Biogeosciences (Aug 2024)

Compound soil and atmospheric drought (CSAD) events and CO<sub>2</sub> fluxes of a mixed deciduous forest: the occurrence, impact, and temporal contribution of main drivers

  • L. Scapucci,
  • A. Shekhar,
  • S. Aranda-Barranco,
  • A. Bolshakova,
  • L. Hörtnagl,
  • M. Gharun,
  • N. Buchmann

DOI
https://doi.org/10.5194/bg-21-3571-2024
Journal volume & issue
Vol. 21
pp. 3571 – 3592

Abstract

Read online

With global warming, forests are increasingly exposed to “compound soil and atmospheric drought” (CSAD) events, characterized by low soil water content (SWC) and high vapour pressure deficit (VPD). Such CSAD events trigger responses in both ecosystem and forest-floor CO2 fluxes, which we know little about. In this study, we used multi-year daily and daytime above-canopy (18 years; 2005–2022) and daily forest-floor (5 years; 2018–2022) eddy covariance CO2 fluxes from a Swiss forest site by the name of CH-Lae (a mixed deciduous montane forest). The objectives were (1) to characterize CSAD events at CH-Lae, (2) to quantify the impact of CSAD events on ecosystem and forest-floor CO2 fluxes, and (3) to identify the major drivers and their temporal contributions to changing ecosystem and forest-floor CO2 fluxes during CSAD events and CSAD growing seasons. Our results showed that the growing seasons of 2015, 2018, and 2022 were the three driest at CH-Lae since 2005 (referred to as the CSAD years), exhibiting similar intensity and duration of the CSAD events but considerably different pre-drought conditions. The CSAD events reduced daily mean net ecosystem productivity (NEP) in all 3 CSAD years by about 38 % compared to the long-term mean, with the highest reduction observed during 2022 (41 %). This reduction in daily mean NEP was largely due to decreased gross primary productivity (GPP; > 16 % below the long-term mean) rather than increased ecosystem respiration (Reco) during CSAD events. Furthermore, forest-floor respiration (Rff) decreased during the CSAD events in 2018 and 2022 (with no measurements in 2015), with a larger reduction in 2022 (41 %) than in 2018 (16 %), relative to the long-term mean (2019–2021). Using data-driven machine learning methods, we identified the major drivers of NEP and Rff during CSAD events. While daytime mean NEP (NEPDT) during the 2015 and 2018 CSAD events was limited by VPD and SWC, respectively, NEPDT during the 2022 CSAD event was strongly limited by both SWC and VPD. Air temperature had negative effects, while net radiation showed positive effects on NEPDT during all CSAD events. Daily mean Rff during the 2018 CSAD event was driven by soil temperature and SWC but was severely limited by SWC during the 2022 CSAD event. We found that a multi-layer analysis of CO2 fluxes in forests is necessary to better understand forest responses to CSAD events, particularly if the first signs of NEP acclimation to CSAD events – evident in our forest – are also found elsewhere. We conclude that CSAD events have multiple drivers with different temporal contributions, making predictions about site-specific CSAD events and long-term forest responses to such conditions more challenging.