Introduction #

The Amazon rainforest, of critical importance for global biodiversity and climate regulation, is facing unprecedented threats from deforestation. Nature crime, which can be defined as “illegal activities that cause significant damage to the environment, ecosystems, and biodiversity”¹, is both accelerating this destruction and fueled by it. Most deforestation across the Amazon is illegal, occurring in areas where the removal of forest cover is prohibited or requires an authorization that has not been given or even requested. While deforestation is criminalized in and of itself in most countries of the Amazon, it is often associated with other crimes such as money laundering, trafficking and corruption.

In this report, we provide a comprehensive analysis of the drivers of deforestation within the Amazon’s Protected Areas (PAs) (national, state, and municipal levels) and Indigenous Territories (ITs) (here defined as indigenous lands with a formal title) across the Amazon basin, with a particular focus on the record-breaking year of 2024, and how they relate to illegal behaviors that could indicate potential manifestations of Nature Crime. The focus on PA and IT, where large scale removal of forest cover is generally prohibited, is here used as a proxy for illegality of deforestation. By using newly available data on drivers of deforestation, we reveal how these deforested areas relate to a wide range of illegal activities beyond the mere removal of forest cover and start painting a picture of the extent and distribution of various manifestations of Nature Crime across the region.

For the year 2024 alone, we record a staggering loss of primary forest in both PAs and ITs, with a significant portion attributable to activities linked to Nature Crime. Our analysis reveals that while wildfires, often deliberately set, are a major contributor to forest loss, a substantial and quantifiable area has been cleared for permanent and shifting agriculture, hard commodity extraction (primarily gold mining), and illegal logging. Since these activities are frequently carried out by organized criminal networks, the implications go beyond their environmental impacts, undermining the public safety and governance of Amazonian jurisdictions.

This report quantifies the scale of this destruction, identifying the specific PAs and ITs most affected by these illicit activities. We present a country-by-country breakdown of the data, highlighting the varying patterns of deforestation across the Amazon basin. For instance, while wildfires are the dominant driver in Brazil and Bolivia, agriculture-related deforestation is more prevalent in Colombia. Furthermore, we pinpoint specific hotspots of gold mining deforestation in several countries, providing clear evidence of the encroachment of organized crime into protected and indigenous lands.

Further below, in Section 3, we focus on the strategic state of Mato Grosso in eastern Brazil for a more detailed look at legal vs illegal deforestation events across multiple sectors.

Methodology #

This analysis is based on the latest data from the University of Maryland (UMD), which provides a detailed breakdown of primary forest loss from 2001 to 2024. Our methodology involved an in-depth examination of this data in relation to the geographic boundaries of over 4,200 PAs and ITs across the Amazon biome. To ensure the accuracy of our findings, we further corroborated the UMD data with high-resolution satellite imagery from Planet, allowing for a granular assessment of deforestation drivers on the ground.

Following an initial comparison of the two main Amazon-wide datasets – University of Maryland (featured on Global Forest Watch) and Mapbiomas – we found that only the former was updated to 2024. Given the importance of the 2024 season, and the goal to provide as timely information as possible, we focused on the University of Maryland (UMD) data to provide an initial and timely estimate of Nature crimes for 2024. See Annex I for a more detailed comparison of UMD and Mapbiomas for 2023.

The UMD data indicates loss of primary forest due to both fire and non-fire causes. It extends from 2001 to 2024 and covers the entire Amazon biome and watershed. We analyzed this data in relation to PA and IT (data source RAISG 2024)², yielding a large initial sample size of over 4,200 areas and territories overall.

1. Forest loss in Protected Areas and Indigenous Territories #

Total Area covered by report:

PA total area = 230.9 M ha

IT total area = 199.7 M ha

Broken down by country like this:

Brazil has the most with over 128M ha PAs and 115 M ha ITs

For PAs, Brazil distantly followed by Venezuela (31.1 M ha), Bolivia (23.8M ha), Peru 20.9 M ha)

For ITs, Brazil followed by Peru (32.2 M ha) and Colombia (27.9 M ha

We analyzed entirety of all these areas for primary forest loss

Graph 1.

