Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Diverging responses of terrestrial ecosystems to water stress after disturbances

Nature Climate Change volume 15, pages 73–79 (2025)Cite this article

Subjects

Abstract

Terrestrial ecosystems are major carbon (C) pools, sequestering ~20% of anthropogenic C emissions. However, increasing frequency and intensity of climate-sensitive disturbances (for example, drought and wildfire) threaten long-term C uptake. Although direct effects of disturbances are well-documented, indirect effects remain unknown. Here we quantify changes in the sensitivity of terrestrial gross primary production to water stress before and after severe droughts and fires. We find divergent changes across the globe, where dry regions have increased sensitivity, while wet regions have decreased sensitivity. Water availability, solar radiation, nutrient availability and biodiversity are the main drivers mediating these changes. Sensitivity takes ~4–5 years to recover after disturbances, but the increasing frequency of disturbances threatens this recovery. Our results reveal strong cross-system discrepancies in ecosystem responses to disturbances, highlighting the vulnerability of dryland ecosystems in future climates.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on SpringerLink
  • Instant access to the full article PDF.

USD 39.95

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Large spatial heterogeneities in the change in drought sensitivity after disturbances.
Fig. 2: Drought sensitivity in dry regions increased significantly after disturbances.
Fig. 3: Climate is the main driver of the change in drought sensitivity.
Fig. 4: Drought sensitivity recovers ~4–5 years after disturbances.

Similar content being viewed by others

Data availability

The GLASS GPP data are from http://www.glass.umd.edu/Download.html. The EC-LUE GPP data are available via figshare at https://doi.org/10.6084/m9.figshare.8942336.v3 (ref. 50). The NIRv GPP data are available via figshare at https://doi.org/10.6084/m9.figshare.12981977.v2 (ref. 51). The P-model GPP data are available via Zenodo at https://doi.org/10.5281/zenodo.1423483 (ref. 52). The BESS (v.2.0) GPP data are from https://www.environment.snu.ac.kr/data. The TRENDY-v11 GPP data are from https://blogs.exeter.ac.uk/trendy/. The historical PDSI data and the climatic data (temperature, precipitation and downward shortwave solar radiation) are from TerraClimate (https://www.climatologylab.org/terraclimate.html). Future PDSI data are from NCAR (https://rda.ucar.edu/datasets/ds299.0/) and CarbonPlan (https://carbonplan.org/). The GFED4.1s burned area data are available at https://www.globalfiredata.org/index.html. Global data for the aridity index are available via figshare at https://doi.org/10.6084/m9.figshare.7504448.v6 (ref. 53). Global soil moisture (0–100 cm) data are downloaded from ERA5-land (https://cds.climate.copernicus.eu/datasets). GIMMS LAI4g data are available via Zenodo at https://doi.org/10.5281/zenodo.7649107 (ref. 54). CO2 data are available via Zenodo at https://doi.org/10.5281/zenodo.5021360 (ref. 55). Global data for soil total nitrogen and soil organic carbon (0–100 cm) are from https://daac.ornl.gov/SOILS/guides/IGBP-SurfaceProducts.html. Soil organic phosphorus (0–50 cm) data are from https://daac.ornl.gov/SOILS/guides/Global_Phosphorus_Dist_Map.html. Soil cation exchange capacity and nitrogen deposition data are from https://daac.ornl.gov/NACP/guides/NACP_MsTMIP_Model_Driver.html. AGB data are available via Zenodo at https://doi.org/10.5281/zenodo.4161693 (ref. 56). The biodiversity data are from https://anthroecology.org/anthromes/plantbiodiversity/.

Code availability

All analysis was done in the open-source software R v.4.1.3. The code is available via figshare at https://doi.org/10.6084/m9.figshare.25482928 (ref. 57).

References

  1. Friedlingstein, P. et al. Global carbon budget 2022. Earth Syst. Sci. Data 14, 4811–4900 (2022).

    Google Scholar 

  2. Ciais, P. et al. Europe-wide reduction in primary productivity caused by the heat and drought in 2003. Nature 437, 529–533 (2005).

