Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Journal of Experimental Botany, Vol. 53, No. 367, pp. 195–200, February 2002 Hydraulic and chemical signalling in the control of stomatal conductance and transpiration Jonathan P. Comstock1 Boyce Thompson Institute for Plant Research, Tower Road, Ithaca, NY 14853, USA Received 25 September 2001; Accepted 23 October 2001 Abstract Abscisic acid (ABA) transported in the xylem from root to shoot and perceived at the guard cell is now widely studied as an essential regulating factor in stomatal closure under drought stress. This provides the plant with a stomatal response mechanism in which water potential is perceived in the root as an indication of soil water status and available water resources. There is also ample evidence that stomata respond directly to some component of leaf water status. This provides additional information about water potential gradients developing between root and shoot as the result of water transport, allowing for a more stable regulation of shoot water status and better protection of the transport system itself. The precise location at which leaf water status is sensed, however, and the molecular events transducing this signal into a guard cell response are not yet known. Major questions therefore remain unanswered on how water stress signals perceived at root and leaf locations are integrated at the guard cell to control stomatal behaviour. Key words: ABA, drought response, humidity response, hydraulic conductance, leaf water potential, stomatal regulation, xylem cavitation. Introduction The past decade has seen considerable progress in understanding both chemical and hydraulic signals in the regulation of stomatal conductance, and this parallel growth has also led to some debate on the relative importance of the two signal types in controlling transpiration. Although sometimes framed in eitheruor 1 E-mail: [email protected] ß Society for Experimental Biology 2002 terms, the positive results in both areas suggest that the important questions are when, where and how the two types of signal are integrated in the regulatory process. Foliage, in which photosynthesis and 95% or more of the plant’s water consumption will occur, is distantly removed from the sources of water in the soil, but is connected to the source pool by an elaborate xylem transport system and the absorptive surfaces of the roots. This system is wonderfully efficient, extracting water from the soil matrix and maintaining plant hydration despite frequently high transpiration rates. While the hydration states of different tissues are very sensitive to the magnitude of the hydraulic conductance, direct physical control of transpiration itself resides almost entirely in environmental conditions of temperature and humidity and the stomatal regulation of gas-phase diffusion between leaf air spaces and the atmosphere (van den Honert, 1948). While the opening and closing of the stomatal pore is actually achieved by turgor regulation in the guard cell, these are generally actively controlled processes of osmotic regulation and not passive consequences of bulk tissue water potential (Buckley and Mott, 2001). Tremendous importance has therefore been placed on understanding the regulatory signals that govern stomatal behaviour. These signals must generate behaviour that protects the plant from dehydration while still allowing ample diffusive exchange for CO2 capture. Guard cell regulation has become an important model system in plants because of the central importance and accessibility of guard cells, the fact that several distinct signal transduction mechanisms are known to exist and thus provide a model to study the processes of signal integration, and because the phenotypic responses of turgor regulation and aperture control are relatively easy to score (Schroeder et al., 2001). This signal transduction matrix allows an integration of both the opening signals related to the need for carbon capture and closing signals 196 Comstock related to water management and hydraulic constraints. This enormous topic can only be touched upon in this commentary. Therefore the focus will be on the stomatal closure response to water deficits with an evaluation of (1) root-based assessment of soil water status (Qs), transmitted to the guard cells via long-distance chemical signals such as abscisic acid (ABA), (2) sensing of the proximal leaf water potential (Ql) near or at the guard cell, and (3) the integration of these signals in guard cell behaviour. Evidence for chemical signalling Reports on long-distance chemical signalling have pointed out that declining stomatal conductance during drought is consistently correlated with changes in Qs, while it is only weakly or only in some species wellcorrelated with Ql. Conceptually, it can be argued that soil moisture represents the available resource pool, and the status of that resource should logically govern the modulations of growth and plant water use, including stomatal closure during drought (Davies and Zhang, 