Annals of Botany 78 : 399–407, 1996
BOTANICAL BRIEFING
Sap Ascent in Vascular Plants : Challengers to the Cohesion Theory Ignore the
Significance of Immature Xylem and the Recycling of Mu$ nch Water
J O H N A. M I L B U R N
Department of Botany, UniŠersity of New England, Armidale, N.S.W. 2351, Australia
Received : 31 January 1996
Accepted : 16 May 1996
In recent years the cohesion theory has been attacked on the grounds that direct measurements made with the pressure
probe indicate that sap tensions are much less (maximum tension approx. 0±7 MPa) than indicated by parallel
measurements made with the more conventional methods : osmotic methods, pressure bomb, or psychrometer. It has
also been claimed that other direct methods do not support the cohesion theory. Thus a re-examination using the
Renner technique indicated sap tensions of approx. 2±5 MPa. Also an independent method based on mercury
penetrometry provides evidence that sap tensions of at least 2±0 MPa can be demonstrated directly implying, that
serious limitations arise from the pressure probe method itself. Without tensions exceeding 2±0 MPa mangroves
would be unable to extract fresh water for transpiration from seawater. It is suggested that the pressure probe is
susceptible to bias because it investigates the least mature xylem conduits while they are still under varying degrees
of turgor pressure and only partially interconnected with the main xylem system. This supposition is supported by
claims that the xylem sap sampled by the probe contains significant concentrations of solutes. Additionally water,
supplied by reverse osmosis from the sieve tubes (‘ Mu$ nch water ’), is continually being liberated in the vicinity of the
outermost xylem vessels hydrating them to an atypical degree which can explain several of the discrepancies claimed.
These results, which are supported by the work of others, demonstrate that the challenges to the cohesion theory for
the ascent of sap are ill-founded. The release of water from the phloem can explain not only some discrepancies
claimed by the cohesion challengers, but also explain the refilling of cavitated xylem conduits : a hitherto unsuspected
role for the phloem transport system.
#1996 Annals of Botany Company
Key words : Cohesion theory, sap ascent, cavitation, pressure probe, xylem transport, vessel development, recycled
water, reverse osmosis.
INTRODUCTION
The mechanism of sap ascent via the xylem has been a
controversial subject for over 150 years. Anatomically, the
system is complex and the physico-chemical mechanisms at
work are generally unfamiliar because they cannot be
utilized very readily in our macroscopic world ; nevertheless
the mechanism, in all its complexity, functions remarkably
well for most of the time with an impressive, silent,
efficiency.
Recent measurements of the negative pressures in xylem
elements using the pressure probe have failed to show the
tensions (pressures more negative than 0±7 MPa) required to
draw water from roots to shoots at the fluxes commonly
observed in transpiring plants (reviewed by Zimmermann,
Meinzer and Bentrup, 1995, in an earlier Botanical Briefing).
As interpreted, these measurements necessitate the total
rejection of the orthodox theory of long distance transport,
as proposed by Dixon (1914). They also throw doubt upon
the validity of the wealth of information gleaned over many
decades from osmotic techniques (e.g. many careful
measurements by Dixon, 1914 ; Ursprung and Blum, 1919,
etc.) and furthermore the legitimacy of modern conventional
methods to measure water pressures and potentials, e.g.
pressure chambers and psychrometers. The theory requires
the presence of considerable negative pressures to drive
0305-7364}96}100399­09 $18.00}0
water flow through xylem by exploiting the strong cohesion
between water molecules caused by hydrogen bonding.
Water in plant tissues (fern sporangia) has been demonstrated experimentally to withstand tensions up to 45 MPa,
and some vascular xylem has been claimed to continue
function at tensions around 10 MPa. So the pressure probe
results, indicating a maximum tension of 0±7 MPa, are
clearly at variance with these findings. Several authors have
already reacted to this challenge to orthodoxy and given
their reasons for continuing to hold the conventional view
(Passioura, 1991 ; Sperry and Saliendra, 1994 ; Holbrook,
Burns and Field, 1995 ; Pockman, Sperry and O’Leary,
1995). In this article I give my own reasons, based on
experimental results, for continuing to support the conventional cohesion theory. These include a new interpretation which explains why the very negative pressures
expected were not detected by the pressure probe. A
mercury injection method has been used to demonstrate
directly that negative pressures of at least 2 MPa, demanded
by orthodox cohesion theory and considerably greater than
tensions measured by the pressure probe, are indeed
demonstrable in transpiring plant tissue.