    CAS  Google Scholar 

  3. van der Woude, A. M. et al. Temperature extremes of 2022 reduced carbon uptake by forests in Europe. Nat. Commun. 14, 6218 (2023).

    Google Scholar 

  4. Anderegg, W. R. L. et al. Tree mortality from drought, insects, and their interactions in a changing climate. New Phytol. 208, 674–683 (2015).

    Google Scholar 

  5. Anderegg, W. R. L. et al. A climate risk analysis of Earth’s forests in the 21st century. Science 377, 1099–1103 (2022).

    CAS  Google Scholar 

  6. Penuelas, J. Decreasing efficiency and slowdown of the increase in terrestrial carbon-sink activity. One Earth 6, 591–594 (2023).

    Google Scholar 

  7. Dai, A. Drought under global warming: a review. WIREs Clim. Change 2, 45–65 (2011).

    Google Scholar 

  8. Cooper, L. A., Ballantyne, A. P., Holden, Z. A. & Landguth, E. L. Disturbance impacts on land surface temperature and gross primary productivity in the western United States. J. Geophys. Res. Biogeosci. 122, 930–946 (2017).

    Google Scholar 

  9. Frank, D. et al. Effects of climate extremes on the terrestrial carbon cycle: concepts, processes and potential future impacts. Glob. Change Biol. 21, 2861–2880 (2015).

    Google Scholar 

  10. Anderegg, W. R. L. et al. Pervasive drought legacies in forest ecosystems and their implications for carbon cycle models. Science 349, 528–532 (2015).

    CAS  Google Scholar 

  11. Fu, Z. et al. Atmospheric dryness reduces photosynthesis along a large range of soil water deficits. Nat. Commun. 13, 989 (2022).

    CAS  Google Scholar 

  12. Flack-Prain, S., Meir, P., Malhi, Y., Smallman, T. L. & Williams, M. The importance of physiological, structural and trait responses to drought stress in driving spatial and temporal variation in GPP across Amazon forests. Biogeosciences 16, 4463–4484 (2019).

    CAS  Google Scholar 

  13. Zhang, Y. et al. Canopy and physiological controls of GPP during drought and heat wave. Geophys. Res. Lett. 43, 3325–3333 (2016).

    Google Scholar 

  14. Detto, M. & Pacala, S. W. Plant hydraulics, stomatal control and the response of a tropical forest to water stress over multiple temporal scales. Glob. Change Biol. 28, 4359–4376 (2022).

    CAS  Google Scholar 

  15. Liu, M., Trugman, A. T., Peñuelas, J. & Anderegg, W. R. L. Climate-driven disturbances amplify forest drought sensitivity. Nat. Clim. Change 14, 746–752 (2024).

    CAS  Google Scholar 

  16. Palmer, W. C. Meteorological Drought (US Department of Commerce Weather Bureau, 1965).

  17. Beguería, S., Vicente-Serrano, S. M., Reig, F. & Latorre, B. Standardized precipitation evapotranspiration index (SPEI) revisited: parameter fitting, evapotranspiration models, tools, datasets and drought monitoring. Int. J. Climatol. 34, 3001–3023 (2014).

    Google Scholar 

  18. Breiman, L. Random forests. Mach. Learn. 45, 5–32 (2001).

    Google Scholar 

  19. Li, B. et al. BESSv2.0: a satellite-based and coupled-process model for quantifying long-term global land–atmosphere fluxes. Remote Sens. Environ. 295, 113696 (2023).

    Google Scholar 

  20. Stocker, B. D. et al. P-model v1.0: an optimality-based light use efficiency model for simulating ecosystem gross primary production. Geosci. Model Dev. 13, 1545–1581 (2020).

    Google Scholar 

  21. Li, F. et al. Global water use efficiency saturation due to increased vapor pressure deficit. Science 381, 672–677 (2023).

    CAS  Google Scholar 

  22. Wang, S., Zhang, Y., Ju, W., Qiu, B. & Zhang, Z. Tracking the seasonal and inter-annual variations of global gross primary production during last four decades using satellite near-infrared reflectance data. Sci. Total Environ. 755, 142569 (2020).

