find pesticides in a third of Australian frogs tested. Did these cause mass deaths?

Unveiling nature's custodians: Study highlights crucial role of scavengers in wetlands

Let us know if there is a problem with our content.

Use this form if you have come across a typo, inaccuracy or would like to send an edit request for the content on this page. For general inquiries, please use our contact form . For general feedback, use the public comments section below (please adhere to guidelines ).

Please select the most appropriate category to facilitate processing of your request

Thank you for taking time to provide your feedback to the editors.

Your feedback is important to us. However, we do not guarantee individual replies due to the high volume of messages.

E-mail the story

Your email address is used only to let the recipient know who sent the email. Neither your address nor the recipient's address will be used for any other purpose. The information you enter will appear in your e-mail message and is not retained by Phys.org in any form.

Newsletter sign up

Get weekly and/or daily updates delivered to your inbox. You can unsubscribe at any time and we'll never share your details to third parties.

More information Privacy policy

Donate and enjoy an ad-free experience

We keep our content available to everyone. Consider supporting Science X's mission by getting a premium account.

E-mail newsletter

Study: Conservation actions highly effective at halting, reversing biodiversity loss

A new study involving ASU researchers and published in Science provides evidence that nature conservation is successful, and that scaling up conservation interventions would be transformational for halting and reversing biodiversity loss. Courtesy photo

A new study, led and contributed to by Arizona State University faculty, provides the strongest evidence to date that not only is nature conservation successful, but that scaling up conservation interventions would be transformational for halting and reversing biodiversity loss — a crisis that can lead to ecosystem collapse and a planet less able to support life — and reducing the effects of climate change.

The study was published today, April 25, in the journal Science .

These findings of the impact of conservation action are crucial, as more than 44,000 species are currently documented as being at risk of extinction , the result of which would cause tremendous consequences for the ecosystems that stabilize the climate and provide billions of people around the world with clean water, livelihoods, homes and cultural preservation, among other ecosystem services.

To address this risk, governments recently adopted new global targets to halt and reverse biodiversity loss, making it even more critical to understand whether conservation interventions are working.

“If you look only at the trend of species declines, it would be easy to think that we’re failing to protect biodiversity, but you would not be looking at the full picture,” said Penny Langhammer , lead author of the study, an adjunct professor of biology in the School of Life Sciences , affiliate researcher with the Center for Biodiversity Outcomes and executive vice president of Re:wild .

“What we show with this paper is that conservation is, in fact, working to halt and reverse biodiversity loss. It is clear that conservation must be prioritized and receive significant additional resources and political support globally, while we simultaneously address the systemic drivers of biodiversity loss, such as unsustainable consumption and production.”

Although many studies look at individual conservation projects and interventions, and their impact compared with no action taken, these papers have never been pulled into a single analysis to see how and whether conservation action is working overall.

The co-authors conducted the first-ever meta-analysis of 186 studies — including 665 trials — that looked at the impact of a wide range of conservation interventions globally, and over time, compared with what would have happened without those interventions.

The studies covered over a century of conservation action and evaluated actions targeting different levels of biodiversity: species, ecosystems and genetic diversity.

“We have a lot of science-based evidence for successful conservation actions that can directly mitigate threats while working to improve species populations,” said Beth Polidoro , an associate professor of environmental chemistry and aquatic conservation in the School of Mathematical and Natural Sciences , who also contributed to the study.

Polidoro also serves as associate director of biodiversity assessment in the Center for Biodiversity Outcomes, which is jointly housed in the School of Life Sciences and the Global Institute of Sustainability and Innovation .

“However, there are still many gaps in understanding the specific combination of contexts that are needed for these actions to be successful over the long term — namely the economic, social, technical and other factors that are important for facilitating sustainable conservation outcomes,” she said.

The meta-analysis found that conservation actions — including the establishment and management of protected areas, the eradication and control of invasive species, the sustainable management of ecosystems, habitat loss reduction and restoration — improved the state of biodiversity or slowed its decline in the majority of cases (66%) compared with no action taken at all.

And when conservation interventions work, the paper’s co-authors found that they are highly effective.

For example:

Management of invasive and problematic native predators on two of Florida’s barrier islands, Cayo Costa and North Captiva, resulted in an immediate and substantial improvement in nesting success by loggerhead turtles and least terns, especially compared with other barrier islands where no predator management was applied.
In the Congo Basin, deforestation was 74% lower in logging concessions under a forest management plan (FMP) compared with concessions without an FMP.
Protected areas and Indigenous lands were shown to significantly reduce both the deforestation rate and fire density in the Brazilian Amazon. Deforestation was 1.7 to 20 times higher and human-caused fires occurred four to nine times more frequently outside the reserve perimeters compared with inside.
Captive breeding and release boosted the natural population of Chinook salmon in the Salmon River basin of central Idaho, with minimal negative impacts on the wild population. On average, fish taken into the hatchery produced 4.7 times more adult offspring and 1.3 times more adult second-generation offspring than naturally reproducing fish.

Even in the minority of cases in which conservation actions did not succeed in recovering or slowing the decline of the species or ecosystems that they were targeting, compared with taking no action, conservationists benefited from the knowledge gained and were able to refine their methods.

For example, in India, the physical removal of invasive algae caused the spread of the algae elsewhere because the process broke the algae into many pieces, enabling its dispersal. Conservationists could now implement a different strategy to remove the algae that is more likely to be successful.

This might also explain why the co-authors found a correlation between more recent conservation interventions and positive outcomes for biodiversity. In other words, conservation is likely getting more effective over time. Other potential reasons for this correlation include an increase in funding and more targeted interventions.

In some other cases in which the conservation action did not succeed in benefiting the target biodiversity compared with no action at all, other native species benefitted unintentionally instead.

For example, seahorse abundance was lower in protected sites because protected marine areas increase the abundance of seahorse predators, including octopus.

“It would be too easy to lose any sense of optimism in the face of ongoing biodiversity declines,” said study co-author and Associate Professor Joseph Bull from the University of Oxford’s Department of Biology. “However, our results clearly show that there is room for hope. Conservation interventions seemed to be an improvement on inaction most of the time; and when they were not, the losses were comparatively limited."

To put this in perspective, in 2022, global fossil fuel handouts — which are destructive to nature — were $7 trillion . This is 13 times the highest amount needed annually to protect and restore the planet. Today, more than $121 billion is invested annually into conservation worldwide, and previous studies have found the cost-benefit ratio of an effective global program for the conservation of the wild is at least 1:100.

“Conservation action works — this is what the science clearly shows us,” said Claude Gascon, co-author and director of strategy and operations at the Global Environment Facility . “It is also evident that to ensure that positive effects last, we need to invest more in nature and continue doing so in a sustained way. This study comes at a critical time where the world has agreed on ambitious and needed global biodiversity targets that will require conservation action at an entirely new scale.

"Achieving this is not only possible, it is well within our grasp as long as it is appropriately prioritized.”

The paper also argues that there must be more investment specifically in the effective management of protected areas, which remain the cornerstone for many conservation actions. Consistent with other studies, this study finds that protected areas work very well on the whole. And what other studies have shown is that when protected areas are not working, it is typically the result of a lack of effective management and adequate resourcing. Protected areas will be even more effective at reducing biodiversity loss if they are well resourced and well managed.

Moving forward, the study’s co-authors call for more rigorous studies, in more countries, that look at the impact of conservation action versus inaction for a wider range of conservation interventions, such as those that look at the effectiveness of pollution control, climate change adaptation and the sustainable use of species.

“For more than 75 years, IUCN has advanced the importance of sharing conservation practice globally,” said Grethel Aguilar, IUCN director general. “This paper has analyzed conservation outcomes at a level as rigorous as in applied disciplines like medicine and engineering, showing genuine impact and thus guiding the transformative change needed to safeguard nature at scale around the world. It shows that nature conservation truly works, from the species to the ecosystem levels across all continents.

"This analysis, led by Re:wild in collaboration with many IUCN members, commission experts, and staff, stands to usher in a new era in conservation practice.”

This work was conceived and funded through the International Union for Conservation of Nature (IUCN) by the Global Environment Facility.

Writing contributions include Lindsay Renick Mayer and Devin Murphy.

More Environment and sustainability

Students pitch in to help solve plastic problem in Ethiopian national park

This weekend, 30 students from Arizona State University’s Ira A. Schools of Engineering will take an 18-hour flight to northern Ethiopia. Their mission? To tackle a damaging plastic problem that has…

Photo of flowering plant at ASU Polytechnic campus

Barrett Honors College to host nature walks for science, relaxation

Barrett, The Honors College at Arizona State University is gearing up to participate in the City Nature Challenge (CNC) for the fourth consecutive year. This annual event, taking place April 26–29,…

Knowledge Exchange for Resilience Executive Director Patricia Solís welcomes attendees to the inaugural Extreme Heat Policy Innovation Summit in Washington D.C.

Arizona adapting to heat crisis with initiatives featured in ASU report

Arizona State University's Knowledge Exchange for Resilience, also known as KER, released its Recommendations Report on Extreme Heat Preparedness earlier this April during a summit in the nation's…

Deforestation and its Impacts in the Case-Study Areas

Cite this chapter.

Solon L. Barraclough &
Krishna B. Ghimire

Part of the book series: International Political Economy Series ((IPES))

96 Accesses

This chapter reports some of the findings of the field studies that examined processes directly leading to deforestation and the impacts on different social groups. The four case-studies are discussed in separate sections. This is done in order better to suggest inte.ractions and linkages among processes, policies and more stable relationships influencing deforestation and its consequences in each area.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Available as PDF
Read on any device
Instant download
Own it forever

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Unable to display preview. Download preview PDF.

You can also search for this author in PubMed Google Scholar

Copyright information

About this chapter

Barraclough, S.L., Ghimire, K.B. (1995). Deforestation and its Impacts in the Case-Study Areas. In: Forests and Livelihoods. International Political Economy Series. Palgrave Macmillan, London. https://doi.org/10.1057/9780230375802_3

Download citation

DOI : https://doi.org/10.1057/9780230375802_3

Publisher Name : Palgrave Macmillan, London

Print ISBN : 978-0-333-62890-4

Online ISBN : 978-0-230-37580-2

eBook Packages : Palgrave Economics & Finance Collection Economics and Finance (R0)

Share this chapter

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Publish with us

Policies and ethics

Find a journal
Track your research

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

View all journals
My Account Login
Explore content
About the journal
Publish with us
Sign up for alerts
Open access
Published: 19 January 2024

Mapping the diversity of land uses following deforestation across Africa

Robert N. Masolele 1 ,
Diego Marcos 2 na1 ,
Veronique De Sy 1 na1 ,
Itohan-Osa Abu 3 ,
Jan Verbesselt 1 ,
Johannes Reiche 1 &
Martin Herold 4 na1

Scientific Reports volume 14 , Article number: 1681 ( 2024 ) Cite this article

6267 Accesses

1 Citations

111 Altmetric

Metrics details

Climate change
Computational science
Environmental impact
Scientific data

African forest are increasingly in decline as a result of land-use conversion due to human activities. However, a consistent and detailed characterization and mapping of land-use change that results in forest loss is not available at the spatial-temporal resolution and thematic levels suitable for decision-making at the local and regional scales; so far they have only been provided on coarser scales and restricted to humid forests. Here we present the first high-resolution (5 m) and continental-scale mapping of land use following deforestation in Africa, which covers an estimated 13.85% of the global forest area, including humid and dry forests. We use reference data for 15 different land-use types from 30 countries and implement an active learning framework to train a deep learning model for predicting land-use following deforestation with an F1-score of $84\pm 0.7$ for the whole of Africa. Our results show that the causes of forest loss vary by region. In general, small-scale cropland is the dominant driver of forest loss in Africa, with hotspots in Madagascar and DRC. In addition, commodity crops such as cacao, oil palm, and rubber are the dominant drivers of forest loss in the humid forests of western and central Africa, forming an “arc of commodity crops” in that region. At the same time, the hotspots for cashew are found to increasingly dominate in the dry forests of both western and south-eastern Africa, while larger hotspots for large-scale croplands were found in Nigeria and Zambia. The increased expansion of cacao, cashew, oil palm, rubber, and large-scale croplands observed in humid and dry forests of western and south-eastern Africa suggests they are vulnerable to future land-use changes by commodity crops, thus creating challenges for achieving the zero deforestation supply chains, support REDD+ initiatives, and towards sustainable development goals.

An unexpectedly large count of trees in the West African Sahara and Sahel

Nation-wide mapping of tree-level aboveground carbon stocks in Rwanda

The conservation value of forests can be predicted at the scale of 1 hectare

Introduction.

