The high diffusivity of carbon dioxide in air compared to liquid water provided a strong selective pressure for land plants to evolve more than 400 million years ago (Raven 2002). While it is comparatively easy to obtain carbon dioxide from the atmosphere, terrestrial photosynthesis comes at the cost of transpiration (Wong et al. 1979). To maintain the hydration of internal tissues, plants have evolved a suite of traits including a water impermeable cuticle, stomata to regulate gas exchange, and an internal water transport system of xylem elements to supply aerial tissue with water from the soil. All of these anatomical adaptations are essential for determining the rate of water loss from a plant, yet the vulnerability of the xylem to invasion by air under negative tension (i.e. embolism) alone determines the lethal water potential threshold for land plants (Brodersen et al. 2011; Brodribb et al. 2016, 2017; Choat et al. 2012, 2016; Cochard et al. 2015).

In a recent study, Lamarque et al. (2018) elegantly illustrate the importance of xylem resistance in determining a lethal water potential threshold in the Mediterranean tree Laurus nobilis L. This observation, as their title suggests, is somewhat inconvenient for a conventional wisdom that suggests xylem cavitation-induced embolism is a regular, non-lethal occurrence. This idea stems from indirect and destructive methods such as bench dehydration, air injection and centrifugation (Torres-Ruiz et al. 2017) for determining the percentage of embolised vessels in the xylem. The conclusion from most of these studies, many based on observations in L. nobilis, has been that a high percentage of vessels experience daily cycles of embolism and refilling (Hacke and Sperry 2003, Salleo et al. 2004, Trifilò et al. 2014). As a result, these studies could mistakenly lead one to conclude that xylem vulnerability to embolism plays little or no role in determining the threshold water potential at which plants die under water stress.

There has been much debate in recent years about the validity of these indirect and destructive methods, particularly those that involve observations in species with long vessels (Choat et al. 2010; Cochard et al. 2010; Hacke et al. 2015; Jacobsen and Pratt 2012; Torres-Ruiz et al. 2017). With knowledge of the methodological limitations and the long vessels of L. nobilis, Lamarque et al. (2018) employed two non-invasive methods, i.e. MicroCT and the optical method (Brodribb et al. 2016, 2017), in leaves and stems to construct vulnerability curves, as well as centrifugation in branch segments that exceeded the length of the longest vessel. Contrary to 25 years of results using methods prone to artefacts, this species was found to have xylem that is extremely resistant to the formation of embolism, with the water potential at which 50% of the xylem is embolised (P50) being − 7.9 MPa in stems and − 8.3 MPa in leaves. These results are unsurprising given that during drought, L. nobilis can survive with little or no leaf damage until − 9.0 MPa (Lamarque et al. 2018). Furthermore, during summer, field-grown plants still have open stomata at a water potential of − 2.4 MPa (Salleo et al. 2009), a negative water potential previously believed to cause around 90% embolised vessels (Hacke and Sperry 2003).

A remarkable result emerging from the work of Lamarque et al. (2018), and other recent studies using direct methods to quantify xylem vulnerability (Brodribb et al. 2017; Rodriguez-Dominguez et al. 2018; Skelton et al. 2017; Torres-Ruiz et al. 2017), is that highly resistant xylem is commonly observed in species native to dry or arid environments. This observation supports the idea that resistant xylem is adaptively relevant for plant survival during periods of water deficit and for the radiation of species into dry environments. In the conifers, for instance, P50 is a well-known trait associated with species evolution into dry environments (Brodribb et al. 2014; Larter et al. 2017). Many species of derived Cupressaceae and Taxaceae live in very dry environments and have xylem that is highly resistant to embolism (Brodribb et al. 2014; Larter et al. 2017). It is in this group of conifers that the species with the most resistant xylem so far recorded, Callitris tuberculata, native to arid areas of southwest Australia, evolved, with a P50 of − 18.8 MPa (Larter et al. 2015).

