The best scientific estimate puts the number at roughly 390 billion individual trees, a figure derived from a massive 2013 study that combined field data from more than a thousand forest plots scattered across the Amazon basin. That number is staggering on its own, but what surprised researchers even more was the lopsided way those trees are distributed among species: a tiny fraction of the Amazon’s estimated 16,000 tree species accounts for about half of every trunk standing in the forest.
How Scientists Arrived at 390 Billion
Nobody walked the entire Amazon counting trees. The estimate came from a collaboration of hundreds of scientists who pooled data from 1,170 study plots spread throughout the basin, cataloguing and identifying more than 500,000 individual trees by hand.1Nature. Counting trees in the Amazon Each plot was a small, carefully measured patch of forest, typically about a hectare in size, where every tree above a certain trunk diameter was tagged, measured, and identified to species. The research team, led by Hans ter Steege and colleagues, then used statistical models to extrapolate from those plots to the entire 5.5-million-square-kilometer forest. The resulting estimate of about 390 billion trees was published in Science in 2013 and remains the most widely cited figure for the Amazon’s tree population.2Science. Hyperdominance in the Amazonian Tree Flora
That number carries real uncertainty. The plots are not evenly distributed; some regions of the Amazon, particularly remote areas in the western basin and parts of the Guiana Shield, have very few plots. The extrapolation assumes that the sampled plots are reasonably representative of the surrounding forest, which is a strong assumption given the Amazon’s enormous variation in soil, rainfall, and topography. Still, 390 billion is the best-grounded number available, and the margin of error is unlikely to shift the answer by an order of magnitude in either direction.
A Forest Dominated by a Few Species
Perhaps the most striking finding of the 2013 census was not the total tree count but how unevenly those billions of trees are distributed among species. The Amazon is home to an estimated 16,000 tree species, making it the most species-rich forest on Earth.3Mongabay. Amazon’s vast rainforest dominated by few tree species Yet just 227 of those species, roughly 1.4% of the total, turned out to be so common that they account for half of all individual trees across the basin. Researchers dubbed these the “hyperdominant” species.2Science. Hyperdominance in the Amazonian Tree Flora
At the other extreme, the rarest 11,000 or so species together make up just 0.12% of all trees.2Science. Hyperdominance in the Amazonian Tree Flora More than half of those rare species have estimated populations of fewer than 1,000 individuals total, scattered across millions of square kilometers.3Mongabay. Amazon’s vast rainforest dominated by few tree species Many of these species have never been collected in the field plots at all; their existence is inferred statistically. That means the Amazon could harbor thousands of tree species that scientists have never directly observed in a study plot, a humbling reminder of how much remains unknown about the forest’s composition.
The dominance pattern extends beyond sheer numbers. Follow-up research found that roughly the same small group of species controls about half of the Amazon’s total carbon stock and half of its woody productivity. In other words, about 1% of tree species are responsible for half of the forest’s carbon cycling, from the carbon locked in their trunks to the new wood they produce each year.4Nature Communications. Hyperdominance in Amazonian forest carbon cycling This concentration has big implications for climate science and conservation. If those few hundred dominant species were selectively lost or stressed, the consequences for the Amazon’s role as a carbon sink would be disproportionate.
Why Tree Density Varies Across the Basin
The Amazon is not one uniform green carpet. It contains radically different forest types, and the number of trees packed into a given hectare shifts depending on terrain, flooding regime, and soil chemistry. In productive tropical forests like much of the Amazon, studies of global tree-density patterns find roughly 500 to 800 trees per hectare on average.5Nature (Communications Biology). Global patterns of tree density are contingent upon local determinants in the world’s natural forests But that range disguises a lot of local variation.