Understanding the dynamics of land-use following deforestation is an important step in climate change mitigation, having a significant effect on forest biomass, biodiversity, and the water cycle 1 . Over the past two decades, Africa has been experiencing a rapid decline in its forest cover or tree cover 2 . Here we define forest cover or tree cover as adopted from 2 as all vegetation taller than 5 meters in height.The implications of which is the decline in species richness, changes in the water cycle, and loss of forest carbon stock 3 . The complexion of these changes can vary depending on the location, intensity, and spatial extent of forest loss. Thus understanding the spatial-temporal extent and patterns of the drivers of forest loss in Africa is crucial to comprehending its negative impacts on the forest ecosystem and its contribution to greenhouse gas emissions.

Previous studies suggest agriculture-related land-use change as Africa’s main cause of deforestation 4 . However, such information is derived from courser thematic maps or sample-based land-use statistics 5 . There is a lack of thematically detailed data both at small and large-scale to link deforestation to its respective drivers 6 . Specifically, the much-needed and highly detailed thematic information on the types of direct drivers causing deforestation in Africa is rarely observed in the literature, except for a few countries (mainly in the tropical humid forests) 5 , 7 . The availability of large-scale, thematic detailed, and high-resolution maps of land use following deforestation in Africa is essential for strategic planning and implementation of deforestation mitigation actions by governments and forest protection agencies 8 .

In addition, most studies focus their driver assessment on the knowledge of localizing specific drivers in specific regions or countries. For example, the latest studies projected an expansion of oil palm, cacao farms in west Africa 9 , 10 , and small-scale agriculture in the Democratic Republic of Congo 7 . The generally accepted view of forest conversion in Africa aligns with the historical expectation of persistence of various states of subsistence agriculture activities, but little conversion to newly introduced or other types of land uses not previously found in the current location 11 . Aligning with the general view that certain land uses are restricted in certain geographical locations, thus overlooking the diverse actual causes of forest loss critical for deforestation monitoring and mitigation efforts at both local and regional scales.

Despite the availability of national or regional statistics to document the trends of forest loss, consistent, detailed, spatially explicit estimates of the continent’s drivers of forest loss area are lacking 2 , 12 , 13 , 14 . Many efforts to map drivers of forest loss rely on expert-based visual interpretation from samples of high-resolution satellite images 7 , 15 . However, the limitation of expert/sample-based visual interpretation is that it lacks detail and consistency in identifying and mapping the drivers over large regions and across time. This, in turn, causes a broad generalization of the drivers of forest loss and, thus, misses most areas and limits the level of detail in pinpointing the exact causes of forest loss useful for government agencies (local, national) responsible for forest monitoring 4 . For example, although the latest study of forest change in tropical forests purportedly assessed agriculture expansion over space and time, its delineation of land uses largely relied on sample-based single-date expert interpretation, and the wall-to-wall-map was not reported 5 .

The absence of detailed systematic monitoring of forest loss drivers complicates assessments of REDD+ efforts and net zero commitments on reducing the impacts of land-use change on forest ecosystems 4 , 8 , 13 , 15 . Generalization and confusion between land-use and land-cover change maps produced at a global scale lead to ‘cryptic forest loss’ and/or overestimates certain land-use and land-cover 4 . This is especially true in Africa, with its rapid deforestation rates and diverse land uses 2 , 16 . National reporting indicates a substantial net increase in forest loss 16 , but keeping track of net changes in the area of land-cover types that undergo frequent disturbances, like forest regrowth, can lead to the underestimation of the occurrence of tree crop expansion as, with time, they become confused or resemble forest cover 16 , 17 . This is especially true in dry forests where tree crops (i.e., cashew) resemble natural forest trees (i.e., Miombo) 2 . Thus, it remains unclear whether natural forest recovery or tree crops are driving expansions in forest cover in Africa. Specifically in countries such as the Ivory Coast, Ghana, and Eastern regions of Tanzania and Mozambique, where commodity crops are dominant.

Assessing the factors that contribute to deforestation consistently over both space and time using remote sensing data is difficult, particularly when relying on medium-resolution satellite imagery, with a spatial resolution of 10 to 30 meters, which is required for comprehensive worldwide coverage 18 . The difficulty stems partially from similarities and differences, both in spectral and spatial aspects, between land-use practices 4 , 19 , 20 . Land uses over large scale are spectrally heterogeneous, with substantial variation in spectral signatures across space, time, elevation, soil types, forest types, and disturbance intensities. These land-use similarities and differences, along with variation in spectral reflectance and phenology across geographic regions, as well as persistent tropical cloudiness, make it difficult to distinguish land uses using satellite imagery consistently. Although variability of land-uses has constrained its classification at larger scales 21 , maps based on automated classification of remotely sensed imagery have successfully monitored drivers of forest loss at regional scales (for example 4 ). However, the spatial, temporal, and thematic detail of the regional maps of land-use after deforestation is limited, and varies widely across different geographies and land-use types.

To tackle these challenges, we undertook a continental assessment of the direct drivers of forest loss here, defined as a human-related land-use conversion or land-use following deforestation. We focused on the tropics (30° N to 30 $^{\circ }$ S) due to the widespread conversion of natural forest to agriculture, mining, settlements, and commodity tree crops across tropical latitudes and high rates of potential carbon sequestration from tropical tree regrowth. We aimed to use high-resolution remote sensing data 22 , deep learning 17 , and active learning (AL) 23 , 24 to accurately identify and map land use following deforestation and assess the trend and hotspots of land-use conversion across countries, and regions in Africa 25 .

Active learning for improving land-use classification

AL could be leveraged to combine heterogeneous data sources with limited labels for the task of semantic segmentation 23 , 26 . Here we present the results with and without using AL to map land-use following deforestation at a continental scale. The original reference labels are limited, geographically scattered, heterogeneous and are subject to errors (land use delineation error, mix of different land use in one polygon). Using independent test samples stemming from AL, on two AL rounds, we improved the classification performance on all land-use classes. Specifically, the macro average F1-score improved from 43% with the original data to 54% after the first AL round and to 84% after the second round (Fig. 1 ). The exception to this trend is cacao, which shows high accuracy on all iterations (Fig. 1 ).

The bar chart shows the performance of the attention U-Net model in classifying land-use following deforestation over three active learning cycles in Africa.

Generating land-use map prediction at scale

We present the first wall-to-wall map of land-use following deforestation across Africa covering forest loss from the year 2001 to 2020 (Fig. 2 ). The output predicted map is available at a spatial resolution of 5 m, with 15 land-use classes. The map has a users accuracy of 85%, producers accuracy of 84%, and F1-score of 85% (Fig. 3 ). For detailed visualisation, the predicted output map of land use following deforestation is available at this website: https://robertnag82.users.earthengine.app/view/africalu .

A 5 m resolution land-use following deforestation map in Africa (30 $^{\circ }$ south and 30 $^{\circ }$ north) corresponding to 2020 using planet-NICFI images and Hansen forest loss data between 2000 and 2020 as proxy for forest loss. The zoomed-in maps show ( A ) cacao expansion in Cameroon, ( B ) cacao, oil palm, and rubber expansion in Ghana, ( C ) small-scale cropland expansion in the DRC, ( D ) small-scale cropland expansion in Tanzania, ( E ) cashew expansion in Tanzania, ( F ) a mix of small-scale and large-scale cropland expansion in Zambia, and ( G ) shows a mix of mining along rivers, roads, and small-scale cropland in DRC for the same period.

The users, producer’s accuracy and F1-score of the wall-to-wall map generated in (Fig. 2 ).

Proportion of land-use following deforestation by country

Small-scale cropland was the dominant driver of forest loss in Africa, resulting in 64% of total forest loss from the year 2001 to 2020 (Fig. 4 ). This was also the case for most countries, regardless of their contribution to total forest loss, a notably high proportion of small-scale cropland was observed in Madagascar (88%), followed by (85%) in the Democratic Republic of Congo (DRC), Burundi (81%), Comoros (79%), Malawi (76%), Angola (75%) and Mozambique (74%). Other-land with tree cover (OLWTC) was the second highest driver of forest loss in Africa and contributed to 10% of all forest loss in Africa. The highest proportion was observed in Gabon (34%) and Equatorial Guinea (34%). OLWTC constitutes all forest conversion related to fire, windthrow, lightning, speculative clearings, abandoned croplands, and regrowth 6 . Large-scale cropland was the third highest driver of forest loss in Africa (9%), with the highest proportions by country found in Cape Verde (67%), Gambia (53%), Niger (50%), Sudan (47%), and Nigeria (44%). Likewise, the highest proportion of tea plantation establishments was observed in Kenya (4%) and Rwanda (3%).

The proportion of forest conversion for commodity crops such as cacao, cashew, oil palm, rubber, and coffee accounted for 7% of all forest loss in Africa. By country, the highest proportion of cacao was found in Ghana (25%), Ivory Coast (21%), and Liberia (15%); cashew in Ivory Coast constituted (7%), Ghana, Guinea, and Tanzania each constituted (6%), and Mozambique (5%); on the other hand a high proportion of oil palm was found in Gabon (6%), with Liberia, Ghana, and Ivory Coast each having (2%); a high proportion of rubber was mostly found in Gabon (7%), Ivory Coast and Cameroon were having each (3%), and Liberia (2%); the contribution of coffee was mostly found in in Kenya (1%). Additionally, the highest proportion of pasture was observed in Niger (27%), Somalia (22%), and Kenya (18%).

The conversion of forest settlement was mostly observed in Gambia (10%), Rwanda (8%), and Equatorial Guinea (6%); similarly, roads constituted a majority of forest loss in Equatorial Guinea (14%), Gabon (6%), Congo (5%), and Cameroon (3%), while mining had a higher proportion in Cape Verde (12%), Botswana (7%), and Equatorial Guinea (5%). Water was mostly observed in Niger (14%), with most changes associated with meandering rivers. Not surprisingly, the highest proportion of plantation forests was found in southern African countries such as Eswatini (46%) and South Africa (37%).

Additionally, in appendix A we add a table that shows total area of land use following deforestation in Mega hactare (Mha) per land use and per country.

Proportion of land-use following deforestation 2001–2020 by country across continental Africa 30 $^{\circ }$ south and 30 $^{\circ }$ north.

Trend of land-use following deforestation in Africa

Having predicted the land-use following deforestation across Africa for the year 2001 to 2020, we attempted to estimate the trend of the land-use for the entire study area and across four regions in Africa (Fig. 5 a) based on area and proportion per lustrum. These regions are western, central, eastern, and southern Africa. Our results suggest an increasing trend in the area of all land-use following deforestation, with the exception of pasture and plantation forest (Fig. 5 b). However, when the trend was calculated based on the proportion of each land-use per lustrum, only small-scale cropland showed a positive increasing trend in three regions of western, central, and east Africa with the exception of southern Africa region (Fig. 5 c). This show that although the area of almost every land-use is increasing per lustrum, the change is not proportionate.

Trend of land-use following deforestation in lustrum from 2001-2020 across four regions of Africa, shown in ( a ) are corresponding regions, in ( b ) we show the trend based on the area ( $log10-scsale$ ) for the entire continent and in ( c ) based on the proportion of land-use following deforestation per region per lustrum. LSCP and SSCP stands for large-scale cropland and small-scale cropland, while OLWTC stands for other-land with tree cover.

Hotspots of land-use following deforestation in Africa

We observed a considerable spatial variation of hotspots of land-use following deforestation across continental Africa (Fig. 6 ). The major hotspot locations are small-scale cropland in the Democratic Republic of Congo (DRC), Angola, and Madagascar; large-scale cropland in Nigeria and Zambia; pasture across east Africa but mainly in Tanzania; a major hotspot for cacao is largely along the southern regions of west Africa, but also in central Africa, and form what we call the arc of cacao , (see Fig. 6 ). We also observe hotspots for cashew in northern and central regions of Ivory Coast, Ghana, east-southern Tanzania, and northern Mozambique; hotspots for oil palm were highly observed in Ivory Coast, Ghana, Liberia, Cameroon, and Uganda; coffee in Kenya, Ethiopia; tea plantations in Kenya, Rwanda, and Malawi; rubber in Ivory Coast, Ghana, Liberia, Cameroon, and Gabon; plantation forest in South Africa, and Eswatini; roads in Ghana, Cameroon, Liberia, and Equatorial Guinea; settlements in Ivory Coast, Liberia, Nigeria, and Cameroon; mining in Ghana, Angola and parts of Eastern DRC; hotspots for water were mostly observed along meandering rivers of the coast of Ivory Coast, Ghana, DRC, and along the coast (islands) of Uganda. Of most importance, we observe commodity crops hotspots dominating in western and central African regions and partially forming an arc along the coast of the Atlantic ocean (Fig. 6 ). In general, these hotspot maps conveys an eye catching message to forest conservation agencies, decision and policy makers by showing the exact causes and locations of forest loss, for which more effort need to be directed and prepare best strategies for mitigation actions for future forest conservation.

