Category: Global Land Use Change

Drivers of land use change: food, fiber, forests

Whose Amazon?

“Our House Is On Fire”

“Our house is on fire,” declared French President Macron, describing the fires burning across Brazil’s vast interior.  Satellite imagery revealed clouds of smoke from the thousands of fires obscuring large portions of South America, including the skies of Sao Paulo on Monday, August 20.  News outlets described the Amazon forests as the “lungs of the planet,” and articles warned of the Amazon “tipping” from its present forested state to one in which only savannah ecosystems could survive.  Blame for the fires was laid at the feet of President Bolsonaro, whose anti-environmental, pro-development policies were encouraging rampant conversion of forests to agriculture, mining, timber, and cattle operations. 

Fires in Brazil, August 11 2019
Source: https://earthobservatory.nasa.gov/images/145464/fires-in-brazil

President Bolsonaro’s position on the use of Brazil’s natural resources has been clear: Brazil, not the international community will determine their best use.   Concern regarding the ongoing Amazon fires was highlighted at the just-concluded G7 meeting in Biarritz, where President Macron had declared them a “global emergency,” and the G7 agreed to provide funding to fight the fires and aid in reforestation.  (The Guardian, August 26, 2019) However, in a tweet, President Bolsonaro appeared to reject the G7 proposal, asserting that the G7 was treating Brazil as a colonial entity.  A further exchange of tweets between the French and Brazilian leaders ensued,  which did little to ease the situation. 

Is This Fire Season Different?

The fire season in the southern Amazon runs from June to December, with peak burning activity in September (Global Fire Data)  This website has a great deal of information regarding global fire activity, so let’s take a look at some of the data for the Amazon Region (“Legal Amazon”).  Since 2012, VIIRS satellite data has been available along with the older, somewhat less accurate MODIS data. (The VIIRS data has a resolution of about 375 meters, as compared to about 1 kilometer for MODIS.)

Source: http://www.globalfiredata.org

Here are a few highlights:

1.  As of August 31, the 2019 fire season has the highest count since 2012, when VIIRS data became available. 


Screenshot from Total Legal Amazon August Fire Count 2003-2019 showing Cumulative Monthly Fire Count for August. 2019 is in green.

2.  Fires in 2019 are more intense than in previous years, as measured in terms of radiative power.

3.  There has been a noticeable increase in large, intense, and persistent fires burning along major roads in the central Brazilian Amazon, which is more consistent with land clearing than regional drought.  (NASA Earth Observatory)  As an example, the screenshot below is an enlargement of an unprotected area in Para’ shows the clustering of fires adjacent to existing roads in the middle of the image.   Darker areas are unprotected forests, lighter areas above and below the dark green are National Parks.


Source:  (NASA Fire Information for Resource Management System (FIRMS))

4.  However, if we look at 2019 MODIS Fire Alerts, through August 31 for all of Brazil, 2019 (red line) doesn’t look at all unusual as compared to many other fires seasons.    

Source: https://www.globalfiredata.org/forecast.html#totals

5.  Another representation of the historical data reinforces this impression that the 2019 fire season may be well below many other years. 


Source:  Global Forest Watch Fires Brazil

6.  Fire alerts in Intact Forest Landscape Areas appear to include only 6% of the impacted areas.

It’s Too Early To Draw Conclusions

There is no doubt that August 2019 has seen an historically high number of fires in the Amazon,  but we will not have the full picture until the end of the fire season, when satellite imagery can be compared to pre-2019 data to determine the precise location and true extent of the fires.  The degree to which previously intact tropical forest or other threatened biomes have been transformed by fire won’t be known until this type of analysis can be made.  Satellite imagery clearly shows many fires both within and adjacent to Brazilian National Parks.  For example, the screenshot below (NASA Fire Information for Resource Management System (FIRMS)) shows an area of Para’ with numerous fires in the dark greenish black (unprotected) areas as well as in the protected (lighter green) areas. 