The two broadest categories are terra firme forests, which sit on well-drained uplands and never flood, and várzea forests, which are seasonally inundated by nutrient-rich whitewater rivers. Trees in unflooded terra firme tend to have denser wood than their counterparts in seasonally flooded várzea, even when they belong to the same species.6Forest Ecology and Management. Forest type affects the capacity of Amazonian tree species to store carbon as woody biomass The flooding gradient shapes which species can survive where. Palms offer a clear example: flood-tolerant species like açaí cluster near riverbanks, while other palm species are found exclusively in terra firme, with a transitional zone of shifting abundance within the first hundred meters away from the floodplain.7Acta Amazonica. Palm community transitions along a topographic gradient from floodplain to terra firme in the eastern Amazon
Soil type adds another layer of complexity. White-sand forests in the western Amazon host a notably different community than the clay-soil terra firme forests nearby. Research in the Peruvian Amazon found that the dominant species on white sand tend to be closely related to each other, as if the harsh, nutrient-poor soil filters out all but a few lineages that evolved the right traits to cope. Clay-soil communities, by contrast, show patterns more consistent with recent local speciation, meaning the trees diversified in place over geological time.8Ecography. Phylogenetic community structure and phylogenetic turnover across space and edaphic gradients in western Amazonian tree communities These differences mean that a single hectare of white-sand forest and a single hectare of clay-soil forest, even if they are only a few kilometers apart, can look like entirely different ecosystems in terms of what is growing there.
The Outsized Role of Big Trees
When people picture the Amazon, they tend to imagine enormous emergent trees towering above the canopy. Those giants are real, but they are rare. In studies of eastern Amazonian forests, trees with trunk diameters of 60 centimeters or more made up only about 9% of all stems. Yet those large individuals held nearly half of all the aboveground biomass in the forest.9Forest Ecology and Management. Large trees as key elements of carbon storage and dynamics after selective logging in the Eastern Amazon This means the Amazon’s role as a carbon reservoir depends heavily on a relatively small number of massive trees. If selective logging or climate stress preferentially kills the biggest trunks, the forest can lose a huge fraction of its stored carbon even if the total number of living trees barely changes.
The flip side is regeneration. When a large tree falls, it opens a gap in the canopy and triggers a rush of seedling recruitment on the forest floor. Research on secondary-forest seedlings in the Amazon found that emergence rates can jump dramatically when leaf litter is cleared or soil is disturbed, and that low-lying bottomlands support faster seedling growth than hilltops, likely because water is more available downslope.10Europe PMC. Tree seedling recruitment in Amazon secondary forest: Importance of topography and gap micro-site conditions The implication is that topography shapes not just which species grow where, but how quickly the forest bounces back after a disturbance.
How Human History Has Shaped the Forest
The Amazon is sometimes described as “pristine wilderness,” but that characterization is increasingly recognized as wrong. Indigenous peoples have inhabited and managed parts of the basin for thousands of years, and their influence is still visible in the forest’s composition today. Areas with long histories of pre-Columbian occupation tend to have higher densities of useful tree species like Brazil nut and cacao, suggesting that Indigenous communities selectively encouraged or planted these trees over centuries. In regions where pre-Columbian influence was strong but colonial disruption was relatively light, old-growth forests carrying those ancient ecological legacies are still prevalent.11PLANTS, PEOPLE, PLANET. Indigenous and colonial influences on Amazonian forests
Colonial-era upheaval changed the picture. The catastrophic population collapse of Indigenous peoples after European contact left large areas of previously managed forest to grow without human intervention for centuries. Some of the Amazon’s apparent “wildness” is actually several hundred years of regrowth following the abandonment of managed landscapes. This patchwork of human histories probably plays a meaningful role in explaining why tree species composition varies so much across the basin, beyond what soil and climate alone would predict.11PLANTS, PEOPLE, PLANET. Indigenous and colonial influences on Amazonian forests
Deforestation and What It Means for the Count
That 390 billion figure was calculated from data collected over decades, and the forest has not stood still since. Brazil holds the majority of the Amazon, and by 2020 the Brazilian Amazon had lost somewhere between 17% and 21% of its original forest cover, the largest share among the nine countries that contain Amazonian forest.12Land Use Policy. Amazon deforestation: Drivers, damages, and policies Cattle ranching, soybean agriculture, and road construction are the primary drivers of clearing, concentrated along an “arc of deforestation” at the forest’s southern and eastern margins.