1991). The root is in intimate contact with the soil, and root water potential (Qr) may be the logical point to sense changes in Qs most effectively. Further, numerous and co-ordinated responses to drought are needed, and a chemical signal proceeding from the root throughout the plant may better achieve such co-ordination than independent responses to local water status by scattered diverse tissues located at different positions along the flow path in the plant. These ideas are supported by an extensive and growing database of studies showing that ABA synthesis in roots increases in response to soil water deficits, that this leads to elevated xylem wABAx and increased xylem transport of ABA from root to shoot. A wealth of data has also accumulated on the efficacy of ABA in causing stomata closure, now a standard assay in the study of guard cell physiology. Although the specific receptor(s) for ABA remain to be identified, the cascade of events triggered by ABA that lead to stomatal closure are intensely studied, and at least the broad outlines of these mechanisms are now understood at the molecular level (Schroeder et al., 2001). Particularly important in much of the root-signal research has been the use of splitroot culture, where individual plants are grown with the root system divided among two or more separated soil volumes with independent moisture contents. These have been used to show that having a portion of the roots in dry soil can trigger strong stomatal closure even when shoot water potential does not noticeably decline. Even more importantly, they have sometimes demonstrated that a positive signal is present (an inducible signal added to the transpiration stream under drought), rather than a negative signal (a constitutive signal removed during drought), by showing that stomatal closure responses are much stronger if half the root system is droughted than if half the root system is simply excised (Gowing et al., 1990). Evidence for hydraulic signalling In contrast, studies focused on the hydraulic architecture of plants and the vulnerability of the transport system to failure under stress emphasize that a lack of response to distal shoot water potentials would be dangerous and potentially fatal to plants. Many plants operate close to the hydraulic limits of transport where cavitation of the xylem under excess tension can cause a rapid loss of conducting capacity with a strong potential for positive feedback (Sperry et al., 2001). This would result in total loss of hydraulic conductance in the absence of stomatal regulation of the flux, and the empirically observed margins for safety become even tighter during drought stress. While root-sourced signals might anticipate such stresses in a general manner, they could not respond to specific cavitation events because the progressive failure of the xylem under a constant transpiration rate would at first have no effect at all on upstream Qr, and, in the final stages where xylem continuity was lost entirely, Qr would be more positive. An essential component of this debate is the number of forms that ‘hydraulic signalling’ can take. Much of the literature assumes that hydraulic signals must involve stomatal responses to steady-state changes in Ql, such that there would be a strong correlation between declining stomatal conductance and Ql during drought. This, however, is only one limited possibility. If Ql is regulated homeostatically (Comstock and Mencuccini, 1998; Saliendra et al., 1995), at least in the sense that it is not allowed to drop below some minimum value, then several kinds of ‘hydraulic signals’ are suggested by simply considering the common Ohm’s law analogy for water flux. The water flux must be equal to both the loss through the stomata (gsD) and liquid-phase transport (hcDQ), where D is the leaf to air vapour pressure gradient, gs the stomatal conductance, and hc the hydraulic conductance. Setting these equal at steady-state lets us write: hc (Ys Yl ) (1) gs ¼ D This suggests that under high transpiration conditions when Ql has approached a constant Qmin, stomatal conductance should be directly proportional to hydraulic conductance, drop linearly with Qs during drought, and respond inversely to evaporative demand by the atmosphere. All of these behaviours are frequently reported (Buckley and Mott, 2001; Sperry et al., 2001). Stomatal conductance has also been shown to respond as predicted to experimental manipulations of hydraulic conductance Hydraulic and chemical signalling through wounding or artificially induced cavitation in the stem (Sperry et al., 2001). It is important to note that in this context, many of the split-root experiments showing stomatal closure, despite no reduction in Ql while half the root system is droughted, are entirely consistent with hydraulic signalling through reduced hydraulic conductance. In this view, it is the reduction in hydraulic conductance caused by soil drying, flooding or other perturbation which has required a reduction in stomatal conductance in order to maintain the