In addition to limitations causing breakdown of the
pressure probe method, I also advance two explanations for
the discrepancies not previously discussed by others. I agree
with the challengers that a series of ‘ incompatible obser# 1996 Annals of Botany Company
400
Milburn—Sap Ascent in Vascular Plants
vations ’ (Zimmermann et al., 1995) has indeed been made :
in my view these discrepancies are not entirely attributable
to limitations of the pressure probe method per se. I argue
that because the pressure probe examines predominantly
the outermost xylem conduits the normal sample is biased,
because it includes younger conduits which are atypical on
account of their immaturity. The water balance of the
outermost conduits will also be elevated above the main
body of the xylem by recycled water contributed from the
phloem (‘ Mu$ nch ’ water). The outcome is a better appreciation of how a knowledge of plant physiology,
anatomy, ontogeny and the osmotic properties of cells
actually consolidates the cohesion theory, and indeed
provides a method to restore cavitated conduits. The need
to resort to unsubstantiated forces, such as Marangoni
bubbles (Zimmermann et al., 1995), with a severe energysupply problem, as the motive force for transpiration, is
removed. Thus to a considerable extent the various
viewpoints can be reconciled through a better understanding
of plant physiology, anatomy and development.
THE NATURE OF THE CHALLENGES
By far the most serious challenge has been made by
Zimmermann et al., 1995 (see the numerous references,
dating especially from Balling et al., 1988 ; including Steudle,
1993). Basically, they claim that because their attempts to
measure large negative pressures in the xylem conduits,
using the pressure probe, were generally less negative than
those obtained by conventional methods (osmotica, pressure
bomb, psychrometer, etc.), these latter methods must be
incorrect. If this view is correct then conventional
mechanisms explaining sap ascent must also be wrong
because the tensions necessary to move water at the required
rate are inadequate. Furthermore, by implication, what has
been learned by acoustic analysis about the induction and
reversal of cavitation in xylem of vascular plants (dating
from Milburn, 1966, 1973 a b ; Milburn and Johnson, 1966 ;
Milburn and McLaughlin, 1974, through to Tyree and
Sperry, 1989 and Milburn 1991, 1993) must also be incorrect. The pressure probe measurements of Balling
and Zimmermann (Balling et al., 1988 ; Balling and
Zimmermann, 1990) suggested that xylem was unable to
support sap tensions of more than a few tenths of a MPa (a
few bar) below atmospheric pressure : but as their designs
and technique improved it is now admitted that the xylem is
under more extreme tensions, i.e. pressures approx.
®0±7 MPa (Zimmermann et al., 1995). But these are still
significantly less extreme than required by the cohesion
theory from results of the conventional methods listed
above. Interestingly, even Scholander et al. (1962) made
similar errors in early attempts to measure xylem sap
tensions in mangroves. Later Scholander et al. (1965, 1966,
and Scholander, 1968), after developing the pressure
chamber technique, were able to measure much more
negative values (tensions ranged to at least 6±0 MPa), and
these they confirmed using the Renner technique (see
below).
Apparent support for Zimmermann’s challenge
(Zimmermann et al., 1995) has been provided by Canny
(1995), however Canny’s proposed alternative mechanism
differs very significantly from that of Zimmermann et al.
(1995). He suggests that an externally generated pressure
squeezes the conduits raising sap pressures so aiding sap
ascent. Whether this is steady or intermittently applied is unclear, but it resembles a mechanism suggested long ago by
Bose (1923) and long since dismissed for many excellent
reasons. It is claimed to reduce the necessity for considerable xylem sap tensions. How such an externally applied
pressure, generated by the rays cells, could do this in a
sustained manner during transpiration still needs to be
explained. Also it is far from clear if it is either energetically
feasible or hydrostatically justifiable. In my view, to be tested
by others, Canny’s model must be presented more cogently
and supplemented by working models. He appears to suggest
that xylem sap should spurt from a wound made in mature
xylem of a tree when it is not transpiring. In fact positive
xylem sap pressures are difficult to demonstrate in large
leafy plants or trees at any time. Furthermore, xylem
conduits are strongly lignified and therefore resistant to
externally applied pressurization.
