• Energy Research

Adopting a multiscale perspective that connects forest dynamics from tree stands to landscapes is crucial for understanding how forest ecosystems will evolve under global environmental change. This commentary highlights the significance of Perret et al.'s (2025) study in providing valuable insights into how individual tree plasticity drives community reorganization and ultimately alters ecosystem resilience, via integrating individual tree species into community-level analyses. Their study has important implications for modeling and predicting forest resilience, as well as for informing sustainable management strategies in response to future climate change and increasing disturbances.

The frequency and intensity of droughts have increased over the decades, leading to increased forest decline. The response of forest to drought can be evaluated by both its sensitivity to drought (resistance) and its post-drought recovery rate (resilience). However, it remains uncertain how drought resistance and resilience of forests change over time under climate change. We assessed the spatiotemporal dynamics of forest resistance and resilience to drought over the past century (1901-2015) with global tree ring data records from 2,935 sites, in conjunction with plant trait data. We found that gymnosperms and angiosperms showed different spatial patterns of drought resistance and resilience, driven by variations in eco-physiological traits. Resistance and resilience also varied with drought seasonal timing. Surprisingly, we found that the trade-off between resistance and resilience for gymnosperms, previously reported only spatially, also occurred at the temporal scale. In particular, drought resilience markedly increased, but resistance decreased, for gymnosperms between 1950-1969 and 1990-2009, indicating that previous model simulations assuming invariant resistance may have underestimated the impacts of drought on gymnosperm-dominated forests under future climate change.

ABSTRACT Identifying the thresholds of drought that, if crossed, suppress vegetation functioning is vital for accurate quantification of how land ecosystems respond to climate variability and change. We present a globally applicable framework to identify drought thresholds for vegetation responses to different levels of known soil-moisture deficits using four remotely sensed vegetation proxies spanning 2001–2018. The thresholds identified represent critical inflection points for changing vegetation responses from highly resistant to highly vulnerable in response to drought stress, and as a warning signal for substantial vegetation impacts. Drought thresholds varied geographically, with much lower percentiles of soil-moisture anomalies in vegetated areas covered by more forests, corresponding to a comparably stronger capacity to mitigate soil water deficit stress in forested ecosystems. Generally, those lower thresholds are detected in more humid climates. State-of-the-art land models, however, overestimated thresholds of soil moisture (i.e. overestimating drought impacts), especially in more humid areas with higher forest covers and arid areas with few forest covers. Based on climate model projections, we predict that the risk of vegetation damage will increase by the end of the twenty-first century in some hotspots like East Asia, Europe, Amazon, southern Australia and eastern and southern Africa. Our data-based results will inform projections on future drought impacts on terrestrial ecosystems and provide an effective tool for drought management.

AbstractAccurate quantification of vegetation carbon turnover time (τveg) is critical for reducing uncertainties in terrestrial vegetation response to future climate change. However, in the absence of global information of litter production, τveg could only be estimated based on net primary productivity under the steady‐state assumption. Here, we applied a machine‐learning approach to derive a global dataset of litter production by linking 2401 field observations and global environmental drivers. Results suggested that the observation‐based estimate of global natural ecosystem litter production was 44.3 ± 0.4 Pg C year−1. By contrast, land‐surface models (LSMs) overestimated the global litter production by about 27%. With this new global litter production dataset, we estimated global τveg (mean value 10.3 ± 1.4 years) and its spatial distribution. Compared to our observation‐based τveg, modelled τveg tended to underestimate τveg at high latitudes. Our empirically derived gridded datasets of litter production and τveg will help constrain global vegetation models and improve the prediction of global carbon cycle.

AbstractChanging amplitude of the seasonal cycle of atmospheric CO2 (SCA) in the northern hemisphere is an emerging carbon cycle property. Mauna Loa (MLO) station (20°N, 156°W), which has the longest continuous northern hemisphere CO2 record, shows an increasing SCA before the 1980s (p < .01), followed by no significant change thereafter. We analyzed the potential driving factors of SCA slowing‐down, with an ensemble of dynamic global vegetation models (DGVMs) coupled with an atmospheric transport model. We found that slowing‐down of SCA at MLO is primarily explained by response of net biome productivity (NBP) to climate change, and by changes in atmospheric circulations. Through NBP, climate change increases SCA at MLO before the 1980s and decreases it afterwards. The effect of climate change on the slowing‐down of SCA at MLO is mainly exerted by intensified drought stress acting to offset the acceleration driven by CO2 fertilization. This challenges the view that CO2 fertilization is the dominant cause of emergent SCA trends at northern sites south of 40°N. The contribution of agricultural intensification on the deceleration of SCA at MLO was elusive according to land–atmosphere CO2 flux estimated by DGVMs and atmospheric inversions. Our results also show the necessity to adequately account for changing circulation patterns in understanding carbon cycle dynamics observed from atmospheric observations and in using these observations to benchmark DGVMs.

AbstractTemperature extremes exert a significant influence on terrestrial ecosystems, but the precise levels at which these extremes trigger adverse shifts in vegetation productivity have remained elusive. In this study, we have derived two critical thresholds, using standard deviations (SDs) of growing‐season temperature and satellite‐based vegetation productivity as key indicators. Our findings reveal that, on average, vegetation productivity experiences rapid suppression when confronted with temperature anomalies exceeding 1.45 SD above the mean temperature during 2001–2018. Furthermore, at temperatures exceeding 2.98 SD above the mean, we observe the maximum level of suppression, particularly in response to the most extreme high‐temperature events. When Earth System Models are driven by a future medium emission scenario, they project that mean temperatures will routinely surpass both of these critical thresholds by approximately the years 2050 and 2070, respectively. However, it is important to note that the timing of these threshold crossings exhibits spatial variation and will appear much earlier in tropical regions. Our finding highlights that restricting global warming to just 1.5°C can increase safe areas for vegetation growth by 13% compared to allowing warming to reach 2°C above preindustrial levels. This mitigation strategy helps avoid exposure to detrimental extreme temperatures that breach these thresholds. Our study underscores the pivotal role of climate mitigation policies in fostering the sustainable development of terrestrial ecosystems in a warming world.