observed, stable Ql. On the other hand, experiments discussed above in which a partially severed root treatment behaves more like a fully watered control than the half-droughted treatment, thus indicating the presence of a positive signal, do argue against the sole operation of an hydraulic signal. The reduction of hydraulic conductance would be analogous to a negative signal. Evidence for integration of chemical and hydraulic signals While it is difficult to reject the strong evidence for positive, root-sourced signals in a great many studies, and the likely identity of ABA as at least one of the most common of these important signals, it has also been hard to escape from invoking interactions with Ql to explain the full range of responses in stomatal conductance. For example, to reconcile the dosageu response characteristics of the guard cells to ABA with observed xylem imports over a drought sequence, it is often necessary to assume increased sensitivity to ABA at low Ql, an inherent merging of the earlier concepts of chemical versus hydraulic signalling (Tardieu and Davies, 1993). Further, the very stability of Ql in so many reports is an extremely challenging observation that needs to be fully explained, as illustrated by a simple rearrangement of Equation 1: gs D Yl ¼ Ys (2) hc Numerous reports in the literature detail a wide range of different stress treatments in various experiments which are associated with remarkably constant values of Ql when comparing stressed and control plants. This constancy of Ql is often cited as evidence against a hydraulic signal. Such a conclusion, however, implies that the diverse treatments are (1) as a side effect, causing large alterations in hydraulic conductance, (2) that ABA is having an entirely independent effect on stomatal conductance and (3) that these independent effects on hydraulic and stomatal conductances (weighted by the prevailing humidity, another independent factor) just happen to be so consistently well balanced that they compensate perfectly in Equation 2 and produce no measurable perturbation in Ql. This is possible in some 197 cases, but it seems an improbable explanation of such general behaviour. More likely hypotheses would be either that both chemical and hydraulic signals are operative and being integrated at the level of stomatal regulation, or even that hydraulic conductance itself is somehow being actively regulated. As has been pointed out previously (Tardieu and Davies, 1993; Tardieu and Simonneau, 1998) it is often precisely in those species in which Ql shows the least variation that a component of hydraulic signalling may be most clearly present. Nevertheless, statements that observed homeostatic conservation of Ql during various treatments rules out hydraulic signals are still common in the literature. Recent advances Pressurization of the soil air spaces independently of the shoot environment is now a common treatment in many experiments. When using this technique, both gas and liquid compartments of the soil become equally pressurized, and these altered pressures extend into both respective phases in the root tissues. Because both liquid and gas phases are equally affected, it is argued that turgor–volume relations of the root system and the relative water contents of its tissues are not altered. Thus the root is not thought to perceive these treatments as altering its water status or that of the soil. When xylem fluids pass out of the soil pressure chamber into the shoot environment, however, the water pressure is still elevated, but perceived by the shoot against a background of normal ambient air pressure. The soil airspace pressurization is thus experienced as an increased total water potential, turgor, and relative tissue water content in the shoot, but not the root (Comstock and Mencuccini, 1998; Passioura, 1980). The technique is also now one of the preferred approaches to obtaining xylem sap under active gas-exchange conditions in intact plants, because the soil can be pressurized until the shoot xylem develops a positive pressure and exudes into a sample vial. It has also been extensively used to evaluate direct responses of stomatal conductance to Ql. In just one example of the importance of time scale, in several long-term treatments lasting for days or weeks, stomatal conductance of continuously pressurized plants declined in a similar manner to unpressurized controls (Passioura, 1988a; Schurr and Schulze, 1996), but in the great majority of cases where soil pressurization treatments lasted for a few hours only, large stomatal opening effects could be initiated by elevating Ql (Comstock and Mencuccini, 1998; Else et al., 2001; Fuchs and Livingston, 1996; Matzner and Comstock, 2001; Mencuccini et al., 2000; Saliendra et al., 1995). In many of these studies, the elevation of Ql was a transient phenomenon, and, at steady-state, variations in soil pneumatic pressure 198 Comstock ultimately translated into variation in stomatal conductance at relatively constant Ql, consistent with hydraulic