Another source of support for the challenge to conventional theory has been derived from experiments in
which tensions are generated centrifugally in liquids spinning
in a Z shaped tube. These liquids cavitated rather easily at
slight tensions, cited in an incomplete review of cohesional
tensions by Smith (1994). Despite many repetitions, tensile
measurements were considerably less than those of Briggs
(1950) who developed this method. Smith’s results suggest
that his preparatory techniques for removal of nucleation
sites, on which cavitation bubbles can form, have been
seriously defective. If these nuclei are not removed adequately before tests, mere repetition is quite pointless.
Smith did not consider the quite unambiguous experiments
of Renner, Ursprung and others working on fern sporangia
or, e.g. Milburn (1970), where sap in fungal spores, placed
under tension using osmotica, showed clearly that very
great tensions can indeed be generated and sustained in
plant cells, some far outstripping Smith’s centrifugal
estimates based on physical systems.
Such considerations have led others (e.g. Passioura, 1991 ;
Sperry and Saliendra, 1994), to attribute the results of
challengers to the cohesion theory to mere errors in
measuring techniques. Apparently some determinations
have been made with insufficient rigour. For example
Balling and Zimmermann (Balling et al., 1988 ; Balling and
Zimmermann, 1990) apparently did not appreciate the
importance of allowing their leaves to equilibrate before
conducting pressure bomb determinations. My own interpretation is that there may indeed be some validity in
their measurements ; however these have been seriously
misinterpreted, essentially because the wrong tissue was
presented to the pressure probe.
THE PRESSURE PROBE: COMPARISON
WITH OTHER TECHNIQUES
The pressure probe measures the pressure in cells individually, or when fused, as in xylem vessels. When a
Milburn—Sap Ascent in Vascular Plants
needle is driven into a cell, small fissures form around the tip
penetrating the ruptured wall, rendering it prone to leakage :
in positively pressurized living cells cytoplasmic blockage
from within reduces this danger. However when the needle
is driven into a lignified, non-living cell under negative
pressure it is very likely to first permit entry of externally
released sap and then air through the fissures, causing
disruptive cavitation. Much can be learned by simply
listening, without electronic amplification, to strong steel
needles being driven into (no detectable sound), then
withdrawn from the large mature outermost vessels of oak
(Quercus rubra). As the needle is withdrawn an audible hiss
is followed by a cadenza of ticking sounds. These are caused
by sap, released from ruptured cells (including sieve tubes)
surrounding the needle, causing inflowing air to form a
noisy series of bubbles. Thus the pressure probe can be
expected to apparently work best on these most accessible
exterior cells of the vascular cylinder. The requirement for
‘ good readings ’ produces, therefore, a very biased and
atypical sample of xylem conduit pressures. Furthermore,
living (i.e. immature) cells will seal better, because they still
contain cytoplasmic debris, are incompletely lignified, and
under either positive pressures, or pressures considerably
less negative than the main functional population of mature
lignified vessels. The fact that the greatest tension measured
successfully by the pressure probe and widely cited (about
0±7 MPa negative, relative to atmospheric pressure) was
sustained for only a few minutes, can be attributed to the
time required for a slow influx of external sap (released from
living cells damaged by the tip of the needle) via the tiny
fissures surrounding the needle tip into the selected conduit.
However when this external liquid had been consumed, air
followed, precipitating cavitation and disrupting tension.
COHESION HYPOTHESIS: ARGUMENTS
FOR OPPOSING
Zimmermann et al. (1995) claim that the hydrostatic
gradients in trees have frequently been observed to be
reversed. Reasons for this apparent anomaly are provided
below (see Solutes in xylem conduits and ‘ Mu$ nch water ’
explanation).
At one time it seemed that vascular plants were far more
capable than physical systems of generating and sustaining
water-columns under tension without cavitation ; more
recently this view has been progressively modified. The
Berthelot technique (in which liquid is thrown under tension
by sealing it within a thick-walled glass tube and then
cooling it) has now been greatly extended (tensions now
generated exceed 20 MPa, Henderson and Speedy, 1987).
Even that technique has been exceeded by claims of very
severe tensions causing cavitation in water trapped within
geological specimens (140 MPa, Zheng et al., 1991).
