New research shows that forest fragmentation silently erodes carbon uptake – even when total forest area stays the same
Imagine two landscapes. Both contain exactly the same amount of forest. In one, the trees form a vast, connected expanse. In the other, that same forest has been broken into hundreds of smaller patches, scattered across the land.
According to current climate policy, these two landscapes are equivalent. The forest area is identical, so the carbon sequestration should be too.
New research published in Nature Ecology & Evolution suggests this assumption is wrong, and the implications for how we protect forests are significant.
What the research found
Led by Yibiao Zou and developed in the Crowther Lab at ETH Zürich, the study analysed 17 million forest patches across the United States, asking the question: does it matter how forest is arranged, not just how much of it there is?
After five years of research, the answer was clear.
A hectare of forest embedded within a large, connected forest is on average 38% more productive per year than an isolated hectare under otherwise identical conditions – same climate, same soil, same tree cover.
That difference adds up. The researchers calculated that existing fragmentation has already reduced annual forest carbon uptake across the US by 0.16 billion tonnes of carbon per year, or roughly 14% less than if that same forest were intact and connected.
Why does forest patch size matter?
The key lies at the forest edges.
Every forest patch has a boundary – a place where trees meet open land. At these edges, conditions are harsher: temperatures are higher, air is drier, winds are stronger, and trees are more vulnerable. Small, fragmented patches are dominated by edge environments. A large, connected forest, by contrast, has vast interior zones where trees are sheltered, stable, and far more productive.
Additionally, forest patch size matters for biodiversity. Larger forests tend to support more species, and greater diversity of plant life generally means more efficient use of resources and higher overall productivity.
Zou et al. confirmed these patterns held not just across the US, but globally – with the relationship between patch size and carbon uptake particularly strong in tropical forests, which are among the most biodiverse ecosystems and sensitive to fragmentation.
The policy gap
This finding has a direct challenge embedded in it.
Many corporate, non-government, and government commitments to forests are framed around “net zero deforestation” – the idea that clearing forest in one place can be offset by protecting or replanting forest elsewhere, as long as the total area stays the same. But if fragmentation is already reducing carbon absorption across forests by 14% without a single tree being lost, then area alone is a poor measure of what is actually being protected.
The research doesn’t suggest that protecting small forest patches is worthless. Every fragment matters for biodiversity and local ecosystems. But it does suggest that forest configuration — not just forest cover — needs to be part of how we measure, protect, and restore nature.
Keeping forests whole and reconnecting ones that have been fragmented may be one of the most effective and undervalued strategies we have for maintaining the carbon sink that forests provide.
Find the paper published in Nature Ecology & Evolution
Zou, Y., Smith, G.R., Lauber, T. et al. Larger forest patches have greater per-area productivity. Nat Ecol Evol (2026). https://doi.org/10.1038/s41559-026-03075-5