signalling as discussed above in relation to Equation 1. Other studies also provide strong evidence of rapid changes in local foliar water status because of altered flux dynamics in the surrounding tissues being rapidly translated into stomatal responses (Buckley and Mott, 2000). A great deal of progress has been made in elucidating the nature and dynamics of ABA signalling. For rootbased ABA signals targeting the guard cell, the concentration of ABA in xylem sap rather than total ABA delivery appears to be the primary signal (Jia and Zhang, 1999). Nonetheless, xylem-borne ABA must move through the leaf apoplast in contact with mesophyll and epidermal tissues before reaching the guard cell. Variable rates of ABA sequestration and metabolism by these tissues can have a strong impact on concentrations at the guard cell apoplast where it will be transduced into a closing signal (Hartung et al., 1998). While this process creates many possibilities for interactions dependent on Ql, most current results indicate control by a complement of xylem-borne signals rather than regulatory integration with leaf water status (Wilkinson and Davies, 1997). Thus, at low Qs, the leaf apoplast has a more alkaline pH which reduces uptake of ABA by intervening tissues and increases the concentration near the guard cell, but if it is solely the change in xylem sap pH controlling the leaf apoplast there is no regulatory role for Ql. Other studies, however, have found evidence for a direct influence of Ql on ABA partitioning within a leaf ( Popova et al., 2000), and considerable room for signal integration exists in this area, which needs exploration in future studies. Transpiration itself, independent of stomatal closure owing to lowered soil or plant water status, is most often varied experimentally by varying ambient humidity. Most species show considerable stomatal closure as humidity decreases (Popova et al., 2000), and, at moderate to high water vapour gradients between leaf and air, such closure may be sufficient to compensate fully for variation in humidity, holding transpiration (and Ql) constant. The role of ABA in these diurnal modulations is much less clear than in soil drying. Increased water flux per se should dilute and lower xylem wABAx, but to the degree that the root tissues feel a flux-dependent depression in Qr, increased synthesis and release of ABA might compensate for this in intact plants (Schurr and Schulze, 1995, 1996). Nonetheless, the common reversibility of lowered stomatal conductance at high vapour pressure gradients with soil pressurization, and reports from studies that wild-type humidity responses are seen in ABA-deficient mutants (Assmann et al., 2000) suggests that hydraulic signals and Ql regulation may dominate in these responses. Other recent studies, however, have shown that wABAx of the guard cell apoplast (as well as other chemical and osmotic factors) do in fact increase substantially in response to increased transpiration rates (Zhang and Outlaw, 2001) and this might explain why soil pressurization usually fails to achieve complete stomatal recovery. Here again, multiple concurrent signals are indicated. Issues for the future One of the most limiting issues in assessing these relationships is that current knowledge of ABA signal transduction has greatly outstripped information on exactly how and where Ql is sensed when generating stomatal responses. It is not known to what degree these dynamics are intimately linked with the ABA story, or if there is a completely distinct signal carried to the guard cell. Buckley and Mott recently concluded that Ql was likely to be sensed in the mesophyll or epidermis but not the guard cell, because of the dramatic turgor-fluctuations associated with guard cells (Buckley and Mott, 2001). By contrast, Assman et al. concluded that numerous conflicting observations on behaviour might be most easily reconciled if the guard cells were themselves the site of Ql sensing (Assmann et al., 2000). These contrasting conclusions are still speculative, however, and indicate the great need to identify the actual signal transduction events associated with responses to Ql. The possibility exists, however, that, rather than adjusting stomatal conductance in response to Ql as hydraulic conductance is altered by treatments, hydraulic conductance might itself be under dynamic adjustment to control Ql under the prevailing conditions of transpiration. This runs counter to the most commonly accepted paradigm (Passioura, 1988b), in which hydraulic conductance is viewed to be somewhat variable at very low values of transpiration only, but more stable and fixed by anatomy at high transpiration until xylem cavitation begins at excessive flux rates and depression of Ql. This paradigm of stable hydraulic conductance is consistent with many datasets (Comstock and Mencuccini, 1998; Sperry et al., 2001) in which hydraulic conductance is clearly constant as stomatal conductance is regulated in response to both humidity and moderate soil drying, or in which variation