For decades, the prevailing narrative was seductively simple: more carbon dioxide in the atmosphere means more plant growth, which means forests will obligingly soak up our emissions and buffer the worst of climate change. This CO2 fertilization hypothesis became a cornerstone of optimistic carbon budget projections, baked into Earth system models that governments relied on. The forests, we assumed, would do the heavy lifting.

That assumption is now unraveling. A generation of Free-Air CO2 Enrichment (FACE) experiments, flux tower networks, and remote sensing analyses has revealed a far more nuanced picture. Elevated CO2 does stimulate photosynthesis—but the downstream consequences for forest carbon cycling are tangled in nutrient feedbacks, hydraulic tradeoffs, and allocation shifts that can dampen, redirect, or even reverse the expected gains. The carbon cycle, it turns out, doesn't respond to a single variable in isolation.

What emerges from this body of research is not a story of simple fertilization but one of rewiring—a fundamental reorganization of how carbon moves through forest ecosystems. Trees under elevated CO2 don't just grow more; they grow differently, use water differently, and interact with soil biogeochemistry in ways that reshape the entire system. Understanding these shifts is critical for anyone projecting terrestrial carbon sinks, designing forest-based climate mitigation strategies, or modeling the coupled carbon-climate system through the coming century.

The basic physiology is sound. At elevated CO2 concentrations, the enzyme RuBisCO—which catalyzes the first step of carbon fixation—operates closer to substrate saturation. Photorespiration declines. Gross primary productivity increases. Early chamber studies in the 1980s and 1990s documented photosynthetic enhancements of 30–60% under doubled CO2, and the fertilization narrative gained momentum. Models projected that terrestrial ecosystems would absorb a substantial fraction of anthropogenic emissions indefinitely.

Then the long-term experiments caught up. The Duke FACE experiment, running from 1996 to 2010 in a loblolly pine plantation, initially showed a striking 23% increase in net primary productivity under elevated CO2. But the enhancement progressively declined. By the experiment's final years, the stimulation had narrowed considerably. The culprit was nitrogen. The trees were pulling nitrogen from the soil faster than microbial mineralization could replenish it, and without adequate nitrogen, they couldn't sustain the proteins and chlorophyll required to capitalize on extra carbon.

The EucFACE experiment in Australia delivered an even starker lesson. In a mature Eucalyptus woodland growing on phosphorus-poor soil, elevated CO2 produced no significant increase in aboveground biomass over multiple years. Phosphorus limitation imposed a hard ceiling on growth. The extra carbon that trees fixed was largely shunted belowground—into root exudates and mycorrhizal networks—representing a metabolic investment in nutrient acquisition rather than structural growth or long-term carbon storage.

These results forced a recalibration. The global carbon cycle community now recognizes that nutrient co-limitation—particularly nitrogen and phosphorus—fundamentally constrains the CO2 fertilization effect in most real-world forests. The response magnitude depends heavily on soil fertility, mycorrhizal type, and the stoichiometric flexibility of the dominant tree species. Forests on young, nitrogen-rich soils may sustain a fertilization response for decades; those on ancient, phosphorus-depleted substrates may show almost none.

This has profound implications for Earth system models. Many current-generation models still overestimate the terrestrial carbon sink because they inadequately represent nutrient cycling. Incorporating coupled carbon-nitrogen-phosphorus dynamics reduces projected land carbon uptake by 25–50% in some model frameworks. The fertilization dividend, while real, comes with a nutrient tax that most ecosystems cannot indefinitely afford.

Takeaway

CO2 fertilization is not a free lunch—it's constrained by the nutrients available to build the biological machinery that uses that carbon, and most forest soils are already running a deficit.

When CO2 concentrations rise, plants can afford to be less generous with their stomata. Each stomatal pore admits CO2 for photosynthesis but loses water vapor in the process—a fundamental tradeoff that has governed plant gas exchange for 400 million years. Under elevated CO2, the same photosynthetic rate can be maintained with reduced stomatal conductance, typically declining 20–30% across C3 species. The result is improved leaf-level water-use efficiency: more carbon fixed per molecule of water transpired.

At first glance, this looks like an unambiguous benefit—drought-resilient forests that photosynthesize efficiently. And indeed, there is evidence for it. Stable carbon isotope analyses (δ¹³C) in tree rings globally show a trend toward increased intrinsic water-use efficiency over the industrial period, consistent with a CO2-driven stomatal response. Some modeling studies project that elevated CO2 could partially offset drought stress in water-limited forests, extending their viability under warming scenarios.