    Google Scholar 

  23. Liang, S. et al. The Global Land Surface Satellite (GLASS) Product Suite. Bull. Am. Meteorol. Soc. 102, E323–E337 (2021).

    Google Scholar 

  24. Zheng, Y. et al. Improved estimate of global gross primary production for reproducing its long-term variation, 1982–2017. Earth Syst. Sci. Data 12, 2725–2746 (2020).

    Google Scholar 

  25. van der Werf, G. R. et al. Global fire emissions estimates during 1997–2016. Earth Syst. Sci. Data 9, 697–720 (2017).

    Google Scholar 

  26. Ellis, E. C., Antill, E. C. & Kreft, H. All is not loss: plant biodiversity in the Anthropocene. PLoS ONE 7, e30535 (2012).

    CAS  Google Scholar 

  27. Sheil, D. Disturbance and distributions: avoiding exclusion in a warming world. Ecol. Soc. www.jstor.org/stable/26270356 (2016).

  28. Feeley, K. J., Davies, S. J., Perez, R., Hubbell, S. P. & Foster, R. B. Directional changes in the species composition of a tropical forest. Ecology 92, 871–882 (2011).

    Google Scholar 

  29. Isbell, F. et al. Biodiversity increases the resistance of ecosystem productivity to climate extremes. Nature 526, 574–577 (2015).

    CAS  Google Scholar 

  30. Zhang, Y. et al. Increasing sensitivity of dryland vegetation greenness to precipitation due to rising atmospheric CO2. Nat. Commun. 13, 4875 (2022).

    CAS  Google Scholar 

  31. Li, W. et al. Widespread increasing vegetation sensitivity to soil moisture. Nat. Commun. 13, 3959 (2022).

    CAS  Google Scholar 

  32. Wang, L. et al. Dryland productivity under a changing climate. Nat. Clim. Change 12, 981–994 (2022).

    Google Scholar 

  33. Anderegg, W. R. L., Martinez-Vilalta, J., Mencuccini, M. & Poyatos, R. Community assembly influences plant trait economic spectra and functional trade-offs at ecosystem scales. Proc. Natl Acad. Sci. USA 121, e2404034121 (2024).

    CAS  Google Scholar 

  34. Pimm, S. L. The complexity and stability of ecosystems. Nature 307, 321–326 (1984).

    Google Scholar 

  35. Stocker, B. D. et al. Drought impacts on terrestrial primary production underestimated by satellite monitoring. Nat. Geosci. 12, 264–270 (2019).

    CAS  Google Scholar 

  36. Yuan, W. et al. Global estimates of evapotranspiration and gross primary production based on MODIS and global meteorology data. Remote Sens. Environ. 114, 1416–1431 (2010).

    Google Scholar 

  37. Abatzoglou, J. T., Dobrowski, S. Z., Parks, S. A. & Hegewisch, K. C. TerraClimate, a high-resolution global dataset of monthly climate and climatic water balance from 1958–2015. Sci. Data 5, 170191 (2018).

    Google Scholar 

  38. van der Schrier, G., Barichivich, J., Briffa, K. R. & Jones, P. D. A scPDSI-based global data set of dry and wet spells for 1901–2009. J. Geophys. Res. Atmos. 118, 4025–4048 (2013).

    Google Scholar 

  39. Anderegg, W. R. L., Trugman, A. T., Badgley, G., Konings, A. G. & Shaw, J. Divergent forest sensitivity to repeated extreme droughts. Nat. Clim. Change 10, 1091–1095 (2020).

    Google Scholar 

  40. Dai, A. Dai Global Palmer Drought Severity Index (PDSI). NCAR https://doi.org/10.5065/D6QF8R93 (2017).

  41. Anderegg, W. R. L. et al. Future climate risks from stress, insects and fire across US forests. Ecol. Lett. 25, 1510–1520 (2022).

