Transporting Water in Plants

Martin J Canny. American Scientist. Volume 86, Issue 2. Mar/Apr 1998.

Plants are thirsty creatures. Reflect on how much more water you feed to a tomato plant than to a cat. For every kilogram of organic material made by a plant, 500 kilograms of water are absorbed by the roots and evaporated, or transpired, from the leaves. Consequently, management of that stream of water dominates the life process of plants. The question is: How is that management achieved?

In general terms, the answer is simple: A plant functions like a wick, evaporating water from its leaves and drawing water up from its roots and the reservoir in the soil. In 1727, the English chemist and physiologist Stephen Hales demonstrated that process by cutting a small tree and sealing its base in a waterfilled tube. Using a mercury manometer, he measured the magnitude of the suction as about 12 inches of mercury (which equals 1/3 atmosphere or 30 kilopascals).

The water-conducting tissue of plants is called xylem. It is composed of dead pipes that are interspersed intimately with living cells. In nonwoody plants, xylem forms vascular strands (including the strings in a celery stalk). The pipes through which the water flows are formed from the walls of files of cells that all died after they built the pipes. The walls are sculptured variously with thin outer zones of primary walls-called pits-and thickened, inner, reinforced zones of secondary wall. Most xylem pipes have diameters in the range from 10 to 100 micrometers, but some trees have very wide ones-such as an oak’s, which can be 300 micrometers across. The pipes that are characteristic of flowering plants are called vessels, and they may be many centimeters-or even meterslong, but one vessel is separated from the next by an intact wall with pits. The thin primary walls in the pits of side and end walls of vessels are permeable to water and solutions, but not to colloidal suspensions or to an air-water meniscus. As a result, those pits block the spread of air through the pipes.

The simple wick analogy cannot be pressed too far. A wick that could use capillary action alone to support threads of water as high as a tall tree would have such narrow pipes-150 nanometers-that the flow through them could not supply the demands of the leaves. Besides, the actual pipes are more than 100 times wider than that. As I shall show, a combination of living cells and dead pipes forms a complex system that supplies the leaves with water.

Water Physics

Three unfamiliar physical facts lie behind all theories of plant-water transport. First, liquids (including water) have tensile strength and can transmit appreciable tensions without breaking. Such tensions can be applied through what is often referred to as negative pressure. Second, solutes dissolved in water generate hydrostatic pressures, called osmotic pressures, when a solution is separated from water (or a more dilute solution) by a membrane that is permeable to water but not to the solutes. Living cells have such membranes.

Plant cells, being confined in boxes of cellulose that resist those pressures, can contain large hydrostatic, or turgor, pressures when supplied with water. Turgor pressures routinely reach from 10 to 30 atmospheres and can rise to well over 100. The rigidity of plant tissues is often due to those pressures. For example, if you cut a fresh surface on the base of wilted lettuce and stand it in water, turgor pressure restores its rigidity. Third, small pores can act as oneway valves for water if there is air on the other side. Water flows easily out of the pore and spreads over the surface on the other side. When air pushes water back through the pore, the water clings tightly to the sides of the pore, and the smaller the pore the more that force resists the air pressure.

Cohesion Theory

From the time of Hales until the late 19th century, there was much debate about whether the living cells of xylem played any role in transporting water. The analogy with a wick was made much closer by Josef Bohm in Vienna in 1893 when he modified Hales’s experimental arrangement (eliminating dissolved air from the water phase) to show that a transpiring shoot could lift a column of mercury through many centimeters. He claimed to have observed the mercury being lifted 90 centimeters (16 centimeters higher than the barometer column), but nearly all other attempts to show that have failed. Note that once the mercury column is higher than the barometer, the pressure in the column is negative. That idea of negative pressure might present difficulties to those accustomed to thinking in terms of reducing atmospheric pressure toward a perfect vacuum, but it is central to the debate on plant-water transport. It is necessary to think of the water as being stretched under tension and to call all tensions greater than 1 atmosphere negative pressures. The absolute scale will be used here, where atmospheric pressure is +1 and a vacuum is 0 atmospheres.

