Leaf hydraulic vulnerability to drought is linked to site water availability across a broad range of species and climates

Abstract

Background and aims: Vulnerability of the leaf hydraulic pathway to water-stress-induced dysfunction is a key component of drought tolerance in plants and may be important in defining species' climatic range. However, the generality of the association between leaf hydraulic vulnerability and climate across species and sites remains to be tested.

Methods: Leaf hydraulic vulnerability to drought (P50leaf, the water potential inducing 50 % loss in hydraulic function) was measured in a diverse group of 92 woody, mostly evergreen angiosperms from sites across a wide range of habitats. These new data together with some previously published were tested against key climate indices related to water availability. Differences in within-site variability in P50leaf between sites were also examined.

Key results: Values of hydraulic vulnerability to drought in leaves decreased strongly (i.e. became more negative) with decreasing annual rainfall and increasing aridity across sites. The standard deviation in P50leaf values recorded within each site was positively correlated with increasing aridity. P50leaf was also a good indicator of the climatic envelope across each species' distributional range as well as their dry-end distributional limits within Australia, although this relationship was not consistently detectable within sites.

Conclusions: The findings indicate that species sorting processes have influenced distributional patterns of P50leaf across the rainfall spectrum, but alternative strategies for dealing with water deficit exist within sites. The strong link to aridity suggests leaf hydraulic vulnerability may influence plant distributions under future climates.

Keywords: Leaf hydraulic vulnerability; aridity; climate change; drought; interspecific variation; rainfall; species distribution.

Fig. 1.

The relationship between P50_leaf and site-specific climate parameters related to water availability: (A) mean annual precipitation (MAP) and (B) aridity index (AI). Levels of significance for regressions through log-transformed data are indicated (see insets; ***P < 0·0001). The sites from where species were sampled are colour-coded as indicated in the key in (B).

Fig. 2.

The relationship between leaf hydraulic vulnerability to drought-induced dysfunction (P50_leaf) and climate parameters related to water availability across species distributions: (A) mean annual precipitation; (B) the 5th percentile of mean annual precipitation; (C) the 5th percentile of aridity index. Significant relationships using log-transformed data (***P < 0·0001) are fitted with least-squares regressions. The sites from where species were sampled are colour-coded as indicated in the key in (C).