But the consequences ripple outward in ways that complicate this optimism. Reduced transpiration means less evaporative cooling at the canopy surface. Forests are powerful climate regulators not just through carbon sequestration but through their hydrological cycling—tropical forests can cool their surroundings by 2–4°C through transpiration alone. If stomatal closure reduces this cooling service, the biophysical climate benefit of forests diminishes even as their carbon uptake potentially increases. We may gain a slightly larger carbon sink at the cost of amplified regional warming.

The hydrological consequences extend to the watershed scale. Forests with reduced transpiration return less water to the atmosphere, potentially altering precipitation recycling in regions like the Amazon basin, where a significant fraction of rainfall is transpiration-derived. Simultaneously, reduced water uptake by trees means more water remains in the soil and runs off into streams. Some catchment-scale studies have already detected increased streamflow trends consistent with CO2-driven reductions in canopy transpiration—a shift that could affect water resources, flood dynamics, and aquatic ecosystems downstream.

The emerging picture is one of hydraulic reshuffling rather than simple improvement. Trees under elevated CO2 may indeed survive drought events that would otherwise be lethal—but the systemic feedbacks on regional climate, precipitation patterns, and watershed hydrology introduce new risks. The water-carbon tradeoff doesn't just operate at the leaf; it propagates through the entire Earth system, coupling forest physiology to atmospheric dynamics in ways that defy simple accounting.

Takeaway

Improved water-use efficiency at the leaf level can degrade climate regulation at the landscape level—a reminder that optimizing one part of a coupled system often destabilizes another.

Not all carbon is created equal from a climate mitigation perspective. A ton of carbon locked into a tree trunk persists for decades to centuries. The same ton routed into fine roots or root exudates may cycle back to the atmosphere within months through respiration and decomposition. Under elevated CO2, the critical question is not just how much additional carbon forests fix, but where they put it.

The evidence consistently points toward a belowground shift. Meta-analyses of FACE experiments show that elevated CO2 tends to disproportionately increase fine root production, mycorrhizal colonization, and root exudation relative to aboveground woody biomass. In the EucFACE experiment, virtually all of the extra carbon fixation went underground. At the Oak Ridge FACE site, elevated CO2 stimulated fine root turnover but did not significantly increase stem wood production in later years. Trees, it appears, invest their carbon windfall in nutrient foraging rather than structural growth.

This allocation shift has a metabolic logic. When carbon is abundant but nutrients are scarce, the optimal strategy is to invest in the infrastructure that acquires the limiting resource. Increased root exudation stimulates microbial activity in the rhizosphere—the priming effect—which can accelerate decomposition of soil organic matter and release mineral nutrients. This is beneficial for the tree in the short term but troubling for the carbon budget: the priming effect can mobilize ancient, stable soil carbon pools, potentially turning soils from net carbon sinks into net sources.

The implications for long-term carbon storage are significant. If elevated CO2 primarily enhances ephemeral carbon pools—fine roots, labile exudates, fast-cycling microbial biomass—rather than recalcitrant wood and stable soil organic matter, then the effective residence time of the additional sequestered carbon is far shorter than models typically assume. The gross carbon uptake may increase while the net long-term storage barely changes, or in extreme cases, decreases due to soil carbon priming.

Recent work on mycorrhizal networks adds another layer. Trees associated with ectomycorrhizal fungi, which dominate boreal and many temperate forests, appear to sustain larger CO2 responses than those with arbuscular mycorrhizae, partly because ectomycorrhizal fungi can directly access organic nutrient pools. This means the allocation response—and its carbon storage consequences—varies not just with soil chemistry but with the belowground symbiotic community. Predicting forest carbon dynamics under elevated CO2 requires understanding not just tree physiology but the entire rhizosphere economy, a system we are only beginning to map at the scales that matter.

Takeaway

A forest can fix more carbon without storing more carbon—the distinction between gross uptake and durable sequestration is the gap where optimistic climate projections quietly collapse.

The story of forests in a CO2-rich world is not one of simple enhancement but of systemic reorganization. Carbon flows are being redirected—toward belowground investment, away from long-lived wood; through narrower stomata that fix carbon efficiently but cool landscapes less effectively; into nutrient-mining strategies that may liberate ancient soil carbon in the process.

For climate policy, the implications are sobering. Terrestrial carbon sink projections that rely on sustained fertilization effects without accounting for nutrient constraints, allocation shifts, and biophysical feedbacks are likely overestimating the service forests can provide. Nature-based climate solutions remain valuable, but their carbon arithmetic needs to be grounded in the complex biogeochemistry that FACE experiments and flux networks have revealed.

The forests are responding to our altered atmosphere. They're just not responding the way we hoped.