Deforestation removes trees wholesale, but the forest also loses trees in subtler ways. Degradation from selective logging, fire, and edge effects thins the canopy without clearing the land entirely. Drought stress is an emerging threat, particularly in the drier southern fringes of the basin. Severe drought kills trees while they are still standing, through a process that disrupts water transport inside the trunk. In the wettest parts of the forest, drought plays a smaller role, but along the dry southern edge, water-deficit tolerance is a significant predictor of which trees live and which die.13Nature Communications. Tree mode of death and mortality risk factors across Amazon forests In southwestern Amazonia, where drought intersects with dense stands of bamboo, tree-mortality patterns are shaped by both water stress and competition for light and space.14Trees, Forests and People. Primary modes of tree mortality in southwestern Amazon forests
All of this means the real-time tree count is almost certainly lower than 390 billion today, though no one has repeated the full census at the same scale. Deforestation has removed tens of billions of trees since the baseline data were collected, and degradation has thinned many more. How much of that loss has been offset by regrowth in abandoned agricultural areas is an open question without a firm answer yet.
Why Counting Remains So Difficult
You might assume that satellite imagery and modern remote sensing have made tree counting straightforward. They have not, at least not in a dense tropical forest. Satellites see the canopy surface, not individual trunks, so they are good at measuring canopy cover and detecting clearing but poor at resolving one tree from the next. Airborne laser scanning, known as LiDAR, can peer through gaps in the canopy and map the three-dimensional structure of the forest, but identifying individual trees in the resulting point cloud is still a major technical challenge.
A study testing automated tree-detection algorithms against field-verified counts in Amazonian forest found that the best-performing method correctly detected only about 58% of trees, and even the most accurate algorithm in terms of avoiding false detections still missed a substantial share of real trees.15Cerne. AUTOMATED INDIVIDUAL TREE DETECTION IN AMAZON TROPICAL FOREST FROM AIRBORNE LASER SCANNING DATA The dense, multi-layered canopy of tropical forest means that smaller understory trees are hidden beneath taller neighbors, and their crowns overlap in ways that confuse detection algorithms. Every method tested showed substantial bias at the individual-plot level. This is a far cry from the precision you might expect in an age of satellite everything, and it is a major reason why the 390 billion estimate still relies fundamentally on human fieldwork rather than automated remote sensing.
The practical upshot is that the tree count will remain approximate for the foreseeable future. Improvements in LiDAR resolution and machine-learning algorithms are slowly closing the gap, but the Amazon’s sheer complexity, with its layered canopy, enormous area, and staggering species diversity, makes a precise census impractical with current technology.
How the Amazon Compares Globally
A widely cited 2015 study estimated that Earth holds roughly three trillion trees in total. If the Amazon’s count of around 390 billion is in the right ballpark, the forest contains something like 13% of all trees on the planet, packed into about 3.5% of the global land area. Tree density in the Amazon is high by global standards, but not the highest possible; some temperate forests can pack in more stems per hectare because tropical trees tend to grow larger and crowd each other out through competition for light.
Research on global tree-density patterns shows that in the most productive forest plots, tree density sits in a relatively narrow band of roughly 500 to 800 trees per hectare regardless of latitude. But in plots dominated by large trees, the tropics pull ahead of higher latitudes because the warm, wet climate allows big trees to coexist at higher densities than they can in boreal forests, where harsh conditions thin out the largest individuals more aggressively.5Nature (Communications Biology). Global patterns of tree density are contingent upon local determinants in the world’s natural forests The Amazon, in other words, is not only home to more trees than any other single forest, but its trees are on average bigger, which makes the forest disproportionately important for storing carbon relative to its share of global tree count alone.
The Trees We Have Not Found Yet
One of the more quietly astonishing details from the 2013 census is that the statistical models predicted roughly 11,000 tree species that were never actually detected in any of the 1,170 plots. More than half of those missing species are thought to have total populations of fewer than 1,000 individuals each.3Mongabay. Amazon’s vast rainforest dominated by few tree species For perspective, a species with 1,000 trees spread across an area the size of the contiguous United States would be virtually undetectable by random sampling.
This means the Amazon almost certainly harbors thousands of tree species that have not been formally described or even knowingly encountered by a scientist. New species are named from the region every year, often from remote areas that have only recently become accessible or from herbarium specimens that were collected decades ago and never fully identified. The pace of deforestation and forest degradation raises the uncomfortable possibility that some of these species will be lost before they are ever documented. When conservationists talk about “unknown unknowns” in biodiversity, the Amazon’s phantom trees are a prime example: species that exist in the models, that likely exist in the forest, but that no researcher has yet held a leaf of.