For this full range of data (2001-2024), we estimate 3.9 million hectares of primary forest loss due to non-fire causes in PA across the Amazon (1.6% of the total protected areas system), plus an additional 2 million hectares from fires (see Table 1). Note that 2024 had the highest totals, by far, for both categories.

Table 1

Similarly, we calculated 2.4 million hectares of primary forest loss due to non-fire causes in IT (1.2% of the total area), plus an additional 2.6 million hectares from fires (see Graph 2). Note that 2024 had the highest totals, by far, for both categories.

Table 2

We then focused on the record-breaking 2024 data for a detailed analysis of drivers across these areas. This data includes 685 PA (national, state, and municipal levels) and 3,605 titled IT.

For the PA, we estimate the primary forest loss caused by other drivers (non-fire) of 390,341 hectares, plus an additional impact of 783,218 hectares from fires. For the IT, we estimate the non-fire primary forest loss of 421,811 hectares, plus an additional impact of 1,102,518 hectares from fires.

For PAs, Bolivia had the most forest loss (175,290 ha), followed by Brazil (172,471 ha).

Similarly for ITs, Bolivia had the most forest loss (229,211 ha), followed by Brazil (105,670 ha) – note more than double.

Standardizing for area, for PAs Bolivia had the highest deforestation rate by far (0.74%).

For ITs, Bolivia also had the highest deforestation rate (1.82%), closely followed by Venezuela (1.68%)

To better focus the analysis on the most urgent cases, we then applied two filters: 1) For the PA, we focused just on the national parks, as the most emblematic category (and most likely to not allow forms of sustainable use that could be misconstrued as manifestations of Nature Crime); and 2) for both the national parks and IT, we focused on those that experienced more than 1,000 hectares of primary forest loss in 2024.

The results from these two filters include 23 national parks (covering 37.9 million hectares) and 86 IT (covering 94.6 million hectares). Of the 23 most impacted parks, 12 were in Brazil, 4 in Bolivia, 3 Colombia, 3 in Venezuela and 1 in Peru (see purple areas in Figure 1). And of the 86 most impacted IT, 43 were in Brazil, 25 in Bolivia, 6 in Guyana, 8 in Venezuela, 3 in Colombia, and 1 in Peru (see purple territories in Figure 2).

Figure 1. Map of National Parks affected by Forest Loss in 2024

Figure 2. Indigenous territories affected by forest loss in 2024

2. Analysis of drivers of deforestation for 2024 #

For the next step of the analysis, we analyzed the new forest loss drivers data from the University of Maryland. This dataset further breaks down the primary forest loss into seven categories: Permanent agriculture, Shifting cultivation, Hard commodities, Logging, Wildfire, Settlements and infrastructure, and Other natural disturbances, that can help identify activities associated with Nature Crime. For analysis purposes, we combine Permanent agriculture and Shifting cultivation into an overall Agriculture category, and Hard commodities represents gold mining. These categories, plus Logging, represent the most direct measures ofNature crimes. In the table below, we detail how each of these categories of drivers of forest loss relate to potential manifestations of Nature crime.

