In vivo magnetic resonance imaging of xylem ... - Wiley Online Library

Blackwell Science, LtdOxford, UKPCEPlant, Cell and Environment0016-8025Blackwell Science Ltd 2003? 2003 26?12051214 Original Article Magnetic resonance imaging of xylem contents M. J. Clearwater & C. J. Clark

Plant, Cell and Environment (2003) 26, 1205–1214

In vivo magnetic resonance imaging of xylem vessel contents in woody lianas M. J. CLEARWATER1 & C. J. CLARK2 1

Horticulture and Food Research Institute of New Zealand, Te Puke Research Center, RD 2 Te Puke, New Zealand and Horticulture and Food Research Institute of New Zealand, Ruakura Research Center, East Street, Private Bag 3123, Hamilton, New Zealand

2

ABSTRACT

INTRODUCTION

Previous reports suggest that in some plant species the refilling of embolized xylem vessels can occur while negative pressure exists in the xylem. The aim of this experiment was to use non-destructive nuclear magnetic resonance imaging (MRI) to study the dynamics of xylem cavitation and embolism repair in-vivo. Serial 1H-MRI was used to monitor the contents of xylem vessels in stems of two dicotyledonous (Actinidia deliciosa and Actinidia chinensis, kiwifruit) and one monocotyledonous (Ripogonum scandens, supplejack) species of woody liana. The configuration of the horizontal wide bore magnet and probe allowed the imaging of woody stems up to 20mm in diameter. Tests using excised stems confirmed that the image resolution of 78 mm and digital image subtraction could be used to detect the emptying and refilling of individual vessels. Imaging was conducted on both intact plants and excised shoots connected to a water supply. In the case of Ripogonum the excised shoots were long enough to allow the distal end of the shoot, including all leaves, to be exposed to ambient conditions outside the building while the proximal end was inside the MRI magnet. In total, six stems were monitored for 240h while the shoots were subjected to treatments that included light and dark periods, water stress followed by re-watering, and the covering of all leaves to prevent transpiration. The sudden emptying of water-filled vessels occurred frequently while xylem water potential was low (below -0.5MPa for Actinidia, -1.0MPa for Ripogonum), and less frequently after xylem water potential approached zero at the end of water-stress treatments. No refilling of empty vessels was observed at any time in any of the species examined. It is concluded that embolism repair under negative pressure does not occur in the species examined here. Embolism repair may be more likely in species with narrower xylem vessels, but further experiments are required with other species before it can be concluded that repair during transpiration is a widespread phenomenon.

Much recent attention has been focused on the possible occurrence of daily variation in the proportion of embolized vessels in living, transpiring plants (e.g. Salleo & Logullo 1989; Grace 1993; Canny 1997; Zwieniecki & Holbrook 1998; Holbrook & Zwieniecki 1999; Melcher et al. 2001). Cavitation and the subsequent formation of emboli in xylem conduits was previously thought to be irreversible, or reversible only over long time scales and when pressure in the xylem was close to or above atmospheric pressure (Sperry et al. 1994; Magnani & Borghetti 1995; Sperry 1995). Recent studies using a variety of direct and indirect techniques, described below, have provided compelling evidence that in some species embolized vessels can refill in minutes or hours, and that refilling may occur during transpiration while the sap in neighbouring vessels is under significant tension. If the existence of a mechanism for the refilling of embolized vessels under tension can be demonstrated it will have important consequences for our understanding of long-distance transport and water use by plants (Holbrook & Zwieniecki 1999; Tyree et al. 1999). However, because of the difficulty of accessing the xylem without disrupting its contents, few reports have provided unequivocal evidence of refilling under tension in intact plants. Nuclear magnetic resonance imaging (MRI) is one method that provides the opportunity to non-destructively image the contents of xylem vessels in vivo (Callaghan, Clark, & Forde 1994; Holbrook et al. 2001). In this study we use MRI to non-destructively monitor changes in the contents of individual xylem vessels in three species of woody liana. What evidence is there for refilling under tension? Indirect methods are commonly used for estimating the proportion of embolized xylem conduits. These include measurement of variation in hydraulic conductivity (Salleo et al. 1996; Tyree et al. 1999; Zwieniecki et al. 2000; Melcher et al. 2001), measuring the volume of gas bubbles in sap aspirated from the xylem (Pate & Canny 1999), and recording the proportion of stained vessel walls after stems are excised and supplied with dye (Salleo & Logullo 1989; Zwieniecki & Holbrook 1998). In some examples refilling conforms with physical laws governing the dissolution of gas from bubbles within the xylem when tension is very low (Borghetti et al. 1991; Tyree & Yang 1992; Yang & Tyree

Key-words: Ripogonum; Actinidia; cavitation; embolism repair; negative pressure; transpiration; water transport. Correspondence: Michael J. Clearwater. Fax: + 64 75733871; e-mail: [email protected] © 2003 Blackwell Publishing Ltd

