Traditional global vegetation models treat forest mortality during droughts with a high degree of abstraction. They often rely on empirical statistical relationships—such as killing a fixed percentage of trees if soil moisture drops below a certain threshold for too long. However, trees do not die because of a statistical lottery; they die because their internal physical systems fail.
To build more resilient projections of global carbon sinks under climate change, a new paradigm has emerged. This approach replaces abstract mortality modules with process-based, mechanistic models of tree hydraulics. By simulating the physical flow of water through a tree's internal architecture, these models capture exactly how and when severe droughts cause a tree's plumbing system to permanently break.
The Tree as a Hydraulic Circuit
To understand why a tree dies during a drought, we must first understand how it lives. A tree operates as a continuous hydraulic circuit that spans from the deep soil, through the roots and stem, and out into the atmosphere via the leaves. Water is not pumped upward; rather, it is pulled by the atmosphere. As water evaporates from the leaves during transpiration, it creates a negative pressure (a tension, or negative water potential) that yanks a continuous column of water upward through the xylem tissue.
[Atmosphere: High Tension / Negative Potential]
▲
│ (Transpiration)
[Leaves]
▲
│ (Stomatal Regulation)
[Stem] ◄─── [Internal Storage]
▲
│ (Xylem Transport)
[Roots]
▲
│
[Soil: Variable Moisture & Potential]
In Forest fluxes and mortality response to drought: model description (ORCHIDEE-CAN-NHA r7236) and evaluation at the Caxiuanã drought experiment, Yao et al. and their co-authors formalized this physics by explicitly modeling water potentials and flow rates across distinct plant organs. Instead of treating the tree as a single bucket of water, the authors simulated the dynamic balance between water supply from the soil and water demand from the atmosphere, accounting for the resistance to flow within the roots, stems, and leaves, as well as the capacitive water storage within the tree's own tissues.
Stomatal Defense and Decoupling
When a drought hits, the soil dries out, holding onto its remaining water more tightly. This forces the tree to pull with even greater tension to extract water. If the tension in the water column becomes too extreme, the column can snap. This phenomenon, known as cavitation, introduces air bubbles (embolisms) into the xylem, blocking water transport.
Trees have a primary line of defense against cavitation: they can close their stomata (the microscopic pores on their leaves) to restrict transpiration and preserve internal water potential. This strategy is known as isohydric behavior.
However, this defense mechanism introduces complex physiological trade-offs. As explored by Zhao et al., when environmental conditions shift or transpiration demands increase, trees with strong isohydric behavior tightly regulate their stomata to maintain stable leaf water potentials. The authors demonstrated that this behavior effectively decouples whole-tree sap flux from raw leaf-level transpiration demands. While this protects the tree's hydraulic integrity, closing the stomata also cuts off the intake of carbon dioxide, starving the tree of the fuel needed for metabolic maintenance if sustained over long periods.
Furthermore, trees dynamically adjust their operational strategies during prolonged dry spells. As observed by Tang et al., co-occurring species in seasonal drought environments shift their water budgets significantly, sometimes relying heavily on nocturnal transpiration and alternative water-use strategies to survive periods of intense water stress without triggering catastrophic failure.
The Breaking Point: Percentage Loss of Conductance
When a drought is exceptionally severe or prolonged, stomatal closure is no longer enough to protect the tree. The tension in the xylem inevitably crosses a critical threshold, causing widespread cavitation. As more conduits fill with air, the tree loses its ability to transport water. This state is quantified as the Percentage Loss of Conductance (PLC).
The core breakthrough of recent mechanistic modeling is the direct, algorithmic linkage between PLC and demographic tree mortality. In the ORCHIDEE-CAN-NHA model introduced by Yao et al., the authors implemented a novel empirical algorithm where a tree's risk of death is triggered when stem conductance drops below a critical threshold. Rather than an instantaneous death sentence, this mechanism tracks a cumulative drought exposure threshold and applies a daily mortality fraction, capturing the reality that trees often succumb to hydraulic failure gradually under sustained stress.
Extended Drought ──► Extreme Xylem Tension ──► Cavitation & Embolism
│
▼
Tree Mortality ◄─── Exceed Critical Threshold ◄─── High PLC
By embedding this physical breaking point into regional and global simulations, the predictive power of land surface models changes fundamentally. In How drought events during the last century have impacted biomass carbon in Amazonian rainforests, the authors demonstrated that explicitly modeling these hydraulic and demographic structures is essential for accurately capturing historical variations in forest biomass carbon. Simplified models that ignore these interconnected processes fail to reflect how severe droughts prune specific cohorts of trees.
Scaling Up to Ecosystem Projections
Why does this granular focus on plant plumbing matter for global climate change? Because it determines whether massive ecosystems, like the Amazon rainforest, will remain carbon sinks or become carbon sources.
When applied to future climate scenarios in Future Drought‐Induced Tree Mortality Risk in Amazon Rainforest, Yao et al. and their co-authors utilized a tree mortality sub-model that scales hydraulic failure across different tree size classes. Because taller trees must fight gravity to pull water to their canopies, they experience higher baseline xylem tensions, making them disproportionately vulnerable to cavitation.
By capturing this size-dependent vulnerability through the lens of stem conductance loss, the authors projected that parts of the Amazon could experience widespread hydraulic failure and transition to a net carbon source by the mid-21st century. This macro-level ecosystem shift cannot be reliably predicted by tracking rainfall alone; it can only be understood by modeling the physical limits of the microscopic water columns rising through the trunks of the trees themselves.
Key Foundational Work
Forest fluxes and mortality response to drought: model description (ORCHIDEE-CAN-NHA r7236) and evaluation at the Caxiuanã drought experiment (2022) — contributed 9 concept elements to this direction.
How drought events during the last century have impacted biomass carbon in Amazonian rainforests (2022) — contributed 6 concept elements to this direction.
Future Drought‐Induced Tree Mortality Risk in Amazon Rainforest (2024) — contributed 4 concept elements to this direction.