Dark green areas denote unprotected lands, lighter green are National Parks. Numerous fire counts are present in both areas

However, cumulative monthly fire counts (January-August 31) for 2019 in Para’ are well below many other years (next figure), a further indication that it is simply too soon to draw conclusions and issue condemnations about the overall extent of fire damage.  


Source:  Global Fire Data Para’ (accessed August 31, 2019)

Global Demand Drives Local Change

To return to President Bolsonaro’s assertion that the disposition of Brazil’s forest resources  are a Brazilian, not international issue, this is a much more complicated issue than the President’s statement would indicate.  Surging global demand for soy has been met by Brazil, Argentina and the United States. As of 2018, it is likely that Brazil will surpass the United States as both the largest producer and exporter of soy.(TRASE Yearbook 2018)  Brazil has produced soy first by converting vast undeveloped subtropical regions in its south, then into tropical areas, into the Cerrado (largest savanna region in South America, largely unprotected) in the mid-1990’s, and now into the agricultural frontier area of Matopiba.  The conversion of undisturbed forests and other biomes in Brazil to the production of soy as well as other agricultural products (e.g., sugarcane, beef, timber) has been well-documented and ongoing for many years. Soybean exports are now valued at over USD 20 billion, making them Brazil’s most valuable export commodity. (TRASE Yearbook 2018)

An estimated 1.8 million ha of soy in the Amazon in 2016 and 3.5 million ha of soy in the Cerrado in 2015 were undeveloped in the year 2000—amounting to about 40% and 20% of the total area of soy in each biome (TRASE Yearbook 2018 Chapter 3)

 A complex network of producers, export and import entities links local land use change across Brazil to global consumers. The screenshots from the website (TRASE) illustrates some of these linkages.  Soy is used as feed for pigs and chickens, and is exported in vast quantities to China, the world’s largest producer and consumer of pork. (Brazil is also the world’s largest exporter of chickens.) The pig population in China is estimated to be nearly 500,000,000 and China doesn’t have the land to supply soy for this plethora of pork.  Instead, it has reduced the amount of land planted to soy and become the world’s largest consumer of soy (around 60% of global exports), primarily from Brazil. 

Soy flows from Brazil’s biomes through a complex network of export/import entities to end users across the globe
Source: TRASE
Soy flows from states across Brazil to many countries, but China is by far the largest end-user
Source: TRASE

This dependence on Brazilian soy will likely increase due to the ongoing and escalating trade war between the United States and China.  Prior to the imposition of tariffs, Chinese soy demand had also been met by the United States, but the tariff war is likely to incentive the Brazilians to increase soy production, as China shifts from America to Brazil to meet its soy needs.  In a grim analysis of the possible deforestation consequences of such a shift, a report in Nature (Trade War Disaster for the Amazon) in March estimated that “soya-bean production in Brazil could increase by up to 39%, to 13 million hectares.”   Following the historical pattern of Brazilian soy production, the ready availability undeveloped land will lead to agricultural extensification, rather than intensification. 

U.S. Farmers Expand Production

Just as their Brazilian counterparts, American farmers respond to global and domestic demands agricultural commodities by expanding production. They have planted more soy for export, and they have planted more corn in response to biofuel mandates by the federal government.  This has come largely at the expense of previously intact grasslands.*  In the 8 year period between 2008 and 2016, 10 million acres (4,047,000 ha) of grassland, shrubland, wetland and forestland were converted to crop production in the United States, more than half of which was planted in corn and soy.    80% of new cropland came from grassland ecosystems, of which 2.2 million acres were intact grasslands, defined as “those which had not been previously planted or plowed and are most likely to contain native species and sod.”  The rate of land conversion has continued at nearly 1 million acres per year. 

The conversion of grassland between 2008-2012 released more than 14 million metric tons of carbon per year—equivalent to yearly emissions from 13 coal-fired power plants.