1205

1206 M. J. Clearwater & C. J. Clark 1992; Sobrado, Grace, & Jarvis 1992). In other cases apparent increases in the proportion of water-filled conduits have been observed at xylem tensions much higher than those that would allow an explanation by any conventional theories concerning xylem water transport (e.g. Salleo & Logullo 1989). In most examples measurements are destructive, within-treatment variability is high, and there are sometimes other possible explanations for the observed variation in hydraulic conductivity, such as changes in sap ion concentration or variation in sap viscosity with temperature (Cochard et al. 2000b; Zwieniecki, Melcher, & Holbrook 2001). Refilling may also be observed in experiments that make use of excised and dehydrated shoots that are then bagged, thus preventing further evaporation and normal transpirational flow (Salleo et al. 1996; Tyree et al. 1999). In addition to indirect measurements there are also two methods for directly observing the contents of xylem conduits – cryo-scanning electron microscopy (cryo-SEM) and MRI. Canny and others have provided numerous examples of daily cycles of vessels emptying and refilling by rapidly freezing stems and roots of transpiring plants using liquid N2 and imaging the xylem using the SEM (Canny 1997; Buchard, McCully, & Canny 1999; Melcher et al. 2001; Facette et al. 2001). However, this method is destructive and has been strongly criticized over the possibility that the emptying of vessels may be an artifact caused by freezing of xylem under tension (McCully et al. 2000; Cochard et al. 2000a; Canny, McCully, & Huang 2001; Richter 2001). While debate over the validity of the cryo-SEM method continues, MRI provides a non-destructive alternative, allowing the serial imaging of xylem in intact, transpiring plants. MRI has recently been used to measure bulk flow velocities in plant vascular tissue (Kockenberger et al. 1997; Rokitta et al. 1999; Wistuba et al. 2000; Scheenen et al. 2002), and to study the refilling kinetics of resurrection plants (Wagner et al. 2000). In the first attempt to observe the refilling of individual vessels in a transpiring woody plant, Holbrook et al. (2001) observed embolism formation and refilling in a Vitis stem using MRI. Refilling was observed only after the lights were switched off and measured xylem tension had approached zero. While no root exudation could be detected, the authors were unable to eliminate the possibility that refilling was the result of positive root pressure. Their results clearly demonstrate the potential value of MRI for advancing our understanding of embolism formation and repair. While there are some drawbacks, including high operating costs and the difficulty of obtaining sufficient image resolution to observe individual vessels, there is now an urgent need for further work using MRI to determine what conditions are required for refilling of embolized vessels (Holbrook et al. 2001). In particular, whether embolism repair occurs while significant tension exists in the xylem. The aim of this study was to examine the dynamics of embolism formation and repair using MRI. The configuration of the MRI system meant that the in-plane image resolution was lower than that of Holbrook et al. (2001) (78

versus 20 mm), but the diameter of stems that could be imaged was larger (20 versus 6mm). The associated higher total leaf area of the plants also meant that xylem pressure potential could be measured more frequently using excised leaves and a pressure chamber without significantly altering the overall water balance of the plant. Given the lower image resolution we first needed to confirm that this configuration could resolve the contents of individual vessels, and we tested the use of digital image subtraction to detect changes in vessel contents within an image plane that included large numbers of water- and gas-filled vessels. The system was then used to image vessel contents in stems of intact and excised shoots subjected to a variety of treatments. The species used were selected for large vessel diameter (>100 mm), for a liana growth habit, and for contrasting monocotyledon and dicotyledon xylem anatomy. Actinidia deliciosa (A. Chev.) C.F. Liang et A.R. Ferguson var. deliciosa ‘Hayward’ and Actinidia chinensis Planch. var. chinensis ‘Hort16A’ (Actinidiaceae; green- and yellow-fleshed kiwifruit, respectively) are dicotyledonous, deciduous and have diffuse-porous or semi-ring porous secondary xylem anatomy with numerous vessels embedded in a matrix of fibre-tracheids and multiseriate rays (Ferguson 1990; Condon 1992; Dichio et al. 1999; Fig.1). Ripogonum scandens Forst. (Smilacaceae; supplejack, kareao), in contrast, is a monocotyledon with subterranean branching rhizomes and indeterminate, cataphyll-bearing, woody climbing stems up to 20mm in diameter and 20m or more in length. Vascular bundles are scattered, smaller at the periphery of the stem than in the centre, and usually contain two large collateral meta-xylem elements (Simpson & Philipson 1969; Macmillan 1972; Fig.1). Collenchyma sheaths associated with the bundles and confluent around the periphery make the stems extremely durable and flexible (Macmillan 1972). Ripogonum is an unusual and distinctive feature of native New Zealand forest where its abundant, intertwining climbing stems often make travel on foot difficult.

