Evidence for soil water control on carbon and water dynamics in European forests during the extremely dry year: 2003

André Granier, Markus Reichstein, Nathalie Bréda, Ivan A. Janssens, Eva Falge, Philippe Ciais, Thomas Grünwald, Marc Aubinet, Paul Berbigier, Christian Bernhofer, Nina Buchmann, Osvaldo Facini, Giacomo Grassi, Bernard Heinesch, Hannu Ilvesniemi, P. Keronen, Alexander Knohl, Barbara Köstner, Fredrik Lagergren, Anders Lindroth, Bernard Longdoz, Denis Loustau, J. Mateus, Leonardo Montagnani, Claude Nys, Eddy Moors, Dario Papale, Marianne Peiffer, Kim Pilegaard, Gabriel Pita, Jukka Pumpanen, Serge Rambal, Corinna Rebmann, Abel Rodrigues, G. Seufert, John Tenhunen, Timo Vesala, Q. WangAgricultural 2007 · DOI

TL;DR

The exceptionally dry and hot summer of 2003 provided a unique opportunity to study the impact of drought on European forests. This study analyzed carbon and water fluxes across 12 forest sites, revealing that soil water availability is a critical control on forest functioning, significantly reducing both carbon uptake and water transpiration. The findings highlight the vulnerability of forests to extreme climate events and suggest that future climate change may lead to reduced productivity and even a shift from carbon sinks to sources.

What Problem Does This Paper Address?

Forest ecosystems play a crucial role in the global carbon cycle, acting as significant carbon sinks. However, their functioning is highly sensitive to climatic conditions, particularly water availability. While the influence of average climate on forest productivity is well-established, the specific impacts of extreme events like severe droughts on carbon and water dynamics are less understood. The summer of 2003 in Europe was characterized by an unprecedented heatwave and drought, offering a valuable natural experiment to investigate these impacts. This paper aims to quantify the effects of this extreme drought on carbon and water fluxes across a range of European forest types, to understand the underlying mechanisms, and to assess the sensitivity of different forest ecosystems to water stress.

Methods

The study employed a multi-faceted approach combining direct flux measurements with sophisticated modeling.

  • Eddy Covariance Technique: Continuous measurements of carbon dioxide (CO2) and water vapor fluxes were conducted at 12 European forest sites using the eddy covariance technique. This standard method quantifies the exchange of gases between the ecosystem and the atmosphere.
  • Water Balance Modeling: To overcome the limitation of soil water content measurements being available at only a few sites, a daily water balance model (BILJOU) was implemented. This model estimated key water balance terms, including transpiration, interception, drainage, and crucially, soil water content. The model was driven by daily climate data and site-specific parameters (leaf area index, phenology, soil characteristics, root distribution).
  • Drought Characterization: The water balance model allowed for the calculation of water stress indicators: the onset date of water stress (when relative extractable water, REW, dropped below 0.4), the duration of stress, and its intensity.
  • Flux Partitioning: Net ecosystem exchange (NEE) was partitioned into gross primary production (GPP) and ecosystem respiration (TER) using a robust algorithm that accounts for temperature dependence and other environmental factors.
  • Statistical Analysis: Non-linear regression models were used to relate daily TER and GPP variations to climatic factors (air temperature, global radiation) and the modeled soil water content (REW).

Datasets

The study utilized data from 16 European forest sites, 12 of which had continuous eddy covariance flux measurements. Climate data (rainfall, air temperature, humidity, wind speed, global radiation) were collected at these sites. Soil water content was measured using TDR probes at a subset of sites. Tree growth data were available from three sites. The study also incorporated data from previous years (2002) for comparison to assess the impact of the 2003 drought.

