For decades, scientists have expected that rising carbon dioxide levels in the atmosphere would make forests grow faster. After all, trees absorb carbon dioxide, combine it with water using sunlight, produce sugars for growth, and release oxygen. More carbon in the air should mean more growth, more carbon storage, and a natural brake on climate change.
But long-term measurements from real forests have stubbornly refused to follow that logic.
A new study led by researchers from Duke University and Wuhan University offers a compelling explanation for this puzzle. The research shows that carbon dioxide alone does not control how fast trees grow. Water, and how trees manage it, plays an equally critical role.
Over the past several decades, atmospheric carbon dioxide has steadily increased, yet tree growth across the world has shown inconsistent patterns. In some forests, growth increased slightly. In others, it stayed flat or even declined. These mixed results have left scientists questioning why the so-called “CO₂ fertilisation effect” has not delivered the gains many climate models predicted.
The new study reframes the problem by focusing on a daily decision trees must constantly make. Trees open tiny pores on their leaves, known as stomata, to take in carbon dioxide. But when those pores open, water escapes at the same time. Every day, a tree must balance gaining carbon with avoiding dangerous water loss.
The researchers built a model that treats this balance as an optimisation problem. The goal for the tree is to maximise carbon gain while keeping water loss below a level that could damage its internal water transport system. This approach, inspired by engineering principles, closely matches what scientists have observed in real forests over decades.
Professor Gaby Katul of Duke University, one of the study’s authors, said earlier assumptions focused too heavily on carbon dioxide in isolation. Controlled experiments showed that while higher CO₂ can increase growth under ideal conditions, other environmental stresses often cancel out those benefits.
To test their ideas, the team relied on data from two rare and long-running experiments. At Duke University, a forest plot was exposed to elevated carbon dioxide levels for 16 years. At ETH Zurich, researchers conducted detailed experiments by increasing air humidity around trees. Both studies tracked growth, carbon uptake, leaf behaviour and environmental conditions.
The results were striking. Trees exposed to extra carbon dioxide did not store as much additional carbon as simple models had predicted. The key reason lay in how stomata respond to heat and dryness.
In carbon-rich air, stomata can open less while still absorbing enough carbon dioxide. In theory, this should improve water efficiency and boost growth. But rising temperatures and drier air change the equation. Hot, dry conditions increase evaporation, forcing trees to close their stomata to prevent excessive water loss. When stomata close, carbon intake drops as well.
This response protects the tree’s internal plumbing, known as the xylem, which carries water from roots to leaves. If water tension becomes too high, air bubbles can form and break the flow, a potentially fatal failure that becomes more likely as trees grow taller and climates grow hotter.
Using detailed measurements from individual leaves exposed to changing temperature, humidity and carbon dioxide, the researchers calibrated their model and tested it against real-world data. The model successfully reproduced the muted growth response seen in the Duke experiment and explained why higher humidity allowed trees to take in more carbon in the Zurich study.
When the researchers applied this framework to decades of tropical forest data from around the world, many contradictions disappeared. In regions where warming and atmospheric dryness increased, trees closed their stomata more often, cancelling out potential carbon gains. In wetter regions, forests were more likely to show growth increases.
The findings do not suggest that higher carbon dioxide never boosts tree growth. Instead, they show that the effect depends on local conditions, especially the balance between carbon availability and water stress. The response is controlled at the scale of microscopic leaf pores but shaped by regional climate trends.
The researchers stress that this mechanism is only one part of a complex system. Nutrient availability, soil moisture, tree species, forest age, pests and seasonal shifts also influence growth. Still, the study explains a major piece of the puzzle that has challenged climate scientists for years.
For climate policy and modelling, the message is clear. Rising carbon dioxide does not guarantee that forests will automatically absorb more carbon. On hot, dry days, trees prioritise survival over growth. Protecting forests’ ability to store carbon therefore requires protecting their access to water and reducing heat stress, not just counting on carbon fertilisation.
The study, published in Nature Climate Change highlights a simple but powerful truth. Between carbon in the air and carbon stored in wood lies a living system of trade-offs, evolved to keep trees alive rather than make them grow faster. If forests are to help slow climate change, the conditions that support their survival must come first.