Minimum leaf conductance during drought: unravelling its variability and impact on plant survival

Abstract

Introduction

In the last few decades, a large number of studies have brought to light the particular threat of drought and increasing temperatures on plant survival (Allen & Breshears, 1998; Carnicer et al., 2011; Brodribb et al., 2019; Hammond et al., 2022). One of the main consequences of increasing environmental drought stress is a negative impact on the hydraulic function of plants (Choat et al., 2012; Arend et al., 2021). Indeed, during prolonged drought, dehydration causes large water potential differences between soil and leaves, which can result in hydraulic dysfunction due to embolism formation in the xylem conduits. Such hydraulic failures, induced by sharp drops in water potential, can be avoided through stomatal closure (Creek et al., 2020), which plays a major role in plant survival under drought (Martin-StPaul et al., 2017). Despite stomatal closure being a key reaction to reduce significant plant water losses, water is still lost through imperfectly closed stomata and the leaf cuticle. This process can be quantified by the minimum leaf conductance (g_min) (Duursma et al., 2018). Although the rates of whole-plant water conductance are greatly diminished, g_min can be sufficient to deplete the plant water reserves during stress. Therefore, under prolonged drought, continued water loss via g_min can lead to catastrophic hydraulic failure and substantial tissue dehydration that contribute to organ and plant death (Urli et al., 2013; Mantova et al., 2023; Petek-Petrik et al., 2023). Consequently, g_min has been highlighted as an important trait in predicting whole-plant transpiration and water status under severe stress (Barnard & Bauerle, 2013; Kala et al., 2016). More recently, g_min has been advanced as a key trait to predict the time taken by a plant to reach hydraulic failure (THF) and subsequent mortality during drought (Cochard et al., 2021; Ruffault et al., 2022a; Petek-Petrik et al., 2023) and to predict other current hazards to vegetation such as wildfire incidence (Ruffault et al., 2022b; Torres-Ruiz et al., 2024).

Synthetic studies gathering g_min data showed that estimations of residual transpiration in the literature come from a wide variety of experimental techniques (Kerstiens, 1996; Duursma et al., 2018). While cuticular conductance on isolated cuticles has been measured since early studies (Stålfelt, 1956; Schönherr & Mérida, 1981; Pearcy et al., 1989), only a few protocols to measure g_min on a full leaf have been tested. Generally, measurement protocols rely on evaluating water loss through balance-based mass estimations over time and selecting the slope of the relationship between mass and time after the point of stomatal closure, assuming that water losses through the cuticle are steady as the leaf dehydrates (Sack & Scoffoni, 2010) or through water vapor conductance measurements using custom gas exchange systems (Boyer et al., 1997; Márquez et al., 2022). Automated devices have recently been developed in order to provide continuous measurements of water loss during dehydration under constant environmental conditions (Billon et al., 2020). Yet, to date, no study has estimated potential shifts in leaf water conductance during the entire desiccation process. Furthermore, intrinsic factors can provoke shifts in water conductance estimations. Among these factors, leaf shrinkage or leaf rolling during dehydration can significantly impact the estimation of leaf hydraulic traits (Scoffoni et al., 2014); hence, variations in leaf area need to be considered in conductance calculations. Additionally, water potential decline during leaf dehydration can also slightly influence conductance rates by affecting the water vapor pressure (vapor pressure deficit, VPD) inside the leaf and therefore the driving force for transpiration between the leaf and the atmosphere (Nobel, 2009).

Using a dynamic approach to estimate residual water losses during the entire dehydration process, and including phenomena that are often neglected in leaf conductance estimations, such as leaf shrinkage and water potential-induced VPD changes, this study aimed to assess the dynamics of residual leaf water conductance during dehydration. By measuring its variability during dehydration, this work also aimed to strictly distinguish residual conductance (g_res) as a dynamic water loss during dehydration after stomatal closure, and g_min as conductance values bounded by physiology-informed boundaries. We therefore capture variable residual conductance by measuring g_min at specific thresholds. Because of the key role of g_min in the depletion of the last vital water reserves of the plant after stomatal closure, there is a need to use repeatable and reliable protocols for its estimation. Here, we describe a complete methodology to investigate g_min based on continuous measurements of relative water content (RWC) over time. Moreover, we use water status- and physiology-based thresholds to determine g_min. We provide detailed information on the device and open-source software used to analyze data. By applying this new technique to a selection of nine species with various phenologies, water use strategies, and resistances to drought, we will (i) test the hypothesis that residual conductance after stomatal closure varies along a gradient of declining water potential and RWC, (ii) assess how this variation affects the estimation of g_min at physiologically relevant thresholds distributed along a temporal sequence, and (iii) investigate how the range of g_min estimated along this drought physiological time sequence can influence model outcomes of plant survival under drought stress.