The greatest sap tensions found in vascular plants seem to
develop within the specially thickened annular cells of fern
sporangia (approx. 44±6 MPa, reviewed by Zimmermann,
1983, p. 46). In plant cells capable of generating such
negative pressures, the walls must be so finely porous as to
exclude the passage of both air interfaces and solute
401
molecules, but nevertheless allow the passage of water
molecules. If the solute molecules are large they are generally
too insoluble to generate appropriately extreme osmotic
pressures required for measurements. If the solute molecules
are very small they penetrate the walls and any osmotic
extraction of water, which causes the tensions, ceases
rapidly. The thickened walls of fungal spore cells of Sordaria
excluded sodium chloride (Milburn, 1970). Ritman and
Milburn (1990) measured tensions of 28±0 MPa in Cyathia
australis using glycerol solutions. Most work on fern
sporangia has utilized very concentrated sucrose solutions
because smaller particles (e.g. sodium chloride ions)
penetrate the walls. Unquestionably, experiments in which
plant cells or tissues (e.g. fungal spores or sporangial cells)
are submerged in osmotica, until an entrained bubble forms
spontaneously, demonstrate that caŠitation may occur
endogenously and is not always initiated by sucking gas from
the exterior. Unfortunately the porosity of xylem conduit
walls is too great to allow these simple and direct osmotic
methods to be applied to measure their tendency to cavitate
at known tensions by microscopic observation.
There is now little doubt however that in higher plants
cavitation is usually induced because gas (air) is drawn
through the walls (e.g Crombie, Hipkins and Milburn,
1983 ; Sperry and Saliendra, 1994). In some systems
especially, e.g. xylem conduit transport in banana, where
cavitation can be induced at quite low tensions (0±4 MPa,
reviewed Milburn, 1991) gas must penetrate rather easily,
probably because the conduit wall microfibrils are less
tightly woven in comparison with conduits of ‘ tougher ’
plants. However we must be alert to the existence of other
relatively impermeable barriers in higher plants, e.g.
endodermis, or vascular sheaths in leaves (Canny, 1990),
where dyes (molecular mass 500³200) and other solutes
may be arrested. This certainly does not indicate impermeability to water (molecular mass approx. 18) but these
barriers must restrict hydraulic transport and at the same
time restrict the entry of a gas interface from the external
atmosphere, hence reducing the susceptibility of any
enclosed xylem to cavitation.
RECENT EXPERIMENTS CONFIRMING
THE COHESION HYPOTHESIS
A number of interesting experiments have been reported
recently supporting the cohesion mechanism and rejecting
the challenge, these are merely outlined below. I will also
add some summarized, unpublished, results of my own.
Air pressure to disrupt xylem conduction
Cruiziat (1995) reported that pressurized air applied
around branches using a pressure sleeve (devised originally
by Salleo et al., 1992) stopped sap conduction through gas
invasion of the conduits. The pressures needed to achieve
this (0±7 to 2±5 MPa) were much greater than predicted by
the challengers of cohesion and are entirely in accordance
with the cohesion hypothesis.
402
Milburn—Sap Ascent in Vascular Plants
Vapour equilibration of water in conduits. Richter and
Kikuta (1995) reported that cavitation, detected ultrasonically in specimens of Juniperus wood, occurred at
tensions, established in water vapour controlled by known
osmotica, approached approx. 8 MPa. This value greatly
exceeds the small tensions claimed to be operative in
conducting xylem elements from pressure probe measurements.
Sap
Water
Water
Mercury
Tensions sustained under extreme centrifugation
Holbrook et al. (1995) and Pockman et al. (1995) reported
that woody twigs dehydrated by centrifugation (akin to the
spinning Z tube method outlined above) had the same water
potential values as those predicted by calculating the applied
centrifuge tensions, as measured with a pressure bomb.
Thus, by independent confirmation, the pressure bomb
gives valid results. Also the tensions sustained by the xylem
sap were about 1±5 MPa and 2±0 MPa, respectively ; considerably more severe than the limits proposed by
Zimmermann et al. (1995) on the basis of pressure probe
determinations.
Renner experiments : maximum tensions determined
It has been suggested by Zimmermann et al. (1994 b) that
Renner’s experiments (1915 etc.) indicated that only small
tensions were operative in xylem transport (Renner’s
greatest tensions were in fact 1±0–2±0 MPa below atmospheric pressure). The method is based on measuring the rate
of uptake of water through an applied constriction in a leafy
shoot, with the rate at which water can be driven through
the same constriction using vacuum. The ratio of the two
rates of uptake can be used to deduce that an extremely
negative pressure is generated by the leaves and its
magnitude can be estimated.