This extensification of agricultural production has occurred in the Dakotas, Iowa, Missouri, Kansas, Oklahoma, and Texas, Kentucky and Tennessee, as well as areas bordering Canada in the Northern Great Plains.  As in the case of Brazil, extensification has converted previously intact ecosystems, which provide valuable environmental services, including protection of water quality, critical habitat for bird species, pollination, prevention of soil and nutrient loss, and carbon sequestration.  In the case of grassland ecosystems as well as tropical forests, carbon sequestration is particularly important as a means of buffering continued accumulation of anthropogenically sourced carbon dioxide in the atmosphere. 

*Source: Gibbs Lab

A Telecoupled World

Conversion of intact biomes to agroecosystems is not unique to Brazil or the United States; agriculture occupies about 38% of Earth’s terrestrial surface, making it the largest use of land on the planet (Solutions for A Cultivated Planet).  Flows of energy, resources, information, etc., couple human socioeconomic systems and environmental systems, forming a telecoupled system (Framing Sustainability in a Telecoupled World, one of the hallmarks of the Anthropocene. The relationship between China and Brazil exemplifies this system, driving both extensification and intensification of soy production in Brazil, as vast quantities of soy product flow back to China.  Smaller quantities flow to many other nations, including members of the G7. Through the work of TRASE researchers, the linkages between import/export entities and deforestation have been brought into the open, and it has become clear that a handful of enormous, largely privately held companies dominate these flows.  Land use decisions in Brazil are thus determined by both Brazilian governmental decisions, as well as those of these often vertically integrated transnational agricultural entities. 

In a telecoupled world, it is increasingly difficult to disentangle local land use decisions from global economic forces.  Thus, President Bolsinaro’s claim that Brazilian resources are to be disposed of only by Brazil, is not really that simple. Brazilian resource decisions can be influenced by end-users, mediated by a very complex interplay of actors. Ultimately making a transition to sustainability in the Amazon and elsewhere will be very challenging.  For example, despite the much larger volume of soy exports from Brazil to China, the sourcing of soy from Brazil to Europe actually exposes European nations to higher deforestation risk than China (TRASE 2018 Annual Report)

We Are All Complicit

In a telecoupled world of nearly 8 billion, conversion of vast ecosystems matters in ways that weren’t apparent in earlier eras.  In the plow up of the Great Plains grassland of the United States in the 19th and early 20th century, the near extinction of the buffalo, decimation and relocation of indigenous peoples wasn’t an issue of global concern.  Now, when Brazil is treating its vast frontier regions in much the same fashion as did the United States, it does matter. 

We in the developed world still have our hands dirty; be it grassland conversion in the United States, deforestation of boreal forests in Canada, destruction of the ancient Hambach Forest in Germany for production of lignite—one of the dirtiest of coals.  Why should the Brazilians listen to us? 

Moreover, why should the Brazilians change their behavior?  Perhaps the Chinese should reduce their pork consumption, the Europeans reduce their intake of beef, Americans change their toilet paper purchases from Canadian-sourced pulp to recycled?  In other words, we have outsourced our resource demands from domestic to foreign sources, but want these resources to be extracted on our terms—something we aren’t even doing ourselves.  Can we really have it both ways in a telecoupled world?

President Macron condemns the Brazilians for burning their (“our”) forests.  Who, exactly is lighting the match?

Image of Match: yaoqi-lai-7iatBuqFvY0-unsplash.jpg

Global Land Use Change

A human-dominated Earth System

(Source unless otherwise noted for “Summary of findings” #1-10:   Song, X.-P., Hansen, M. C., Stehman, S. V., Potapov, P. V., Tyukavina, A., Vermote, E. F., & Townshend, J. R. (2018). Global land change from 1982 to 2016. Nature. doi:10.1038/s41586-018-0411-9) 

Globally, land-use change is the direct driver with the largest relative impact on terrestrial and freshwater ecosystems. {2.2.6.2}.  (IPBES)

Land-use change is driven primarily by agriculture, forestry and urbanization, all of which are associated with air, water and soil pollution. {2.1.11} (IPBES)

Summary of Findings

Direct human action on landscapes is found over large areas on every continent, from intensification and extensification of agriculture to increases in forestry and urban land uses, with implications for the maintenance of ecosystem services.

