Unlocking Drought-Induced Tree Mortality: Physiological Mechanisms to Modeling

Abstract

Drought-related tree mortality has become a major concern worldwide due to its pronounced negative impacts on the functioning and sustainability of forest ecosystems. However, our ability to identify the species that are most vulnerable to drought, and to pinpoint the spatial and temporal patterns of mortality events, is still limited. Model is useful tools to capture the dynamics of vegetation at spatiotemporal scales, yet contemporary land surface models (LSMs) are often incapable of predicting the response of vegetation to environmental perturbations with sufficient accuracy, especially under stressful conditions such as drought. Significant progress has been made regarding the physiological mechanisms underpinning plant drought response in the past decade, and plant hydraulic dysfunction has emerged as a key determinant for tree death due to water shortage. The identification of pivotal physiological events and relevant plant traits may facilitate forecasting tree mortality through a mechanistic approach, with improved precision. In this review, we (1) summarize current understanding of physiological mechanisms leading to tree death, (2) describe the functionality of key hydraulic traits that are involved in the process of hydraulic dysfunction, and (3) outline their roles in improving the representation of hydraulic function in LSMs. We urge potential future research on detailed hydraulic processes under drought, pinpointing corresponding functional traits, as well as understanding traits variation across and within species, for a better representation of drought-induced tree mortality in models.

Keywords: carbohydrates; drought; functional traits; hydraulic failure; land surface models; plant hydraulics; tree mortality.

Figure 1

Key physiological processes following reductions in plant water potential as outlined by the biphasic framework of drought-related tree mortality (panel A). Physiological functions including stomatal conductance (g_s, cyan), percentage loss of hydraulic conductivity in leaves (PLC_Leaf, red) and stems (PLC_Stem, blue), as well as branch relative water content (RWC, orange) are shown as percentage of maximum. Vertical dashed line indicates the leaf turgor loss point (TLP). Lethal water potential thresholds (P_Lethal) for leaves and stems are indicated by red and blue circles. Transition from Phase I to Phase II occurs when stomata are fully closed, which theoretically coincides with the turgor loss (broken dashed line). Panel (B) shows the observed variation of these physiological processes from Eucalyptus sideroxylon during a dry-down experiment conducted in a common garden (Blackman et al., 2019), with shaded regions surrounding the lines denoting the 95% confidence interval of fitted curves. Similarity in the two panels indicate that the biphasic framework is generally supported by the experimental evidence. Note that TLP in panel (B) occurred prior to complete stomatal closure, and leaf shedding was initiated when leaf xylem was completely embolized, indicating these traits may not be robust for predicting the timing of these physiological adjustments (see text for detail).

Figure 2

Conceptual diagram summarizing current understanding regarding the relationship among functional traits and illustrating the general approach for simulating of tree dynamics in response to water availability with hydraulic traits in process-based models. Light blue boxes indicate traits that are directly involved in the occurrence of hydraulic failure during drought stress but are not well represented in current TBMs, either due to lack of trait values or insufficient knowledge regarding the variation at spatial or temporal scales. Traits presented in the diagram are maximum rooting depth (RD_max), water potential of soil, root, stem, and leaf (Ψ_soil, Ψ_root, Ψ_stem, and Ψ_leaf, respectively), relative water content (RWC), hydraulic capacitance (C_p), minimum conductance (g_min), water potential threshold of stomata respond to drought (P_gs), hydraulic conductivity (K), percentage loss of hydraulic conductivity (PLC), stomatal conductance (g_s), evaporation (E), leaf area index (LAI), leaf photosynthesis (A), non-structural carbohydrates (NSCs), lethal threshold (P_lethal), gross (GPP), and net primary productivity (NPP).

Figure 3

The relationships between stem water potential at 50% loss of xylem hydraulic conductivity (P₅₀) and two easily measured traits, sapwood density (WD, panel A), and leaf turgor loss point (TLP, panel B), at global scale. P₅₀ data were obtained from Choat et al. (2012). Data for WD were sourced from Zanne et al. (2009), and TLP data were compiled from Bartlett et al. (2012) and Zhu et al. (2018). Regression formula for panel (B) y = 1.34x + 0.25 (R² = 0.22, p < 0.001).