METHODS Plant material Two grafted plants of A. deliciosa and one rooted cutting of A. chinensis were grown in the ground for 3 years in a 0.39ha block of vines at the HortResearch Te Puke Research Orchard near Te Puke, New Zealand. The rootstocks of the two A. deliciosa plants were grown from seedlings of open pollinated A. deliciosa var. deliciosa ‘Bruno’, in keeping with normal commercial practice. All three vines were grown inside 0.08m3 polyester geo-textile fabric rootcontrol bags (Ronabee Tree Farms, Victoria, Australia) as part of an earlier experiment. By digging around and under the root bags it was possible to remove entire plants from the ground with minimal disturbance to the roots. The shoots were trained onto a pergola structure 1.9m high with a permanent central leader and 1-year-old fruiting laterals. Immature fruit were present at the time of measurements. Individuals were selected for the presence of 1-year-old

© 2003 Blackwell Publishing Ltd, Plant, Cell and Environment, 26, 1205–1214

Magnetic resonance imaging of xylem contents 1207

Figure 1. Light microscope and MRI micrograph comparisons of transverse sections of Actinidia deliciosa and Ripogonum scandens stems. (a) Transverse light micrograph of 1-year-old A. deliciosa stem, showing detail from the pith to the bark. (b) Transverse MRI image of the same stem as in (a), showing the whole stem at lower magnification. (c) Transverse light micrograph of Ripogonum stem, showing detail of vascular bundles near the center of the stem. (d) Detail from a transverse MRI image of the same stem as (c), arrows highlight the same metaxylem vessels in each image. Scale bars: (a) 0.5mm; (b) 2mm; (c) and (d) 0.25mm; pi, pith; sx, secondary xylem; sp, secondary phloem; co, cortex; p, phantom (see Methods for explanation).

shoots that were longer than the bore of the MRI magnet (1.2m) and the remainder of the crown was pruned to a manageable size 2 weeks prior to lifting from the ground in November 2001. They were then supported and transferred to a nursery near the MRI facility at Massey University, Palmerston North. The plants were kept well watered and fertilized in the nursery until they were used for MRI imaging in December 2001 or January 2002. Ripogonum can be found growing in forests throughout New Zealand but adult plants are difficult to remove whole without causing damage to the roots and subsequent wilting. Instead, on the day that MRI imaging was to start, whole shoots were cut from wild plants at ground level before dawn, immediately re-cut underwater, then connected to a supply of deionized water via latex tubing. The shoots had stems at least 15mm in diameter and 10m long, and were selected if they could be easily removed from supporting vegetation without damage. They were then coiled and transported by car to the MRI facility. In total, two shoots were collected from regenerating forest on private property near the Te Puke Research Orchard and one shoot from forest near the Massey University campus. For two imaging sessions the stems of one plant each of A. deliciosa and A. chinensis were imaged in vivo with the entire plant transported to the MRI facility and positioned with a 1-year-old shoot inserted through the bore of the magnet. Leaves on laterals at the distal end of the shoot were clear of the bore so that transpirational flow occurred through the stem inside the bore of the magnet and through the imaging plane. The remainder of the plant crown with the majority of leaves was supported at the proximal end of the bore. Banks of 500W tungsten–halogen floodlamps and 100W halogen spotlights were positioned to light the foliage on both sides of the magnet, with natural lighting also provided through skylights in the roof. With the light-

ing switched on the irradiance measured at the leaf surface using a quantum sensor (Licor Inc, Lincoln, NE, USA) varied between 100 and 200 mmol-2 s-1. For a third imaging sequence a 1-year-old shoot was cut from a well-watered A. deliciosa plant at midnight in the nursery, re-cut under water and connected to a water supply using latex tubing. This shoot was then inserted through the magnet without disconnecting the water supply and the foliage at the distal end lit as described above. Three additional imaging sequences were conducted using excised Ripogonum shoots. The length of these vines meant that the proximal end of the main stem could be inserted through the bore of the magnet while the distal end was positioned across the floor of the MRI suite and out through the external door, a distance of approximately 5m. All of the lateral branches and leaves were therefore outside and exposed to ambient light, temperature and humidity. The orientation of the building meant that during the day the leaves were exposed to direct sunlight for at least 8h per day.

Magnetic resonance imaging The theory and application of NMR techniques to biological systems is well covered in the literature. The publications by Ishida, Koizumi & Kano (2000) and Gadian (1995), in particular, provide useful background for those unfamiliar with these approaches. Serial 1H-MRI (200MHz) was performed in a 4.7Tesla horizontal wide-bore (89mm diameter) cryomagnet (Oxford Instruments, Oxford, UK) using custom-designed hardware and software. The 40mm birdcage coil, for example, was a split design. This meant both halves of the coil could be separated, and then rejoined as a collar around the stem, without having to damage the fragile vegetative tissues by forcing them through a narrow


1208 M. J. Clearwater & C. J. Clark orifice. Transverse images of the stem were acquired sequentially during the course of experiment using a Hahn spin-echo pulse sequence with a repetition time (TR) of 2000ms, echo time (TE) of 26ms, bandwidth of 10kHz, and n = 2 acquisitions per slice. Images were acquired as 256 ¥ 256 pixel data arrays with a field-of-view of 20mm (inplane resolution = 78 mm) and a slice thickness of 2mm. Once the imaging parameters (and others such as the gain) were set, they were fixed for the remainder of a study to facilitate comparison of image contrast from one time to the next. Each image required 17min scanning time and they were usually acquired at hourly or half-hourly intervals throughout an experiment. Images were analysed with the public-domain software packages NIH IMAGE (ver. 1.61) and IMAGEJ (ver. 1.27) (National Institutes of Health, Washington, DC, USA). This involved subtracting each image from its predecessor using image math tools and looking for black spots indicative of a high contrast signal change. As a test, transverse images were acquired for excised segments of Actinida and Ripogonum stems that had first been connected to tubing and perfused with water. Further images were acquired after air at a slight positive pressure had been supplied through the tubing without moving the stem from inside the magnet, thus aspirating some of the vessels. In the figures presented, a small high-contrast (white) circular feature is also observed. This is a glass tube containing 200mM MnSO4 (referred to as a phantom) that was attached to the stem to intersect the image plane. The intensity of this feature was used as an internal standard to normalize image intensity and remove any instrument drift during a run.

