The Amazon is often described in terms of how much carbon it stores. It also moves vast amounts of water. Its trees draw moisture from the soil and return it to the air through transpiration. Some rain caught by the canopy also evaporates back into the atmosphere. That moisture falls again, some of it inside the basin and some of it far downwind. For much of South America, the forest helps sustain rainfall, farming, hydropower, and urban water supplies.

A recent paper in Nature, by Nico Wunderling and colleagues, gives this hydrological role greater weight in the climate case for the Amazon. The authors used a dynamical systems model and atmospheric moisture tracking to estimate how deforestation and warming interact across the basin. Without deforestation, their model finds a critical global warming threshold of about 3.7 to 4.0 degrees Celsius, beyond which up to a third of the Amazon forest risks losing stability. When deforestation is included, the risk becomes much larger at a lower level of warming. Under deforestation of 22% to 28% of the biome and warming of 1.5 to 1.9 degrees Celsius, the model finds a near system-wide transition affecting 62% to 77% of the forest.

Those numbers should be treated carefully. The Amazon is not a single ecological unit. Western forests, southern forests, flooded forests, upland forests, secondary forests, and fire-damaged edges differ in species composition, rainfall, soils, and drought tolerance. A model cannot capture every local response. Even so, the paper puts numbers around an important relationship: forest loss weakens the water recycling that helps the remaining forest endure heat and drought.

The mechanism is basic to how the forest functions. Part of the Amazon’s rainfall comes from outside the basin. Part is generated inside it. Trees move water from deep soil layers into the atmosphere, especially late in the dry season, when pastures, crops, and degraded lands return less moisture to the air. Through repeated cycles of rain and evapotranspiration, the forest helps maintain the wet conditions to which it is adapted. When enough forest is removed or damaged, the remaining forest has less moisture to draw on during hot and dry periods.

Why forest condition matters

The condition of the forest therefore matters as much as its extent. A more intact Amazon enters a drought with more canopy cover, greater humidity, deeper rooting systems, fewer dry edges, and fewer places where fire can take hold. A degraded forest enters the same drought with thinner canopy, hotter edges, more fuel, and more exposure to ignition from nearby farms and pastures. Climate change increases drought and heat stress across the basin, and degradation makes affected forests less able to recover.

The 2023 and 2024 droughts suggest that recovery time is becoming a central concern. A 2026 PNAS study used long-term radar observations to measure forest moisture and biomass dynamics during those events. Across intact Amazonian rainforests, the radar signal fell to its lowest level since 1992. More than a quarter of the forest pixels examined reached their three-decade minimum during the event, mostly in eastern Amazonia. The authors projected that, seven years after the droughts, less than half of affected areas would recover to predrought conditions.

That finding is important because the study focused on intact rainforest. Areas already damaged by fire, logging, fragmentation, or repeated edge effects are likely to be more vulnerable. Severe droughts have occurred roughly every seven years over the past three decades, according to the paper. If the interval between major droughts is shorter than the time needed for recovery, parts of the forest may face each new event before they have fully recovered from the last one.

El Niño raises the fire risk

Current El Niño forecasts make that concern more immediate this year. In parts of the Amazon, El Niño is associated with hotter and drier conditions, although the effects vary across the basin. Southern and eastern Amazonia, including parts of the arc of deforestation, are often among the areas of greatest concern. El Niño does not usually cause people to clear forest. Clearing is driven by land speculation, ranching, agriculture, logging, roads, weak enforcement, and political incentives. Its more direct effect is to make already-cleared or degraded landscapes more likely to burn.

During dry years, fires set to manage pasture, burn cut vegetation, or clear land are more likely to escape. Once fire enters standing forest, it can move through the understory, killing trees slowly and opening the canopy. Much of this damage may not appear immediately in deforestation statistics. It can still reduce carbon storage, fragment habitat, and make the next drought more damaging. The 2015-16 El Niño and the 2023-24 drought showed how fire and drought can surge even when annual deforestation is far below the levels seen in the early 2000s.

South American agriculture depends on Amazon moisture

The same moisture system also matters for agriculture outside the basin. Another 2026 PNAS paper, by Hao Li and colleagues, examined how Amazon deforestation affects precipitation and soybean yields across Brazil. Using atmospheric moisture tracking and crop modeling, the researchers found that tree evaporation contributes about one-third of growing-season precipitation across Brazilian soybean states. Recent deforestation reduced seasonal precipitation by 6% to 30%, depending on location and crop stage. The largest absolute production loss was estimated in Rio Grande do Sul, more than 1,500 kilometers from the main deforestation frontier.

Brazil’s agricultural economy depends heavily on rain. The South American monsoon draws moisture inland from the tropical Atlantic. The South American low-level jet helps carry water vapor south toward central and southern Brazil and the La Plata basin. Farmers in Mato Grosso, Paraná, Rio Grande do Sul, Paraguay, Argentina, and Uruguay may experience the Amazon as weather rather than forest. The connection is not simple or uniform, but it is economically relevant.

The soybean study estimated cumulative production losses from deforestation-driven rainfall declines at about 700,000 metric tons. That number does not overturn the economics of Brazilian agriculture by itself. It does show the direction of the risk. Clearing forest may increase cropland in one place while reducing yields in established agricultural regions elsewhere. Lower yields can then increase pressure to open more land, which can further weaken rainfall.

Protection first, restoration where it counts

Amazon conservation increasingly belongs in discussions of food security and climate adaptation. Keeping forest standing supports climate mitigation, biodiversity, and resilience in South America’s food system. Restoring degraded areas, reducing fire, protecting Indigenous and community-managed territories, enforcing forest laws, and maintaining protected areas all support the moisture system on which the forest and surrounding economies depend.

