By Satyabrat Borah
Fifty-six million years ago, Earth experienced one of the most abrupt and intense episodes of global warming in its entire geological history. Known to scientists as the Paleocene-Eocene Thermal Maximum, or PETM, this event saw average global temperatures rise by approximately six degrees Celsius in a geologically brief span, probably no more than five to ten thousand years. Vast quantities of carbon, equivalent to several times the amount humans have released since the Industrial Revolution, flooded the atmosphere and oceans, driving a hyperthermal state that lasted over 150,000 years. What makes the PETM especially relevant today is not only the scale of the warming but also the discovery, published in 2024 in Nature Communications, that the world’s vegetation played a critical and previously underestimated role in prolonging the crisis.
Plants are among the planet’s most powerful natural mechanisms for removing carbon dioxide from the air. Through photosynthesis they draw down CO₂, convert it into sugars and structural tissues, and lock carbon away in leaves, trunks, roots, and soils for decades, centuries, or even millennia. In normal times this process acts as a thermostat, dampening the effects of carbon releases from volcanoes or other natural sources. During the PETM, however, that thermostat appears to have broken down, at least across large parts of the globe, turning a potential buffer into an amplifier of warming.
To understand what happened, researchers built a sophisticated Earth-system model that couples plant evolution, migration, physiology, and carbon cycling. They then tested the model against an extraordinary archive: fossil pollen preserved in ancient sediments from three very different locations. The Bighorn Basin in Wyoming, a mid-latitude continental interior site; the North Sea region, representing northern Europe; and a high-latitude site near the Arctic Circle. Pollen is ideal for this kind of work because plants produce it in enormous quantities, it travels long distances on wind and water, and its tough outer wall survives millions of years of burial. By identifying which species were dominant before, during, and after the PETM, and by measuring traits such as leaf mass per area in fossil leaves, the team could reconstruct how vegetation changed and how those changes affected the carbon cycle.
The results revealed a stark north-south divide. In the mid-latitudes, the story was one of collapse. Before the PETM, the Bighorn Basin supported diverse forests with many deciduous broad-leaved trees. As temperatures soared and rainfall patterns shifted toward greater seasonality, these forests gave way to very different plant communities. Pollen records show a sharp increase in palms, ferns, and other small, drought-tolerant species. Leaf fossils reveal that surviving plants developed thicker, denser leaves, an adaptation that reduces water loss but also reduces the surface area available for photosynthesis per unit of carbon invested. At the same time, soil carbon stocks plummeted. The large, long-lived trees that had once stored massive amounts of carbon in their trunks and root systems were replaced by shorter-lived, lower-biomass vegetation that locked away far less carbon both above and below ground. In effect, the mid-latitude land surface switched from being a strong carbon sink to something closer to carbon-neutral or even a weak source.
In the high latitudes, by contrast, the response was almost the opposite. The Arctic, once clothed in cool-temperate conifer forests, warmed dramatically and became wetter. Pollen assemblages document the arrival of broad-leaved swamp forests and even subtropical elements such as palms that briefly flourished near the North Pole. Plant height and overall biomass increased, and the new vegetation appears to have photosynthesized more vigorously than what it replaced. For a time, the high latitudes compensated for some of the carbon-storage losses farther south, but the area of land capable of this enhanced uptake was limited. The vast tropical and temperate zones, where most terrestrial carbon is normally stored, had largely stopped functioning as effective sinks.
The consequences were profound. Model simulations suggest that the widespread failure of mid-latitude vegetation reduced global terrestrial carbon sequestration by a significant fraction for seventy to one hundred thousand years. Because less carbon was being pulled out of the atmosphere, excess CO₂ lingered far longer than it otherwise would have, stretching what might have been a shorter-lived warming event into a prolonged hothouse episode. Only after tens of thousands of years, as new plant communities slowly re-evolved or migrated and soil carbon pools began to rebuild, did the biosphere regain its capacity to draw down carbon at pre-PETM rates. By then, the oceans had absorbed most of the excess carbon through silicate weathering, a much slower geological process that eventually restored cooler conditions.
The PETM is often described as the closest geological analogue to modern anthropogenic warming, and the parallels are unsettling. The amount of carbon released during the PETM was larger than our current fossil-fuel reserves, but the rate of release appears to have been slower, perhaps ten times slower than the rate at which we are emitting CO₂ today. During the PETM, six degrees of warming over several thousand years was enough to push much of the world’s vegetation beyond its adaptive limits. Today we are approaching two degrees of warming in little more than a century, with projections of four degrees or more by 2100 under high-emission scenarios. Plants can migrate and evolve, but not at anything like the speed required to keep pace with twenty-first-century rates of change.
Modern observations already hint at the kind of disruption seen in the PETM fossil record. Satellite data show that while some regions, especially boreal forests, have greened in recent decades because of longer growing seasons, many tropical forests are losing biomass and shifting toward faster-growing, lower-density species. Drought stress in the Amazon, prolonged heat in Mediterranean ecosystems, and repeated wildfires in Australia and western North America are all pushing plant communities toward compositions that store less carbon. Experiments that expose forests to elevated CO₂ and temperature suggest that the initial boost in photosynthesis often fades within years as nutrients become limiting and heat stress takes its toll.
The PETM teaches us that vegetation is not simply a passive victim of climate change; it can become an active participant in a feedback loop that makes warming worse and longer-lasting. When forests are replaced by scrub or grassland, not only is less carbon stored in living biomass, but vast amounts of soil carbon built up over millennia become vulnerable to decomposition and release. The loss is double: less uptake in the future and more emission from the past. Breaking that feedback requires preserving existing carbon-rich ecosystems and allowing damaged ones enough time and connectivity to recover. Yet time is the one thing rapid modern warming denies them.
Fifty-six million years ago, Earth eventually cooled again, but it took over 150,000 years, far longer than the lifespan of any human civilization. The organisms that lived through the PETM, including early primates that were our distant ancestors, endured a world of strange warmth, acidified oceans, and shifting continents of vegetation. Many lineages went extinct; others radiated into newly opened niches once the crisis passed. Today we have the power to influence whether our own species faces a similar trial by heat. The pollen grains locked in ancient rocks carry a clear message: if we push the planet’s vegetation past its breaking point, the climate will not simply return to normal once we stop adding carbon. The green world that has sustained us may take far longer to heal than we can afford to wait.