1.  Global tree cover has increased 7.1% relative to 1982

2.  Net gain is the result of greater tree cover gain in the subtropical, temperate and boreal climate zones relative to net loss in the tropics

3.  Global bare ground cover has decreased by 3.1%, primarily in agricultural regions in Asia

4.  60% of all land changes are associated with direct human activities; 40% with indirect drivers such as climate change

            a.  Direct human activities–36% for bare ground

            b.  Direct human activities–70% for tree canopy loss

            c.  Direct human activities—86% of changes in Europe, 66% South America, 62% Asia, 50% Africa

5.  Across all climate domains:  montane systems have gained tree cover, many arid and semi-arid ecosystems have lost vegetation cover

6.  On a global scale, the growth of urban areas accounts for a small fraction of all land changes.

7Changes were unevenly distributed across biomes

            a.  Tree canopy loss (relative to 1982):  tropical dry forest biome

            (-8%), moist deciduous forest (-2%)

            b.  Tree canopy gains:  temperate continental forest (+33%), boreal coniferous forest    (+12%), subtropical humid forest (+18%) 

            c.  Short vegetation loss:  temperate continental forest (-14%), boreal coniferous forest (-10%), subtropical humid forest (-9%)

            d.  Short vegetation gain:  tropical shrubland (+10%), tropical dry forest (+5%)

            e.  Bare ground loss:  tropical shrubland (-10%),

            f.  Bare ground gain:  subtropical desert (+4%), subtropical steppe (+5%)

Expansion of the agricultural frontier

Nearly 40% of the terrestrial environment is devoted to agriculture. 

Source:  World Bank Data Agriculture

Source:  World Bank Data Agriculture

8.  Expansion of the agricultural frontier is the primary driver of deforestation in the tropics

            a.  South America—The three countries with the largest area of net tree cover loss during 1982–2016 are all located in South America: Brazil (−385,000 km2, −8%), Argentina (−113,000 km2, −25%) and Paraguay (−79,000 km2, −34%)

                  (i.)  Brazil: Clearing for industrial agriculture in the Cerrado and Gran Chaco; largest gain in short vegetation (+12%)  mainly due to expansion of agricultural frontiers into natural ecosystems

Tree cover loss 2001-2008 Brazil
Global Forest Watch. 2014. World Resources Institute. Accessed on (July 24, 2019). www.globalforestwatch.org.
Location of the Cerrado; deforestation 2002-2008
Source: https://commons.wikimedia.org/wiki/File:Desmatamento_no_Brazil,_por_bioma,_de_2002_a_2008..jpg
Português: Mapa do Desmatamento no Brasil, de 2002 a 2008. Fontes: Prodes (INPE) e Monitoramento por Biomas (IBAMA). Obs: O monitoramento não cobre as áreas de vegetação de Cerrado e Campinarama localizadas no Bioma Amazônico. Autor: Vitor Vieira Vasconcelos (2012)Location of the Cerrado; deforestation 2002-2008

Deforestation in Gran Chaco dry forest.  Landsat 8 image August 14, 2016.  About 20% (55,000 square miles) of the Gran Chaco has been deforested since 1985.  17,000 square miles of forests have been lost in Paraguay between 1987-2012.  NASA Earth Observatory image by Michael Taylor, using Landsat data from the USGS.
Inset showing location of Gran Chaco
https://commons.wikimedia.org/wiki/File:Desmatamento_no_Brazil,_por_bioma,_de_2002_a_2008..jpg

(ii) Argentina and Paraguay

Tree cover loss in Argentina and Paraguay 2001-2018

Global Forest Watch. 2014. World Resources Institute. Accessed on (July 24, 2019). www.globalforestwatch.org.

8b. Queensland, Australia.

