Biodiversity buffers forest ecosystems from compound climate extremes
Xuetao Qiao · Jakob Zscheischler · Michel Loreau · Peter B. Reich · Qingqing Chen · Michael Bahn · Neha Mohanbabu · Hannah J. White · Mingyu Luo · Shaopeng Wang
期刊Nature Communications / Nat Commun
类型Article
发表2026-07-13 (online)
卷期页17(1): 8601
收录2026-10-09
被浏览…
关键词compound extreme eventsheat and droughtforest recoverytree species diversityforest stabilitysynergistic effectforest carbon
标签—
Forest ecosystems provide key ecological, social, and climate benefits, yet their functions are potentially threatened by rising compound extremes characterised by simultaneous heat and drought or moisture excess. Here, we explore how forests resist and recover from compound heat and moisture extremes and how tree diversity mediates these responses, using U.S. forest inventory data along with satellite-based stability estimates. We find that compound extremes result in lower forest resistance and recovery than individual extremes. Compound extremes tend to exhibit synergistic effects, causing greater negative impacts than the sum of individual extremes. Compound heat and drought have slightly stronger destabilising effects than compound heat and moisture excess. Higher tree diversity consistently improves resistance and recovery under both individual and compound extremes. Our findings suggest that focusing solely on individual climate extremes is likely to underestimate the impacts of compound extremes and to increase uncertainty in projections of terrestrial carbon-cycle feedbacks, while also highlighting the critical role of tree diversity in mitigating these threats under future climate conditions.

@article{qiao2026,
author = {Xuetao Qiao and Jakob Zscheischler and Michel Loreau and Peter B. Reich and Qingqing Chen and Michael Bahn and Neha Mohanbabu and Hannah J. White and Mingyu Luo and Shaopeng Wang},
title = {Biodiversity buffers forest ecosystems from compound climate extremes},
journal = {Nature Communications},
year = {2026},
volume = {17},
number = {1},
pages = {8601},
doi = {10.1038/s41467-026-75370-5},
publisher = {Springer Science and Business Media LLC},
}· END ·