Embolism spreading may not rely on a fixed pressure difference threshold, but on pressure-driven diffusion
The observation of embolism spreading under highly variable xylem water potentials in experiment 1 and 2 indicates that embolism formation may not always rely on a certain threshold of the pressure difference between a functional and embolised conduit, as frequently assumed based on the air-seeding hypothesis, and quantitatively estimated by the Young-Laplace equation (Sperry and Tyree, 1988; Choat et al. , 2008). Even if some of our xylem water potential measurements may not be fully accurate due to a pressure gradient heterogeneity (Bouda et al. , 2019), the > 1 MPa difference in PEP50 values between detached leaves and leaves connected to stems provides solid evidence that embolism formation in xylem tissue from the same organ of a species may occur under different xylem water potentials. This finding is not entirely new, and in line with earlier differences in embolism resistance between intact plants and cut plants (Choat et al. , 2010; Torres-Ruiz et al. , 2015; Lamarque et al. , 2018). For instance, a ca. 4 MPa difference in P50 was found for Laurus nobilisbased on microCT observation of cut branches and intact seedlings (Nardini et al. , 2017; Lamarque et al. , 2018). Similar to our results, Skelton et al. (2018) compared the vulnerability curves of cut branches and intact plants based on the optical method forQuercus wislizenii , and found a 1.5 MPa difference in P50 between leaves attached to a long, cut branch, and leaves from an intact plant. The finding that cut plant material can be more vulnerable to embolism spreading than intact plants raises concerns about embolism resistance measurements of plant material samples with pre-existing embolism, the possible induction of embolism due to cutting, and the application of the bench dehydration method on cut plant material (Sperry and Tyree, 1988).
If embolism spreading for given species may not rely on a certain pressure threshold, which mechanism does then trigger embolism? We believe that this is a highly important question in our understanding of water transport, even though this question cannot be fully answered based on the available evidence. We speculate that gas diffusion across interconduit pit membranes plays a role in determining the amount of gas dissolved in xylem sap, which may affect embolism nucleation. Although mass flow is theoretically 105 times faster than diffusion, gas diffusion in xylem is overall much faster and more common than mass flow. The main reason seems to be that gas diffusion takes place continuously over very large areas, while mass flow requires gas movement through multi-layered, tiny pore constrictions of mesoporous pit membranes (Kaack et al. , 2020). However, the continuous nature of gas diffusion and the high amounts of gas (CO2and O2) surrounding conduits (Spicer and Holbrook, 2005; Teskey et al. , 2008) do not mean that gas concentration in xylem sap is always under equilibrium with gas in embolised conduits, as has frequently been assumed (Hammel 1967; Yang and Tyree, 1992; Wheeleret al. , 2013; Schenk et al. , 2016). Gas solubility of xylem sap is affected by pressure and temperature (Mercury et al. , 2003; Schenk et al. , 2016; Lidon et al. , 2018). Due to a nearly constant atmospheric pressure in a cut-open vessel, but a considerably variable liquid water pressure in sap-filled conduits, the driving force for gas diffusion is the unbalanced, anisobaric situation between embolised conduits and functional conduits. We ignore here the effect of temperature on gas solubility of xylem sap as all experiments were conducted in an air-conditioned lab under similar temperature (Schenk et al. , 2016). Moreover, oversolubility of gas may occur in nanoporous cell walls and mesoporous pit membranes, as is known for nanoconfined environments in general (Pera-Titus et al. , 2009; Hoet al. , 2015; Coasne & Farrusseng, 2019), which could contribute to concentration gradients of gas dissolved in xylem sap.
A recently embolised intact vessel is not immediately filled with gas under atmospheric pressure, but eventually achieves equilibrium of its gas pressure with the atmosphere, depending on how fast gas is attracted from surrounding gas sources via diffusion. While cut-open vessels are immediately filled with air and reach atmospheric pressure immediately, intact vessels that embolise are initially filled with water vapour (ca. 2.4 kPa). It has been modelled that gas diffusion takes from 20 min to several hours to obtain atmospheric pressure (ca. 101.3 kPa) in embolised, intact vessels, which may depend on the distance to the nearest gas phase, and the interconduit pit membrane area for gas diffusion (Yang & Tyree, 1992; Wang et al. , 2015a, b). Although gas diffusion happens across conduit cell walls and pit membranes, the micropores (< 2 nm) in hydrated walls are much smaller than the 5 to 50 nm dimensions of pit membrane pores (Donaldson et al. , 2019; Kaack et al. , 2019). Therefore, it is reasonable to assume that gas diffusion across hydrated, 200 to 600 nm thick pit membranes is much faster than across the much thicker layers of secondary cell walls (Yang et al. , submitted). Indeed, axial gas diffusion in wood is found to be about one to two orders of magnitude larger than radial diffusion (Sorz & Hietz, 2006). Moreover, end-wall resistivity of conduits has been suggested to be proportional to lumen resistivity (Sperry et al. , 2005; Hacke et al. , 2006), and conductance of gas increases to the 4th power with conduit diameter or pore diameter according to Hagen-Poiseuille’s equation.
Further research is clearly needed to investigate how gas diffusion may contribute to the very high gas solubility of xylem sap (Schenk et al. , 2016), whether oversolubility occurs due to nanoconfined spaces such as pit membranes and cell walls (Pera-Titus et al. , 2009; Hoet al. , 2015; Coasne & Farrusseng, 2019), how gas-xylem sap interfaces are affected by the dynamic surface tension of xylem sap lipids (Yang et al. , 2020), and how surfactant-coated nanobubbles may contribute to the gas concentration of xylem sap and embolism formation (Schenk et al. , 2015, 2017; Jansen et al. , 2018; Park et al. , 2019).