Larger forest patches have greater per-area productivity

[Geoengineering News]

https://www.nature.com/articles/s41559-026-03075-5

Authors: Yibiao Zou, Gabriel Reuben Smith, Thomas Lauber, Joe Wan, Haozhi Ma, Noel Gorelick, Constantin M. Zohner & Thomas W. Crowther 

15 May 2026

Abstract

Forest fragmentation could reduce carbon sequestration beyond losses caused by declining forest area alone, if smaller patches are intrinsically less productive per unit area than larger ones. Here we analyse 17 million forest patches across the conterminous USA and show that per-area net primary productivity increases systematically with patch size. A hectare embedded within a ~100,000 km2 forest is, on average, 38% more productive than an isolated hectare under comparable environmental conditions. Counterfactual analyses indicate that existing fragmentation has already reduced total annual forest productivity across the conterminous USA by 0.16 GtC yr−1 or 14% relative to an upper-bound configuration of large, contiguous forests. Random forest models identify patch size as a stronger predictor of net primary productivity than topographic and soil variables. Extending the analysis globally with coarser-resolution data, we find consistent positive relationships between patch size and per-area productivity for both net and gross primary productivity. These results demonstrate that forest fragmentation can reduce carbon sequestration without net forest-area loss, highlighting the need to account for fragmentation—not just forest cover—in climate mitigation strategies.

Source: Nature Ecology & Evolution