Main

A prominent theme in discussions during the development of the post-2020 Global Biodiversity Framework was how conservation can be undermined by societal issues, such as insecure land tenure, conflict and crime. Although illegal harvesting, collecting and trade are well-recognized threats to biodiversity, we know less about the conservation impacts of other transnational crimes and associated law enforcement activities.

Narco trafficking is one example of a social problem that accelerates forest loss—directly through the illegal construction of roads and landing strips to move shipments and indirectly through both the expansion of cattle pastures to launder money and control territory and the subsidization of land grabbing and resource extraction in previously remote regions, including protected areas and indigenous homelands1,2,3,4. Researchers have used remote sensing to identify unique patterns of deforestation that can be attributed to narco trafficking and found that 15–30% of annual deforestation within Nicaragua, Honduras and Guatemala can be attributed to the movement of cocaine alone4,5. Deforestation hotspots also correspond spatially to trafficking nodes1, although these hotspots are continuously shifting as traffickers adapt to counter-drug interdiction operations and move into increasingly remote, forested areas that conceal clandestine activities. While some of this dynamic may be unavoidable, the supply-side emphasis of domestic and international drug policies amplifies the cycle by neglecting the roles of unclear tenure regimes, weak governance and under-resourced rural communities that facilitate trafficking and promote land-based money laundering and agri-business expansion6,7.

In this Brief Communication, we focus on the intersection of birds, deforestation and counter-drug interdiction efforts to illustrate why responding effectively to the biodiversity crisis requires knowledge of complex socioecological dynamics, without which even well-intentioned efforts to address social problems may exacerbate biodiversity loss. Birds are a salient example because they play important roles in ecological communities, provide ecosystem services and are powerful economic engines via ecotourism and birdwatching. Deforestation represents an serious threat to Central America’s resident and migratory birds, half of which have declined since 19708. Focusing on movement rather than cultivation of drugs, we quantified the spatial congruence of important landscapes for resident and migratory forest birds, as well as forest cover, with lands that are likely to attract cocaine traffickers when displaced elsewhere by counter-narcotic activities.

Our work leverages Magliocca et al.7, who estimated changes in landscape suitability for cocaine trafficking following peaks in interdiction success (that is, cocaine seizures) in Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua and Panama between 2007 and 2018 (Supplementary Fig. 1). We considered interdiction in this context as operations on land (for example, domestic patrols, airport security and highway checkpoints) and sea (for example, interception of boats delivering drugs and refuelling vessels) intended to disrupt supply chains and seize cocaine, as documented in the US Consolidated Counterdrug Database (CCDB)9. ‘Suitability’ was defined as the likelihood that an area will become a future space for drug trafficking after peak counter-narcotic activities based upon socioecological attributes known to attract trafficking activities (for example, isolation and sparse human populations). For example, after intense police–military enforcement, the suitability of an area for future trafficking may change markedly.

Important landscapes (hereafter termed ‘important bird landscapes’ (IBLs); Supplementary Fig. 2) were identified for 196 forest-associated resident and migratory species using high-resolution data products from eBird Status and Trends (ref. 10; Supplementary Information), which estimate weekly relative abundance at 3 km × 3 km resolution based on millions of checklists submitted to the global participatory science project eBird11 and subjected to rigorous quality control and modelling procedures (refs. 12,13; Supplementary Information). We also estimated the most important breeding areas in North America and percentage of global populations of migratory birds using areas that are increasingly susceptible to narco-trafficking activities.

Following peak interdiction activities, suitability for narco trafficking increased within 69% of IBLs for tropical residents (the total area of IBLs is 93,253 km2) and 62% for migratory birds (the area of IBLs is 124,129 km2), whereas only 36% and 30% of IBLs, respectively, declined in suitability (Fig. 1). IBLs that increased in suitability included areas within Mesoamerica’s Five Great Forests—the Maya Forest in Guatemala, Indio-Maiz in Nicaragua, Tortuguero in Costa Rica and the Honduran Moskitia.

Fig. 1: Changes in narco-trafficking suitability in relation to birds and forest.
figure 1

a,b, Spatial overlap between IBLs for 67 migratory forest birds (top) and 129 tropical forest residents (bottom) (a) and differences in 2018 forest cover in areas of increased and decreased suitability (b).

Full size image

Landscapes that were comparably more suitable for narco trafficking had slightly greater forest cover (63.2% versus 57.7%) but encompassed much larger forest patches (Fig. 1). Patch sizes of the 20 largest forests within areas of increased suitability for narco trafficking were over twice as large (mean of 880,242 ha; standard deviation of 1,267,768; 4,421,732 ha max) as those within areas of decreased suitability (mean of 351,772 ha; standard deviation of 511,368; 1,979,304 ha max) (Supplementary Fig. 3).

Over half of migratory species had >25% of their global population—and 11 species, including the federally-listed Golden-cheeked warbler (Setophaga chrysoparia), had >50%—within IBLs that became more suitable for narco trafficking (Supplementary Table 1). Breeding areas of migratory species using areas of increased suitability were most concentrated in the eastern United States and Canada, particularly in the upper Midwest, Appalachians and lower Mississippi River Valley (Fig. 2).