Hotspots of different land-use following deforestation in Africa along 30 $^{\circ }$ south and 30 $^{\circ }$ north, for the year 2001 to 2020. The colorbar shows the estimated pixel density of each land-use in a given location.

This study was carried out within the African continent along 30 $^{\circ }$ south and 30 $^{\circ }$ north, which covers countries in western, central, eastern, and southern Africa (Fig. 7 ) . The region is characterized by humid forests as well as dry forests. The study area was defined based on the coverage of high-resolution Planet-NICFI data 22 .

Map showing study location in the African continent along 30 $^{\circ }$ south and 30 $^{\circ }$ north. ( A ) is the study area map with forest loss locations between 2001 and 2020, and ( B ) shows the locations of the training data over three active learning training cycles.

Reference land-use data

We identified 15 land-use following deforestation classes as the main direct drivers of deforestation in Africa using information from 5 , 9 , 15 , 17 , 21 , 27 . The land-use data from 5 was obtained via crowd-sourcing using citizen science while 17 , 21 annotated the land-use data using high-resolution images in collaboration with stakeholders on the respective country. The data from 15 , 27 is based on the FAO global Remote Sensing Survey for 2010, where they used a systematic sampling along latitude and longitude with grids spaced 10km by 10km. Other reference data was retrieved from open source data available via online searching 28 , 29 , 30 , 31 . The land-use classes identified and annotated for this study are small-scale cropland, large-scale cropland, pasture, mining, roads, other-land with tree cover, plantation forest, coffee, settlement, tea plantation, water, oil palm, rubber, cashew, and cacao (refer appendix D for class definitions). However, it is important to highlight that the majority of reference labels are available as point vectors, or their polygon contains a mix of land-use classes and does not accurately delineate the land-use borders, making it challenging for direct use in the semantic segmentation task. Thus, extra annotation was implemented using an AL process described in Sect. " Models and implementation details ".

Satellite imagery

We used annual mosaics of high-resolution planet-NICFI images with 4.77 m $\approx $ 5 m resolution to train a deep learning model and map drivers of deforestation in continental Africa (30 $^{\circ }$ south and 30 $^{\circ }$ north) 22 . The image mosaics are made available as a result of Norway’s International Climate & Forests Initiative (NICFI) program, which aim to help protect forest and biodiversity and reduce the impact of climate change 22 . The images used to create the mosaics are acquired from the Planet company, which operates a low-orbit constellation of satellites with a revisit frequency of 1 day. The images are accessible as biannual mosaics from December 2015 to August 2020 and as monthly mosaics from September 2020 onwards. The image mosaics come at a spatial resolution of 4.77 m and have low cloud cover, thanks to the daily acquisition frequency, which has proven useful for forest monitoring over the whole of the tropics. The mosaics consist of four spectral bands, namely blue, green, red, and near-infrared (nir). Since the image mosaics are created from a combination of many sensors and cover various regions in Africa, they have variability in spectral and visual appearance. To correct for this, all the images were normalized such that the range of pixel values is between 0 and 1. This was done for blue, green, red, and near-infra-red bands. Additional vegetation indices such as the normalized difference vegetation index-(NDVI) given as $ndvi=(nir$ − $red)/(nir+red)$ , soil-adjusted vegetation index − (SAVI) given as $savi=(nir-red)/(nir + red +0.5))*(1.5)$ , and the normalized difference moisture index - (NDMI) given as $ndmi=(green $ - $ nir)/(green + nir)$ were created to enhance models’ capability to segregate land-use following deforestation 17 , 21 . The inclusion of these indices was considered following results from the test experiments and literature 17 , 21 . Adding these indices improved the deep learning model performance in identifying land use following deforestation when compared to solely using the typical four bands (blue, green, red and near infra-red) from Planet-NICFI. In addition, in the literature 20 , 32 , 33 , 34 , 35 the NDVI, SAVI, and NDMI are the most widely used indices for forest change detection, crop phenological monitoring, moisture content, and for land cover classification making them the best possible additional bands for this task. Also, since we only have four bands from Planet-NICFI images, these are the best possible uncorrelated indices we can derive from the data.

Data preprocessing

For this study, we created three data pools: (1) a pool of annotated training data, (2) a pool of unannotated training data, and (3) independent test data (Fig. 9 ). In total, we had 2357 images acquired from all across Africa (refer Sect. " Data ", of which only 895 images were having annotations. 80% of the 895 images were placed in a pool of training data, while 20% was left out as independent test data. The remaining 1462 images were placed in a pool of unannotated training data. The unannotated data was later used in Sect. " Models and implementation details " in an active learning cycle where model uncertainty was used to decide which images are useful for improving the model performance. The distribution of land-use classes for initial model training, second, and third active learning iteration is displayed in Fig. 8 .

Shows the number of labels in pixel counts used for the model training in the 1st, 2nd, and 3rd rounds of active learning.

Research design

In this study, we used high-resolution (4.77 m) satellite images from Planet-NICFI to characterize the spectral, temporal and spatial patterns of land-use following deforestation. We used the Hansen forest loss dataset between 2001 and 2020 2 as a proxy for forest loss to estimate land-use following deforestation. Based on this dataset, we selected images $\ge 3.8$ ha in size with detected forest loss in the period covered by the dataset (n = 1821). Over each resulting patch containing forest loss pixels, we extracted Planet-NICFI imagery corresponding to the year 2022 and created reference labels via visual interpretation. We then used these labeled training data and deep learning to predict land-use for 2022 as a proxy for the driver of deforestation from 2001 to 2020. The output was a 5 m resolution classified map of land-use following deforestation.

Additionally, we assessed the spatial hotspots and temporal extent of each of the land-use following deforestation across Africa (30 $^{\circ }$ north and 30 $^{\circ }$ south) based on the output map predicted using the planet-NICFI data.

Models and implementation details

In this study, we adopted the Attention U-Net model developed by 17 for the classification of land-use following deforestation in Ethiopia, since this architecture had provided high performance in identifying drivers of forest loss from satellite data. The model was used to upscale the characterization of land-use after forest loss in Africa in high thematic detail (fifteen classes) and spatial coverage (continental Africa and Madagascar (30 $^{\circ }$ north and 30 $^{\circ }$ south), as opposed to nine land-use classes and the national scale for which the base model was developed. Although the model had shown to perform well in classifying land-use following deforestation at a country scale, the limited availability of training data and heterogeneity of land-uses across Africa would prevent it from being usable at the continental scale. To counteract this, we incorporated AL in the training process to inform data selection and optimize the labeling process required for the model to learn new and informative features from diverse data sets across regions in Africa. We tested the ability of the attention U-Net model to effectively learn from multiple new data-set across different regions in the continent and be applied to different classes from another set of regions. In addition, we performed an independent assessment of the resulting maps of drivers of forest loss in the whole of tropical Africa.

Model details

We trained the attention U-Net model for 200 epochs, with a batch size of 64, using Adam optimization with a learning rate of $10^{-4}$ . In order to help dealing with the class imbalance we used the Focal Loss (FL) 36 ,

to solve the multi-class segmentation problem. Where $p_{i}$ stands for probability of predicting class i, $\gamma $ for Gamma , and $\alpha $ for alpha. We adopt FL as one of the mathematical function or loss function that is used in a deep learning models to calculate the deviation/error of the predicted value from the true value during the model training process with the goal of reducing this deviation by learning the trainable parameters, such as weights and biases 36 . There are many loss functions used in deep learning analyses, however focal loss has shown to perform well on data with imbalanced classes 17 , 36 . During model training the focal loss forces the model to gives more importance to rare/hard classes than majority/easy classes. This is achieved by adding $\gamma $ , and $\alpha $ , and class weights in a function as modulating factor.

We also performed a padding operation after each convolutional layer in order to ensure that the size of the output stays equal to the input. The feature maps between the convolutional layers were normalized using BatchNorm, and dropout with a rate of 0.1 was used to improve the generalization of the model. These models for classifying drivers of forest loss were implemented and run in the Sepal geospatial analysis platform (SEPAL 2.0) 37 , a cloud-based computing environment offered by FAO, with instance type g8, which provides NVIDIA Tesla M60 GPUs with 32GB of VRAM. The model was implemented using the Keras library 38 with TensorFlow 39 as the backend.

Evaluation of model performance

The study presents accuracy metrics based on the F1-score, which is calculated as $F1=2(P*R)/(P+R)$ 21 . P and F stand for precision and recall. The class-wise F1-score measures the model’s capability to identify every single class of land-use following deforestation. The average of all classes F1-scores, also known as macro-averaged F1-score, is used to show the overall classification accuracy of the model. This provides the average of all class-wise F1-score values and compensates for any class imbalance, since it gives the same inportance to all classes, thus increasing the weight of samples from rare classes. While the micro-average F1-score computes the aggregated contribution of all classes by using precision and recall values averaged across all classes, thus putting emphasis on the common classes in the data since it gives each sample the same importance 21 , 40 .

Active learning

One challenge of employing a deep learning methodology is that it tends to require a large amount of data for training 41 , 42 . In reality, there exists a limited amount of training labels to cover the variability of all land-use classes 41 , and the task of labeling all the required land-use data using satellite imagery on a continental scale can be expensive 26 . Indeed, during initial model training using the annotated data described in Sect. " Data ", we achieved an unsatisfactory model classification performance on some of the land-uses on independent datasets, with F1-score of 0.05, 0.1, 0.41, 0.25, and 0.14 for mining, roads, settlement rubber, and cashew, respectively. To be able to improve the classification performance of the model for these land-uses we had to adopt AL 23 , 24 to identify images where the model provides highly uncertain predictions to identify the most informative images and annotate more labels from these images as an addition in the training pool.

Iterative pool-based learning We first train our model using a set of annotated data and assess its accuracy using the independent test data. The first training is done on what is called a pool of annotated data. We then use the trained algorithm to select a set of images from a pool of unannotated data to be annotated by a human annotator. For image selection, we employ entropy

as an uncertainty measure for unlabeled images, not in the training set. $\sum _{i=1}^{k}$ stands for the sum of images’ possible values, given a discrete class membership Y and probability p of class i . Only images with entropy ( $>0.6$ ) were assigned for manual annotation and then later added to the training pool. The process was repeated two times until the best accuracy was obtained for all the classes. In total, 716 labeled images from the training pool were used in the initial training. In the second cycle using AL, we annotated 372 images which were then added to the training pool for second round of model training, followed by 554 images for the third AL training cycle, Figure 9 . The final model performance assessment for each cycle was done on 179 separate independent test data, not in either training pool.

The process of active learning with a pool of annotated data for initial model training, a pool of unannotated data for iterative labeling, and test data for independent testing.

Wall-to-wall mapping

We used FAO and SURFSARA cloud computing platforms 37 , 43 to run inference of the fifteen land-use following deforestation classes across the entire African continent, 30 $^{\circ }$ south, 30 $^{\circ }$ north 22 . The inference was run on 5 m resolution freely available planet-NICFI imagery using the attention U-Net model trained in Sect. " Models and implementation details ". We used planet-NICFI imagery for the year 2022 as a proxy for predicting land-use following deforestation from the year 2000 to 2020. The inference was only applied to forest loss areas identified in 2 .

Assessment of the wall-to-wall map accuracy

To estimate the number of samples needed for evaluating the map of forest loss drivers, we employed a stratified evaluation of area and accuracy based on 44 . The number of samples and accuracy were calculated based on four lustra (2001–2005, 2006–2010, 2011–2015, and 2016–2020) in order to be able to assess temporal variations in the map accuracy and proportion of forest loss drivers. For each stratum (class), sample estimation weights were calculated based on the area estimation of each driver of forest loss for each time period. The calculated weights were utilized to determine the quantity of samples necessary to evaluate the map’s accuracy for each driver over a five-year period. Afterward, the accuracy of the wall-to-wall map was computed through the use of both the user’s and producer’s accuracy measures.