UMD dataset class description	Relation to potential Nature crime
Permanent agriculture: Long-term, permanent tree cover loss for small- to large-scale agriculture.	Permanent agriculture, ranging from small-scale farms to large industrial plantations, is a primary driver of deforestation in the Amazon. When agricultural expansion occurs on land without clear use rights and without the potential approval requirements for vegetation removal, it becomes an illegal activity, passible of administrative or criminal sanctions. Organized criminal groups across the Amazon are known to engage in large-scale land grabbing, illegally clearing vast tracts of forest to establish cattle ranches and agricultural operations. These activities are often used to launder profits from other illicit enterprises, such as drug trafficking and illegal mining. Sources: The ecosystem of environmental crime in the Amazon; Cattle Supply Chains and Deforestation of the Amazon
Hard commodities: Loss due to the establishment or expansion of mining or energy infrastructure.	The extraction of hard commodities, particularly mining for gold, while relatively small in terms of scale of area of forest cover lost, causes wide-ranging environmental and social impacts. it has This sector is heavily infiltrated by organized crime, and operations often involve the use of toxic substances like mercury, which contaminates rivers and soil, poisoning local communities and ecosystems. Criminal groups control mining sites, using violence and intimidation to displace Indigenous communities and exploit workers. The illicitly extracted minerals are then laundered into legal supply chains, making it difficult to trace their criminal origins. The expansion of energy infrastructure, while sometimes legal, can also be associated with Nature crime, as it can open up previously inaccessible areas of the forest to illegal colonization and resource extraction. Sources: Nature Crime Fuels Deforestation in the Amazon; Inside the fight against illegal mining in the Amazon
Shifting cultivation: Tree cover loss due to small- to medium-scale clearing for temporary cultivation that is later abandoned and followed by subsequent regrowth of secondary forest or vegetation.	Small-scale, traditional shifting cultivation by Indigenous and local communities is a sustainable and licit practice within Indigenous Territories. In National Parks with strict conservation status, it is likely to be illegal. In either land categories, shifting cultivation footprint may be associated with illicit crops, such as coca. This makes it challenging to distinguish between traditional, sustainable practices and criminal exploitation of this driver of forest cover loss. Source: The Unstoppable Rise of Environmental Crime
Logging: Forest management and logging activities occurring within managed, natural or semi-natural forests and plantations, often with evidence of forest regrowth or planting in subsequent years.	Illegal logging is a major form of Nature crime in the Amazon, often operating under the guise of legal forest management. Criminal organizations systematically harvest valuable timber from PA, IT, and public lands. This illicitly sourced timber is then frequently laundered into legal supply chains through fraudulent permits and corruption, making its way to domestic and international markets. The logging industry in the Amazon is rife with violence, with criminal groups using intimidation and force against environmental defenders, Indigenous communities, and law enforcement officials who stand in their way. The construction of illegal roads to access remote logging sites further exacerbates deforestation and opens up new frontiers for other criminal activities. Sources: Stolen Amazon: Environmental Crime in the Tri-Border; Illegal Logging, Wildlife Trafficking, and Enforcement Tech
Wildfire: Tree cover loss due to fire with no visible human conversion or agricultural activity afterward. Fires may be started by natural causes (e.g. lightning) or may be related to human activities (accidental or deliberate).	Wildfires do not naturally occur in the Amazon, and most instances of fire are deliberately set by human hand. Arson is widely used to clear large areas of forest for illegal activities such as agriculture, cattle ranching, and land speculation. Organized criminal groups orchestrate these fires to expand their territorial control and eliminate evidence of illegal logging. The use of fire as a weapon of deforestation is a clear manifestation of Nature crime, transforming vast landscapes and contributing to a cycle of environmental destruction and violence. Sources: Nature Crime Fuels Deforestation in the Amazon; Amazon rainforest fires 2022: Facts, causes, and climate impacts
Settlements and infrastructure: Tree cover loss due to expansion and intensification of roads, settlements, urban areas, or built infrastructure (not associated with other classes).	The expansion of settlements and infrastructure, such as roads and urban areas, is often a precursor to and facilitator of Nature crime in the Amazon. While some infrastructure development is legal and planned, illegal road construction is a common tactic used by criminal organizations to gain access to remote, previously inaccessible areas of the forest for illegal logging, mining, and agricultural expansion. These illegal roads act as arteries for environmental destruction, opening up new frontiers for deforestation and creating corridors for the transportation of illicit goods. Source: How the Amazon crisis is reshaping responses to environmental crime
Other natural disturbances: Tree cover loss due to other non-fire natural disturbances (e.g., landslides, insect outbreaks, river meandering). If loss due to natural causes is followed by salvage or sanitation logging, it is classified as logging.	N/A

Results in protected areas #

For the national parks, the vast majority (84.4%) of the forest loss detected was due to escaped human-caused fires (defined as wildfires by UMD). An additional 2.4% was due to natural causes(such as landslides, windstorms, and shifting rivers). Thus, our initial estimate is that 13.2% (31,306 hectares) of the area lost in national parks analyzed (that lost more than 1,000 ha in 2024) is linked to agriculture (permanent and shifting), gold mining, and logging, potential indicators of Nature crime.