    Google Scholar 

  42. Aitken, A. C. On least squares and linear combination of observations. Proc. R. Soc. Edinb. 55, 42–48 (1936).

    Google Scholar 

  43. Cheng, W. et al. Global monthly gridded atmospheric carbon dioxide concentrations under the historical and future scenarios. Sci. Data 9, 83 (2022).

    CAS  Google Scholar 

  44. Zomer, R. J., Xu, J. & Trabucco, A. Version 3 of the global aridity index and potential evapotranspiration database. Sci. Data 9, 409 (2022).

    Google Scholar 

  45. Xu, L. et al. Changes in global terrestrial live biomass over the 21st century. Sci. Adv. 7, eabe9829 (2021).

    CAS  Google Scholar 

  46. Cao, S. et al. Spatiotemporally consistent global dataset of the GIMMS leaf area index (GIMMS LAI4g) from 1982 to 2020. Earth Syst. Sci. Data 15, 4877–4899 (2023).

    Google Scholar 

  47. Olson, D. M. et al. Terrestrial ecoregions of the world: a new map of life on Earth. BioScience 51, 933–938 (2001).

    Google Scholar 

  48. Štrumbelj, E. & Kononenko, I. Explaining prediction models and individual predictions with feature contributions. Knowl. Inf. Syst. 41, 647–665 (2014).

    Google Scholar 

  49. Swann, A. L. S., Hoffman, F. M., Koven, C. D. & Randerson, J. T. Plant responses to increasing CO2 reduce estimates of climate impacts on drought severity. Proc. Natl Acad. Sci. USA 113, 10019–10024 (2016).

    CAS  Google Scholar 

  50. Zheng, Y. et al. Improved estimate of global gross primary production for reproducing its long-term variation, 1982–2017. figshare https://doi.org/10.6084/m9.figshare.8942336.v3 (2020).

  51. Wang, S., Zhang, Y. & Ju, W. Long-term (1982-2018) global gross primary production dataset based on NIRv. figshare https://doi.org/10.6084/m9.figshare.12981977.v2 (2020).

  52. Stocker, B. GPP: site-scale and global model outputs from P-model used for Stocker et al. (2019) Nature Geosci. (v1.0) [Dataset]. Zenodo https://doi.org/10.5281/zenodo.1423484 (2018).

  53. Zomer, R. & Trabucco, A. Global aridity index and potential evapotranspiration (ET0) database: version 3 [Dataset]. figshare https://doi.org/10.6084/m9.figshare.7504448.v6 (2019).

  54. Cao, S. et al. Spatiotemporally consistent global dataset of the GIMMS Leaf Area Index (GIMMS LAI4g) from 1982 to 2020 (V1.0) [Dataset]. Zenodo https://doi.org/10.5281/zenodo.7649108 (2023).

  55. Cheng, W. et al. Global monthly distributions of atmospheric CO2 concentrations under the historical and future scenarios [Dataset]. Zenodo https://doi.org/10.5281/zenodo.5021361 (2021).

  56. Xu, L. et al. Dataset for "Changes in global terrestrial live biomass over the 21st century" (2.0) [Dataset]. Zenodo https://doi.org/10.5281/zenodo.4161694 (2021).

  57. Liu, M. et al. Responses of global terrestrial ecosystems to water stress. figshare https://doi.org/10.6084/m9.figshare.25482928 (2024).

Download references

Acknowledgements

This study was supported by the Wilkes Center at the University of Utah, with thanks to the Anderegg laboratory. We also thank the TRENDY team. J.P. was supported by the PID2022-140808NB-I00 and TED2021-132627 B–I00 grants funded by MCIN of Spain and the European Union NextGeneration EU/PRTR and by the Catalan Government grant AGAUR2023 CLIMA 00118. A.T.T. acknowledges funding from National Science Foundation grant nos. 2003205, 2017949 and 2216855, from the USDI Park Service Award nos. P24AC00910 and P24AC01425, from the University of California Laboratory Fees Research Program Award no. LFR-20-652467 and from the Gordon and Betty Moore Foundation grant no. GBMF11974. W.R.L.A. acknowledges support from the David and Lucille Packard Foundation and US National Science Foundation grant nos. 1802880, 2003017 and 2044937, as well as the Alan T. Waterman award IOS-2325700. G.V.G. acknowledges support from the NOAA Climate and Global Change postdoctoral fellowship administered by UCAR Cooperative Programs for the Advancement of Earth System Science under the NOAA Science Collaboration Program Award no. NA21OAR4310383.