Hotspot analysis of the wall-to-wall map

Hotspot analysis is an interesting visualization technique to get insight into the data that can inform targeted actions, resource allocation and decision-making for forest conservation 45 . We used the kernel density estimation technique (KDE) to estimate the hotspot of the sixteen land-use following deforestation classes predicted in our wall-to-wall map. KDE uses a bandwidth of a specified size to estimate the density of pixels of a land-use within a location to create heatmaps for each land-use. By using the KDE package in python 46 , we run a kernel of size 0.1 $^{\circ }$ × 0.1 $^{\circ }$ to predict the hotspot of each of the fifteen land-use following deforestation classes presented in this study. The output is a smoothed-continuous map of pixel size 0.1 $^{\circ }$ × 0.1 $^{\circ }$ , with each pixel representing the density of a given land-use.

New approach for large-scale mapping of land-use following deforestation

Our results in Sect. " Active learning for improving land-use classification " achieve generalizability at a continental scale and show that it is possible to map land-use following deforestation at a large-scale with high spatial resolution images (5 m) and high thematic detail. In order to bridge across multiple regional reference datasets we applied AL for iterative training, data selection and labeling 23 . The use of AL was fundamental to improve the performance of the deep learning model to classify land-use following deforestation for continental Africa over the traditional model training approaches (see Fig. 1 ). Particularly in cases where the training data is limited, heterogeneous or is prone to labeling errors (i.e. shape correctness, a mix of land uses), the model using the original data is not able to generalize across regions. In such cases AL is shown to be required to achieve higher accuracy in the prediction process 47 , 48 . In our study, the only class that obtains good performance using the original data is cacao, which maintains this high accuracy on the two AL iterations (Fig. 1 ). The reason for this might be attributed to the fact that the reference data for cacaocame from same region (western Africa), which exhibit similar patterns of cultivation, weather and climatic conditions. Thus there is a lower variability than for most other classes, which were retrieved from almost all geographical regions in Africa, with large variations in climate conditions, topography and soil types.

Similar approach and results have been observed in 48 , but only on a small scale applications. Other, similar improvements using AL are also been reported in 49 , 50 , 51 , 52 , 53 , 54 , emphasizing the importance of having human in a loop (AL) when developing models for land use classification task. We hope these model performance gains, coupled with an iterative process of AL, will complement existing visual interpretation methods, and land use mapping tasks, thus accelerating tracking land-use following deforestation at a global scale on a wall-to-wall and more frequently.

Effects of training data quality

After the first training on the original data, (Fig. 1 ), the initial model achieved a macro averaged F1-score of 43%, compared to successive models obtaining 54% and 84% after one and two rounds of addition of training data using the AL approach, respectively. We qualitatively assess the reason for the low performance of the initial model as:

Limited representation of training data from all spatial location In appendix B we show the different data sources used in this study. Here we observe the geographical biases of certain land-uses in the training data. A majority of commodity crops data such as cacao, rubber, cashew were retrieved from datasets in countries from western Africa i.e., Ghana, Ivory Coast, and Nigeria, while data for the remaining land use classes i.e. small-scale croplands, large scale croplands, pasture, oil palm shows a much broader distribution in all regions. Another exception is data for coffee and tea classes which mostly comes from Ethiopia and Kenya.

Errors in training data In appendix C are example tiles showing some level of mismatch of the polygon used in model training. The polygon shows a mix between land-uses, possibly due to temporal mismatches, hence creating confusion for the model during training. These errors are an indication of course delineation of land-use boundaries critical for land-use segmentation task.

Limited labels Lack of enough labels for each class are partially responsible for the low performance of the initial model. From Fig. 8 and appendix B we can also observe the under-representation of some of the land-use classes i.e. coffee, tea, roads, rubber. This problem leads to imbalanced classification results as the model tend to put more weight on the majority classes.

Land-use following deforestation and its implications for forest monitoring

Our finding on wall-to-wall land-use prediction (Sects. " Generating land-use map prediction at scale ", " Proportion of land-use following deforestation by country ") indicates that small-scale cropland is the dominant driver of forest loss in continental Africa. This is similar to findings reported in 4 . A notably high proportion of small-scale cropland was found in Madagascar and DRC. According to 7 an increase in population and political conflicts are the indirect cause of increasing forest loss due to small-scale cropland in DRC, while in Madagascar, is solely associated with population growth along the western coast 55 . Interestingly, small-scale cropland was the only land-use class that showed an increasing trend of change in relation to other land uses per stratum from the year 2001 to 2020. This increase was observed in western, central, and eastern Africa. Conversely, however, when we analyze the trend based on area change, the area of every land-use increased per stratum with the exception of plantation forest and pasture. This is due to the fact that plantation forests are cleared and re-planted on rotational bases, mostly in exact same locations, which may result in false positive detections of forest loss by the algorithms used in 2 . While for pasture, this might be related to confusion with other-land with tree cover during the classification process.

Forest conversion to commodity crops was another important direct driver of forest loss in Africa (Fig. 6 ). Using the kernel density estimation method, we identified distinct hotspot patterns of commodity crops in areas of western and central Africa, specifically cacao, cashew, oil palm, and rubber, with other hotspots for cashew in Tanzania and Mozambique. According to 56 , 57 , the favorable climate condition is the reason for the increasing expansion of commodity crops in these locations. Conversely, land-uses such as tea plantation, coffee, and pasture dominates in east Africa 58 , 59 , 60 . Its proximity to the equator creates favorable conditions for these land-uses 61 . For example, the increase in the establishment of tea and coffee plantations in Rwanda, Kenya, Ethiopia, and Uganda ensures an all-year-round supply of fresh tea to the global markets, which would be challenging in other regions 58 , 59 , 60 , 61 . Previously these commodities were for export; however, the current increase in domestic consumption has created demand within the region, thus the need for more plantations 62 . On the other hand, east Africa is known to host nomadic communities with animal grazing/pasture as their main land-use activity 63 , 64 , 65 . Specifically, Ethiopia, Tanzania, and Kenya are known to have a large number of cattle per household 65 . However, the lack of grazing areas and water due to drought and global warming has forced the pastoral communities to move to forested areas where they can get forage for their cattle as well as an opportunity to diversify their practices by farming (silvopastoral) 64 . Thus putting more pressure on remaining forest. This is commonly observed in Ethiopia, Kenya, and Tanzania.

Another highlight is the number of access roads and settlements detected, mostly in western and central Africa (Fig. 6 ). Growth in the number of roads and settlements in Africa is closely linked to the expansion of agricultural activities 56 , 66 , 67 . Specifically in west Africa, the increase in commodity crops has caused an increase in settlements and roads, which are essential to provide accommodation and accessibility for farming communities 56 . Conversely, in central African countries, a majority of newly developed roads are associated with an increase in logging activities (Congo basin) 68 . However, our analyses also indicate that logging roads disappear with time as a result of abandonment and regrowth. Similar studies report an increasing number of roads in central Africa linked to logging activities 68 . Additionally, mining is most prominent in Ghana, eastern to southern DRC, and Angola, with artisanal mining as the main driver of forest loss along river lines of Ghana and eastern DRC (Fig. 6 ). The presence of mining along rivers has caused not only the loss of forests and wildlife habitat but also a decrease in the quality of water 69 . This poses a health problem to surrounding communities as they become exposed to a toxic chemical used for extracting minerals. Thus a successful forest conservation action would not only save forests but also save communities from health hazards posed by being exposed to mining activities.

Institutional involvement in the expansion of commodity crops

As indicated above, in our analysis, western African countries, specifically Ghana, Ivory Coast, and Liberia, have the greatest rate of forest conversion for commodity crop production. In Western Africa, the growth in the production of commodity crops such as cocoa, cashew, oil palm, and rubber is attributable to unique climatic conditions 56 , 57 . During colonial authority, cashew, cocoa, and rubber were introduced, and seedlings were brought from Latin America 56 . Forest zones were the primary producers of cocoa, rubber, and oil palm, whereas savannas were suitable for groundnut. During and after the fall of the Atlantic slave trade, slave labor played a crucial role in the rise of commodity crop production in many regions of Western Africa 70 .

These commodity crops fared remarkably well in West Africa because millions of smallholder farmers were able to manage their fields using short-term intercropping and intercropping depending on space and soil fertility to respond to the growing demand for commodity crops and poor soils 71 . Several African research and extension institutions, such as the West Africa Agricultural Productivity Program (WAAPP), the Cocoa Research Institute of Nigeria (CRIN), the Rubber Research Institute of Nigeria (RRIN), the Nigerian Institute of Oil Palm Research (NIFOR), and the Ghanaian Ministry of Food and Agriculture (MOFA), have been established in order to address constraints in the production and supply of cash crop seeds and seedlings and to provide credit facilities 71 , 72 . The primary objective of genetic modification initiatives in Africa has been to raise the agricultural yields required by African consumers and producers of commodity crops. The increased cultivation of these crops, however, has come at the loss of tropical forests.

The “commodity crop revolution” brought up new geographical disparities and increased existing ones, which resulted in bigger migratory flows than ever before 70 , 73 . Population migrations have a long pre-colonial history related to slave raids as well as free cyclical migration 74 . Even though many people moved from Nigeria to Ghana, the vast majority of them worked as farmhands in colonial Nigeria. At the time, Nigeria was experiencing a rise in exports of cocoa beans from the southwest, oil palm from the southeast, and groundnuts from parts of the central north. Ghana was one of the primary destinations for these migrants 73 . During the dry season of 1952–53 in Nigeria, some 190,000 migrants were recorded moving southward from the northern part of the country. In addition, research has demonstrated that the cultivation of commodity crops has a significant positive effect on household income by alleviating poverty in many African communities, and this has an effect on household migration decisions in many west African countries, including Burkina Faso, Ghana, the Ivory Coast, and Nigeria 73 , 75 . Due to infrastructure development, rising cash crop output (coffee, cocoa, groundnut), mining sector growth, and the exploration of crude oil, the area has seen an increase in labor migration 74 , 76 . Many people from the Sahel, including some with their families, moved to commodity crop farms in Ghana, Senegal, Côte d’Ivoire, and Nigeria. Ghana and Côte d’Ivoire largely drew Malians, Chadians, Burkinabes, and Nigeriens to their cocoa plantations, while Senegal and the Gambia supplied labour on their cotton and groundnut farms 73 , 76 .

Comparison with related land use following deforestation work

Compared to previous works on land use following deforestation 4 , 7 , 9 , 17 , 77 , 78 , 79 , we advance the land use assessment in multiple aspects. For example, 4 assessed land use following deforestation on a global scale at a 10 km $\times $ 10 km resolution and on six classes, while 9 assessed oil palm plantation at 30 m resolution, and 78 used U-Net model to map four classes of land use following deforestation in Indonesia. The originality of this paper resides on the fact that it goes beyond traditional benchmarking tasks on mapping land use following deforestation 4 , 9 , 78 , 79 , where either a few classes are accounted for, or are produced at a lower resolution than what would make them useful for both large and local scale applications. This study includes more thematic detail than these previous work, with a specific focus on commodity crops, which dominate in most parts of western Africa and have recently received increased attention for targeted conservation and mitigation actions on deforestation 80 .

Furthermore, this study assesses land use following deforestation at higher spatial resolution (5 m) than previous work on a large scale, making it useful for local scale applications, especially on distinguishing between commodity crops and small-scale/large-scale agriculture, where other studies struggle 4 . For example 4 estimated shifting agriculture in Africa to account for 92% of forest loss. However, the study overestimates shifting agriculture as it faces the challenge of separating it from commodity crops. This overestimation may be attributed to the coarse scale (10 km $\times $ 10 km) at which the prediction is done. This leads to commodity crops being predicted as shifting agriculture, especially along the southern arc of western to central Africa. Additionally, the algorithms in 9 , 78 are methodically akin to the U-Net model used in this work and are closely comparable with our model in terms of architecture used. However, their works maps only a few classes, using Landsat satellite images (30 m), thus hindering a direct performance comparison. This work can thus support a more detailed spatial and temporal assessment of where forests are lost, and the land use activities driving this loss. This will allow the targeting of EUDR and REDD+ mitigation efforts towards specific proximate deforestation drivers in order to achieve more impact.

Another study by 79 used a sample-based citizen science approach for identification of land use following deforestation, while 7 , 15 , 27 used a sample-based approach to interpret land use following deforestation visually. One drawback of these approaches is that they are not wall-to-wall, and they may miss out on many deforestation drivers in unsampled locations, reducing their usefulness for deriving global statistics or getting a general overview 7 , 27 . However, these efforts are important for obtaining data useful for developing models used in deriving wall-to-wall maps of land use following deforestation that map every location where deforestation occurred. Specifically, in combination with deep learning and AL, we show that these data sources have great potential to contribute to obtaining maps for monitoring land use following deforestation.