For the 23 national parks analyzed in this section, Venezuela had the most area (14.2 M ha), followed by Brazil, Colombia, and Bolivia.

Bolivia had the most forest loss (125,946 ha), mostly due to wildfires. Similarly, most of Brazil’s forest loss was wildfire.In contrast, most of Colombia’s limited forest loss was due to permananet agriculture.

There were important variations by country.

In both Brazil and Bolivia, most of the impact in National parks was due to wildfires (88% and 96%, respectively). The most fire-impacted park by far was Bolivia’s Noel Kempff Mercado (110,000 ha; see Figure 3), followed by Brazil’s Mapinguari and Pacaás Novos and Bolivia’s Isiboro Securé (each with over 10,000 ha).

Figure 3

Other areas in Brazil and Bolivia, however, had >1,000 ha impact from agriculture and gold mining, indicating more direct relationship with potential Nature crimes. For example, Bolivia’s Isiboro Securé and Brazil’s Jamanxim (see Figure 4; note this area also had extensive fires) and Mapinguari each had over 1,000 ha of agriculture related primary forest loss in 2024.

Figure 4

In Colombia, the pattern differed, with less impact from fires (1.8%) and much more on agriculture (87.7%). Three parks had over 1,000 ha impact from Nature crimes (agriculture and logging): Tinigua (Figure 5), Macarena, and Chiribiquete.

Figure 5

In Peru, an important feature of data was highlighted: exonerating areas with high levels of natural forest loss (such as landslides and wind storms). Manu had 1,385 ha primary forest loss in 2024, indicating a major problem in one of the world’s most emblematic parks. But the drivers data revealed that this was mostly natural loss (Figure 6).

Figure 6

Results in indigenous territories #

There were similar findings for the IT in 2024. The vast majority (90%) of this loss was due to escaped fires (wildfires), mostly in Brazil and Bolivia, and an additional 1% was due to natural causes.

Thus, our initial estimate is nearly 9% (119,580 hectares) of deforested area in IT analyzed (that lost more than 1,000 ha in 2024) being potentially linked to Nature crimes involving illegal conversion of natural ecosystems for agriculture (permanent (5.3%) and shifting (0.7%)), gold mining (0.6%), and logging (2%). Specifically, 19 territories had over 1,000 ha agriculture loss and 3 had over 1,000 ha mining loss. For example, the data indicated extensive gold mining deforestation in Sararé Indigenous Territory in Brazil (Figure 7) and agriculture deforestation in  Llanos del Yarí Yaguara II Indigenous Territory in Colombia (Figure 8).

For the 86 ITs analyzed in this section, Brazil had the most area, followed by Venezuela and Bolivia. For both Brazil and Bolivia, most of this loss was wildfire. For Venezuela, it was wildfire followed by permanent agriculture.

Figure 7

Figure 8

Detailed analysis of agriculture and gold mining-related forest loss #

For the final part of the 2024 analysis, we focused on forest loss caused by agriculture and gold mining in the national parks and IT. For agriculture, we lowered the threshold to 250 hectares per area, and for gold mining down to 100 hectares.

In addition, for each of the identified areas above these levels, we confirmed the forest loss and driver using high resolution satellite imagery (Planet). This step proved important as we identified a number of false positives in the data, mostly from fires classified as agriculture.

We confirmed 12 national parks with over 250 hectares of agriculture-related deforestation in 2024 (Figure 9). Most notable were Colombia’s Tinigua and Macarena (see Figure 10 for example of the satellite imagery confirmation), with over 4,000 ha and 3,000 ha, respectively. Other confirmed areas include: Jamanxim (Figure 11), Mapinguari, Serra do Pardo, Juruena, and Serra do Divisor in Brazil; Chiribiquete La Paya, and Picachos in Colombia; and Carrasco in Bolivia.

Figure 9

Figure 10

Figure 11

We determined that forest loss in some park areas, such as this example (figure 12) of windstorm damage in Bolivia’s Isiboro-Secure, were natural, and thus not likely to be associated with Nature crime.

