Skip to main content
Dryad

Large leaf hydraulic safety margins limit the risk of drought-induced leaf hydraulic dysfunction in Neotropical rainforest canopy tree species

Data files

Mar 20, 2023 version files 16.36 KB

Abstract

The sequence of key water potential thresholds from the onset of water stress to mortality, and the timing of stomatal closure with regard to leaf xylem embolism formation are essential to characterizing plant adaptive strategies to drought. This constitutes a critical knowledge gap for tropical rainforest species, which may be less vulnerable to drought than previously thought.

We recorded key leaf and stem water potential thresholds, leaf hydraulic safety margins (HSMleaf), leaf stomatal safety margins (SSMleaf) and estimated native embolism levels during a normal-intensity dry season across 18 Neotropical rainforest tree species. We also solved a sequence of key water potential thresholds. Additionally, we provide a cross-biome analysis of SSMleaf encompassing 97 species from four major biomes based on a literature survey.

In the studied rainforest species, leaf turgor loss point, used as a surrogate for stomatal closure, typically occurred before the onset of leaf xylem embolism. Most species exhibited positive HSMleaf and SSMleaf, with contrasting values across species and nearly absent embolism levels during the dry season irrespective of the experienced midday leaf water potentials. Our results point out that leaf xylem embolism is not routine for Neotropical rainforest tree species.

Based on our proposal of the water potential sequence for tropical rainforest trees, we argue that leaf xylem embolism is a rare event for these species. This was supported by the literature survey, indicating that across biomes, most woody species have rather large SSMleaf and that leaves of tropical rainforest trees are not necessarily more vulnerable than in other biomes. However, we found evidence that some tropical rainforest species may be more vulnerable than others to ongoing climate change. Our data provide an opportunity to parametrize tree-based or land-surface models for tropical rainforests.