Limitation and future opportunities

In this study, we used high resolution Planet-NICFI images available for the year 2022 as a proxy for mapping land-use following deforestation from the year 2001 to 2020. The Planet-NICFI data was used because of its high performance when used in detecting subtle changes in land use. However, we acknowledge that the detected land uses on images from the year 2022 might not indicate the primary cause of deforestation in earlier deforestation years. In fact, we might be identifying a secondary land use or regrowth. Thus, care must be taken when using these data to compute statistics of land use for forest loss for the years 2001 to 2015, when Planet-NICFI mosaics were not available. For example, areas where logging roads existed in earlier years, say 2001 to 2015, will now have regrown or be covered by trees in 2022 and hence be classified as other-land with tree cover. Likewise, one specific area can be labeled as commercial cropland in 2022, while being small-scale cropland five years before.

Despite this limitation, this study and derived data fall in line with the recent global efforts to stop deforestation, specifically the recent European Union Deforestation Free Regulation (EUDR) aimed to stop the import of all commodities linked to deforestation after December 2019, including cacao, cashew, coffee, oil palm, wood, soy, and cattle 80 . In addition, it provides a potential pathway for the European Union (EU) as it embarks on achieving its carbon neutrality ambitions by 2050, as set out in its European Green Deal Investment Plan 81 . Importantly, this study showcases the potential of the dataset derived using high resolution satellite datasets and deep learning methods to detect and monitor land use following deforestation, specifically commodity crops, which is critical for the success of the EUDR in an open, transparent, and accessible manner. Likewise, the use of deep learning methods provides an opportunity for repeated monitoring of follow-up land uses on forest loss over successive years.

In this work, we chose to use the Hansen forest loss data as a proxy for deforestation because of its ease of accessibility and download, extensive documentation, including error reporting, single global definition of tree cover, global coverage, relatively high accuracy, and consistency since 2001 2 . Specifically, it is the only consistent large-scale forest loss data that captures loss to some extent in parts of dry forest in Africa, which is the most challenging landscape to track deforestation in. According to a study by 82 , the Hansen forest loss data achieves comparable results to national-derived forest loss data when forest definitions are matched. However, it has also been reported that the Hansen forest loss data does have errors 2 , 82 , 83 , 84 , 85 . Milodowski et al. 84 reports that small scale disturbances less than 2 ha could be underestimated by up to 50%. Similarly, the Hansen forest loss data in the tropics is reported to have a 13 % false positive rate and a 17 % false negative rate. In Africa, the lowest accuracy was found in sub-Saharan Africa, with false negatives of up to 48%, indicating an underestimation of forest loss in this region, specifically in dry forests 2 . Thus, care must be taken when interpreting this data. Nonetheless, these errors are within acceptable range and may not impact the task of identifying land use following deforestation; on the contrary, the data are useful to gain valuable insight, especially at regional and global scale 2 . We argue that relying on existing deforestation data makes the task of identifying causes of deforestation more tractable without compromising its applicability 3 , 17 . The recent efforts to integrate different forest loss product provides an opportunity to minimize errors arising from using only one deforestation product 86

Additionally, despite the success of planet-NICFI data in classifying land-use following deforestation in continental Africa. Persistence cloud cover provided challenges in mapping some parts of the Congo basin and West Africa. We acknowledge the existence of misclassification in parts where planet-NICFI images are covered with cloud cover and haze 34 , 87 . The inclusion of synthetic aperture radar data (SAR) offers a complementary advantage in detecting land-use in these areas. SAR has the characteristics of imaging through clouds, haze, day and night. It is expected that future missions of SAR data, such as the National Aeronautics and Space Administration (NASA) and the Indian Space Research Organization (ISRO) SAR (NISAR) expected to be launched in 2023 88 will bridge this gap by providing SAR data continuously in all weather conditions useful for forest monitoring. The NISAR data will be acquired at L-band with a 12-day revisit time, range, and azimuth resolution of 3–10 m and 7 m, respectively. One advantage of L-band SAR is that it can penetrate through a forest canopy, thus useful even in detecting subtle land-uses such as logging roads and artisanal mining in areas where optical data fail to do so due to canopy coverage 89 .

Data availability

The reference data are freely available at the citations reference in this Manuscript and upon request to the corresponding author. The satellite data can be accessed via https://www.planet.com/pulse/nicfi-tropical-forest-basemaps-now-available-in-google-earth-engine/ . For interactive exploration, the predicted output map of land use following deforestation is available at this website: https://robertnag82.users.earthengine.app/view/africalu .

Spracklen, B. D. et al. A global analysis of deforestation in moist tropical forest protected areas. PLoS ONE 10 (12), e014,3886. https://doi.org/10.1371/JOURNAL.PONE.0143886 (2015).

Article CAS Google Scholar

Hansen, M. et al. High-resolution global maps of 21st-century forest cover change. Science 342 (6160), 850–853. https://doi.org/10.1126/science.1244693 (2013).

Article ADS CAS PubMed Google Scholar

Zeng, Z. et al. Highland cropland expansion and forest loss in Southeast Asia in the twenty-first century. Nat. Geosci. 11 (8), 556–562. https://doi.org/10.1038/s41561-018-0166-9 (2018).

Article ADS CAS Google Scholar

Curtis, P. G., Slay, C. M., Harris, N. L., Tyukavina, A. & Hansen, M. C. Classifying drivers of global forest loss. For. Ecol. 1111 (September), 1108–1111 (2018).

Google Scholar

Fritz, S. et al. A continental assessment of the drivers of tropical deforestation with a focus on protected areas. Front. Conserv. Sci. https://doi.org/10.3389/FCOSC.2022.830248 (2022).

Article Google Scholar

Pendrill, F. et al. Disentangling the numbers behind agriculture-driven tropical deforestation. Science https://doi.org/10.1126/SCIENCE.ABM9267 (2022).

Article PubMed Google Scholar

Tyukavina, A. et al. Congo Basin forest loss dominated by increasing smallholder clearing. Sci. Adv. 4 (11), eaat2993 (2018).

Article ADS PubMed PubMed Central Google Scholar

IPCC, Summary for Policymakers. In: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Masson-Delmotte, V., P. Zhai, A. Pirani, S. L. Connors, C. Péan. Tech. rep., Intergovernmental Panel on Climate Change (2021). https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Full_Report.pdf

Descals, A. et al. High-resolution global map of smallholder and industrial closed-canopy oil palm plantations. Earth Syst. Sci. Data 13 , 1211–1231. https://doi.org/10.5194/essd-13-1211-2021 (2021).

Article ADS Google Scholar

Abu, I. O., Szantoi, Z., Brink, A., Robuchon, M. & Thiel, M. Detecting cocoa plantations in Côte d’Ivoire and Ghana and their implications on protected areas. Ecol. Ind. 129 , 107863. https://doi.org/10.1016/J.ECOLIND.2021.107863 (2021).

Fisher, B. African exception to drivers of deforestation. Nat. Geosci. 3 (6), 375–376. https://doi.org/10.1038/ngeo873 (2010).

FAO, State of the World’s Forests 2016. Forests and agriculture: land-use challenges and opportunities. Tech. rep., Food and Agriculture Organization of the United Nation, Rome (2016). http://www.fao.org/3/a-i5588e.pdf

FAO, GLOBAL FOREST RESOURCES ASSESSMENT 2010. Tech. rep., Forestry Department Food and Agriculture Organization of the United Nations, Rome (2010). www.fao.org/forestry/fra http://www.fao.org/3/al501e/al501e.pdf

De Sy, V. et al. Land use patterns and related carbon losses following deforestation in South America. Environ. Res. Lett. 10 (12), 124004. https://doi.org/10.1088/1748-9326/10/12/124004 (2015).

De Sy, V. et al. Tropical deforestation drivers and associated carbon emission factors derived from remote sensing data. Environ. Res. Lett. 14 (9), 094022. https://doi.org/10.1088/1748-9326/ab3dc6 (2019).

FAO, Global Forest Resources Assessment 2020 Main report. Tech. rep., FAO, Rome, Italy (2020). https://doi.org/10.4060/ca9825en .

Masolele, R. N. et al. Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia. GISci. Remote Sens. 59 (1), 1446–1472. https://doi.org/10.1080/15481603.2022.2115619 (2022).

Finer, M. et al. Combating deforestation From satellite to intervention. Science 360 (6395), 1303–1305. https://doi.org/10.1126/science.aat1203 (2018).

Pelletier, C., Webb, G. I. & Petitjean, F. Temporal convolutional neural network for the classification of satellite image time series. Remote Sens. 11 (5), 523 (2019).

Pandey, B., Zhang, Q. & Seto, K. C. Time series analysis of satellite data to characterize multiple land use transitions: A case study of urban growth and agricultural land loss in India. J. Land Use Sci. 13 (3), 221–237. https://doi.org/10.1080/1747423X.2018.1533042 (2018).

Masolele, R. N. et al. Spatial and temporal deep learning methods for deriving land-use following deforestation: A pan-tropical case study using Landsat time series. Remote Sens. Environ. 264 , 112600. https://doi.org/10.1016/J.RSE.2021.112600 (2021).

NICFI. NICFI Program - Satellite Imagery and Monitoring, Planet (2021). https://www.planet.com/nicfi/

Rawat, S. et al. How useful is image-based active learning for plant organ segmentation?. Plant Phenomics https://doi.org/10.34133/2022/9795275 (2022).

Article PubMed PubMed Central Google Scholar

Joshi, A. J., Porikli, F., & Papanikolopoulos, N. Multi-class active learning for image classification. pp. 2372–2379 (2010). https://doi.org/10.1109/CVPR.2009.5206627

UNEP-WCMC and IUCN. Protected Planet: The World Database on Protected Areas (WDPA) and World Database on Other Effective Area-based Conservation Measures (WD-OECM) [Online], November 2022, Cambridge, UK: UNEP-WCMC and IUCN (2022). https://www.protectedplanet.net/region/AF

Tasar, O., Tarabalka, Y. & Alliez, P. Incremental learning for semantic segmentation of large-scale remote sensing data. IEEE J. Sel. Top. Appl. Earth Observ. Remote Sens. 12 (9), 3524–3537. https://doi.org/10.1109/JSTARS.2019.2925416 (2019) arXiv:1810.12448 .

FAO and JRC, Global forest land-use change 1990–2005 (2012).

Sirko, W., et al. Continental-Scale Building Detection from High Resolution Satellite Imagery (2021). https://doi.org/10.48550/arxiv.2107.12283 arXiv:abs/2107.12283v2 .

WRI. Kenya GIS Data, World Resources Institute (2017). https://www.wri.org/data/kenya-gis-data

IPIS. Open Data - IPIS (2022). https://ipisresearch.be/home/maps-data/open-data/

GRID. Welcome - GRID3 (2022). https://grid3.gov.ng/datasets

Verbesselt, J., Hyndman, R., Newnham, G. & Culvenor, D. Detecting trend and seasonal changes in satellite image time series. Remote Sens. Environ. 114 (1), 106–115. https://doi.org/10.1016/j.rse.2009.08.014 (2010).

Sinha, S., Kant Sharma, L. & Singh Nathawat, M. Improved Land-use/Land-cover classification of semi-arid deciduous forest landscape using thermal remote sensing. Egypt. J. Remote Sens. Space Sci. https://doi.org/10.1016/j.ejrs.2015.09.005 (2015).

Reiche, J., Hamunyela, E., Verbesselt, J., Hoekman, D. & Herold, M. Improving near-real time deforestation monitoring in tropical dry forests by combining dense Sentinel-1 time series with Landsat and ALOS-2 PALSAR-2. Remote Sens. Environ. 204 , 147–161. https://doi.org/10.1016/J.RSE.2017.10.034 (2018).

Sousa Da Silva, V. et al. Methodological evaluation of vegetation indexes in land use and land cover (LULC) classification. Geol. Ecol. Landsc. 4 (2), 159–169. https://doi.org/10.1080/24749508.2019.1608409 (2020).

Lin, T. Y., Goyal, P., Girshick, R., He, K. & Dollar, P. Focal loss for dense object detection. IEEE Trans. Pattern Anal. Mach. Intell. 42 (2), 318–327. https://doi.org/10.1109/TPAMI.2018.2858826 (2018).

FAO, SEPAL, a big-data platform for forest and land monitoring (2021). https://www.fao.org/publications/card/en/c/CB2876EN/

F. Chollet, others. Keras (2015). https://github.com/fchollet/keras

Abadi, M., et al. in Proceedings of the 12th USENIX Symposium on Operating Systems Design and Implementation, OSDI 2016 (USENIX Association, 2015), pp. 265–283. https://doi.org/10.5555/3026877.3026899 . https://www.tensorflow.org/

Zhang, C. et al. Joint Deep Learning for land cover and land use classification. Remote Sens. Environ. 221 (May 2018), 173–187. https://doi.org/10.1016/j.rse.2018.11.014 (2019).