In 2010, Queensland had 9.06Mha of natural forest, extending     over 5.3% of its land area. In 2018, it lost 21.5kha of natural forest. Global Forest Watch. 2014. World Resources Institute. Accessed on (July 24, 2019). www.globalforestwatch.org

Global Forest Watch. 2014. World Resources Institute. Accessed on (July 24, 2019). www.globalforestwatch.org.
Tree Loss Australia, inset showing Queensland

8c. Southeast Asia Source: Global Forest Watch. 2014. World Resources Institute. Accessed on (July 24, 2019). www.globalforestwatch.org

Southeast Asia
(i) Myanmar
Myanmar Tree Cover Loss 2001-2018
(ii) Vietnam
(iii) Indonesia
  • In 2010, Indonesia had 137Mha of natural forest, extending over 73% of its land area.
  • In 2018, it lost 1.22Mha of tree cover, equivalent to 480Mt of CO₂ of emissions. 340kha of this loss occurred within primary forests and 907kha within natural forest.
  • Indonesia lost 6 Mha of old-growth and selectively logged natural forests between 2000 and 2012, and surpassed Brazil in the rate of its forest clearance in 20121
  • Indonesia is the world’s largest palm oil producer; 80% of its palm oil is exported, of which 66% is shipped to India, China, Pakistan, Malaysia, Italy, Egypt and the United2 Kingdom.
Tree Cover Loss in Indonesia 2001-2018
Tree Cover Loss in Indonesia 2001-2018

1 Gaveau, et al., Rapid conversions and avoided deforestation:  examining four decades of industrial plantation expansion in Borneo, Scientific Reports 6(32017), 2016.

2 Rulli, et al., Interdependencies and telecoupling of oil palm expansion at the expense of Indonesian rainforest, Renewable and Sustainable Energy Reviews,Volume105,May2019,499-512. https://doi.org/10.1016/j.rser.2018.12.050

(iv) Borneo (https://www.cifor.org/map/atlas/)

Borneo, Earth’s third largest island with 73.7 million ha (Mha) feeds the world with palm oil, a multi-billion-dollar business encompassing cosmetics, processed food and biofuels. With 8.3 Mha of industrial oil palm plantations, about half of the estimated global planted area of 18 Mha, Borneo is perhaps the world‘s largest center of palm-oil production. Pulpwood plantations (1.3 Mha) – mainly fast growing Acacia and Eucalyptus – make a major contribution to the global production of wood pulp.

Borneo Forested Area 1973

Borneo is shared by Indonesia, Malaysia, and Brunei. It has the largest deforestation rates in the world, with an average 350,000 ha cleared every year between 2001 and 2016.


Borneo Forested Area 2016

76%, or 55.8 million hectares, of Borneo was old-growth rainforest (green) in 1973. Old-growth forest ecosystems are intact and include many old (>500 years) closed-canopy evergreen trees.

About 18.7 Mha of old-growth forest area was destroyed between 1973 and 2015 by fire, agricultural expansion, mining and hydropower dams. By 2015, 50% of the island shared by Malaysia, Indonesia and Brunei remained forested, with 28% old-growth rainforest, 22% logged-over (still in good condition), and 12% covered in industrial plantations (9.2 Mha).

Borneo, Indonesian Borneo in particular, suffered large-scale forest loss prior to the expansion of Industrial plantations. This cleared land permitted the development of some large-scale industrial plantations without necessarily causing additional forest loss. Nonetheless, in the last decade plantations have become the primary cause of direct deforestation.

3 Gaveau, et al., Rapid conversions and avoided deforestation:  examining four decades of industrial plantation expansion in Borneo, Scientific Reports 6(32017), 2016

8d. Sub-Saharan Africa: smallholder agriculture and commodity cultivation
(i) Congolian rainforest
  • (second largest tropical rainforest in the world after Amazonia*)
  • Unpaved logging roads used by timber firms, as well as paved and unpaved public roads, have expanded greatly. Comparing old (before 2003) and new (2003–2018) road datasets derived from Landsat imagery, … the total length of road net-works inside logging concessions in Central Africa has doubled since 2003, whereas the total length of roads outside concessions has increased by 40%.*
  • Annual deforestation rates between 2000 and 2017 near (within 1 km) roads increased markedly and were highest for old roads, lowest for abandoned roads and generally higher outside logging concessions.*

