SEARCH

SEARCH BY CITATION

Keywords:

  • air seeding;
  • bordered pit;
  • cavitation;
  • embolism;
  • torus–margo pit membranes;
  • xylem

The question of what structural features underlie differences in resistance to xylem cavitation is a long-standing issue fundamental to our understanding of water transport in plants. Plants routinely face xylem tensions great enough to cause cavitation and embolism, which may result in significantly increased hydraulic resistance, limitations on leaf gas exchange and ultimately carbon starvation and plant death (Tyree & Zimmermann, 2002; McDowell et al., 2008). The relative resistance of a plant to embolism is a major determinant of species distribution and the ability of plants to survive in the face of environmental stresses such as drought and freezing (Stuart et al., 2007). The xylem consists of a highly compartmentalized network of conduits in which emboli can be isolated while water transport continues in adjacent conduits. The continued function of this network depends to a large degree on the nano-porous primary cell walls (pit membranes) that separate conduits from one another. Pit membranes function as safety valves in the xylem, allowing the free passage of water between cells as it moves from the roots to the leaves, but limiting the spread of gas or pathogens. However, the fine porosity of pit membranes also results in significant hydraulic resistance, with pit hydraulic resistance accounting for a large proportion of total xylem hydraulic resistance (Zwieniecki et al., 2001; Choat et al., 2006). The structure and function of pit membranes is therefore of great importance in both the hydraulic efficiency of the xylem and cavitation resistance (Choat et al., 2008). Although there is a great breadth of diversity in bordered pit structure across higher plants, pit membranes can generally be divided into two major forms: homogeneous pit membranes, typical of angiosperm species; and margo–torus pit membranes of tracheid-bearing conifers (Fig. 1). In this issue of New Phytologist, two exciting studies extend our understanding of the relationship between xylem structure and resistance to cavitation: Christman et al. (2009, pp. 664–674) examine the anatomical underpinning of cavitation resistance in angiosperm species, while Hacke & Jansen (2009, pp. 675–686) report a detailed investigation of margo–torus pit structure and its influence on cavitation resistance in three conifer species.

image

Figure 1. Variation in pit structure. (a) A homogeneous pit membrane of an angiosperm species, Acer negundo, and (b) a margo–torus type pit membrane of a conifer, Calocedrus decurrens. Homogeneous pit membranes, typical of angiosperm species, have a relatively uniform array of microfibrils, whereas in margo–torus pit membranes of tracheid-bearing conifers, the conductive and protective functions of the membrane are spatially distinct as a porous outer region (margo) that allows for movement of water between conduits and a central thickened plug (torus).

Download figure to PowerPoint

‘It is obvious from this work that rare, leaky pits have dramatic consequences for the ability of plants to sustain water transport as water stress and xylem tension increases.’

Cavitation resistance in angiosperms

  1. Top of page
  2. Cavitation resistance in angiosperms
  3. Control of cavitation resistance in conifers
  4. Future directions for research
  5. References

Pores of homogenous pit membranes are of sufficiently small dimensions that they will prevent gas being drawn into an adjacent water-filled vessel until a critical threshold is reached (e.g. a 100-nm-diameter pore will prevent gas penetration up to a pressure difference of 2.88 MPa across the pit membrane). The potential for the spread of embolism between vessels and throughout the xylem is therefore dictated by the porosity of pit membranes and the minimum value of xylem water potential (negative hydrostatic pressure in the xylem fluid). Species with smaller pit-membrane pores are predicted to have greater cavitation resistance, and thus to tolerate greater degrees of water stress, than those with larger pit-membrane pores. However, although the relationship between pit-membrane pore size and cavitation resistance has sound theoretical underpinnings, it has been difficult to confirm this empirically by matching observed pore sizes to measured cavitation resistance across a range of species, with many studies failing to find pores large enough to be responsible for air seeding at realistic pressures (Wheeler, 1983; Shane et al., 2000; Choat et al., 2003). One explanation for this discrepancy is that the pores responsible for air seeding are actually extremely rare. Because air seeding will always occur first at the largest pore, it is only required that there be one large pore present in all of the many thousands of pit connections between two vessels. A rare, large pore may therefore escape detection by electron microscopy or particle-exclusion experiments. Support for this idea is provided by the work of Wheeler et al. (2005), which shows a strong correlation between cavitation resistance and the average area of pit overlap between vessels. This suggests that cavitation resistance might be determined stochastically, with the probability of having a rare, large pore increasing with the area of contact between vessels.