Reichstein, M. et al. Deep learning and process understanding for data-driven Earth system science. Nature 566 , 204. https://doi.org/10.1038/s41586-019-0912-1 (2019).

Zhu, X. X. et al. Deep learning in remote sensing: A comprehensive review and list of resources. IEEE Geosci. Remote Sens. Mag. 5 (4), 8–36. https://doi.org/10.1109/MGRS.2017.2762307 (2017).

SURFsara. SURFsara (The Netherlands), ESCAPE (2022). https://projectescape.eu/partners/surfsara-netherlands

Olofsson, P. et al. Good practices for estimating area and assessing accuracy of land change. Remote Sens. Environ. https://doi.org/10.1016/j.rse.2014.02.015 (2014).

Huggins, X. et al. Hotspots for social and ecological impacts from freshwater stress and storage loss. Nat. Commun. 13 (1), 1–11. https://doi.org/10.1038/s41467-022-28029-w (2022).

PyPI. spatial-kde · PyPI (2022). https://pypi.org/project/spatial-kde/

Elmes, A. et al. Accounting for training data error in machine learning applied to earth observations. Remote Sens. 12 , 1034. https://doi.org/10.3390/rs12061034 (2020).

Li, J., Huang, X. & Chang, X. A label-noise robust active learning sample collection method for multi-temporal urban land-cover classification and change analysis. ISPRS J. Photogramm. Remote. Sens. https://doi.org/10.1016/j.isprsjprs.2020.02.022 (2020).

Wang, K., Zhang, D., Li, Y., Zhang, R. & Lin, L. Cost-effective active learning for deep image classification. IEEE Trans. Circuits Syst. Video Technol. 27 (12), 2591–2600. https://doi.org/10.1109/TCSVT.2016.2589879 (2017).

Bengar, J.Z. et al. Class-balanced active learning for image classification. Proceedings - 2022 IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2022 pp. 3707–3716 (2021). https://doi.org/10.48550/arxiv.2110.04543 . arxiv:https://arxiv.org/abs/2110.04543v1

Wu, J. et al. Multi-label active learning for image classification. in 2014 IEEE International Conference on Image Processing , ICIP 2014 pp. 5227–5231 (2014). https://doi.org/10.1109/ICIP.2014.7026058 .

Tuntiwachiratrakun, P. & Vateekul, P. Applying active learning strategy to classify large scale data with imbalanced classes. in 2016 International Conference on Control, Automation and Information Sciences , ICCAIS 2016 pp. 100–105 (2017). https://doi.org/10.1109/ICCAIS.2016.7822443

Ma, T., Ge, J. & Wang, J. Combining active learning and semi-supervised for improving learning performance. ACM Int. Conf. Proc. Ser. https://doi.org/10.1145/2093698.2093871 (2011).

Kim, H. G. et al. Active learning for large-scale object classification: From exploration to exploitation. in HAI 2015 - Proceedings of the 3rd International Conference on Human-Agent Interaction pp. 251–254 (2015). https://doi.org/10.1145/2814940.2814989 .

Moser, C. An economic analysis of deforestation in Madagascar in the 1990s. Environ. Sci. 5 (2), 91–108. https://doi.org/10.1080/15693430801912170 (2008).

Roessler, P., Pengl, Y. I., Marty, R., Titlow, K. S. & van de Walle, N. The cash crop revolution, colonialism and economic reorganization in Africa. World Dev. https://doi.org/10.1016/j.worlddev.2022.105934 (2022).

Jalloh, A. et al. West African agriculture and climate change. International Food policy Research Institute p. 408 (2013). https://doi.org/10.2499/9780896292048 . http://ebrary.ifpri.org/cdm/ref/collection/p15738coll2/id/127444

Karuri, A.N. Adaptation of Small-Scale Tea and Coffee Farmers in Kenya to Climate Change. African Handbook of Climate Change Adaptation (2021). https://doi.org/10.1007/978-3-030-45106-6_70 .

Abdalla Juma, S. & Administration, P. The black tea industry in East Africa: History, culture, trends, and opportunities. J. Econ. Sustain. Dev. 10 (6), 160–170. https://doi.org/10.7176/JESD/10-6-19 (2019).

Hylander, K., Nemomissa, S., Delrue, J. & Enkosa, W. Effects of coffee management on deforestation rates and forest integrity. Conserv. Biol. 27 (5), 1031–1040. https://doi.org/10.1111/COBI.12079 (2013).

Patay, E. B., Bencsik, T. & Papp, N. Phytochemical overview and medicinal importance of Coffea species from the past until now. Asian Pac. J. Trop. Med. 9 (12), 1127–1135. https://doi.org/10.1016/J.APJTM.2016.11.008 (2016).

Article CAS PubMed Google Scholar

Dufrêne, B. Africa Dominates World Tea Exports, While Still Evolving - Tea & Coffee Trade Journal (2019). https://www.teaandcoffee.net/feature/21743/africa-dominates-world-tea-exports-while-still-evolving/

Wynants, M. et al. Drivers of increased soil erosion in East Africa’s agro-pastoral systems: Changing interactions between the social, economic and natural domains. Reg. Environ. Change 19 (7), 1909–1921. https://doi.org/10.1007/S10113-019-01520-9/FIGURES/2 (2019).

Homewood, K., Coast, E. & Thompson, M. In-migrants and exclusion in East African rangelands: Access, tenure and conflict. Africa J. Int. Afr. Ins. 74 (4), 567. https://doi.org/10.2307/3556842 (2004).

Fratkin, E. East African pastoralism in transition: Maasai, Boran, and Rendille cases. Afr. Stud. Rev. 44 (3), 1–25. https://doi.org/10.2307/525591 (2001).

Moriconi-Ebrard, F., Harre, D. & Heinrigs, P. Urbanisation dynamics in West Africa 1950–2010. OECD https://doi.org/10.1787/9789264252233-EN (2016). https://www.oecd-ilibrary.org/development/urbanisation-dynamics-in-west-africa-1950-2010_9789264252233-en

Herrmann, S. M., Brandt, M., Rasmussen, K. & Fensholt, R. Accelerating land cover change in West Africa over four decades as population pressure increased. Commun. Earth Environ. 1 (1), 1–10. https://doi.org/10.1038/s43247-020-00053-y (2020).

Kleinschroth, F., Laporte, N., Laurance, W. F., Goetz, S. J. & Ghazoul, J. Road expansion and persistence in forests of the Congo Basin. Nat. Sustain. 2 (7), 628–634. https://doi.org/10.1038/s41893-019-0310-6 (2019).

Edwards, D. P. et al. Mining and the African Environment. Conserv. Lett. 7 (3), 302–311. https://doi.org/10.1111/CONL.12076 (2014).

Austin, G. Cash crops and freedom: Export agriculture and the decline of slavery in colonial West Africa. Int. Rev. Soc. Hist. 54 (1), 1–37. https://doi.org/10.1017/S0020859009000017 (2009).

Article MathSciNet Google Scholar

Bekunda, M., Sanginga, N. & Woomer, P. L. Restoring soil fertility in sub-Sahara Africa. Adv. Agron. 108 (C), 183–236. https://doi.org/10.1016/S0065-2113(10)08004-1 (2010).

Babu, S. C. et al. Strategies for restructuring the Agricultural Research Council of Nigeria: Process, opportunities, and lessons - IFPRI Publications Repository - IFPRI Knowledge Collections (2017). https://ebrary.ifpri.org/digital/collection/p15738coll2/id/131230

de Haas, M. & Travieso, E. Cash-crop migration systems in East and West Africa: Rise, endurance, decline. migration in Africa pp. 231–255 (2022). https://doi.org/10.4324/9781003225027-16

Adepoju, A. Reflections on international migration and development in sub-Saharan Africa. Afr. Popul. Stud. 25 (2), 298–319. https://doi.org/10.11564/25-2-233 (2011).

Bryceson, D. F. African rural labour, income diversification & livelihood approaches: A long-term development perspective. Rev. Afr. Polit. Econ. 26 (80), 171–189. https://doi.org/10.1080/03056249908704377 (2007).

Quartey, P., Setrana, M. B., & Tagoe, C. A. Migration in West and North Africa and across the Mediterranean - Chapter 21 (2020).

De Sy, V. et al. Synergies of multiple remote sensing data sources for REDD+ monitoring. Curr. Opin. Environ. Sustain. 4 (6), 696–706. https://doi.org/10.1016/j.cosust.2012.09.013 (2012).

Irvin, J. et al. in 34th Conference on Neural Information Processing Systems (Vancouver, 2020), p. 10. https://stanfordmlgroup.github.io/projects/forestnet

Fritz, S. et al. A global dataset of crowdsourced land cover and land use reference data. Sci. Data 4 (1), 1–8. https://doi.org/10.1038/sdata.2017.75 (2017).

EUROPEAN COMMISSION. Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL on the making available on the Union market as well as export from the Union of certain commodities and products associated with deforestation and forest degradation and repealing Regulation (EU) No 995/2010 (2022). https://eur-lex.europa.eu/legal-content/EN/ALL/?uri=CELEX:52021PC0706

EUROPEAN COMMISSION. COMMUNICATION FROM THE COMMISSION :The European Green Deal (2020). https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:52019DC0640&from=EN

Galiatsatos, N. et al. An assessment of global forest change datasets for national forest monitoring and reporting. Remote Sens. 12 (11), 1790. https://doi.org/10.3390/RS12111790 (2020).

Bovolo, I. & Donoghue, D. N. M. Has regional forest loss been underestimated?. Environ. Res. Lett. 12 , 111003. https://doi.org/10.1088/1748-9326/aa9268 (2017).

Milodowski, D. T., Mitchard, T. A. & Williams, M. Forest loss maps from regional satellite monitoring systematically underestimate deforestation in two rapidly changing parts of the Amazon. Environ. Res. Lett. 12 , 94,003. https://doi.org/10.1088/1748-9326/aa7e1e (2017).

Tropek, R. et al. Comment on “High-resolution global maps of 21st-century forest cover change’’. Science https://doi.org/10.1126/science.1248753 (2014).

Berger, A., Schofield, T., Pickens, A., Reiche, J., & Gou, Y. Explore GFW’s New Integrated Deforestation Alerts (2022). https://www.globalforestwatch.org/blog/data-and-research/integrated-deforestation-alerts/

Bae, S. et al. Radar vision in the mapping of forest biodiversity from space. Nat. Commun. 10 (1), 1–10. https://doi.org/10.1038/s41467-019-12737-x (2019).

Rosen, P. A., et al. Global persistent SAR sampling with the NASA-ISRO SAR (NISAR) mission. 2017 IEEE Radar Conference, RadarConf 2017 pp. 0410–0414 (2017). https://doi.org/10.1109/RADAR.2017.7944237

Reiche, J. et al. Forest disturbance alerts for the Congo Basin using Sentinel-1. Environ. Res. Lett. 16 , 24,005. https://doi.org/10.1088/1748-9326/abd0a8 (2021).

Download references

Acknowledgements

We acknowledge the Food and Agriculture Organization of the United Nations (FAO) for providing access to its big data cloud computing infrastructure, known as SEPAL. We also thank SURFsara- the Netherlands national HPC and e-science support center, for providing an extra computational resource for this research.

The research was in part supported by funding from the International Climate Initiative (IKI) of the German Federal Ministry for the Environment, Nature Conservation, Building, and Nuclear Safety (BMUB) 20_III_108. The other funding came from the European Commission Horizon Europe project “Open-Earth- Monitor” (Grant number 101059548), and the CGIAR initiative MITIGATE+. Additional funding was provided through Norway’s Climate and Forest Initiative (NICFI), the US Government’s SilvaCarbon program.

Author information

These authors contributed equally: Diego Marcos, Veronique De Sy and Martin Herold.

Authors and Affiliations

Laboratory of Geo-Information Science and Remote Sensing, Wageningen University and Research, Droevendaalsesteeg 3, 6708, Wageningen, PB, The Netherlands

Robert N. Masolele, Veronique De Sy, Jan Verbesselt & Johannes Reiche

Inria, University of Montpellier, Montpellier, France

Diego Marcos

Department of Remote Sensing, Julius-Maximilians-University, Oswald-külpe-Weg, 97074, Würzburg, Bayern, Germany

Itohan-Osa Abu

GFZ, German GeoResearch Center, Potsdam, Germany

Martin Herold

You can also search for this author in PubMed Google Scholar

Contributions

R.N.M.: Conceived this study, collected the input datasets, developed the method, analyzed the results, and led the writing and editing of the article. D.M.: Helped with the development of the deep learning methods and editing of the article. V.S.: Conceived this study and reviewed the article. I.-O.A.: Contributed to data collection and writing a discussion section. J.V.: Helped with reviewing the article. Johannes Reiche: Helped with reviewing the article following a round of reviewers comments, and provided project funding. M.H.: Conceived this study, provided inputs to the methodology, reviewed the article, and provided project funding.