I have recently performed a number of Renner-type
experiments. The technique serves to demonstrate cohesion,
but it lacks precision, because tissues take 1–4 h to reach the
desired equilibration. Thus in one recent experiment a gas
pressure 0±072 MPa below atmospheric drew water past a
xylem constriction in a privet (Ligustrum Šulgare) shoot 37
times more slowly than the leafy shoot itself, indicating that
the operative xylem sap tension was 2±66 MPa. An equilibrated pressure bomb evaluation of the xylem tension in the
shoot gave a value of 2±85 MPa. Both these values greatly
exceed cohesion challengers ’ limitations but accord well
with results from research on mangroves and other plants
by Scholander et al. (Scholander, 1968) in which the Renner
method was also utilized. The indicative value of the Renner
technique was only approximately correct because minor
errors in the measurement of rates under ‘ vacuum ’ are
multiplied many times. Pressure bomb values are much
more dependable.
Recent mercury injection experiments
In 1965, a xylem sap measuring technique was demonstrated to me personally by Dr Franz Floto in Copenhagen,
Scale
Xylem
Conduit
F. 1. The principle of the ‘ Floto gauge ’ for measuring tensions using
the traverse of a bead of mercury, drawn by sap tension into tapered
glass tubing of decreasing radius, to estimate the magnitude of sap
tensions in the xylem.
Denmark. It consisted of a tapered capillary tube containing
water and a small bead of mercury (Fig. 1). Mercury resists
being drawn into narrow tubes, so the extent of its travel can
be used to gauge negative pressures directly. With this
device Floto demonstrated xylem sap tensions of 0±6 MPa in
stems of potted Phaseolus seedlings. He never published his
technique, believing it to have been eclipsed by the recently
developed pressure bomb.
Fortunately there is a simpler and more direct way to use
mercury to demonstrate considerable xylem sap tensions
within the conduits than that used by Floto. If mercury is
drawn by sap tensions into narrowing tapered xylem
conduits it is possible to obserŠe microscopically the extent
of this penetration. By measuring the diameters of these
conduits, it is possible to estimate directly the tensions
generated utilizing the physical laws of surface tension. To
support the orthodox cohesion theory, all that needs to be
demonstrated is that significantly greater tensions are
present in the xylem than the limits claimed to exist by the
challengers of the cohesion theory (i.e. tensions greater
than 0±7 MPa). Mercury injections have demonstrated (Fig.
2 A–D) how xylem sap tensions can draw mercury into pits
and tapering conduits so fine that accurate measurements
extend beyond the limits of resolution of the normal optical
microscope. Thus mercury was found to penetrate conduits
of 1 µm diameter or less, corresponding with tensions of
about 2±0 MPa. Hence the power of cohesion has been
demonstrated within vascular tissues by a direct method.
Arguments that the cohesion theory cannot operate beyond
tensions of 0±7 MPa are therfore untenable.
SOLUTES IN XYLEM CONDUITS: IMPACT
OF IMMATURE XYLEM
During their anatomical development, the accepted view is
that the xylem conduit cells (i.e. vessel elements) fill with
Milburn—Sap Ascent in Vascular Plants
403
F. 2. A, Mercury injection into the xylem conduits of Pinus sylŠestris. The pool of mercury is seen below where it enters wounded conduits in
which it can be seen in silhouette. Low power (¬4) objective. B, Detail of Fig. 2 : mercury injection into the xylem conduits of P. sylŠestris. Dry
mount : mercury is being drawn by xylem sap tension into tapering conduit terminations : some taper to a point. (Objective ¬20, Bar ¯ 20 µm).
C, Mercury injection into the xylem conduits of P. sylŠestris seen laterally. Pits were injected via the pit pores before mercury retracted. (Objective
¬20). D, Detail of xylem tracheids showing pits and pit apertures in surface view. Scale and experimental materials as in (B) and (C). E, Exudation
from the xylem of the manna ash Fraxinus ornus in May 1979, Messina, Sicily. The yellowish dried exudate is clear proof of the high solute
concentration of the immature xylem sap. It closely resembles the more familiar ‘ manna ’ which exudes in September from the severed sieve tubes
in the bark, often in ‘ commercial ’ quantities. Scale as in (F). F, Sap freshly exuded from immature xylem in May 1979 from a grid of pinpricks
in the shaved bark of F. ornus. Its refractive index indicates a concentration about 1}3 of phloem sap. White square is 1 cm#. [(E) and (F) from
Milburn, Lo Gullo and Tyree, unpubl. res.)]. G, Phloem exudation from sieve tubes of F. ornus. during annual commercial tapping in September,
at Castelbuono, Sicily. Note the similarity of the dried exudate with that from the xylem [(E) and (F)] on account of the high mannitol content.
