Reforestation is gaining global momentum as a climate solution, but scientists warn that planting trees indiscriminately isn’t effective. Its success depends on how and, crucially, where it’s done.

A new study in Nature Communications, co-authored by scientists from The Nature Conservancy (TNC) and other institutions, seeks to advance global reforestation efforts by mapping the locations where tree planting and forest regrowth are most likely to deliver climate, biodiversity and community benefits. The authors compiled their findings into a web-based tool called the Reforestation Hub.

“Reforestation is a readily-available, scalable, cost-effective carbon removal solution, but we have neither time nor resources to restore trees everywhere,” said study co-author Susan Cook-Patton, lead reforestation scientist at TNC. “Our goal with this research was to help accelerate action by highlighting the places where reforestation holds the greatest promise for people, nature, and climate,” she added in an email interview that Mongabay conducted with the TNC team.

Their research builds on previous maps of tree cover restoration potential, refining them by excluding areas where planting trees could do more harm than good, such as non-forested and fire-prone grasslands and savannas.

Kurt Fesenmyer, the study lead author and TNC’s forest spatial data scientist, said earlier maps have been critiqued for three main reasons: how they defined “forest”; the data quality; and the lack of safeguards to ensure positive outcomes for people and the environment.

“In developing our new map, we sought to address those critiques while building off the strengths of previous maps,” he said.

They concluded that 195 million hectares (482 million acres) of land worldwide — an area the size of Mexico — is available for tree restoration, a figure 71-92% smaller than previous estimates. Still, it holds the potential to remove more than 2.2 billion metric tons of carbon dioxide equivalent (CO2e) per year, or roughly 5% of total global fossil fuel and land-use change emissions in 2022, if the “right” areas are prioritized.

Mapping the right places

To create their new map, Fesenmyer and his team reviewed 89 reforestation maps of varying scales published since 2011. Deploying high-resolution 2020 land cover data, they included only areas naturally suited for closed-canopy forests. To avoid negative trade-offs, they excluded croplands, fire-prone grasslands, snow-covered terrain, and areas with high ecological or social risks.

“There is opportunity for reforestation in all countries with forest ecosystems, including in tropical, temperate and boreal forest biomes,” Fesenmyer said, noting the importance of the tropics due to the capacity of trees there to absorb carbon dioxide at the highest rates.

Of the 195 million hectares identified as reforestation hotspots, 83% lie close to existing forests, where natural regeneration and biodiversity can be enhanced, Fesenmyer added. However, only 90 million hectares (222 million acres) meet all social safeguard criteria, such as political stability, secure land tenure, and minimal dependence on the land for food or fuel.

Such areas include large swaths of land in Brazil, Colombia, the U.S., Canada and Western Europe, according to Fesenmyer.

The new tool, Reforestation Hub, allows decision-makers to explore hotspots worldwide, down to the country or province level, and see how much carbon can be captured annually by tree planting or natural regrowth in these areas. The tool also presents various scenarios based on different restoration goals.

A debate over forest definitions and restoration goals

The paper’s findings differ sharply from a 2019 study led by Jean-François Bastin, at the time a postdoctoral researcher at the Crowther Lab at ETH Zürich. That earlier study estimated 900 million hectares (2.22 billion acres) of global tree restoration opportunity — an area larger than Brazil — with the potential to store 205 billion metric tons of CO2e in total. According to Bastin’s paper, reaching this maximum restoration potential could reduce a considerable portion of the excess carbon in the atmosphere to date, estimated at approximately 300 billion metric tons by the U.N.’s Intergovernmental Panel on Climate Change (IPCC).

However, studies like Bastin’s have faced criticism for using broad definitions of forest that include grassy ecosystems, where planting trees may actually harm biodiversity and compromise ecosystem services, especially water availability.

“We addressed [such] definition critiques by focusing on ecosystems that support closed-canopy forests,” Fesenmyer said. “We do not identify reforestation opportunities in savannas and other ecosystems with naturally sparse tree cover, or in places with frequent wildfire.”

Bastin, currently a professor of ecology and remote sensing at Gembloux Agro-Bio Tech at the University of Liège, Belgium, praised Fesenmyer’s study as a “very valuable contribution.” But he added he still supports a broader view of restoration potential.

“I believe it’s important to look beyond a forest/non-forest dichotomy and to embrace a more inclusive understanding of ecosystems,” Bastin told Mongabay in an email. “That was our intention in 2019, by focusing on trees and not on forest [restoration potential], and it remains our direction.”

According to Bastin, their team mapped all land that could support any tree cover, based on climate and soil parameters, across a range of ecosystems, including grasslands. He said restrictive forest definitions risk excluding drylands and other important ecosystems from restoration agendas — such as Africa’s Great Green Wall, where tree planting is promoted to combat desertification and poverty.

“In some regions, shifting climate and socio-economic conditions may mean past ecosystems are no longer viable, making afforestation a potential practical option,” Bastin said. Climate mitigation shouldn’t be the only goal of nature restoration, he added.