Corresponding author

Correspondence to Robert N. Masolele .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary information., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Masolele, R.N., Marcos, D., De Sy, V. et al. Mapping the diversity of land uses following deforestation across Africa. Sci Rep 14 , 1681 (2024). https://doi.org/10.1038/s41598-024-52138-9

Download citation

Received : 01 February 2023

Accepted : 14 January 2024

Published : 19 January 2024

DOI : https://doi.org/10.1038/s41598-024-52138-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

Explore articles by subject
Guide to authors
Editorial policies

Sign up for the Nature Briefing: Anthropocene newsletter — what matters in anthropocene research, free to your inbox weekly.

Protecting endangered monkeys from poachers, habitat loss

Scientists make a case for stepped-up biodiversity conservation.

The Tai Forest Monkey Project has operated a research field station in west Africa's Ivory Coast for 30 years, but on the one day since its opening that the site was unstaffed because of conflict in nearby Liberia, poachers took advantage -- and killed 18 endangered monkeys.

The anecdote is a telling example, scientists say, of how thousands of field stations studying primates in forests around the world not only generate knowledge about these threatened species, but also contribute to biodiversity conservation by deterring poaching, deforestation and illegal extraction of natural resources.

The 17 species of red colobus monkey living across Africa, including in the Tai Project field station region, have been singled out by scientists in an April 30 Conservation Letters article as a priority conservation target. Protection of these monkeys, they assert, can be expected to produce benefits throughout tropical African forests where hunting and climate change have led to dramatic wildlife and habitat loss.

W. Scott McGraw, professor and chair of anthropology at The Ohio State University and co-director of the Tai Monkey Project operating out of the field station in Ivory Coast, is a co-author of the article -- as well as of a companion Conservation Letters article published in March. The two papers call for scientific communities, governments and funding agencies to support red colobus conservation efforts and step up financial backing of research field stations.

"One of the primates at our field station is a red colobus species that is endangered," McGraw said. "If we were to stop our work and pull out of the forest, I would not be surprised to see the numbers of all animals just crash. By simply being at the field station, we are protecting what we're studying in addition to generating knowledge. Every one of those species of red colobus in Africa is in danger. And the places where they're protected the most are those that have research field stations."

The authors, from almost 20 institutions in the United States, Europe and Africa, focused on priority action areas that include national and international designation of red colobus as priority conservation species; ecological surveys to determine populations in need of protection; greater investment in protected area creation and management; and engaging with people who live in close proximity to red colobus monkeys.

The monkeys' threatened status has already led to one likely extinction: McGraw was among a team of primatologists who documented the disappearance of a species called Miss Waldron's red colobus in the late 1990s -- the first primate to have gone extinct in 500 years.

"People would say, well, there are lots of other primates -- who cares if we lose one? But we argued that this could be the beginning of a wave of extinction that could make its way across all of Africa," McGraw said. "This was the tip of the iceberg suggesting we were on the edge of catastrophe. What's next? Then the leopards go, then the hippos go.

"As sad as documenting the disappearance of a primate is, we hoped it would be the canary in the coal mine, if you will, and people would pay attention -- for example, to what field stations do."

Broadening understanding of field stations' cost-effective protection of biodiversity resources was the point of the earlier article co-authored by McGraw and 172 other scientists. The paper outlined quantitative estimates of field station-related conservation of wildlife species and forests based on a survey of the leaders of 157 field stations in 56 countries.

The international effort was largely prompted by the COVID-19 pandemic's heavy financial blow to research field stations, many of which haven't recovered. Half of the station directors surveyed reported that in the four years since the lockdown began, their sites have been functioning with less -- or much less -- funding than they had in 2019. About one-quarter have remained partially or completely closed since the pandemic began.

"Most of these field stations are located in biodiversity hotspots, and operate at a relatively low cost," McGraw said. "We work in a well-known national park, but every day we hear gunshots around us. We've collected good evidence over 30 years that the density of animals drops off dramatically immediately outside of our primary study area. So just by being there, we are affording protection."

Most survey respondents reported that the presence of a field station staffed full-time -- by institutional scientists and, in nearly every case, local employees -- reduced hunting and improved enforcement of poaching laws in areas serving as habitats for over 1,200 threatened animal species. The authors estimated that each station has a direct biodiversity impact on almost 50 square miles of habitat around its physical site, and noted that more than half operate with annual budgets of less than $50,000.

Spatial analysis showed that forest cover loss was significantly lower near field stations, showing almost 18% less deforestation overall mainly driven by sites throughout Africa, where field stations' presence is associated with 22% less deforestation.

McGraw and his colleagues have seen the risks to protected lands and field stations' conservation impact firsthand. The worldwide demand for chocolate has made cocoa an extremely valuable plant, and some residents in Ivory Coast -- the largest producer of cocoa beans -- have turned protected areas into illegal cocoa farms and hunting grounds.

Ivory Coast researchers, including McGraw, found that patrolling the grounds of two forest reserves helped reduce illegal activity by well more than half between 2012 and 2016. The team had previously reported, almost 10 years ago, that 75% of land in 23 protected areas had been transformed into cocoa production -- and the loss of forest habitat led to dramatic decreases in primate populations.

"In the eastern part of Ivory Coast, where there should be forest, it looks like a bomb has dropped," McGraw said. "So much of the forest has been cut down, and that which remains is under increasing pressure from us. The human population growth is so steep and is taking place in areas often containing the largest number of threatened taxa. People are, in many cases, eking out a living right next to a forest containing biodiversity that they've grown up with.

"It's a real conservation crisis."

Endangered Animals
Ecology Research
Endangered Plants
New Species
African Wild Dog
Common Chimpanzee
Deforestation
Black Rhinoceros
Global warming
Conservation ethic

Story Source:

Materials provided by Ohio State University . Original written by Emily Caldwell. Note: Content may be edited for style and length.

Journal Reference :

Joshua M. Linder, Drew T. Cronin, Nelson Ting, Ekwoge E. Abwe, Florence Aghomo, Tim R. B. Davenport, Kate M. Detwiler, Gérard Galat, Anh Galat‐Luong, John A. Hart, Rachel A. Ikemeh, Stanislaus M. Kivai, Inza Koné, William Konstant, Deo Kujirakwinja, Barney Long, Fiona Maisels, W. Scott McGraw, Russell A. Mittermeier, Thomas T. Struhsaker. To conserve African tropical forests, invest in the protection of its most endangered group of monkeys, red colobus . Conservation Letters , 2024; DOI: 10.1111/conl.13014

Cite This Page :

Explore More

Far-Reaching Effects of Exercise
Hidden Connections Between Brain and Body
Novel Genetic Plant Regeneration Approach
Early Human Occupation of China
Journey of Inhaled Plastic Particle Pollution
Earth-Like Environment On Ancient Mars
A 'Cosmic Glitch' in Gravity
Time Zones Strongly Influence NBA Results
Climate Change and Mercury Through the Eons
Iconic Horsehead Nebula

Causes of rainforest deforestation in Malaysia

Causes of Rainforest Deforestation in Malaysia

Introduction to malaysia’s rainforest.

Malaysia is located in southeast Asia. The country is split between two areas; Peninsular Malaysia borders Thailand, and East Malaysia is located on the island of Borneo.

At 192,838 km², the Malaysian rainforest is the 24th largest in the world.

What are the rates of rainforest deforestation in Malaysia?

Deforestation is the cutting down of trees, often on a vast scale. Hardwood timber is a high-value export. Once land is cleared of trees, it can be used for other profit-making activities such as cattle ranching, rubber and palm oil production, and commercial farming.

Between 2000 and 2012, Malaysia had the highest rate of deforestation in the world, according to a global forest map developed in partnership with Google. Malaysia’s total forest loss during the period amounted to 14.4 per cent of its 2000 forest cover. The loss translates to 47,278 square kilometres (18,244 square miles), an area larger than Denmark.

Between 2012 and 2015, the rate of tropical rainforest deforestation fluctuated, as shown in the graph below.

Between 2016 and 2020, the rate of tropical rainforest deforestation in Malaysia has steadily declined from 185200 to 73000 hectares per year.

Deforestation rates are reflected in the share of land covered by forests in Malaysia.

Deforestation is occurring on a significant scale in Malaysia. Between 2001 and 2021, Malaysia lost 17 per cent of its rainforest cover. The highest amount of deforestation occurred in the Sarawak region, on the island of Borneo.

What are the causes of rainforest deforestation in Malaysia?

Commercial Farming

Commercial agriculture involves producing food for profit. Economically, one of Malaysia's most important agricultural products is palm oil.

Palm oil is editable vegetable oil from the fruit of the oil palm tree. According to the WWF , "Palm oil is in nearly everything – it's in close to 50% of the packaged products we find in supermarkets, everything from pizza, doughnuts and chocolate, to deodorant, shampoo, toothpaste and lipstick. It's also used in animal feed and as a biofuel in many parts of the world (not in the UK, though!). "

The oil palm tree was introduced to Malaysia by the British and to Indonesia by the Dutch in the mid-1800s and was first planted as an ornamental tree.

The palms start bearing fruit about 30 months after planting and are productive for the next 20 to 30 years. They produce four to 10 times more oil than other vegetable oil crops per unit of cultivated land.

Malaysia is the second-largest producer of palm oil in the world.

Since the 1970s, vast rainforest areas have been cleared and transformed into oil palm plantations, as shown in the graph below. Despite a rapid increase in palm oil production between the 1970s and 2010s, palm oil production in Malaysia has recently levelled off despite the continued worldwide growth in production.

This large-scale, unsustainable method of deforestation has destroyed habitats for animals such as orangutans.

KUCHING, MALAYSIA - MAY 16 2015: Deforestation. Photo of tropical rain forest in Borneo being destroyed to make way for oil palm plantation.

Population Pressure

Between 1956 and 1980, the urban poor were encouraged to migrate to rural areas from the rapidly growing cities of Malaysia. This transmigration policy led to around 15000 hectares of rainforest being felled for the settlers, many of which set up plantations.

Energy Development

Hydropower is poised to play an increasingly important role in meeting Malaysia's energy and climate goals. The share of hydropower in the country's electricity generation is around 11 per cent.

Most of Malaysia's electricity generation capacity is natural gas-fired and coal. Still, the government is seeking to achieve a more balanced portfolio of electricity generation over the coming years to meet its growing demand and reduce its dependency on fossil fuels.

Due to its high rainfall and geography, the state of Sarawak on the island of Borneo is expected to experience the lion's share of new developments.

After five decades of delays, the Bakun Dam in Sarawak started to generate electricity in 2011. At 205m, the dam is the largest in Asia, outside of China. The dam's reservoir flooded over 700km² of forests and farmland. The dam supplies energy to industrialised Peninsular Malaysia.

Bakun Hydroelectric Power Plant in Sarawak, Malaysia

More HEP schemes are planned to boost energy supplies in Malaysia.

Mineral Extraction

Mining for bauxite is common in Peninsular Malaysia. Bauxite is the mineral used in the production of aluminium. Bauxite exports have increased recently, following the end of a ban in 2019.

KUANTAN, Malaysia - Bauxite at a mining site in Bukit Goh

Drilling for oil and gas has recently started in Borneo.

During the 1980s, Malaysia was the world's largest exporter of tropical wood. Clear felling led to the destruction of forest habitats. Selective logging has now replaced clear-felling, where only fully-grown trees are cut down. Ecologically valuable trees are left.

Road Building

Roads have been constructed to provide access to energy projects, new settlements and mining areas. However, few rainforest roads are mapped in Malaysia, so deforestation may be underestimated.

Road construction in the rainforest in Sabah, Malaysia

Subsistence Farming

Subsistence farming involves farmers producing enough food for themselves and their families. Traditionally, local communities would hunt and gather food from the rainforest and grow some crops in small areas of cleared forest. This method of farming is small-scale and sustainable.

Subsistence farming can involve slash and burn, which is the process of using fire to clear the land. The ash provides valuable nutrients to the soil and helps crops grow. However, large areas of rainforests can be destroyed by slash and burn if these fires get out of control.