404
Milburn—Sap Ascent in Vascular Plants
solutes, inducing considerable turgor pressures within the
still unthickened immature cells. This turgor is harnessed in
the massive expansion required to enlarge the diameter of
vessels while also overcoming opposing turgor pressures
from surrounding parenchymatous cells. Only later are they
integrated into the mature negative-pressure system. During
this time of growth and maturation the xylem cells are living
and turgid with a functional plasmalemma within thin
unlignified walls. If such xylem initials are wounded they
may even exude solute-rich sap like sieve tubes (see review
by Milburn and Kallarackal, 1991), but this seems to depend
on wall elasticity (Milburn, 1972) and also the extent of
lignification, which proceeds basipetally. In the case of
Fraxinus ornus, the manna ash, this xylem exudation was
detected easily, not only by removing the bark, showing that
the exudation must arise from the xylem, not the phloem,
but also because it occurs much earlier in the season
(May–June in Sicily) than phloem exudation (beginning
August–September). Evaporation produced a manna-like
deposit (see Fig. 2 E and F). True manna is the dried exudate
derived from the sieve tubes (Fig. 2 G). Both are rich in mannitol. Apparently, during their positive hydrostatic
pressure phase, xylem initials are filled with sap very similar
in composition to sieve tube sap but more aqueous. Later
they become, through digestion of membranes and materials
within the perforation plates, linked and integrated with
other mature conduits sharing the burden of the negativepressure irrigation system. I suggest that experiments using
a pressure probe are very prone to bias when encountering
such immature xylem conduits, because they seal well and
Vacuolation
+ ve
Time
P
– ve
Integration
F. 3. Diagram illustrating the hypothesized time-course of progressive
pressurization as xylem initials become mature conduits. The solute
composition would probably coincide with the positive pressure curve
only, because pressurization is osmotically driven : when the plasmalemma degrades, positive pressurization ceases and the solute
composition would become diluted by the transpiration stream until
almost pure water.
also because they form a readily accessible sheath around a
tree trunk. These facts both produce invalid samples of the
whole xylem conducting system. The extent of the initial
pressurization in young xylem vessels can be gauged from
the enormous increase in volume created through vacuolation. Standard plant anatomy texts generally omit these
issues, but Priestley and Scott (1955) give a clear account
(see also Milburn, 1979, p.83). Hence the presence of solutes
in xylem sap sampled by the pressure probe does not
indicate some mysterious new mechanism at work
(Zimmermann et al., 1994 a) transporting solute rich xylem
sap upwards within mature conduits. Instead it is probably
derived from immature developing xylem elements, confirming the explanations offered above. Pressure probe
measurements on these elements (see Fig. 3) could register
positive pressures initially and then negative pressures
(relative to atmospheric pressure) as the elements mature,
quite dissimilar from the main body of the conducting
xylem. Results might not indicate the true hydrostatic
gradients which could even seem to be ‘ inverted ’.
Sap transport in mangroŠes
Zimmermann et al. (1994 a, b) claimed that high solute
concentrations and minor xylem sap tensions occur in
mangroves, which is at variance with conclusions reached
by Scholander et al. (1962 and Scholander, 1968). To avoid
misleading results it is very important to ascertain the
osmotic pressure of water available to the roots from which
sap must be extracted. Recently (January 1996) I repeated
some seasonal measurements on the grey mangrove
(AŠicennia maritima) at Urunga, N.S.W., Australia. Pressure
bomb readings on shoots indicated, as usual, xylem sap
tensions between 2±8 and 3±2 MPa during sunlight (n ¯ 5).
The seawater had an osmotic potential of ®2±07 MPa.