in hydraulic conductance is clearly explained by altered environmental conditions such as temperature fluctuations (Matzner and Comstock, 2001). There are, however, several recent reports that variability in hydraulic conductance may be seen even at high transpiration and at various points in the flow pathway (Clarkson et al., 2000; Makoto and Tyree, 2000; Steudle, 2001; Zwieniecki et al., 2001). This behaviour is reported to be under reversible physiological control via Hydraulic and chemical signalling the gating of aquaporins, changes in the ionic composition of the xylem or other unspecified mechanisms. It is far from clear yet how common these phenomena are in creating variable hydraulic conductance at high transpiration, but they raise the possibility that hydraulic conductance could be adjusted secondarily, after chemical signals have regulated stomatal conductance, perhaps in order to hold Ql constant and maintain stable osmotic relationships in the leaf. This could possibly be an important aspect for the stability of other signalling mechanisms at the guard cell, but such behaviour is not yet well described or understood. In this context it is worth noting that the water flow pathway leaving the xylem network at the leaf veins is, like the radial, extraxylary flow of water in roots, one of the largest single resistances in the flow pathway for many plants (Comstock and Mencuccini, 1998; Matzner and Comstock, 2001; Tiekstra et al., 2000), and likely to involve a symplastic step in the bundle sheath parenchyma around veins (Canny, 1990). Variable hydraulic conductance from aquaporin expression and gating could conceivably act here to stabilize Ql at low to moderate transpiration without adding to the tensions and cavitation vulnerability of the xylem network, but such regulatory effects are wholly speculative and have not been documented in any extensive studies. It seems unlikely that hydraulic conductance at high transpiration, when damaging levels of tension are approached in the xylem, would be allowed to remain at less than its physiological maximum in any event. Of increasing interest in the next several years will be broader comparative studies on the differences of behaviour in different species. Anisohydric versus isohydric species, herbaceous versus woody with their contrasting capacitances and xylem volumes, and plants with a range of different water-use patterns may all have strongly contrasting patterns of the relative importance of chemical versus hydraulic signals in the responses to different environmental variables. Timecourses of responses are also poorly understood, but often suggest that not only are multiple signals present in the induction anduor maintenance of stomatal closure, but they may have a strong temporal variation in which signals predominate under different conditions (Else et al., 2001; Liang and Zhang, 1999). The increasing availability of mutants and other genetic tools to probe specific responses and interactions will be important approaches in resolving some of these issues (Assmann et al., 2000; Borel et al., 2001). Conclusions Both hydraulic and chemical signalling are probably important in the stomatal regulation of most plants. The 199 tight regulation of Ql is best viewed as evidence for hydraulic signaling and not the lack of it. Even a more positive value of Ql during a stress treatment is only conclusive evidence that other signals are present and may have increased in importance, and not necessarily that hydraulic signalling is absent at that time. Rootsourced and hydraulic signals are likely to have different kinds of information in them about environmental conditions as well as the internal state of the plant and its transport system, and therefore it is expected that both would be important components in regulating stomatal behaviour. These dual components are supported by strong evidence in both cases, and guard cells are apparently influenced by both, as well as numerous other signals in their integrated responses. References Assmann SM, Snyder JA, Lee YRJ. 2000. ABA-deficient (aba1) and ABA-insensitive (abi1-1, abi2-1) mutants of Arabidopsis have a wild-type stomatal response to humidity. Plant, Cell and Environment 23, 387–395. Borel C, Frey A, Marion-Poll A, Tardieu F, Simonneau T. 2001. Does engineering abscisic acid biosynthesis in Nicotiana plumbaginifolia modify stomatal response to drought? Plant, Cell and Environment 24, 477–489. Buckley TN, Mott KA. 2000. Stomatal responses to non-local changes in pfd: evidence for long-distance signaling. Plant, Cell and Environment 23, 301–309. Buckley TN, Mott KA. 2001. Stomatal water relations and the control of hydraulic supply and demand. Progress in Botany (in press). Canny MJ. 1990. What becomes of the transpiration stream. New Phytologist 114, 341–368. Clarkson DT, Carvajal M, Henzler T, Waterhouse RN, Smyth AJ, Cooke DT, Steudle E. 2000. Root hydraulic conductance: diurnal aquaporin expression and the effects of nutrient stress. Journal of Experimental Botany 51, 