Key Insights and Ideas

  • REW as a Critical Threshold: The study consistently found that when relative extractable soil water (REW) dropped below approximately 0.4, both water vapor fluxes (transpiration) and carbon fluxes (GPP) were significantly reduced. This suggests a critical threshold for stomatal closure and reduced photosynthetic activity.
  • Drought's Dual Impact: The 2003 drought not only reduced carbon uptake (GPP) but also ecosystem respiration (TER), though GPP was generally more sensitive. This led to a complex impact on the net carbon balance (NEE).
  • Species-Specific Sensitivity: Broadleaved temperate and Mediterranean forests showed a higher sensitivity to drought-induced reductions in GPP compared to coniferous forests. Coniferous stands appeared more resilient.
  • Lag Effects on Growth: Tree growth measurements revealed that the impact of the 2003 drought extended into the following year (2004), indicating a lag effect on radial growth and potentially on overall tree health. This highlights the long-term consequences of extreme events.
  • Heatwave Confounding Factor: The study acknowledges that the 2003 event was a combined heatwave and drought, making it challenging to fully disentangle the effects of each. However, the strong correlation with soil water deficit suggests water availability was a primary driver.

Key Results

  • Reduced Carbon Uptake: Across the investigated sites, the 2003 drought led to a significant reduction in annual NEE compared to 2002, with some sites becoming carbon sources for periods. For example, at Tharandt and Le Bray, NEE decreased by 150-300 g C m⁻² year⁻¹.
  • Water Vapor Flux Reduction: Water vapor fluxes were reduced when REW dropped below 0.4, with a marked decrease occurring below this threshold.
  • Species Differences: Beech forests and Mediterranean species showed a more pronounced reduction in GPP (45-60%) under severe water stress compared to coniferous forests (16-38%).
  • Growth Impairment: Tree growth was significantly reduced in 2003. At Hesse, annual NEE was reduced by 18%, while biomass increment decreased by 22%. At Tharandt, NEE reduction was 24%, and biomass increment decreased by 44%.
  • Model Performance: The water balance model (BILJOU) showed good agreement with measured soil water content where available. The models for TER and GPP also captured variations well, with GPP models generally performing better than TER models.

Limitations and Future Work

  • Soil Water Measurement Limitations: The study acknowledges that the relatively poor knowledge of soil-root properties and the temporal variation of LAI at some sites introduced uncertainty into the drought quantification. Future work should focus on more comprehensive soil and root characterization at eddy flux sites.
  • Disentangling Heat and Drought: The authors note that it was difficult to completely separate the effects of high temperature from drought. Future studies could aim to isolate these factors more effectively.
  • Long-Term Growth Effects: While lag effects on growth were observed, more long-term observations are needed to determine if extreme droughts lead to tree dieback or complete recovery, and to understand the underlying physiological mechanisms (e.g., carbohydrate reserves, xylem dysfunction).
  • Autotrophic vs. Heterotrophic Respiration: The study could not definitively separate the contributions of autotrophic and heterotrophic respiration to the overall TER reduction. Further research is needed to understand the specific responses of these components to drought.
  • Improved Modeling: The authors suggest that incorporating more detailed site parameters, such as physical soil properties and root distribution, would improve drought modeling accuracy. Additionally, exploring the influence of direct-to-diffuse radiation ratio and vapor pressure deficit on GPP could enhance model performance.

Summary updated

BibTeX
@inproceedings{granier2007evidence,
  author = {André Granier and Markus Reichstein and Nathalie Bréda and Ivan A. Janssens and Eva Falge and Philippe Ciais and Thomas Grünwald and Marc Aubinet and Paul Berbigier and Christian Bernhofer and Nina Buchmann and Osvaldo Facini and Giacomo Grassi and Bernard Heinesch and Hannu Ilvesniemi and P. Keronen and Alexander Knohl and Barbara Köstner and Fredrik Lagergren and Anders Lindroth and Bernard Longdoz and Denis Loustau and J. Mateus and Leonardo Montagnani and Claude Nys and Eddy Moors and Dario Papale and Marianne Peiffer and Kim Pilegaard and Gabriel Pita and Jukka Pumpanen and Serge Rambal and Corinna Rebmann and Abel Rodrigues and G. Seufert and John Tenhunen and Timo Vesala and Q. Wang},
  title = {{Evidence for soil water control on carbon and water dynamics in European forests during the extremely dry year: 2003}},
  booktitle = {Agricultural},
  year = {2007},
  doi = {10.1016/j.agrformet.2006.12.004},
}
Figure