Author information

Authors and Affiliations

  1. School of Biological Sciences, University of Utah, Salt Lake City, UT, USA

    Meng Liu, German Vargas G, Linqing Yang & William R. L. Anderegg

  2. The Wilkes Center for Climate Science and Policy, University of Utah, Salt Lake City, UT, USA

    Meng Liu, German Vargas G, Linqing Yang & William R. L. Anderegg

  3. CREAF, Cerdanyola del Vallès, Barcelona, Spain

    Josep Peñuelas

  4. CSIC, Global Ecology Unit CREAF-CSIC-UAB, Bellaterra, Barcelona, Spain

    Josep Peñuelas

  5. Department of Geography, University of California Santa Barbara, Santa Barbara, CA, USA

    Anna T. Trugman

  6. Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR, USA

    German Vargas G

  7. Department of Forest Ecosystems & Society, Oregon State University, Corvallis, OR, USA

    German Vargas G

Authors
  1. Meng Liu
    View author publications

    Search author on:PubMed Google Scholar

  2. Josep Peñuelas
    View author publications

    Search author on:PubMed Google Scholar

  3. Anna T. Trugman
    View author publications

    Search author on:PubMed Google Scholar

  4. German Vargas G
    View author publications

    Search author on:PubMed Google Scholar

  5. Linqing Yang
    View author publications

    Search author on:PubMed Google Scholar

  6. William R. L. Anderegg
    View author publications

    Search author on:PubMed Google Scholar

Contributions

M.L. and W.R.L.A. conceptualized, designed and improved the study with input from all co-authors. M.L. wrote the initial draft and J.P., A.T.T., G.V.G., L.Y. and W.R.L.A. discussed the design, analyses and results and provided extensive and valuable comments and revisions.

Corresponding author

Correspondence to Meng Liu.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Climate Change thanks the anonymous reviewers for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Change in drought sensitivity using the five remote sensing GPP products separately.

The change in drought sensitivity after disturbances is derived from each GPP product (a-j) using PDSI to indicate water stress. The distributional maps are aggregated to a resolution of 1° for visual display. ‘Mean’ is the mean change in sensitivity derived from GLS models. *, p < 0.05 (two-sided) based on the GLS models. Multiple comparisons are not applicable. Basemaps from Natural Earth (https://www.naturalearthdata.com/downloads/110m-physical-vectors).

Extended Data Fig. 2 Change in drought sensitivity based on SPEI and TRENDY GPP data.

(a-b) The changes in drought sensitivity after (a) severe droughts and (b) fires using SPEI to indicate water stress (remote sensing GPP regressed by SPEI). (c-d) The changes in drought sensitivity using TRENDY GPP (regressed by PDSI). The distributional maps are aggregated to 1° for visual display. ‘Mean’ is the mean change in sensitivity derived from GLS models. *, p < 0.05 (two-sided) based on the GLS models. Multiple comparisons are not applicable. Basemaps from Natural Earth (https://www.naturalearthdata.com/downloads/110m-physical-vectors).

Extended Data Fig. 3 Distribution of the aridity levels.

The aridity levels are defined based on the aridity index (AI): hyperarid (AI < 0.05), arid (AI < 0.2), semi-arid (AI < 0.5), dry sub-humid (AI < 0.65), and humid (AI ≥ 0.65). Basemaps from Natural Earth (https://www.naturalearthdata.com/downloads/110m-physical-vectors).

Extended Data Fig. 4 Changes in drought sensitivity in the aridity levels when using SPEI and TRENDY GPP.