Treatments and measurements For the two whole Actinidia plants, water was withheld in the nursery for 3d before a plant was used for an imaging session. Imaging began in the mid-afternoon and continued for at least 20h and through one dark period before the root bag was re-watered to saturation. Imaging continued for a further two day/night periods (48h), during which a variety of additional treatments were tested for effects on the dynamics of embolism formation or repair. These included misting with water and covering all of the leaves with large plastic bags to prevent transpiration, addition of 800g of pre-dissolved complete fertilizer (Nitrophoska 1615-16.6; BASF New Zealand, Auckland, New Zealand) to the root bag, and near the end of the imaging run, partial and complete cutting of the stem at a position proximal to the imaging plane. With one plant a lateral proximal to the imaging plane was cut near the end of the experiment, the stump connected to tubing and water supplied at a pressure of 140kPa for an hour. For the excised Actinidia shoot imaging began during the night immediately after excision and proceed for 30h. High transpiration rates were induced the following morning using the artificial lighting, reduced during darkness in the evening, then finally brought to zero by misting and enclosing all of the leaves in a plastic bag. Experiments with the three Ripogonum shoots lasted

between 15 and 48h. Periods of elevated xylem tension were induced by disconnecting the water supply tubing underwater, covering the cut end in pre-soaked dialysis tubing (12kDa MW cut-off) and inserting the stem for up to 2.5h in a polyethylene glycol (PEG; 20000MW) solution made up to give a final osmotic potential of 0.5MPa (Edwards & Jarvis 1982). Alternatively the stem was disconnected from the water supply and held in the air. In all cases the stem was then re-cut underwater, the cut end planed with a razor blade and then reconnected to the water supply tubing. At the end of one experiment the cut end was connected to tubing and water supplied at a pressure of 200kPa for 1h. Periodically during imaging a Scholander pressure chamber (Soil Moisture Equipment Corp., Santa Barbara, CA, USA) was used to measure xylem water potential with leaves that had earlier been covered with aluminium foil to prevent transpiration (Meinzer, Clearwater, & Goldstein 2001). Stomatal conductance (gs) was measured with a steady-state porometer (Licor 1600; Licor Inc.). With excised shoots (one A. deliciosa and the three Ripogonum shoots) a 1mL graduated pipette was fitted to a T-junction in the water supply tube and transpiration measured by temporarily diverting flow through the pipette and timing the movement of the meniscus. Periodic checks were made on the Actinidia plants for signs of root pressure, including checking the cut petiole surfaces for signs of exudation and the ‘pricking’ of the main stem near the soil surface using a hypodermic needle (an effective method for detecting root pressure in orchard-grown plants). At the end of all experiments all leaves were counted and a sub-sample of leaves taken for measurement of leaf area using a leaf area meter (Licor Inc.,). The imaged stem was removed from the RF coil and a portion of the stem from the region of the image plane fixed in formalin, acetic acid and alcohol. Transverse sections were cut by hand or a vibrating microtome, stained in toluidine blue and viewed with a light microscope.

RESULTS Comparison of light and MRI micrographs confirmed that this MRI configuration could achieve sufficient resolution to resolve individual water-filled vessels (Fig.1). The widest vessel members in all three species were up to 300 mm in diameter. In the two dicotyledonous Actinidia species water-filled vessels were visible as groups of whiter pixels within the secondary xylem (Fig.1a & b). In the monocotyledonous Ripogonum the two largest metaxylem vessels within each vascular bundle were normally clearly visible (Fig.1c & d). Aspiration of a water-filled piece of Actinidia stem while it was in place inside the MRI magnet confirmed that it was possible to resolve the change from water-filled to air-filled for individual vessels, and that image subtraction could be used to highlight the vessels that had emptied (Fig.2a–c). Similarly, the refilling of previously embolized vessels was clearly observed after water was applied under pressure to the cut ends or a lateral branch of both Actinidia and Ripogonum stems (Fig.2d–f). Vibration or



Figure 2. Serial transverse MRI images showing artificial emptying and refilling of vessels. (a) Excised piece of 2-year-old stem from A. chinensis, before aspiration with air; (b) after aspiration with air; and (c) the result of subtraction of image (a) from (b). (d) Stem of a Ripogonum shoot, before application of water under pressure to the cut base; (e) after application of water; and (f) the result of subtraction of image (e) from (d). Note that in (c) dark spots represent vessels that have emptied, whereas in (f) the opposite subtraction shows vessels that have refilled. Circles highlight the same vessel in each image sequence. Scale bars: 3mm.