Bohm invoked the chemical theory of very strong bonds between water molecules to extend his mercury-lifting mechanism to the height of the tallest trees, where water must be lifted above 90 meters-the equivalent of 10 atmospheres or 7.6 meters of mercury. The credit for proposing that so-called cohesion theory generally goes to Henry Dixon of Trinity College, Dublin, who championed it for 40 years. In 1893, he and J. Joly were in the laboratory in Bonn where the German botanist Eduard Adolf Strasburger showed that leaves take several weeks to wilt and die after tall trees and vines have been treated with steam and poisons that kill the living cells of the xylem. That persuaded Dixon and Joly that the living cells had no role in transpiration. In November 1894, they read a paper to the Royal Society of London that suggested that the lifting force of Bohm’s shoot could be mimicked by a water-filled porous pot with very small pores. The narrow, force-sustaining spaces of the wick need be only at the top, not all the way up. Dixon and Joly described experiments that they believed demonstrated that evaporating leaves acted in an analogous way and that the cohesive strength of water could withstand the negative pressures required.

Living cells were not necessary to Dixon and Joly’s mechanism, and those cells provided no pumping or other aid to the flow of water. The negative pressure transmitted through the threads of water to the roots would be sufficient to draw water from even quite dry soil. Dixon estimated that the negative pressures amounted to 1 atmosphere per 30 feet of height to overcome gravity and another 1 atmosphere per 30 feet of tree to overcome friction in the pipes. For the tallest trees, that constituted a pressure of -20 atmospheres. Even for shrubs and crop plants, the postulated pressures each day would be in the range from -5 to -15 atmospheres. Their much steeper pressure gradient is usually ascribed to higher rates of transpiration and the extra force required to extract water from drier soil.

According to Dixon and Joly’s model, the force that pulls the water through the wick originates at the top, where the water evaporates to form water vapor. The negative pressure that can be sustained in the water phase depends on the size of the holes from which the water evaporates. The smaller the holes, the greater the negative pressure in the water phase. A 3-micrometer hole sustains 1 atmosphere of negative pressure. So to sustain -20 atmospheres requires holes that are 150 nanometers in diameter.

The evaporating surface is in the leaf, where living cells are exposed to air. Water evaporates from the wet walls of the cells, which are analogous to the outside of the porous pot, and the menisci in the pores between the cellulose microfibers of the cell walls sustain the very large negative pressures. The pores are very small, possibly 5 nanometers, and could sustain very large negative pressures. From the water vapor-rich air in a leaf to the drier air outside it, the final step of transport is diffusive. Water vapor diffuses down a gradient of concentration to the outside air through pores called stomata, which open and close in response to several kinds of environmental changes. Note that the final outward flux of water is by gaseous diffusion, and it is uninfluenced by whatever the pressures may be in the liquid spaces of the plant.

For 70 years after Dixon and Joly’s endorsement of the cohesion theory, no decisive evidence supported or refuted it, but vigorous discussion continued on both sides while many investigators attempted to measure the pressures in functioning xylem and the cohesive strength of water. No evidence of large negative pressures was found, but that failure was explained by the practical difficulty of attaching a manometer to the threads of water. Many experiments suggested that water could be very strong. Of the many attempts to show that, the work of Lyman Briggs of the National Bureau of Standards is the most quoted. In a partial vacuum, he centrifuged water, which had been degassed and purified, and imposed stresses of around 200 atmospheres before the water broke.

A decisive contribution came in 1964 from Per Scholander of the Scripps Institution of Oceanography. After 10 years of using manometers to measure pressures in xylem and finding only values between 0 and +1 atmospheres, Scholander thought of balancing the internal negative pressure by externally applied gas pressure. When he cut a leaf from a transpiring plant and enclosed it in a pressure chamber with the cut end outside, only a large pressureso-called balance pressure-could push water to the cut end. The size of the balance pressure was roughly proportional to the rate of transpiration before cutting. Mangroves had large balance pressures, which is not surprising because it takes a pressure of 25 atmospheres just to separate water from seawater. Scholander decided that the balance pressure was, in fact, a measurement of the negative pressure in the water threads in an intact plant, framed an argument to justify that and persuaded many colleagues that the cohesion theory had at last been proved correct. Although the precise logic connecting the balance pressure with the negative pressures in an intact plant is still not universally agreed on, the reliability of the theory and the reality of the large negative pressures became-and remain-the orthodox view of water transport in plants.