Figure 12

Likewise, we confirmed 38 IT with over 250 hectares of agriculture-related deforestation in 2024 (Figure 13). Deforestation for agricultural purposes is not necessarily illegal in indigenous territories, and this data would require further analysis of the area of single events and type of crops, to more reliably distinguish between industrial scale architecture (likely not authorized within these territories) from small-scale subsistence farming, and to identify potentially illicit crops such as coca. Most notable were Ace Catato, Baure, Monte Verde, Takana 1, Takovo Mora, and Guarayo in Bolivia; Cachoeira Seca in Brazil; and Llanos del Yarí- Yaguara II and Nukak – Maku in Colombia; all with over 1,000 ha each. Other confirmed territories include: Mosetén and Ayopaya in Bolivia; Tenharim Marmelos, Trincheira/Bacajá, Sepoti, Sete de Setembro, and Uru-Eu-Wau-Wau in Brazil, Selva De Mataven in Colombia, Yamanunka in Ecuador; and Colpa, Janabeni, Morroyacu, and Muruinia in Peru.

Figure 13

We also confirmed 6 national parks with over 100 hectares of gold mining deforestation in 2024 (Figure 14). These include Campos Amazônicos (Figure 15), Jamanxim, Juruena, and Rio Novo (Figure 16) in Brazil; and Caura and Canaima in Venezuela.

Figure 14

Figure 15

Figure 16

Likewise, we confirmed 15 IT with over 100 hectares of gold mining deforestation in 2024 (Figure 17). Most notable were Kayapó (Figure 18) and Sararé (Figure 19) in Brazil; and Pemon in Venezuela; with over 1,000 ha each. Other impacted areas include: San José de Karene (Figure 20), Barranco Chico, Boca Inambari, Kotsima, and Tres Islas in Peru; Yanomami, Aripuanã, and Mundurucu in Brazil; Baramita and Malali in Guyana; and Akawayo in Venezuela.

Figure 17

Figure 18

Figure 19

Figure 20

Dominant driver in protected areas and indigenous territories #

Finally, for this section, we present a map of forest loss by dominant driver for each national park meeting the thresholds noted above (250 hectares for agriculture and 100 hectares for gold mining). Note, this data has been corrected by our follow-up analysis of satellite imagery (that is, false positives in the UMD data have been corrected in this map). Note that most in Brazil and Bolivia were related to fire, while in Colombia most were related to agriculture (Figure 21).

Figure 21

Of the 78 National Parks across the Amazon, Fires impacted the most of them (26 parks, 33% of the total number of parks, covering 202,704 ha), followed by Natural Causes (20 parks, 26%, covering 7,009 ha) and Permanent Agriculture (17 parks, 22%, covering 23,433 ha). Rounding out the list is Logging (9 parks, 12%, covering 6,836 ha), Shifting Cultivation (4 parks, 5%, covering 4,335 ha), and Mining (2 parks, 3%, covering 1,214 ha) (Figure 22).

Figure 22

Of the more than 3,000 Indigenous Territories, Shifting Cultivation impacted the most of them (1,243 territories, 41% of the total), followed by Permanent Agriculture (915 territories, 30%). Rounding out the list is Natural Causes (334 territories, 11%), Fires (270 territories, 9%), Logging 209 (territories, 7%), and Mining (39 territories, 1%). Note on the map that fires seem more dominant because they impacted the larger territories in Brazil and Bolivia, but agriculture is more widespread impacting the smaller territories.

3. In-Depth estimate of illegal deforestation in the State of Mato Grosso, Brazil. #

This section builds off a four-year collaboration with the Brazilian organization Instituto Centro de Vida (ICV). During this time, Amazon Conservation led monthly monitoring of major cases of new forest loss detected by GLAD alerts (University of Maryland). All cases confirmed as deforestation by analysis of high-resolution (3 meters) imagery, were then entered into a shared spreadsheet. Also, based on the imagery, each case was classified by the driver. ICV would then classify each case as Legal or Illegal based on analysis of land registry records.

