4.5. Vulnerability to embolism

The xylem vulnerability to embolism indicates the risk of loss of water transport during drought. Vulnerability is expressed as the percentage of the water-saturated xylem conductance that is lost at given stem-water potentials. The water stream in the conduits is under tension, which can in some cases become as high as 100 times the atmospheric pressure. Any air entry into a water-conducting element will dissipate the tension in it and quickly expand (becoming an air embolism), effectively blocking water flow through that element. The more such emboli develop, the greater the loss of xylem conductance. The ability of a species to tolerate highly negative water potentials (= high tensions) without embolising varies greatly among species (cf. Section 3.15) and is an important aspect of drought tolerance.

What and how to collect

Follow the above instructions for collecting for measurement of xylem conductivity.


Vulnerability to embolism is quantified by constructing a xylem vulnerability curve. This consists of plotting measured values of xylem conductance, on the y-axis, against the values of stem water potential (Ψ), on the x-axis, at which these conductance values (represented as a percentage of the maximum water-saturated conductance) were obtained. The shape of this curve is usually sigmoid. To characterise this curve by one number, the value of Ψ at its mid-point (50% loss of conductance) is commonly used.

Stem segments with different Ψ values can be obtained by any of three possible methods.

(1) Evaporative dehydration. A large branch that includes some lateral secondary branches is cut from the plant and kept unwrapped during the day, or longer, for its xylem to partially dehydrate, and develop tension as a result of transpiration from its leaves. Ψ is determined periodically, by the pressure chamber method (cf. Section 3.15), on secondary branches that are removed at intervals for this purpose. When a target Ψ has been reached, conductance of a segment of the main axis near the last-removed lateral is measured. Water is then flushed through this segment briefly under a pressure high enough to displace all air embolisms from it (see cited references), and its conductance is re-measured to obtain the maximum conductance of the segment. This affords a value for the percentage loss of conductance at that particular Ψ. Separate branches are similarly tested to obtain conductance-loss values for other values of Ψ.

(2) Centrifugation. Attaching a short stem segment horizontally (and symmetrically) across the top of a centrifuge rotor and spinning it generates tension in the water within the xylem conduits. From the centrifugal force that was applied, a corresponding (negative) Ψ or P can be inferred. Advantages of the method are that stem segments can be brought to different effective water potentials very quickly. However, the technique is difficult for stem segments longer than ~20 cm, or ones that cannot be reliably attached to an available rotor, and it cannot be used for segments longer than the width of the rotor chamber. Details needed for actually employing this method are given in some of the cited references.

(3) Air injection. This simple method is based on the principle of substituting internal tension by external positive air pressure, which is applied to a stem segment that is located within a pressure chamber. The procedure uses a special pressure chamber designed for a stem segment to pass completely through it, allowing a conductance measurement while the external pressure is applied. This type of chamber is commercially available from the PMS Instrument Co., Albany, Oregon, USA (http://pmsinstrument.com).

References on theory, significance and large datasets: Tyree and Sperry (1989); Davis et al. (2002); Brodribb et al. (2003); Maherali et al. (2004); Holbrook and Zwieniecki (2005); Choat et al. (2007); Feild and Balun (2008); Sperry et al. (2008a).
References on methods: Cochard et al. (1992); Sperry et al. (1988, 2008b); Alder et al. (1997); Pammenter and Vander Willigen (1998).