(a) The change in drought sensitivity using SPEI to indicate water stress (remote sensing GPP regressed by SPEI) (left to right, N = 58, 2893, 3194, 1044, 2982 for drought; N = 65, 2077, 1868, 532, 670 for fire). (b) The change in drought sensitivity using TRENDY GPP data (regressed by PDSI) (left to right, N = 37, 2209, 3291, 1437, 4412 for drought; N = 66, 2133, 2223, 884, 1171 for fire). The height of each bar indicates the mean change in sensitivity derived from GLS models, and the error bar shows one standard error. *, p < 0.05 (two-sided) based on the GLS models. Multiple comparisons are not applicable.

Extended Data Fig. 5 Response of the change in drought sensitivity to the most important predictors for severe droughts.

The eight most important predictors (a-h) are shown, where the smoothing curves are fitted by generalized additive models in the ‘ggplot2’ package in R, and the shading represents the 95% confidence interval. The Shapley value indicates the response of the change in sensitivity to the predictors.

Extended Data Fig. 6 Response of the change in drought sensitivity to the most important predictors for fires.

The eight most important predictors (a-h) are shown, where the smoothing curves are fitted by generalized additive models in the ‘ggplot2’ package in R, and the shading represents the 95% confidence interval. The Shapley value indicates the response of the change in sensitivity to the predictors.

Extended Data Fig. 7 Return intervals of severe droughts and fires.

(a-b) Drought return intervals using historical PDSI data in (a) 1982–2018 and (b) 1958–1981. (c) Fire return intervals using the burned area from GFED 4.1 s. The bin width is one year. The median return intervals for panels (a-c) are 9.25, 12.0, and 6.67 years, respectively.

Extended Data Fig. 8 Drought return intervals using future PDSI data in 2061–2100.

Boxplots of drought return intervals based on PDSI from (a) NCAR (left to right, N = 22976, 76923) and (b) CarbonPlan (left to right, N = 7772, 10166) under SSP245 and SSP585. The solid lines indicate the recovery time (five years after severe droughts). Box plot lines represent the interquartile range (IQR) and median, respectively, whereas the whiskers represent 1.5 times IQR (or the minimum/maximum).

Extended Data Fig. 9 The best model at the global scale.

(a) Correlations between the GPP anomalies and PDSI. (b) The distribution of pixels with significant correlations (50% of pixels indicating significant correlations). There are 146210 and 73717 pixels in panels (a) and (b), respectively. p values (two-sided) are based on cor.test in R. Multiple comparisons are not applicable. (c) The best model for the significant pixels, where there are three models: linear (blue), quadratic (orange), and logistic (red) models. The best model is defined as the model with minimum Akaike Information Criterion (AIC) and significant regression coefficients (for example the linear model is used when the quadratic term in the quadratic model is not significant). The linear model is the best for 88% of the pixels. Basemaps from Natural Earth (https://www.naturalearthdata.com/downloads/110m-physical-vectors).

Extended Data Fig. 10 Results of simple linear regression (SLR) and multiple linear regression (MLR) are comparable.

Scatter plot of the coefficients (that is sensitivity) of PDSI from SLR (GPPanomaly ~ PDSI) and MLR (GPPanomaly ~ Sradanomaly + Tanomaly + PDSI) based on data in 1982–2018. Each dot is a pixel, and only pixels with significant correaltions between GPPanomaly and PDSI are shown. There are 84.85% of pixels exhibiting positive correlations between GPP anomalies and PDSI. The p value (two-sided) is based on linear regression. Multiple comparisons are not applicable.

Supplementary information

Supplementary Information (download PDF )

Supplementary Methods, Figs. 1–8 and Tables 1–3.

Reporting Summary (download PDF )

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Liu, M., Peñuelas, J., Trugman, A.T. et al. Diverging responses of terrestrial ecosystems to water stress after disturbances. Nat. Clim. Chang. 15, 73–79 (2025). https://doi.org/10.1038/s41558-024-02191-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • Version of record:

  • Issue date:

  • DOI: https://doi.org/10.1038/s41558-024-02191-z

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing