The information held in annual tree rings offers further insight into the potential of the world's forests to slow global warming (see Nature 523, 20–22; 2015). These data reveal fluctuations in growth rate caused by climatic, physiological and ecological factors, and provide long-term and spatially extensive records.

Tree-ring measurements provide information on functional trade-offs that can affect a tree's future, such as altered hydraulic and growth responses to the long-term rise in atmospheric carbon dioxide (see P. van der Sleen et al. Nature Geosci. 8, 24–28; 2015). And unlike data from remote sensing of forests and monitoring of tree diameters, tree-ring observations extend over centuries.

Such studies also mitigate the problems you mention of investigating small forest patches, because the size and number of research areas can be adjusted.

A relatively small percentage of ring-forming species in tropical forests and the fading record of tree-ring collections further back in time are impediments. Also, most existing tree-ring collections are not representative of trees that lived in the past (F. Babst et al. Oecologia 176, 307–322; 2014).

We suggest that combining retrospective tree-ring analysis, repeated forest inventory measurements and modelling studies could more effectively predict responses of the world's forests to climate change.