The impact of logging on deforestation is partially ameliorated by the nearly fourfold higher rate of road abandonment inside con- cessions, but the overall expansion of

logging roads in the Congo Basin is of broad concern for forest ecosystems, carbon storage and wildlife vulnerable to hunting.*

*Source:  Kleinschroth, et al., Road expansion and persistence in forests of the Congo Basin, Nature Sustainability, Volume 2, July 2019, 628-634

Impacts from climate change

9a.  Western United States:  forest impacts due to insects, wildfires, heat, drought

9b.  Arctic:  warming is facilitating woody vegetation growth in northeastern Siberia,                      western Alaska and northern Quebec

9c.  Central and West Africa:  forest expansion and woody encroachment probably due                   to increased CO2 and precipitation

9d.  Sahel:  greening due extreme high-rainfall anomalies

9e.  Altitudinal biome shifts:  global treeline positions have been advancing since AD 1900

9f.   Montane biomes:  overall bare ground loss, short vegetation loss and bare canopy gain

Political, social, economic factors

10a.  Europe and European Russia:  tree canopy increase of 35% after the fall of the Soviet Union

10b.  China:  +34%  tree canopy gain due to reforestation/afforestation; second largest loss of bare ground due to resource extraction and urban sprawl

10c.  United States:  +15% tree canopy gain mostly in the eastern US

10d.  India:  largest bare ground loss of all countries (-34%); second in short vegetation gain (+9%) primarily due to intensification of existing agricultural lands—continuation of the “Green Revolution.”

Arid and semi-arid drylands:  long-term land degradation

11a.  Large areas of decrease in short vegetation and large areas of increase in bare ground    

11b.  Hotspots of vegetation loss include the southwestern United States, southern Argentina, Kazakhstan, Mongolia, Inner Mongolia, China, Afghanistan and large areas of Australia.

(i.)  Australia:  decrease in short vegetation:  long-term decline in local growing season

(ii.)  Mongolian steppe:  Rising surface temperatures, a reduction in rainfall, and overgrazing caused extensive grassland deterioration

(iii.)  United States (southwest):  degradation of soils and vegetation combined with an increased dominance of invasive species

Intact Forest Landscapes (IFLs)

Source:  P. Potapov, M. C. Hansen, L. Laestadius, S. Turubanova, A. Yaroshenko, C. Thies, W. Smith, I. Zhuravleva, A. Komarova, S. Minnemeyer, E. Esipova, The last frontiers of wilderness: Tracking loss of intact forest landscapes from 2000 to 2013. Sci. Adv. 3, e1600821 (2017).

Definition:  “(I)ntact forest landscape (IFL) (are) a seamless mosaic of forests and associated natural treeless ecosystems that exhibit no remotely detected signs of human activity or habitat fragmentation and are large enough to maintain all native biological diversity, including viable populations of wide-ranging species.”

Signficance:  These areas have the highest conservation value in terms of the range of ecosystem services they provide, including harboring biological diversity, stabilizing terrestrial carbon storage*, regulating hydrological regimes.  The size of these areas and degree of intactness, lack of degradation and fragmentation are critical to their function and conservation value. 

*Although the remaining IFLs comprise only 20% of tropical forest area, they account for 40% of the total aboveground tropical forest carbon.

Status (2000-2013) (See Figure 1 and Table 1)

  • Between 2000-2013 global IFL area decreased by 7.2%
  • Global IFL loss is driven by large losses in tropical South America and Africa, accounting for 60% of global loss
  • Drivers of IFL loss are driven (in order of importance) by industrial timber extraction, agricultural expansion (South America), forest fires, energy production (oil and gas extraction and hydropower)
  • Fragmentation of IFLs by logging and establishment of roads and other infrastructure initiates a cascade of changes that lead to landscape transformation and loss of conservation values.
  • Fragmentation of IFLs by logging and establishment of roads and other infrastructure initiates a cascade of changes that lead to landscape transformation and loss of conservation values.