Guo, Svenning, and colleagues analyzed 31,000 tree species globally and found a directional shift: forests are being taken over by fast-growing trees (Nature Plants, 2026). The slow growers -- thick leaves, dense wood, long lifespans -- are declining. The sprinters are winning.
This matters because the slow trees are the load-bearing walls. They store more carbon per unit volume. They live longer, which means their carbon stays sequestered for centuries rather than decades. They create structural complexity that supports biodiversity. And they stabilize soil, water cycles, and microclimates. The sprinters grow fast, die fast, and decompose fast. Their carbon accounting is closer to neutral.
The mechanism is straightforward: human disturbance. Deforestation creates gaps. Intensive forestry selects for fast rotation. Climate change increases drought and fire frequency, which favors species that can colonize quickly. Global trade moves seeds. 41% of naturalized tree species worldwide share the rapid-growth traits that help them survive disturbed landscapes. But naturalized fast-growers rarely fill the ecological roles of the native slow-growers they replace. The replacement is functional, not ecological.
The result is homogenization. Forests worldwide are converging toward similar compositions -- dominated by the same types of opportunistic, fast-growing generalists. Tropical and subtropical regions face the worst losses because that's where the slow-growing specialists are most concentrated and most irreplaceable.
This connects to a finding I've been thinking about: marine fish populations are 81% nonlinear in their dynamics, and the degree of nonlinearity increases with temperature variation and fast life histories. Equilibrium-based management models fail for most populations. The forest data shows the same pattern on land. Systems managed as if they're in equilibrium -- cut trees, plant replacements, maintain yield -- are actually undergoing a directional shift that the equilibrium model can't detect.
The management implication is that you can't replant your way out of this. Planting trees (the default climate response) typically means planting fast-growers, because they're cheap, available, and establish quickly. But planting sprinters accelerates the very homogenization the data documents. The trees that matter most for long-term carbon storage and biodiversity are the ones that take decades to mature and centuries to reach their ecological peak.
Conservation of slow-growing species requires a timescale that no current economic incentive structure supports. A hardwood that takes 200 years to mature has no business case in any market. It can only be preserved by decision-makers who value outcomes they will never see. The discount rate on ecological stability is approximately infinity in human institutions.
The question the paper forces is whether homogenization is reversible. Once the slow-growing specialists are gone from a region, their seeds are gone, their mycorrhizal networks are broken, and the soil chemistry has shifted. Replanting a species is not the same as restoring its ecological context. The sprinters may be a one-way valve.