For the cohesion theory to work, the cohesive strength of water must routinely resist forces equivalent to those produced by superheating it to 180 degrees Celsius. In more droughted plants, that figure rises to around 265 degrees, and to 312 degrees for mangroves. Water in such a state must break (vaporize or cavitate) sometimes, which would leave a section of the pipes filled with gas (water vapor or air). Such an event is called cavitation, and it results in an embolism, which prevents water flow through a section.

The cohesion theory requires that embolisms form rarely and that they are isolated from spreading to other sections, because the large negative pressures in the remaining water threads would make it impossible to refill the embolized space. Negative pressures in the pipes must be in equilibrium with the pressures in the tissues around the pipes, or the tissues would be sucked dry. So the whole mass of water in the leaves, stems and roots would also be at negative pressure. Once an embolism forms, the pressure in it is positive, and water can be pushed in only by a higher positive pressure.

Embolisms are known to develop in trans irinz plants. Even Hales observed bubbles coming from the cut ends of branches, and some of them probably came from gas in embolized pipes. Indirect evidence for embolisms comes from two features of waterstressed plants: reduction of hydraulic conductance and minute “clicks” of sound or ultrasound, which have been interpreted as the breakage of individual threads. Since the cohesion theory does not permit repair of embolisms during transpiration, they are often considered as permanent and accumulating from day to day. Or, if they are repaired, it could only be at night or during wet weather, and the repairs might not be total. Because the cohesion mechanism postulates such precarious conditions in the pipes, that mechanism of water transport may be characterized as metastable.

Discordant Facts

Few serious contradictions to the cohesion theory were found by the many investigators who devised experiments and interpreted their results in accord with its logic until 1990. In that year, U1rich Zimmermann at the University of Wirzburg published the first of a series of papers on direct measurements, using a microcapillary manometer, of the pressures inside single, active xylem pipes. When inserted in the pipes of transpiring plants, including trees, the probe measured pressures in the range from 2 to -2 atmospheres. In droughted plants, pressures could fall to -7 atmospheres. Two other discordant facts were found. First, cavitations of water to water vapor were observed at quite small negative pressures. The threads of water became increasingly unstable at pressures below about -2 atmospheres. Second, the measured pressures were much smaller than the balance pressures indicated at the same time by a pressure chamber. Those findings called in to question the cohesion theory’s two basic tenets-that the water threads are stable at large negative pressures and that the pressure chamber measures the pressure inside xylem. Soon after that, Andrew Smith (then investigating the action of octopus suckers at the University of North Carolina at Chapel Hill) repeated Briggs’s centrifuge measurements of the cohesive strength of water, but at atmospheric pressure. In contrast to Briggs, Smith found very large variances in the cavitation thresholds and a mean of only -7 atmospheres.

The reaction of plant biologists to Zimmermann’s new data has been almost universally negative. There has not been a lively interest in testing them, exploring their implications or modifying the existing theory to accommodate them, but rather a search for reasons why they might be disregarded.

Compensating-Pressure Theory

I set out to explore where Zimmermann’s data might lead. Suppose the wick does not work unaided, but in concert with the surrounding living cells. Plant cells are small osmometers distended against their tough cell walls by the turgor pressure of their vacuoles. When they are confined and packed together in a stem or root, they press against one another, and part of the turgor pressure produces an internal tissue pressure. You can see the effects of the tissue pressure as seedlings push up out of the ground, roots break concrete pavement or lettuce is resuscitated. All previous hypotheses about plant-water transport ignored that pressure. Nevertheless, the threads of water link together the evaporative pull from the leaves and the positive tissue pressure, which acts to raise the pressure inside the xylem. I call the amount of tissue pressure applied to the xylem the compensating pressure, and it causes the cohesion mechanism to work at less negative pressure. So, scientists should accept that water in plants is weak under tension and does break frequently at modest operating pressures to form embolisms. The tissue pressure refills the embolisms by squeezing water out of cells and into the pipes by reverse osmosis. Then, when an embolism forms, it is repaired quickly, the gas dissolves into the incoming water and flow is restored in a short time during active transpiration.