A literature review reveals that, currently, at least 136 species, from both gymnosperm and angiosperm lineages, have evolved highly resistant xylem (i.e. arbitrarily defined here as a P50 less than − 5.0 MPa) (Fig. 1). From this brief analysis of the literature, there seems to be very little phylogenetic signal in the occurrence of highly resistant xylem, with species being found across the seed plant phylogeny in 62 genera and 20 orders. The large number of convergent evolutionary transitions towards highly resistant xylem so far reported in angiosperms is reminiscent of the number of evolutionary transitions into aquatic habitats (Cook 1999) or deciduousness in this lineage (Axelrod 1966). Ancestral state reconstruction in the conifers suggests that the last common ancestor of this lineage had xylem that was vulnerable to embolism formation (Brodribb et al. 2014; Pittermann et al. 2012). While the xylem physiology of early diverging angiosperms is understudied, many of these lineages are either native to ever-wet, understorey tropical environments or are aquatic, it is likely that the xylem of the common ancestor of all angiosperms was likewise highly vulnerable (Feild and Holbrook 2000; Hacke et al. 2007).

Fig. 1
figure 1

Phylogenetic distribution of orders (displayed in red) and associated genera containing plant species with highly resistant xylem (defined at a P50 < − 5 MPa). Numbers following each genus indicate the number of species so far identified in each genus (see Supporting Table S1 for the dataset used in this phylogenetic analysis)

Full size image

The recent results of Lamarque et al. (2018) add the first species, to this ever-growing list, from the early diverging order Laurales. The family Lauraceae is believed to have had an origin in ever-wet temperate and tropical forests, with an abundance of genera today dominating tropical cloud forests (Chanderbali et al. 2001; Donoghue 2008). It is believed that L. nobilis is a relict of a Tethyan rainforest flora that dominated the Mediterranean basin until the Pliocene (Rodríguez-Sánchez et al. 2009). The evolution of highly resistant xylem in this species may be related to the survival of this lineage in the Mediterranean basin during the acidification over the past 30 million years. Unlike the Lauraceae, which is so far only represented by a single species with highly resistant xylem, some genera contain many species with highly resistant xylem (Fig. 1). One such genus is Quercus with nine species so far identified with highly resistant xylem (Bhaskar et al. 2007; Dietrich et al. 2018; Lobo et al. 2018; Martin-StPaul et al. 2014; Scoffoni et al. 2017; Skelton et al. 2018; Vaz et al. 2012; Vilagrosa et al. 2003). Many of the angiosperms species so far identified are native to particularly dry regions of the globe, such as central and southern Australia, the Mediterranean basin and southwest North America. The broad phylogenetic spread of species so far identified with highly resistant xylem suggests a strong selective pressure on the evolution of this trait (Choat et al. 2012, 2018).

We do not yet know what specific anatomical or physiological traits are responsible for the evolution of highly resistant xylem, although a number have been proposed. These include xylem cell wall thickness (Blackman et al. 2010; Cardoso et al. 2018), pit membrane thickness (Choat et al. 2006; Hacke et al. 2006; Jansen et al. 2009; Pittermann et al. 2010; Wheeler et al. 2005) and the consistency of surfactants in the xylem (Schenk et al. 2015). It could be that each evolutionary transition towards highly resistant xylem was the result of modification to a different trait or set of traits, influencing xylem physiology. In L. nobilis, Lamarque et al. (2018) and Klepsch et al. (2018) suggest pit membrane thickness is the primary driver behind the evolution of highly resistant xylem. Comparative anatomy and physiology focusing on small clades in which highly resistant xylem has a single, recent evolutionary origin could provide a powerful tool to further investigate the key anatomical or physiological traits controlling xylem resistance to embolism formation. Using this approach, we may finally be able to answer questions about the molecular regulation of this critical trait. While many future studies are possible from the use of direct methods for observing embolism formation in the xylem, the simple phylogenetic analysis presented here suggests that further studies in species native to dry environments should ensure that highly resistant xylem becomes a truth universally acknowledged.