Find out about the impacts of deforestation in Malaysia .

Premium Resources

Please support internet geography.

If you've found the resources on this page useful please consider making a secure donation via PayPal to support the development of the site. The site is self-funded and your support is really appreciated.

Topic Home

The impacts of rainforest deforestation in malaysia, share this:.

Click to share on Twitter (Opens in new window)
Click to share on Facebook (Opens in new window)
Click to share on Pinterest (Opens in new window)
Click to email a link to a friend (Opens in new window)
Click to share on WhatsApp (Opens in new window)
Click to print (Opens in new window)

If you've found the resources on this site useful please consider making a secure donation via PayPal to support the development of the site. The site is self-funded and your support is really appreciated.

Search Internet Geography

Latest Blog Entries

Pin It on Pinterest

Click to share
Print Friendly

IMAGES

PPT
Case study deforestation and land grabbing in the palm oil sector by
Amazon Rainforest Deforestation Case Study
tropical rainforest deforestation case study
Deforestation: Case Studies
16 Deforestation Facts

COMMENTS

Case Study: The Amazon Rainforest
Case Study: The Amazon Rainforest The Amazon in context. Tropical rainforests are often considered to be the "cradles of biodiversity." Though they cover only about 6% of the Earth's land surface, they are home to over 50% of global biodiversity. ... Many factors contribute to tropical deforestation, but consider this typical set of ...
Deforestation: Case Studies
Deforestation is putting our planet at risk, as the following case studies exemplify. It is responsible for at least 10 per cent of global greenhouse gas emissions 1 and wipes out 137 species of plants, animals and insects every day 2.The deplorable practice degenerates soil, losing half of the world's topsoil over the past 150 years. 3 Deforestation also leads to drought by reducing the ...
Deforestation and Forest Loss
Explore long-term changes in deforestation, and deforestation rates across the world today. Learn how human activities have reduced forest cover by one-third over the past 10,000 years, and how technological innovations may help slow down the trend.
Deforestation, warming flip part of Amazon forest from carbon sink to
A nine-year study reveals that deforestation and climate change have reduced rainfall and increased temperature in eastern Amazonia, turning it from a carbon sink to a source of CO2 emissions. The research, published in Nature, shows that the more deforested and dry regions were net emitters of CO2, while the wetter and intact regions were neutral or balanced.
Study Shows the Impacts of Deforestation and Forest Burning on
A new study, co-authored by a team of researchers including UConn Ecology and Evolutionary Biology researcher Cory Merow provides the first quantitative assessment of how environmental policies on deforestation, along with forest fires and drought, have impacted the diversity of plants and animals in the Amazon.
How to tackle the global deforestation crisis
"On-site observation is very expensive and logistically challenging, and you're talking about case studies. These satellite-based data sets just open up opportunities to see deforestation at scale, systematically, across the globe." Balboni and Olken have now helped write a new paper providing a road map for thinking about this crisis.
Tropical deforestation causes large reductions in observed ...
Previous work has assessed the impacts of tropical deforestation on precipitation, but these efforts have been largely limited to case studies 2. A wider analysis of interactions between ...
Deforestation and world population sustainability: a quantitative
Deforestation. The deforestation of the planet is a fact 2. Between 2000 and 2012, 2.3 million Km 2 of forests around the world were cut down 10 which amounts to 2 × 10 5 Km 2 per year. At this ...
How to tackle the global deforestation crisis
Deforestation is a major contributor to climate change, producing between 6 and 17 percent of global greenhouse gas emissions, according to a 2009 study. Meanwhile, because trees also absorb carbon dioxide, removing it from the atmosphere, they help keep the Earth cooler. And climate change aside, forests protect biodiversity.
The Brazilian Amazon deforestation rate in 2020 is the ...
The 2020 deforestation rate is 182% higher than the established target of 3,925 km 2 and represents a reduction of only 44% instead of the 80% established in law 3. ... a case study for craneflies
Drivers of Deforestation
A global study of palm-driven deforestation found that in Southeast Asia, 45% of oil palm plantations came from areas that were forests in 1989. 29 In Indonesia this was 54% and in Malaysia, 40%. ... This is not the case for biofuels, yet two-thirds of the EU's imported palm oil goes to bioenergy production.
First-of-its-kind study definitively shows that conservation actions
Download images. A new study published online today, April 25, in the scientific journal Science provides the strongest evidence to date that not only is nature conservation successful, but that scaling conservation interventions up would be transformational for halting and reversing biodiversity loss—a crisis that can lead to ecosystem collapses and a planet less able to support life—and ...
New study says conservation works, providing hope for biodiversity efforts
Representation of different broad categories of conservation impact, with illustrative case studies drawn from the study's dataset. Figure from Langhammer et al 2024 .
Deforestation: A Continuous Battle—A Case Study from Central Asia and
In the Central Asia, around 30% population of. rural areas lives near forests and depends on forest products. Studies show a. tremendous increase in deforestation in this region. As of 2006 ...
Case Studies
Case studies. Lessons from practitioners on conserving forests for nature and people. filter. All Approaches. BACK. clear all apply. wwf approach. ... How Argentina could emerge as a leader in mainstreaming beef free from deforestation. The Markets Institute at WWF - 13 July 2021. Using wood forensic science to deter corruption and illegality ...
PDF Deforestation and its Impacts inthe Case-Study Areas
3 Deforestation and its Impacts inthe Case-Study Areas This chapter reports some ofthe findings ofthe field studies that examined processes directly leading to deforestation and the impacts on different social groups. The four case-studies are discussed in separate sections. This is done in order better to suggest inte.ractions and linkages among
Deforestation opens paths for disease spread from animals to humans
Morand was the lead author of a 2020 study of the links between deforestation and human health, taking into account population growth. The research found that increases in outbreaks of zoonotic ...
The Deforestation of the Amazon
Abstract. In this case study, students examine tropical deforestation in the Amazon from the perspective of three dominant stakeholders in the region: a peasant farmer, logger, and environmentalist. As part of the exercise, students perform a cost-benefit analysis of clearing a plot of tropical forest in the Amazon from the perspective of one ...
PDF DEFORESTATION IN INDIA OVERVIEW AND PROPOSED CASE STUDIES Pankaj ...
DEFORESTATION IN INDIA OVERVIEW AND PROPOSED CASE STUDIES Pankaj SEKHSARIA Kalpavriksh - Environment Action Group, India I. INTRODUCTION India is a vast country - encompassing a large canvas of habitats, and ecological niches; rich in bio-diversity and simultaneously supporting a rich, and vibrant diversity of human cultures.
Deforestation reduces rainfall and agricultural revenues in the
Otimizagro is a spatially explicit model that simulates land use, land-use change, forestry, deforestation and regrowth under various scenarios of agricultural land demand and deforestation ...
Deforestation : A Case study in Behali Reserved Forest, Assam, India
From the year 1984 to 2021, a total of 80 sq. km forest cover was lost in the Behali Reserve Forest. The main causes of deforestation in Behali are found to be Illegal logging, Hunting Fuel wood ...
First-of-its-kind study shows that conservation actions are effective
In the Congo Basin, deforestation was 74% lower in logging concessions under a Forest Management Plan ... This effect has been mirrored across a large proportion of the case studies we looked at."
The positive impact of conservation action
The timespan of datasets analyzed in the different studies was highly variable, with the shortest being 1 month and the longest 110 years (median = 4.7, mean = 7.4) . These date back to 1890 and show that studies mainly focused on protected areas until the 1990s but later diversified to a wider range of interventions.
Study: Conservation actions highly effective at halting, reversing
A new study, led and contributed to by Arizona State University faculty, provides the strongest evidence to date that not only is nature conservation successful, but that scaling up conservation interventions would be transformational for halting and reversing biodiversity loss — a crisis that can lead to ecosystem collapse and a planet less able to support life — and reducing the effects ...
Deforestation and its Impacts in the Case-Study Areas
Abstract. This chapter reports some of the findings of the field studies that examined processes directly leading to deforestation and the impacts on different social groups. The four case-studies are discussed in separate sections. This is done in order better to suggest inte.ractions and linkages among processes, policies and more stable ...
Mapping the diversity of land uses following deforestation across
Masolele, R. N. et al. Spatial and temporal deep learning methods for deriving land-use following deforestation: A pan-tropical case study using Landsat time series. Remote Sens. Environ. 264, 112600.
Protecting endangered monkeys from poachers, habitat loss
Scientists make a case for stepped-up biodiversity conservation ... deforestation and illegal extraction of natural resources. ... 2020 — A new study has found that cultivated foods offer ...
Case study- management of the Malaysian rainforest
Case study- management of the Malaysian rainforest Case study - management of rainforests - the Malaysian rainforest Tropical rainforests can be managed in the following ways to reduce deforestation:
Causes of rainforest deforestation in Malaysia
Deforestation is the cutting down of trees, often on a vast scale. Hardwood timber is a high-value export. Once land is cleared of trees, it can be used for other profit-making activities such as cattle ranching, rubber and palm oil production, and commercial farming. Between 2000 and 2012, Malaysia had the highest rate of deforestation in the ...
Is Sustainability More Than A Corporate Buzzword?
Moreover, businesses are joining together under global agreements to solve large-scale problems like deforestation and climate change. The economic case is vital, too.

Case Study: The Amazon Rainforest

The Amazon in context

How does a rainforest work?

What is driving deforestation in the Amazon?

Deforestation: Case Studies

Deforestation case study: Brazil

History of deforestation in Brazil

Brazil’s deforestation of the Amazon rainforest since 2010

The impact of continued deforestation in Brazil

Impact on climate change

Deforestation case study: The Congo Basin

Deforestation in the Democratic Republic of the Congo (DRC)

Drivers of deforestation in the DRC’s Congo Basin

The impact of deforestation in the Congo Basin

Preserve our forests

Related posts:

Related Posts

10 Facts About Deforestation in 2021

Where Is Deforestation Happening in 2021?

The History of Deforestation: The Past and Origins

The Economic Impact of Deforestation

Silviculture and Timber Production

How Does Planting Trees Affect Climate Change?

10 Causes of Deforestation: The Roots of Forest Degradation

How Much CO2 Does a Tree Absorb?

The Effect of Oil Sands Tailings Ponds on Climate Change

Privacy Overview

Deforestation, warming flip part of Amazon forest from carbon sink to source

The study area, which represents about 20 percent of the Amazon basin, has lost 30 percent of its rainforest

No sign of greenhouse gases increases slowing in 2023

5 science wins from the 2023 NOAA Science Report

Filling A Data Gap In The Tropical Pacific To Reveal Daily Air-Sea Interactions

Scientists detail research to assess viability and risks of marine cloud brightening

UConn Today

Study Shows the Impacts of Deforestation and Forest Burning on Biodiversity in the Amazon

Recent Articles

UConn Staff Raises Awareness for Humane Biomedical Research

Researchers Explore Non-Surgical Approaches to Treat TMD Pain, Free Treatment Available

Students From Public Health House Learning Community Feel Their Best Self

How to tackle the global deforestation crisis

MIT News | Massachusetts Institute of Technology

Departments

Centers, Labs, & Programs

How to tackle the global deforestation crisis

*Terms of Use:

Share this news article on:

Related Topics

Related Articles

A clean alternative to one of the world’s most common ingredients

Keeping humanity central to solving climate change

3 Questions: Benjamin Olken on the economic impact of climate change

The economic cost of increased temperatures

More MIT News

Francis Fan Lee, former professor and interdisciplinary speech processing inventor, dies at 96

Fostering research, careers, and community in materials science

Natural language boosts LLM performance in coding, planning, and robotics

Nuno Loureiro named director of MIT’s Plasma Science and Fusion Center

Studies in empathy and analytics

Science communication competition brings research into the real world

Drivers of Deforestation

Brazil and Indonesia account for almost half of tropical deforestation

Beef, soy and palm oil are responsible for 60% of tropical deforestation

We can tackle a lot of deforestation by focusing on a few key supply chains

Is our appetite for soy driving deforestation in the Amazon?

How has demand for soy changed over time?

Which countries produce the most soy?

More than three-quarters of global soy is fed to animals

Is soy production driving deforestation?

The indirect impacts of soy production

Why boycotting palm oil could do more harm than good

Palm oil production has grown to meet rising demands for vegetable oils

Who uses palm oil and what is it used for?

Where is palm oil grown?

How has land use for palm oil changed over time?

Is palm oil responsible for deforestation?

Palm oil versus the alternatives

What would be the impact of substituting palm oil?

How can we use palm oil without destroying tropical forests?

Cite this work

Reuse this work freely