unintentional movement of the stem between images (e.g. while removing leaves for pressure chamber measurements) sometimes prevented the use of image subtraction for a pair of images. However, ‘stacking’ and viewing animated sequences of images within the image analysis software also effectively highlighted any changes in vessel contents. The sudden emptying of Actinidia vessels while plants were imaged in vivo was frequently observed and it is assumed that this emptying was the result of cavitation (images and data are shown only for the A. chinensis plant; Figs3 & 4). Refilling of vessels in vivo was never observed. Between 9 and 43 vessels cavitated during imaging of each of the three Actinidia stems (Fig.4). A higher number of vessels were already empty at the beginning of imaging (between 10 and 70%; Fig.3). Vessel contents sometimes changed from high to intermediate and finally to low signal intensity over a sequence of three images (not shown), probably as a result of cavitation during acquisition of the second image. Overall signal intensity did sometimes change during imaging. However, signal intensity is susceptible to other processes that induce changes in the observed water content. Change in mem-

brane permeability resulting from osmotic stress and temperature change, for example, influences relaxation properties and hence signal intensity in seedlings (Van der Weerd et al. 2001, 2002). No alternative investigations were pursued that might explain these observations, therefore only black and white differences in the contrast of vessel contents have been used in this study. Cavitation was more frequent while xylem water potential was more negative, and decreased in frequency after the vines were re-watered, during dark periods, or when the leaves were misted and covered to prevent transpiration (Fig.4). There was no clear pattern governing which vessels cavitated. Cavitated vessels were sometimes adjacent and sometimes widely distributed across the sapwood, and they appeared to belong to a range of diameter size classes. Addition of dissolved fertilizer to the root-bag of one plant had no apparent effect on the frequency of cavitation or xylem water potential, while partial or complete cutting of the stem 0.5m proximal to the image plane caused the sudden emptying of between 10 and 50 vessels. Stomatal conductance of the Actinidia plants was difficult to measure reliably because of interference caused by the MRI magnet. A gs of 130mmolm-2 s-1 was measured on one


1210 M. J. Clearwater & C. J. Clark

Figure 3. MRI micrographs of cavitation in 1-year-old A. chinensis stem in vivo. Images (a) and (b) were acquired 30min apart. (c) The result of subtracting image (a) from (b), three cavitated vessels are visible. Scale bar: 2mm.

plant while it was water stressed in the nursery before it was transferred to the MRI suite. During imaging, the average gs was low (40mmolm-2 s-1), reflecting the low light levels in the MRI suite compared to outdoors, and there was an increase in gs (to approximately 60mmolm-2 s-1) after the two intact plants were re-watered. Total leaf areas for the intact A. deliciosa and A. chinensis plants and the detached A. deliciosa shoot were 2.6, 3.1 and 0.5m2, respectively. No signs of root pressure were observed. In contrast to the Actinidia stems, the majority of metaxylem vessels in the monocotyledonous Ripogonum stems were water-filled at the beginning of imaging (Fig.5; images and data are shown for one stem only). Cavitation was observed only when xylem water potential decreased rap-

idly in response to the transfer of a cut stem to PEG solution (Figs5 & 6). Cavitation occurred more frequently towards the centre of the stem and usually only involved one of the two collateral meta-xylem vessels in each vascular bundle (Fig.5). During these treatments stomatal conductance decreased rapidly from 47 to 16mmolm-2 s-1 and wilting of the leaves was observed after 1h, especially when the leaves were exposed to bright sun during the treatments. The more negative xylem water potential reflects the osmotic potential of the PEG solution and resistance to flow through the dialysis membrane and cut stem (Fig.6). After the stem was returned to the normal water supply cavitation ceased, transpiration and xylem water potential recovered quickly and the leaves regained turgor. During

Figure 4. Time course of xylem water potential and number of cavitated vessels for the same A. chinensis stem as in Fig.3. Time is measured from the beginning of serial MRI imaging and cavitated vessels is the cumulative number of vessels that cavitated during imaging. Horizontal bars indicate night-time, and the vertical arrow indicates the time of re-watering. © 2003 Blackwell Publishing Ltd, Plant, Cell and Environment, 26, 1205–1214


Figure 5. MRI micrographs of cavitation in a Ripogonum scandens stem during transpiration. Images (a) and (b) were acquired 15min apart. (c) The result of subtracting image (a) from (b), four cavitated vessels are visible. Scale bar: 2mm.

the night transpiration decreased to below 0.05mmolm-2 s-1 and xylem water potential approached zero (Fig.6). No refilling of Ripogonum vessels was observed at any time, regardless of whether a vessel was already empty at the beginning of imaging or whether it emptied during the PEG treatments. Total leaf area for the three Ripogonum shoots varied between 0.5 and 1.7m2.