61–70. Comstock JP, Mencuccini MM. 1998. Control of stomatal conductance by leaf water potential in Hymenoclea salsola (T&G), a desert subshrub. Plant, Cell and Environment 21, 1029–1238. Davies WJ, Zhang JH. 1991. Root signals and the regulation of growth and development of plants in drying soil. Annual Review of Plant Physiology and Plant Molecular Biology 42, 55–76. Else MA, Coupland D, Dutton L, Jackson MB. 2001. Decreased root hydraulic conductivity reduces leaf water potential, initiates stomatal closure and slows leaf expansion in flooded plants of castor oil (Ricinus communis) despite diminished delivery of ABA from the roots to shoots in xylem sap. Physiologia Plantarum 111, 46–54. Fuchs EE, Livingston NJ. 1996. Hydraulic control of stomatal conductance in Douglas fir wPseudotsuga menziesii (Mirb) Francox and alder wAlnus rubra (Bong)x seedlings. Plant, Cell and Environment 19, 1091–1098. Gowing DJG, Davies WJ, Jones HG. 1990. A positive rootsourced signal as an indicator of soil drying in apple, Malus 3 domestica Borkh. Journal of Experimental Botany 41, 1535–1540. 200 Comstock Hartung W, Wilkinson S, Davies WJ. 1998. Factors that regulate abscisic acid concentrations at the primary site of action at the guard cell. Journal of Experimental Botany 49, 361–367. Jia W, Zhang J. 1999. Stomatal closure is induced rather by prevailing xylem abscisic acid than by accumulated amount of xylem-derived abscisic acid. Physiologia Plantarum 106, 268–275. Liang J, Zhang J. 1999. The relations of stomatal closure and reopening to xylem ABA concentration and leaf water potential during soil drying and rewatering. Plant Growth Regulation 29, 77–86. Makoto T, Tyree MT. 2000. Plant hydraulic conductance measured by the high pressure flow meter in crop plants. Journal of Experimental Botany 51, 823–828. Matzner S, Comstock JP. 2001. Temperature dependence of shoot hydraulic resistance: implications for stomatal behavior. Plant, Cell and Environment (in press). Mencuccini M, Mambelli S, Comstock JP. 2000. Stomatal responsiveness to leaf water status in common bean is a function of time of day. Plant, Cell and Environment 23, 1109–1118. Passioura JB. 1980. The transport of water from soil to shoot in wheat seedlings. Journal of Experimental Botany 31, 333–345. Passioura JB. 1988a. Root signals control leaf expansion in wheat seedlings in drying soil. Australian Journal of Plant Physiology 15, 687–693. Passioura JB. 1988b. Water transport in and to roots. Annual Review of Plant Physiology and Plant Molecular Biology 39, 245–265. Popova LP, Outlaw Jr WH, Aghoram K, Hite DRC. 2000. Abscisic acid: an intraleaf water-stress signal. Physiologia Plantarum 108, 376–381. Saliendra NZ, Sperry JS, Comstock JP. 1995. Influence of leaf water status on stomatal response to humidity, hydraulic conductance and soil drought in Betula occidentalis. Planta 196, 357–366. Schroeder JI, Allen GJ, Hugouvieux V, Kwak JM, Waner D. 2001. Guard cell signal transduction. Annual Review of Plant Physiology and Plant Molecular Biology 52, 627–658. Schurr U, Schulze E-D. 1995. The concentration of xylem sap constituents in root exudate and in sap from intact, transpiring castor bean plants (Ricinus communis L.). Plant, Cell and Environment 18, 409–420. Schurr U, Schulze E-D. 1996. Effects of drought on nutrient and ABA transport in Ricinus communis. Plant, Cell and Environment 19, 665–674. Sperry JS, Hacke UG, Oren R, Comstock JP. 2001. Water deficits and hydraulic limits to leaf water supply. Plant, Cell and Environment (in press). Steudle E. 2001. The cohesion–tension mechanism and the acquisition of water by plant roots. Annual Review of Plant Physiology and Plant Molecular Biology 52, 847–875. Tardieu F, Davies WJ. 1993. Integration of hydraulic and chemical signalling in the control of stomatal conductance and water status of droughted plants. Plant, Cell and Environment 16, 341–349. Tardieu F, Simonneau T. 1998. Variability among species of stomatal control under fluctuating soil water status and evaporative demand: modelling isohydric and anisohydric behaviours. Journal of Experimental Botany 49, 419–432. Tiekstra AE, Else MA, Jackson MB. 2000. External pressures based on leaf water potentials do not induce xylem sap to flow at rates of whole plant transpiration from roots of flooded or well-drained tomato and maize plants. Impact of shoot hydraulic resistances. Annals of Botany 86, 665–674. van den Honert TH. 1948. Water transport in plants as a catenary process. Discussions of the Faraday Society 3, 146–153. Wilkinson S, Davies WJ. 1997. Xylem sap pH increase: a drought signal received at the apoplastic face of the guard cell that involves the suppression of saturable abscisic acid uptake by the epidermal symplast. Plant Physiology 113, 559–573. Zhang SQ, Outlaw Jr WH. 2001. The guard-cell apoplast as a site of abscisic acid accumulation in Vicia faba L. Plant, Cell and Environment 24, 347–355. Zwieniecki MA, Melcher PJ, Holbrook NM. 2001. Hydrogel control of xylem hydraulic resistance in plants. Science 291, 1059–1062.