Based on this methodology, we were able to estimate illegal deforestation across the entire state of Mato Grosso, not just within PA and IT. The identification of deforestation drivers evolved over the four years. In 2022, the monitoring focused on soy and cattle. In 2023, a third category of non-soy crops was added. And in 2024, we incorporated gold mining.

During this monitoring (2022-2025), we confirmed 479 cases of deforestation in Mato Grosso, of which 67% were illegal (Figure 23). The area covered by these cases is 131,702 hectares. Most of this area (61%) was illegal (Nature crimes), indicating issues with governance. 39% of that area was lost to legal deforestation. According to the Brazilian forest code, landowners are allowed to deforest a fraction of their properties, provided they request a prior authorization.

Standardizing for just 2024-2025, the two years with data for all four driver categories, we confirmed 336 cases of deforestation, of which 75% (252) were illegal (Table 3). The majority of these illegal cases (70%) were gold mining.

The area covered by these cases is 86,615 hectares. Most of this area (70%) was illegal (60,553 hectares; see Table 3). Breaking down this illegal deforestation, nearly half (49%) of the area were cattle (29,854 ha), 39% non-soy crops (23,783), 10% gold mining (6,053 ha), and 1% soy (862 ha).

Note the distinction between number of cases and area impacted. For example, the highest number of Nature crime cases were illegal gold mining (70%), but the area impacted (10%) is less than the agricultural cases (Table 3).

Table 3

Figure 23

Comparison of two Amazon-wide datasets: UMD and MapBiomas #

For this section, we go back one year to 2023 to compare two leading annual Amazon-wide forest loss datasets: University of Maryland and Mapbiomas.

There are a number of key differences between them. First, as noted above, the University of Maryland updates the annual data faster. This year, UMD released the 2024 data in May 2025, while Mapbiomas has not yet released 2024 (as of October 2025). Second, Maryland records forest loss of primary forest, while Mapbiomas records any type of land use transition.

Third is spatial resolution. While the annual UMD data spatial resolution is 30 meters, their forest loss drivers dataset is coarser at 1,000 meters (1 km). Mapbiomas data, including drivers, is 30 meters.

Fourth, they break down forest loss into similar, but technically different drivers, categories. As noted above, the University of Maryland has seven major categories: Permanent agriculture, Shifting cultivation, Hard commodities, Logging, Wildfire, Settlements and infrastructure, and Other natural disturbances. Mapbiomas also has seven, but somewhat different, categories: Pasture, Agriculture, Palm Oil, Mosaic of Uses, Urban Area, Mining, Other non-Vegetated Areas.

Importantly, note that Mapbiomas does not specifically include the major forms of forest loss most associated with degradation: fires and logging. Also importantly, Mapbiomas does not appear to specifically distinguish natural vs. human-caused forest loss. On the other hand, Mapbiomas does specifically account for cattle pasture relative to other agriculture.

Related to our overall analysis of Nature crimes in protected areas, for University of Maryland we estimate the loss of 163,795 hectares primary forest due to non-fires and an additional 177,732 ha due to fires (for a total for 341,528 ha) across all 674 protected areas. For Mapbiomas, we estimate the transition of 299,837 hectares across all 650 protected areas. Note that Mapbiomas is considerably higher for estimated deforestation (163,795 ha UMD vs 299,837 ha Mapbiomas) but lower if fire is included (341,528 ha UMD vs 299,837 ha Mapbiomas).

As an additional step for Mapbiomas, we estimated primary forest loss (to be more directly comparable to UMD) by filtering data for transitions from the three main forest categories (Forest, Forest Formation, Flooded forest). With this filter, the primary forest loss estimate dropped to 242,982 ha, bringing it a bit closer to non-fire primary forest loss UMD (163,795 ha as noted above).

For Indigenous territories, the University of Maryland data estimates the loss of 178,798 hectares primary forest due to non-fires and an additional 146,088 ha due to fires (for a total for 324,887 ha) across all 3,602 territories. For Mapbiomas, we estimate the transition of 303,832 hectares across 3,496 territories.

Similar to protected areas, note that Mapbiomas is considerably higher for estimated deforestation (178,798 ha UMD vs 303,832 ha Mapbiomas) but lower if fire is included (324,887 ha UMD vs 303,832 ha Mapbiomas).