DISCUSSION Using this MRI system it was possible to image the contents of individual xylem vessels in woody stems up to 20mm in diameter. Serial images of xylem contents were obtained in vivo for intact Actinidia plants and for Ripogonum shoots with their leaves exposed to full sunlight under outdoor conditions. Changes in the contents of single xylem vessels

within a matrix of large numbers of water- and gas-filled vessels were easily detected using digital image subtraction and stacking. Whereas sudden emptying of water-filled xylem vessels was frequently observed in both the dicotyledonous Actinidia stems and the monocotyledonous Ripogonum, no refilling was observed at any time in any of the stems examined. Cavitation was more frequent when xylem tensions were high and less frequent when water stress was relieved. Although there is no previous information on the vulnerability of xylem to cavitation in these species, cavitation occurred more frequently and at lower xylem tensions in Actinidia than in Ripogonum. In contrast to the refilling observed in Vitis by Holbrook et al. (2001), refilling did not occur even when xylem tensions fluctuated widely, when xylem tension was reduced to less than 0.1MPa, when

Figure 6. Time course of xylem water potential (solid line), transpiration rate (dotted line) and number of cavitated vessels for the same Ripogonum stem as in Fig.5. Transpiration was measured at the cut base of the shoot and is expressed per unit leaf area. The horizontal bar indicates night-time, and the vertical arrow indicates the time the cut base was transferred to a PEG solution to induce water stress. See Fig.4 for further explanation. © 2003 Blackwell Publishing Ltd, Plant, Cell and Environment, 26, 1205–1214

1212 M. J. Clearwater & C. J. Clark plants were in darkness, or when transpiration was prevented by complete covering of the leaves. Refilling in the Vitis stem was observed only after leaf water potentials had risen above -0.25MPa, and root pressure cannot be excluded as a possible cause (Holbrook et al. 2001). It is possible that a very rapid refilling mechanism could refill vessels during the 17min required to acquire images in this study (Martin Canny; personal comm.). However, if this occurred a series of images should show signal intensity for some vessels changing from high, to intermediate and then back to high. Within the limits of the imaging method used in this study, the only clear changes were from high to low. Therefore, using MRI we do not yet have clear evidence for embolism repair in transpiring plants while significant tension exists in the xylem. Similarities can be drawn between Vitis, and Actinidia, although the two genera are taxonomically distinct. Both are deciduous, temperate, woody lianas, and both exhibit positive sap pressure in spring that (at least in Vitis) is thought to result in the refilling of embolized xylem vessels (Sperry et al. 1987). Root pressure in Actinidia, when present, is normally easily detected prior to budburst as copious exudation from cut or wounded stems and roots (Ferguson, Eiseman, & Leonard 1983; Clark, Holland, & Smith 1986). Spring xylem pressures up to 100kPa have been measured in A. deliciosa rootstocks (M Clearwater, unpublished results). Root pressure during summer when leaves are present may only be detected as exudation or guttation when evaporative demand is zero, usually on humid nights. At other times transpiration can continue at night (Green, McNaughton, & Clothier 1989) and no exudation is observed. Positive root pressure is less commonly observed in root-controlled and previously water-stressed Actinidia plants, as used in this study, and we were unable to detect any sign of root pressure during MRI measurements. The large number of embolized vessels present in the Actinidia stems at the beginning of imaging also indicates that these plants had not developed positive root pressure in the recent past. It is not known whether positive root pressure occurs in Ripogonum. It is concluded that refilling of embolized xylem vessels while tension exists in conducting vessels is unlikely to occur in the species examined here. Spring and nocturnal root pressure may instead be an important mechanism for maintaining hydraulic capacity in A. deliciosa and A. chinensis, as has been hypothesized for some other liana species (Ewers, Cochard, & Tyree 1997; Fisher et al. 1997). Ripogonum scandens may simply be less vulnerable to cavitation and more tolerant of drought stress. From these results we cannot exclude the possibility that refilling of embolized xylem vessels under tension occurs in other species. A mechanism for embolism repair requires the movement of water from neighbouring xylem parenchyma into the embolized conduit (Canny 1997; Holbrook & Zwieniecki 1999; Tyree et al. 1999). Refilling may therefore be more likely in species with smaller-diameter xylem vessels or tracheids where the volume of water required to fill the lumen is smaller. Refilling may also be more likely in

organs where the xylem is more closely associated with living parenchyma cells, such as petioles, herbaceous stems and in leaf veins. MRI is more difficult to apply in these situations, but more sophisticated imaging techniques can also be used to measure the amount of flowing and stationary water per pixel, even if image resolution is not sufficient to discriminate individual vessels (Scheenen et al. 2002). Further studies of the dynamics of embolism formation and repair using MRI are needed to unambiguously demonstrate whether embolism repair under tension can occur and to determine what conditions are required for repair. Until this has been done we believe that it is not possible to conclude that embolism repair during transpiration is a widespread phenomenon.

ACKNOWLEDGMENTS This work was supported by a grant from the Marsden fund to M.J.C. We thank N.M. Holbrook, P. Minchin and M. Thorpe for their advice and encouragement. G. Thorp, S. Owen, A. Mandemaker, G. Milne, R. Diack, A. Lang, P. Austin and other HortResearch staff provided assistance with plants and equipment. R. Dykstra and others at Massey University provided assistance in the NMR suite.