Bastin’s latest research, recently also published in Nature Communications, estimates that 43% of the planet’s land could naturally support trees, 39% shrubs and grasses, and 18% bare ground. The study also models how vegetation might shift under varying fire and grazing pressures.

Rather than pinpointing reforestation or afforestation hotspots, Bastin said his latest maps are meant to inform land-use planning by offering insights into the ecological dynamics of each region.

“Ultimately, the best approach varies by landscape, and should reflect the needs and goals of local stakeholders,” he said.

Similarly, Susan Cook-Patton with TNC said their maps don’t dictate where reforestation should occur, but rather where it “can maximize benefits and minimize trade-offs.” She added that “it’s critically important that reforestation is done well.”

From maps to reality

While mapping tools provide valuable guidance, on-the-ground projects offer critical insights into effective tree-planting methods. A notable example is the Sardinilla experiment in Panama, the oldest in the International Network of Tree Diversity Experiments (TreeDivNet), which explores carbon sequestration, biodiversity and ecosystem functioning in planted forests.

Established in 2001 on former pastureland, with the consent of landowners, the experiment’s design includes plots planted with one, two, three or five native tree species. The most diverse plots combine fast-, intermediate- and slow-growing species, mimicking the structure of natural forests.

Florian Schnabel, forest scientist and lecturer at the University of Freiburg, Germany, led one of the most recent studies in Sardinilla, concluding that the planted forests with five tree species captured and stored significantly more aboveground carbon than monocultures. The findings were published earlier this year in the journal Global Change Biology.

“The species-rich forests captured 57% more carbon in aboveground tree biomass than monospecific forests,” Schnabel said via email, noting that these more biodiverse plantings yielded an average annual net uptake of 5.7 metric tons of CO2e per hectare.

“Mixed-species planted forests can be an effective climate mitigation strategy,” Schnabel said.

Location likely played a role in these results. Sardinilla is situated within a tropical ecoregion in eastern Panama and lies close to remaining natural forests, providing highly favorable conditions for tree growth and carbon sequestration. For example, on the Reforestation Hub map, the area is classified as having high climate mitigation potential.

Beyond boosting carbon capture, Schnabel’s study indicates that planting a diversity of native tree species enhances biodiversity and ecological stability, fostering greater resilience to disturbances. Some of the co-authors have also explored how to design reforestation strategies that deliver environmental and livelihood benefits to Indigenous peoples and local communities, particularly the Ipetí Emberá, an Indigenous community living on collectively owned land near Sardinilla.

The Sardinilla experiment offers a key reforestation lesson: how a forest is planted matters as much as where. Still, the research team cautions that tree planting alone won’t solve climate change.

“It would need one-year tree growth on 11 ha [27 acres] of the Sardinilla planted forest to compensate for the emissions of a single one-way flight between Frankfurt and Panama City,” Schnabel said. “Planting trees does … not release us from the urgent need to reduce our carbon emissions.”

Along the same lines, Fesenmyer said planting new forests is no replacement for protecting existing ones.

“Reforestation can be an important restoration strategy,” he said, “but the biggest climate benefits come from protecting the carbon [stored] within existing forests, including both young and old.”

Banner image: A walk through the Sardinilla Experiment.

Citations:

Fesenmyer, K. A., Poor, E. E., Terasaki Hart, D. E., Veldman, J. W., Fleischman, F., Choksi, P., … Cook-Patton, S. C. (2025). Addressing critiques refines global estimates of reforestation potential for climate change mitigation. Nature communications, 16(1), 4572. doi:10.1038/s41467-025-59799-8

Bastin, J.-F., Finegold, Y., Garcia, C., Mollicone, D., Rezende, M., Routh, D., … Crowther, T. W. (2019). The global tree restoration potential. Science, 365(6448), 76-79. doi:10.1126/science.aax0848

Bastin, J. F., Latte, N., Bogaert, J., Garcia, C. A., Berzaghi, F., Maestre, F. T., … Lejeune, P. (2025). Global alternatives of natural vegetation cover. Nature Communications, 16(1), 6484. doi:10.1038/s41467-025-61520-8

Schnabel, F., Guillemot, J., Barry, K. E., Brunn, M., Cesarz, S., Eisenhauer, N., … Potvin, C. (2025). Tree diversity increases carbon stocks and fluxes above—but not belowground in a tropical forest experiment. Global Change Biology, 31(2), e70089. doi:10.1111/gcb.70089

Forgues, K., Carignan, M.-C., Marchena, B., Mancilla, L., Pacheco, C., Pacheco Ortega, E., … Potvin, C. (2024). Comparing carbon offsets and livelihood benefits in a long‐term reforestation project: Agroforestry versus native timber versus enrichment planting. Ecological Solutions and Evidence, 5(3), e12372. doi:10.1002/2688-8319.12372