Christman et al. provide further support for this hypothesis, using an elegant pairing of theory and empirical data. Probability theory was used to model the cavitation threshold of pit membranes in three Acer species that have differing resistances to cavitation. The model incorporates the theory that if there is a normal distribution of pore diameters in any connection between vessels, only the extreme tail of the distribution will be responsible for air seeding. In fact, the model suggests that only one in 10 000 pits would be ‘leaky’ enough to cause air seeding at measured air-seeding thresholds. To test this model, Christman et al. measured air-seeding thresholds on different stem lengths of the three Acer species. This is analogous to a membrane-filter bubble test, where the pore diameter of a filter can be predicted from the pressure required for gas penetration through the filter. The model predicts that short stem segments with fewer vessel end walls should air seed at lower pressures than longer stem sections in which air must penetrate an increasing number of intervessel end walls to move through the entire segment. The empirical data matched the modeled predictions of air-seeding pressures closely. As the stem length increased, air-seeding pressures also increased, indicating that the effects of rare, large pit-membrane pores was masked by the majority of end walls, which lack very leaky pits. In the shortest stem segments, air-seeding pressures were consistently lower than the average cavitation pressures of each species. This evidence confirms that there is wide variation in the porosity of pit connections within each stem, and strongly suggests that a very small variation in the frequency of the rare, large pores can have a significant effect on cavitation resistance, which is independent of the number of pits or the total pit area.

It is obvious from this work that rare, leaky pits have dramatic consequences for the ability of plants to sustain water transport as water stress and xylem tension increases. The important question now becomes how plants would control the frequency of such pits in the xylem. Christman et al. suggest that such pits might be the result of ‘mistakes’ occurring during the development and hydrolysis of the primary walls that make up the pit membranes in secondary xylem. Weak spots in the pit membranes, where the density of cellulose microfibrils is lower, would be particularly susceptible to air seeding if large pressure differences are present across the membrane, as is the case when embolized vessels border water-filled vessels under tension. In this study there was no correlation between the average pit area and the cavitation resistance. Therefore, the frequency of large pores cannot be explained by a uniform stochastic model, as presented by Wheeler et al. (2005). The most likely alternative explanation is that changes in the intrinsic properties of pit membranes (average porosity and thickness) are responsible for differences in cavitation resistance. If pit membranes are, in general, more porous and flimsy, then it is easy to imagine that there would be a greater chance of a large pore developing somewhere in the many pit membranes that connect each vessel to others. This is supported by the large between-species variation in pit membrane properties and evidence that air-seeding pressures of individual vessels are correlated with the average pore diameter of pit membranes, as observed by electron microscopy (Jansen et al., 2009).

Perhaps the most important point to acknowledge is that cavitation resistance will not be controlled exclusively by either tissue-level properties (vessel length, diameter, pit overlap area) or pit-level properties (such as average porosity and thickness). Characteristics at both pit and tissue levels will influence cavitation resistance in plants. It is simply a question as to what extent the selective pressures that act to match cavitation resistance of angiosperm species to their environment have been satisfied by shifts in tissue-level traits rather than variation in microscopic pit structure. It will be most intriguing to see how the relationship between these traits evolves as more data become available.

Control of cavitation resistance in conifers

  1. Top of page
  2. Cavitation resistance in angiosperms
  3. Control of cavitation resistance in conifers
  4. Future directions for research
  5. References