REFERENCES Borghetti M., Edwards W.R.N., Grace J., Jarvis P.G. & Raschi A. (1991) The refilling of embolized xylem in Pinus sylvestris L. Plant, Cell and Environment 14, 357–369. Buchard C., McCully M.E. & Canny M. (1999) Daily embolism and refilling of root xylem vessels in three dicotyledonous crop plants. Agronomie 19, 97–106. Callaghan P.T., Clark C.J. & Forde L.C. (1994) Use of static and dynamic NMR microscopy to investigate the origins of contrast in images of biological tissues. Biophysical Chemistry 50, 225– 235. Canny M.J. (1997) Vessel contents during transpiration – embolisms and refilling. American Journal of Botany 84, 1223–1230. Canny M.J., McCully M.E. & Huang C.X. (2001) Cryo-scanning electron microscopy observations of vessel content during transpiration in walnut petioles. Facts or artefacts? Plant Physiology and Biochemistry 39, 555–563. Clark C.J., Holland P.T. & Smith G.S. (1986) Chemical composition of bleeding xylem sap from kiwifruit vines. Annals of Botany 58, 353–362. Cochard H., Bodet C., Ameglio T. & Cruiziat P. (2000a) Cryoscanning electron microscopy observations of vessel content during transpiration in walnut petioles. Facts or artifacts? Plant Physiology 124, 1191–1202. Cochard H., Martin R., Gross P. & Bogeat-Triboulot M.B. (2000b) Temperature effects on hydraulic conductance and water relations of Quercus robur L. Journal of Experimental Botany 51, 1255–1259. Condon J.M. (1992) Aspects of kiwifruit stem structure in relation to transport. Acta Horticulturae 297, 419–426. Dichio B., Baldassarre R., Nuzzo V., Biasi R. & Xiloyannis C. (1999) Hydraulic conductivity and xylem structure in young kiwifruit vines. Acta Horticulturae 498, 159–164.


Magnetic resonance imaging of xylem contents 1213 Edwards W.R.N. & Jarvis P.G. (1982) Relations between water content, potential and permeability in stems of conifers. Plant, Cell and Environment 5, 271–277. Ewers F.W., Cochard H. & Tyree M.T. (1997) A survey of root pressures in vines of a tropical lowland forest. Oecologia 110, 191–196. Facette M.R., McCully M.E., Shane M.W. & Canny M.J. (2001) Measurements of the time to refill embolized vessels. Plant Physiology and Biochemistry 39, 59–66. Ferguson A.R. (1990) Stems, branches, leaves and roots of the Kiwifruit vine. In Kiwifruit: Science and Management (eds I.J. Warrington & G.C. Weston), pp. 58–70. NZ Soc. for Hort. Sci. & Ray Richards Publisher, Auckland, New Zealand. Ferguson A.R., Eiseman J.A. & Leonard J.A. (1983) Xylem sap from Actinidia chinensis: seasonal changes in composition. Annals of Botany 51, 823–833. Fisher J.B., Angeles G., Ewers F.W. & LopezPortillo J. (1997) Survey of root pressure in tropical vines and woody species. International Journal of Plant Sciences 158, 44–50. Gadian D.G. (1995) NMR and its Application to Living Systems, 2nd edn. Oxford University Press, Oxford, UK. Grace J. (1993) Refilling of embolized xylem. In Water Transport in Plants Under Climate Stress (eds M. Borghetti, J. Grace & A. Raschi), pp. 51–62. Cambridge University Press, Cambridge, UK. Green S.R., McNaughton K.G. & Clothier B.E. (1989) Observations of night-time water use in kiwifruit vines and apple trees. Agricultural and Forest Meteorology 48, 251–261. Holbrook N.M. & Zwieniecki M.A. (1999) Embolism repair and xylem tension: Do we need a miracle? Plant Physiology 120, 7– 10. Holbrook N.M., Ahrens E.T., Burns M.J. & Zwieniecki M.A. (2001) In vivo observation of cavitation and embolism repair using magnetic resonance imaging. Plant Physiology 126, 27– 31. Ishida N., Koizumi M. & Kano H. (2000) The NMR microscope: a unique and promising tool for plant science. Annals of Botany 86, 259–278. Kockenberger W., Pope J.M., Xia Y., Jeffrey K.R., Komor E. & Callaghan P.T. (1997) A non-invasive measurement of phloem and xylem water flow in castor bean seedlings by nuclear magnetic resonance microimaging. Planta 201, 53–63. Macmillan B.H. (1972) Biological flora of New Zealand. 7. Ripogonum scandens J. r. et G. forst. (Smilacaceae) Supplejack, Kareao. New Zealand Journal of Botany 10, 641–672. Magnani F. & Borghetti M. (1995) Interpretation of seasonalchanges of xylem embolism and plant hydraulic resistance in Fagus sylvatica. Plant, Cell and Environment 18, 689–696. McCully M.E., Baker A.N., Shane M.W., Huang C.X., Ling L.E.C. & Canny M.J. (2000) The reliability of cryoSEM for the observation and quantification of xylem embolisms and quantitative analysis of xylem sap in situ. Journal of Microscopy 198, 24–33. Meinzer F.C., Clearwater M.J. & Goldstein G. (2001) Water transport in trees: current perspectives, new insights and some controversies. Environmental and Experimental Botany 45, 239– 262. Melcher P.J., Goldstein G., Meinzer F.C., Yount D.E., Jones T.J., Holbrook N.M. & Huang C.X. (2001) Water relations of coastal and estuarine Rhizophora mangle: xylem pressure potential and dynamics of embolism formation and repair. Oecologia 126, 182–192. Pate J.S. & Canny M.J. (1999) Quantification of vessel embolisms by direct observation: a comparison of two methods. New Phytologist 141, 33–43. Richter H. (2001) The cohesion theory debate continues: the pitfalls of cryobiology. Trends in Plant Science 6, 456–457.