Analisis 2023, UMD vs Mapbiomas #

Related to our specific analysis of Nature crime drivers, for agriculture, we compare Maryland’s Permanent agriculture and Shifting cultivation layers to Mapbiomas’ Pasture, Agriculture, and Palm Oil layers. Figure 24 illustrates the comparison results. Note that both datasets converge (indicated in red) across most of Brazil and parts of Bolivia and Colombia. But Mapbiomas has more extensive coverage in eastern Brazil, while UMD has more extensive coverage in the western and northeastern Amazon.

Figure 24

For mining, we compare Maryland’s Hard commodities layer to Mapbiomas’ Mining layer (Figure 25).

Figure 25

In 2023 across all protected areas, for Mapbiomas the vast majority of the transitions were related to agriculture (71%), especially pasture (66%). Agriculture is followed by Mosaic of Uses (45%), Mining (1%), and Other non-Vegetated Areas (1%). Applying the primary forest filter, agriculture increased (to 79%), especially pasture (to 73%), while Mosaic of Uses decreased (to 19%). Mining stayed the same (1%).

In 2023, across all Indigenous territories, for Mapbiomas the majority of the transitions were related to agriculture (52%), especially pasture (40%). Agriculture is followed by Mosaic of Uses (26%), Mining (1%), and Other non-Vegetated Areas (1%).

We then narrowed the analysis to national parks with over 1,000 ha impact in 2023. For UMD there were 8 areas, vs 4 areas for Mapbiomas. Only one of these areas overlapped between the two datasets (agriculture in Isiboro Securé). Thus, Mapbiomas missed several areas with permanent agriculture and fire-related loss in Bolivia (Noel Kempff), Brazil (Juruena and Araguaia), and Venezuela (Canaima and Caura). On the other hand, UMD missed three areas with Agriculture and Mosaic transitions in Bolivia (Carrasco) and Ecuador (Cayambe Coca and Sangay).

Similarly, we narrowed the analysis to Indigenous territories with over 1,000 ha impact in 2023. For UMD there were 50 territories, vs 52 territories for Mapbiomas. Eighteen of these areas overlapped between the two datasets (Bolivia, Brazil, and Colombia). Focusing on specific Nature crimes drivers, for agriculture, we found 30 territories with more than 250 hectares impact in UMD vs 40 for Mapbiomas. For mining, interestingly we found 5 territories with more than 100 hectares impact in UMD vs zero for Mapbiomas, putting into question the accuracy of the Mapbiomas mining layer. These included the expected territories of Kayapó, Mundurucu, Sararé, and Yanomami.

Conclusion #

The analysis in this report establishes a quantifiable link between deforestation in the Amazon’s PA and IT and potential manifestations of Nature Crime. Such activities are not isolated events but are often part of broader patterns of land use change driven by organized entities.

The results highlight the necessity for consistent and ongoing monitoring of deforestation drivers and their correlation with various forms of Nature Crime. As the methods employed by those engaged in illicit activities evolve, so too must the analytical approaches used to detect and understand them. Continuous monitoring can provide law enforcement and environmental agencies with the data required to identify emerging trends, predict areas of future forest loss, and support targeted interventions.

By operating across national borders and maintaining institutional continuity, civil society can provide consistent, long-term monitoring necessary to track changes in deforestation patterns and detect emerging threats. Furthermore, civil society’s role in data dissemination and transparency can enhance accountability and ensure that monitoring findings informs policy and enforcement decisions.

In the future, consistent time series that analyze drivers of deforestation and their correlation with Nature Crime across the Amazon could also provide useful insights into the effectiveness of enforcement measures taken to address it.

Notes #

1. WRI, 2025, People, Planet, Justice: Understanding and Countering Nature Crime. Available at : https://www.wri.org/research/people-planet-justice-understanding-and-countering-nature-crime

2. The Amazon Network of Georeferenced Socio-Environmental Information is a consortium of civil society organizations from the Amazon countries, supported by international partners, concerned with the socio-environmental sustainability of Amazonia.https://www.raisg.org/