In a structural departure from the homogenous pit membrane of angiosperms, the conifer torus–margo pit membrane combines a high degree of cavitation protection with efficient water transport that allows conifer xylem to achieve hydraulic efficiencies similar to those of angiosperms. In conifer wood, water transport occurs through single-celled conduits, known as tracheids, rather than through the long multicellular vessels characteristic of angiosperms; consequently, as water moves from one tracheid to another at a given segment length, it encounters a much higher frequency of the pitted end-wall regions than it would in vessels. Thus, in the absence of low-resistance end-walls, conifer xylem can potentially represent a hydraulically inefficient, high-resistance network requiring large pressure gradients to drive water transport. Conifers have avoided this problem by developing the torus–margo pit membrane in their tracheid end-walls. This structure reduces the end-wall hydraulic bottleneck because water travels through the net-like margo region of the membrane, a substantially more porous structure than the homogenous pit membrane. Consequently, pit resistance in conifers is almost 60 times lower than in angiosperms, effectively compensating for the frequent end-wall crossings presented by short tracheids (Pittermann et al., 2005; Sperry et al., 2006). Importantly, the hydraulic efficiency of conifer xylem is equivalent to that of angiosperm xylem for a given conduit diameter.

Despite the structural differences, conifers achieve the same degree of cavitation resistance as angiosperms, albeit using a different mechanism (Box 1). This raises the question of which xylem features control cavitation resistance in conifers. Hacke & Jansen use a combination of scanning electron microscopy and transmission electron microscopy to examine the relationship between fine pit structure and cavitation resistance in the root and shoot tissue of three conifer species. Comparisons between root and shoot tissue are ideal because root tissue is consistently more susceptible to cavitation than stem tissue within the same plant.

Hacke & Jansen's study reveals tight correlations between cavitation resistance and discrete features of pit anatomy. Specifically, cavitation resistance was positively related to the ratio of torus to aperture diameters whereby a larger torus for a given aperture diameter resulted in greater resistance to cavitation. This is intuitive because the greater the torus overlap against the pit chamber, the lesser the chance of the torus being pulled through the aperture and allowing air seeding to occur (Pittermann et al., 2006; Domec et al., 2008). In addition, the thickness of the torus and the depth of the pit chamber were inversely related to cavitation resistance, suggesting that a combination of a thinner torus and a shallow pit chamber may form a tighter seal over the pit aperture. The authors suggest that deeper pit chambers may require the margo to stretch further to seal the aperture, thereby predisposing the fibrils to irreparable damage by tearing.

Hacke & Jansen's study arrives on the heels of recent work that underscores the functional significance of conifer pit membranes on tree height. The pit aperture may represent a significant proportion of transport resistance in their xylem, so if apertures shrink to improve cavitation resistance, the resulting decrease in aperture conductance thus represents a clear trade-off in hydraulic efficiency at the pit level. Indeed, a linear relationship between the torus : aperture ratio and cavitation resistance has been observed with increasing height in very tall Douglas-fir trees: at greater branch heights, cavitation resistance increases to compensate for increasing xylem tensions but at the cost of reduced transport efficiency through the pit aperture (Domec et al., 2008). Given the linear relationship between the torus : aperture diameter and height, this compromise in pit structure places an important constraint on the maximum height that these trees can reach. Whether any clear relationship exists between pit architecture and tree height in tall angiosperms remains to be seen.

Future directions for research

  1. Top of page
  2. Cavitation resistance in angiosperms
  3. Control of cavitation resistance in conifers
  4. Future directions for research
  5. References

While great strides have recently been made in our understanding of structure–function relationships in the xylem, important gaps still remain. For example, the spatially complex structure of the angiosperm vessel network has not often been incorporated into measurements of xylem function. The three-dimensional arrangement and connectivity of vessels has enormous potential to influence the efficiency and the propagation of embolism through the xylem. As imaging technology, such as X-ray computed tomography and magnetic resonance imaging, is refined, our ability to resolve flow and propagation of embolism, in three dimensions and in real time, will be greatly improved. We can now measure hydraulic function directly at the pit level, so given the great variation in the structure of interconduit pits and their importance to hydraulic function, further direct measurements are warranted. In conifers, additional evaluation of the margo structure, and its implications for hydraulic trade-offs, should be considered. Because of its delicate nature the margo is often difficult to visualize using scanning electron microscopy. A combination of careful observation of margo structure and improved capability to simulate flow through complex structures should allow an improved resolution of the role that variation in margo structure plays in trade-offs at the tissue and whole-plant levels.

References

  1. Top of page
  2. Cavitation resistance in angiosperms
  3. Control of cavitation resistance in conifers
  4. Future directions for research
  5. References