Fig. 2: Connections between breeding and non-breeding areas affected by narco trafficking.
figure 2

The 50% most important breeding areas for 45 migratory bird species that overwinter (non-breeding season) in IBLs (black polygons), which are predicted to increase in suitability for narco trafficking following peak interdiction.

Full size image

Our work builds upon prior research that demonstrates how counter-drug strategies, which previously shifted the drug transit zone from the Caribbean to Central America, continue to push activities into remote forested areas4,7. Here, we provide evidence that current interdiction strategies may increase narco traffickers’ use of roughly two-thirds of IBLs for forest-associated species in Central America and within the largest remaining forests in Central America—the so-called ‘Five Great Forests’ that are disproportionately inhabited by Indigenous people7. This increased risk may have population-level consequences, especially for migratory birds that breed in eastern North America. One in five forest-associated migratory species, including one endangered species, have >50% of their global populations within IBLs expected to increase in narco-trafficking suitability. As narco deforestation is a threat multiplier4,5, the spatial convergence of important areas for birds, forest and narco trafficking represents a credible threat to biodiversity conservation. In this way, our study reinforces the idea that sociopolitical processes—in this case, narco trafficking and interdiction policies—can have complex and unintended negative consequences for the environment.

Acknowledging the challenges and uncertainties associated with estimating clandestine phenomena7, our goal was not to predict future hotspots for narco trafficking but, rather, to expand the scope of potential collateral damages considered as part of drug interdiction strategies and reform alternatives. Despite known problems associated with the ‘supply side’ drug interdiction strategy long championed by the United States14,15, we recognize that no easy solutions exist. Conventional interdiction actions that emphasize seizures and arrests keep traffickers nimble and profits high, which can corrupt government actors and create incentives for the narco enriched to operate through, and transform, landscapes of high-biodiversity value6.

While US drug policy is not solely responsible, supply-side interdiction throughout the region has led to the spatial displacement of narco-trafficking activities. More holistic strategies that establish trust and build capacity in communities and governments to protect their lands are needed for deterring narco traffickers from establishing in new areas. These strategies include supporting governance institutions that can address trafficking-related corruption, enhancing transparency and the rule of law, improving land tenure rights, reducing poverty within Indigenous and rural communities by supporting forest-based income streams, and strengthening the ability of Indigenous and rural communities to protect forests16. Enhancing the ability of Indigenous groups and rural smallholders to re-assert their territorial control and resource governance norms has been shown to be protective against narco trafficking and other environmental crimes, such as illegal logging or wildlife poaching17. These additional components would benefit conservation as well as support peace, security and economic development, including ecotourism. Our research reinforces the Global Biodiversity Framework message that seemingly unrelated social issues can have serious consequences for cultural and biodiversity conservation.

Methods

Using data from Magliocca et al.7, we identified areas where suitability of narco-trafficking activity increased, decreased or remained unchanged before and after peak interdiction (Supplementary Fig. 1), and compared forest cover and patch size among suitability categories based on 2018 land cover (Supplementary Fig. 2). Their analysis of trends in cocaine trafficking through Central America relied on extensive field experience in the region and on data from the US government’s CCDB that includes all known cocaine trafficking events by air or water through the Western Hemisphere Transit Zone, which includes all of Central America9. Although the CCDB is considered the most reliable and unbiased source of narco-trafficking data, it probably underestimates cocaine flow, particularly in regions not routinely surveilled (for example, Nicaragua, which does not provide domestic data to the CCDB but is covered by radar). Changes in suitability were estimated on the basis of variables known to predict use by narco traffickers (proximity to roads, country borders and international ports, indigenous territories, population density and protected areas) and then used to create polygons where suitability increased, decreased or remained unchanged (Supplementary Fig. 1). Few areas showed an absence of any change, which is consistent with the adaptive capacity of traffickers that utilize Central America as a primary transit zone for illicit drugs.

Important landscapes for forest-associated birds during breeding (129 resident species) or non-breeding (67 Neotropical migratory species) seasons were identified using high-resolution data products from the Cornell Lab of Ornithology’s eBird Status and Trends project, which estimates relative abundance of species based on millions of eBird checklists submitted by volunteers11. Raw eBird data are subjected to rigorous data quality procedures12,13 and analysed with an adaptive spatio-temporal exploratory model that includes environmental predictors and variables related to detectability (for example, effort, time of day and observers10). For each species, we calculated the maximum annual percentage of the global population within each 3 km × 3 km pixel across 52 weeks of the year and, from these, we retained the 20% most important pixels for each species within our study area. We then summed percentages across all species in each pixel and retained the 30% most important areas that were ≥50 km2 (that is, IBLs) separately for tropical residents and Nearctic–Neotropical migrants. Using methodology co-developed by Partners In Flight and the Cornell Lab of Ornithology18, we visualized the key breeding areas in North America for migratory birds that overwintered in areas of increased suitability for narco trafficking. We focused on species with >0.5% of their population within the focal non-breeding areas, defined as the area of overlap between increased suitability for narco trafficking and migratory forest bird IBLs (Supplementary Fig. 3). Relative abundance maps for each species during the breeding season were weighted by the per cent of the global population within the focal non-breeding area and then stacked and summed across all species.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.