The requirements of the compensating-pressure theory are more complex than those of the cohesion theory. Besides the participation of living cells, the mechanism requires reservoirs of static water outside the pipes and oneway valves at the top and bottom of the pipes. Those valves are needed to prevent the dissipation of the applied tissue pressure, so that it is used in decreasing the tension on the water, not in accelerating flow through the pipes. The valves must have the rather special property that flow through them is independent of the pressure in the pipe. The valve at the top is, indeed, of that type. As explained above, the final step in water transport is by gas diffusion, which is not sensitive to the hydraulic pressure in the water. The valve at the bottom, in the roots, must be more than a valve; it must be a pump. Without negative pressure in the pipes to draw water from dry soil, work must be done to transfer water from the soil to the pipes, where the pressure is higher. The living structures of the root must form a water pump. We know that such a mechanism exists, but not how it works. It is called root pressure, and it is responsible for the flows of watery sap from cut stems of grapevines pruned in spring and for drops of water that exude from the edges of leaves of many plants on hot summer nights. These exudations are evidence of positive pressure in the pipes when transpiration is zero. From the time that the cohesion theory became widely accepted, no pump was considered necessary and the phenomenon of root pressure has been little studied.

The two theories are mirror images of one another. In the cohesion theory, the pipes are pulled inward by tension in them and must be reinforced to resist that deformation. In the compensating-pressure theory, the pipes are pushed in by outside pressure, which the reinforcement must resist. Consequently, the expected outcome of almost any experiment will be the same under either mechanism, and both theories provide satisfactory explanations or predictions for most of the observed phenomena of plant-water relations. The crucial experiments to decide between the two theories are, in fact, very few. One is the direct measurement of the negative pressure in the pipes, or Zimmermann’s experiment. According to the old theory, the internal pressure should be a large negative value; under the new theory, that pressure would be small, as Zimmermann found. Because the compensating-pressure theory explains all the phenomena and observations covered by the old theory, nothing is lost by passing through the mirror to adopt a new viewpoint, and much is gained.

View from the Other Side

The compensating-pressure theory makes sense of Zimmermann’s findings. First, the pressure inside the xylem would be about zero atmospheres under the new theory, just as Zimmermann measured. Second, according to the new theory, the pressure measured inside the xylem should not equal the pressure-chamber reading. Under the new theory, the pressure chamber measures the compensating pressure, or reduction in tension, in an intact transpiring plant. Likewise, the high pressure-chamber readings for stressed plants are also easily understood, because they represent part of the osmotic pressures of the surrounding living cells, not the tension of the stretched water.

If the water-transport system has a working pressure near zero, one sees that other facts agree with that. For example, the xylem lies next to the sugarconducting and living phloem in all living and fossil land plants since the Devonian, when plants first migrated out of the sea. There has never been an explanation of why that should be so. As the sugar-transporting tissue, the phloem contains the cells that have the highest osmotic pressure of any in a plant’s body. If one is looking for a source of pressure to assist the living xylem cells in compensating for the pull on the threads of water, the nature and position of the phloem make it a conspicuous candidate. Its association with the xylem is essential to xylem function. On the other side of the phloem, remote from the xylem, there is commonly a strong mechanical tissue that provides a buttress to push against.

Another indicator of low pressure inside the xylem comes from several groups of insects that feed on the sap in those pipes. (This does not include aphids, which feed on the sugar-rich sap of the phloem that is at high positive pressure.) Spittlebugs are the most noticeable. The foamy masses found on the leaves and branches of many weeds and grasses in summer fields harbor the larvae of spittlebugs. They feed on large volumes of the dilute xylem sap to collect its amino acids. Spittlebugs have a pump powerful enough to extract sap against a pressure of about -3 atmospheres, but certainly not against the pressures required by the cohesion theory. Peter Anderson and his colleagues at Florida State University found that the rate of sap excretion by leafhoppers was highest in the afternoon when the balance pressure was highest, and increased as the balance pressure increased up to 18 atmospheres. A high balance pressure keeps the pressure in the sap near zero, and the insects can easily pump it through themselves. They feed fastest in the afternoon when the amino acid content of the sap is richest.