Rokitta M., Peuke A.D., Zimmermann U. & Haase A. (1999) Dynamic studies of phloem and xylem flow in fully differentiated plants by fast nuclear-magnetic-resonance microimaging. Protoplasma 209, 126–131. Salleo S. & Logullo M.A. (1989) Xylem cavitation in nodes and internodes of Vitis vinifera L. plants subjected to water stress: limits of restoration of water conduction in cavitated xylem conduits. In Structural and Functional Responses to Environmental Stresses: Water Shortage (eds K.H. Kreeb, H. Richter & T.M. Hinckley), pp. 33–42. SPB Academic Publishing, The Hague, The Netherlands. Salleo S., Logullo M.A., DePaoli D. & Zippo M. (1996) Xylem recovery from cavitation-induced embolism in young plants of Laurus nobilis: a possible mechanism. New Phytologist 132, 47– 56. Scheenen T., Heemskerk A., de Jager A., Vergeldt F. & Van As H. (2002) Functional imaging of plants: a nuclear magnetic resonance study of a cucumber plant. Biophysical Journal 82, 481– 492. Simpson P.G. & Philipson W.R. (1969) Vascular anatomy in vegetative shoots of Ripogonum scandens Forst. (Smilacaceae). New Zealand Journal of Botany 7, 3–29. Sobrado M.A., Grace J. & Jarvis P.G. (1992) The limits to xylem embolism recovery in Pinus sylvestris L. Journal of Experimental Botany 43, 831–836. Sperry J.S. (1995) Limitations on stem water transport and their consequences. In Plant Stems: Physiological and Functional Morphology (ed. B.L. Gartner), pp. 105–124. Academic Press, San Diego, CA, USA. Sperry J.S., Holbrook N.M., Zimmermann M.H. & Tyree M.T. (1987) Spring filling of xylem vessels in wild grapevine. Plant Physiology 83, 414–417. Sperry J.S., Nichols K.L., Sullivan J.E.M. & Eastlack S.E. (1994) Xylem embolism in ring-porous, diffuse-porous, and coniferous trees of northern utah and interior alaska. Ecology 75, 1736– 1752. Tyree M.T. & Yang S.D. (1992) Hydraulic conductivity recovery versus water-pressure in xylem of Acer saccharum. Plant Physiology 100, 669–676. Tyree M.T., Salleo S., Nardini A., Logullo M.A. & Mosca R. (1999) Refilling of embolized vessels in young stems of laurel. Do we need a new paradigm? Plant Physiology 120, 11–21. Van der Weerd L., Claessens M.M.A.E., Efde C. & Van As H. (2002) Nuclear magnetic resonance imaging of membrane permeability changes in plants during osmotic stress. Plant, Cell and Environment 25, 1539–1549. Van der Weerd L., Claessens M.M.A.E., Ruttink T., Vergeldt F.J., Schaafsma T.J. & Van As H. (2001) Quantitative NMR microscopy of osmotic stress responses in maize and pearl millet. Journal of Experimental Botany 52, 2333–2343. Wagner H.J., Schneider H., Mimietz S., Wistuba N., Rokitta M., Krohne G., Haase A. & Zimmermann U. (2000) Xylem conduits of a resurrection plant contain a unique lipid lining and refill following a distinct pattern after desiccation. New Phytologist 148, 239–255. Wistuba N., Reich R., Wagner H.J., Zhu J.J., Schneider H., Bentrup F.W., Haase A. & Zimmermann U. (2000) Xylem flow and its driving forces in a tropical liana: concomitant flow-sensitive NMR imaging and pressure probe measurements. Plant Biology 2, 579–582. Yang S. & Tyree M.T. (1992) A theoretical-model of hydraulic conductivity recovery from embolism with comparison to experimental-data on Acer saccharum. Plant, Cell and Environment 15, 633–643. Zwieniecki M.A. & Holbrook N.M. (1998) Diurnal variation in xylem hydraulic conductivity in white ash (Fraxinus americana


1214 M. J. Clearwater & C. J. Clark L.), red maple (Acer rubrum L,) and red spruce (Picea rubens Sarg,). Plant, Cell and Environment 21, 1173–1180. Zwieniecki M.A., Hutyra L., Thompson M.V. & Holbrook N.M. (2000) Dynamic changes in petiole specific conductivity in red maple (Acer rubrum L.), tulip tree (Liriodendron tulipifera L.) and northern fox grape (Vitis labrusca L.). Plant, Cell and Environment 23, 407–414.

Zwieniecki M.A., Melcher P.J. & Holbrook N.M. (2001) Hydrogel control of xylem hydraulic resistance in plants. Science 291, 1059–1062. Received 13 November 2002; received in revised form 10 February 2003; accepted for publication 10 February 2003