Restoration has to mean more than putting trees back on cleared land. The most useful recovery includes canopy closure, native species recruitment, lower fire susceptibility, soil moisture recovery, and the return of animals that pollinate plants and disperse seeds. Natural regeneration can do much of this work in some places if cattle, repeated burning, invasive grasses, and renewed clearing are kept out. Other areas will need active planting, enrichment, or long-term maintenance.

A long-running study in southeastern Amazonia shows how recovery can proceed, and where it remains fragile. Researchers found that heavily degraded, grass-invaded forests could recover once burning stopped, especially in the forest interior. Edge forests recovered more slowly. Tree communities shifted toward drought-tolerant generalists, while Amazon forest specialists continued to decline. The forest did not become savanna, but the communities that reassembled remained vulnerable to renewed fire, drought, or windthrow.

Avoided deforestation remains the first priority. Old-growth forest already has the structure, biodiversity, hydrological function, and carbon stocks that restoration tries to rebuild. Restoration can repair strategic gaps, reconnect fragments, protect riparian areas, and reduce fire risk. It is most valuable when it supports protection, rather than compensating for continued clearing.

The policy case is practical

Many of the needed policies already exist, at least on paper. Enforcement against illegal clearing and burning has to be consistent. Protected areas and Indigenous territories need durable finance, staff, monitoring, and political backing. Farmers need credit and technical support to increase production on already-cleared land. Infrastructure planning has to account for the forest loss it may induce. Fire management needs to become a standing public function, especially in dry years.

The politics are harder because the water benefits are widely shared. The costs of keeping forest standing are often local. The benefits are spread across the basin, across national borders, and across agricultural regions far from the deforestation frontier. A municipality may forgo revenue by restricting clearing. A farmer hundreds or thousands of kilometers away may receive part of the rainfall benefit. Public finance, carbon markets, commodity standards, and international support all struggle with that mismatch.

The Nature paper focuses on thresholds, but the policy implications are more immediate. The Amazon is more likely to remain stable if warming is limited, deforestation stops, fire is controlled, and degraded forests recover. Those actions reduce emissions. They also help preserve the rainfall system that supports one of the world’s major agricultural regions. In a hotter climate, the case for keeping the Amazon wet becomes harder to separate from the case for keeping South America productive.

Banner image: Rain over the Amazon. Photo by Rhett Ayers Butler

Citations:

• Bai, H., Liu, X., Yang, H., Chave, J., Ciais, P., Wigneron, J.-P., Saatchi, S., Xiao, J., Le Toan, T., Hu, X., Yang, Z., Wang, L., Fan, L., Yao, Y., Chen, X., Liu, Y., Xue, B., Guo, Q., Tang, Z., … Tao, S. (2026). Unprecedented Amazonian rainforests damage during the 2023–2024 droughts. Proceedings of the National Academy of Sciences, 123(13), Article e2514066123. doi.org/10.1073/pnas.2514066123
• Berenguer, E., Lennox, G. D., Ferreira, J., Malhi, Y., Aragão, L. E. O. C., Barreto, J. R., Espírito-Santo, F. D. B., Figueiredo, A. E. S., França, F., Gardner, T. A., Joly, C. A., Palmeira, A. F., Quesada, C. A., Rossi, L. C., Seixas, M. M. M. D., Smith, C. C., Withey, K., & Barlow, J. (2021). Tracking the impacts of El Niño drought and fire in human-modified Amazonian forests. Proceedings of the National Academy of Sciences, 118(30), Article e2019377118. doi.org/10.1073/pnas.2019377118
• Climate Prediction Center. (2026). Official NOAA CPC ENSO strength probabilities. National Oceanic and Atmospheric Administration. cpc.ncep.noaa.gov/products/analysis_monitoring/enso/roni/strengths/
• Li, H., Lesk, C. S., Zhu, L., Crowther, T. W., Chen, M., Ray, D. K., & Mueller, N. D. (2026). Amazon deforestation reduces precipitation and soybean yields across Southern Brazil. Proceedings of the National Academy of Sciences, 123(24), Article e2525378123. doi.org/10.1073/pnas.2525378123
• Maracahipes, L., Brando, P. M., Silvério, D. V., Maracahipes-Santos, L., Silveiro, A. C., Verona, L., Macedo, M. N., Trumbore, S., Lenza, E., Starinchak, B., Potter, N., Herrera-Ramírez, D., Barros, F. de V., Uribe, M. del R., Byrnes, L., Andrade, A. F. A., Rocha, E. X., Rattis, L., Nunes, T. F., & Oliveira, R. S. (2026). Forest recovery pathways after fire, drought, and windstorms in southeastern Amazonia. Proceedings of the National Academy of Sciences, 123(17), Article e2532833123. doi.org/10.1073/pnas.2532833123
• World Meteorological Organization. (2026, June 2). WMO: Prepare for El Niño. wmo.int/news/media-centre/wmo-prepare-el-nino
• Wunderling, N., Sakschewski, B., Rockström, J., Flores, B. M., Hirota, M., & Staal, A. (2026). Deforestation-induced drying lowers Amazon climate threshold. Nature, 654, 114–120. doi.org/10.1038/s41586-026-10456-0

Cite this article

Rhett Ayers Butler (2026). South America’s farms depend, in part, on a healthy Amazon. Mongabay Conservation news. DOI: https://doi.org/10.66709/news-321533

Credits

Rhett Butler Editor

Topics

AgricultureConservationDroughtEcosystem ServicesEl NinoEnvironmentEnvironmental PolicyEnvironmental ServicesFiresforest degradationForestsHydroelectric PowerNatural CapitalNature-based climate solutionsPrecipitationRainforestsRestorationSaving RainforestsSolutionsSoyThreats To RainforestsTropical DeforestationWaterAmazonBrazilLatin AmericaSouth America

See Topics