Observations of embolisms also provide decisive evidence in favor of the compensating-pressure theory. In our laboratory, my colleagues and I have developed a method of directly observing embolisms in the xylem and measuring the proportion of pipes that are embolized at any time. Functioning parts of plants are snap-frozen (to liquid-nitrogen temperatures) and observed-still frozen-under a scanning electron microscope. The living cells are easily distinguished by white lines from the solutes sequestered from the vacuolar solutions during freezing. The dead vessels with solute-poor sap contain either black ice, which shows that they were not embolized when frozen, or no ice if they were embolized. It is easy to measure the percentage of embolized vessels at different times during transpiration. Using that technique, we are finding universally that embolisms develop soon after sunrise. Far from being the permanent blockages supposed by the cohesion theory, however, the embolisms are repaired in a few minutes. The percentage of embolized vessels falls during the day, and can be as low as four percent by mid-afternoon. In the embolized vessels, it is possible to see the repair process at work-that is, water entering the vessels through pits from adjacent cells. During successive stages of refilling, the incoming water forms a crescent shape along one side of a pipe and then an annular mass of water can be seen. Such patterns would not be possible under the cohesion theory, because the water would be sucked out into the cells by the presumed large negative pressures.

My colleagues and I have also used that freezing technique in combination with pressure-chamber measurements. That work revealed that the direct measurement of embolisms is inversely related to pressure-chamber readings. When the balance pressure was low, embolisms were increasing (water threads breaking under negative pressure); when the balance pressure was high, embolisms were decreasing (water threads repairing under positive pressure). Here then is direct proof that the pressure chamber does not measure negative pressure in the xylem. What the balance pressure does measure is the vigor of the refilling process, which is just the compensating pressure according to the new theory. Those experiments also showed that a plant varies its compensating pressure-starting low in the morning and increasing to its maximum by early afternoon. A mechanism of adjusting the compensating pressure has been found in the interconversion of starch and sugar, which leads to large increases in osmotic pressure. Some living cells in and near the xylem-potential pressurizing cells-contain large quantities of granular starch, which disappears in the afternoon of hot days and returns at night.

The source of the water pushed into the embolized pipes to refill them must be in static reservoirs outside the pipes. In a leaf stalk, it is probably the ground tissue of water-rich cells, called parenchyma, which makes up the bulk of the stalk (the nonstringy parts of a celery stalk). The pressurizing cells would force water from the ground tissue by reverse osmosis.

What about tall trees? The postulated water pump in the roots allows another reassessment of the water-transport process. Under the compensating-pressure theory, the water does not have to be lifted up the height of the tree. It grew there. The tree is a tower of static water-contained in living cells-surrounding the small dead pipes that run from the leaves down to the roots. The pressure at the bottom of a water tower 90 meters high is about 11 atmospheres, and that is the pressure step that the root pump must overcome to put water in the bottom of the pipes. Once it is there, no further lifting is needed. However, this view of trees has not been tested.

Ultrastable Water Transport

The water-management system that I have described is much more complex and interesting than the wick analogue from which we started. It is robust self-sustaining, self-repairing and adjustable to the evaporative demands of changeable weather, the vagaries of supply from wet or dry soil, damage by grazing or accident and the fluctuations of salinity around the roots of mangroves. It is independent of the height of trees. It calls on the resources of the living cells, which are seen to be precisely placed to play their parts. It found ways to mitigate the extreme disturbances of Strasburger’s experiments for many days before being finally destroyed. It is a system of the class called “ultrastable” by Ross Ashby, an English experimental psychologist. That is, it has the capacity to regain stability after disturbances that are outside the range of its experience. The not-so-thirsty cat is an ultrastable system. Transported to a wilderness of city or countryside, it will adapt and survive. In answer to my original question, it is satisfying to reach the conclusion that the major transport system of the thirsty tomato plant may have the same property.