Xylem sap, collected by the well-established method of
Bennet, Anderssen and Milad (1927), and widely used by
Bollard (e.g. 1953) and Pate et al. (e.g. 1990) had osmotic
potentials ranging from ®0±22 to ®0±40 MPa, or less (n ¯
7), at 20 °C. High molecular weight solutes, other than those
which could be explained by experimental contamination,
were undetectable in the sap by refractometry (similar
findings were reported to me by Dr W. Allaway and also Dr
Marilyn Ball, pers. comms, Australian Society of Plant
Physiologists Meeting, Sydney, Australia, September,
1995). I am very confident that Scholander’s analysis is
substantially correct. Mangroves must generate sap tensions
exceeding the osmotic pressure of seawater in order to
extract from seawater the almost-pure water needed for
transpiration. In my experiments this extra tension was
0±73 –1±3 MPa. These mangroves were functioning at xylem
sap tensions several times greater than the limits indicated
by pressure probe measurements. Solute levels in sap were
quite low and apparently not essential for sap ascent.
PRESSURE PROBE RESULTS:
A L T E R N A T I V E ‘ M U$ N C H W A T E R ’
EXPLANATION
When the water relations of xylem conduits are examined in
detail, especially in the outermost regions adjacent to the
Milburn—Sap Ascent in Vascular Plants
cambium, it is apparent that immaturity is only one of the
factors modifying the water relations of the vasculature.
Prior to 1930, Mu$ nch had recognized that wherever solutes
are being deposited from the arriving phloem sap stream,
the surplus water, which originally transported the solutes,
will be released. This water is being discharged continuously
as solutes are transported via the ray system to support the
construction of immature xylem. This recycling of water
was estimated by Mu$ nch (see Bu$ sgen and Mu$ nch, 1929) as
1–3 % of the total xylem transport. However its main point
of release lies in the outermost layers where new xylem is
being deposited. The percentage of water released in this
way may be as high as 45 % of the xylem stream as was seen
in Ricinus communis seedlings kept in humid atmospheres
when the magnitude of the flows was estimated by NMR
imaging (pers. comm., Dr W. Ko$ ckenberger, Beyreuth
University, Germany, at the International Phloem Conference, Canterbury, UK, August 1995).
Water status of xylem : reŠersal of embolisation following
caŠitation
There are several reports of reversal of xylem embolisation
(e.g. Magnani and Borghetti, 1995) which are difficult to
explain by other means (e.g. uptake of rainwater or root
pressure). It is apparent that restoration by ‘ Mu$ nch water ’
released from the phloem, perhaps utilizing supraatmospheric pressures within the intact plant, can explain
these effects. Convincing support was provided by the
neglected Molotovsky (1934, but see Milburn 1975, 1979).
Molotovsky confirmed and measured the flow of water
from flaps of bark precisely (exactly as predicted by Mu$ nch,
1930) and showed in several species that it was diurnal and
maximal at midday, indicative of photo-assimilation. I have
recently confirmed his results on willows (Salix spp.).
Similar fluxes have been measured in developing cowpea
fruits (Pate et al., 1985). Thus ‘ Mu$ nch water ’ may have a role
in preventing or removing any cavitation emboli induced by
excessive tensions, thus helping to maintain the cohesive
flow of water through the plant.
Recently evidence was presented (Salleo et al., 1995)
indicating that xylem conduits could only be refilled,
restoring xylem conduction, if phloem continuity is maintained. This is further powerful evidence that the phloem is
operating to supply Mu$ nch water to the outermost xylem
vessels. Recycled water can thus explain two puzzling
aspects of the conventional cohesional theory (a) the
apparent reversal of hydrostatic gradients and (b) the
restoration of xylem conduction in many plants, and
especially trees, following embolization.
CONCLUSIONS
There is no need to question the functioning of the cohesion
hypothesis (Dixon and Joly, 1895 ; Dixon, 1914) on the basis
of evidence published recently. Mercury injection, studies
on mangroves, and experiments by others, show that xylem
indeed generates sap tensions far greater than those yet
405
measured by the pressure probe. These accord with what
might be expected from vastly documented studies of water
potentials of parenchyma cells abutting the xylem conduits.
Hence there is no need to dismiss the concept of a cohesional
catena drawing water from soil to atmosphere through the
plant. Through application of Occam’s razor it is more
appropriate to rely on known and established facts than to
launch into complex, fanciful, unproven alternatives.
The pressure probe measurements of Zimmermann et al.
(e.g. 1994 a, b, 1995) can be mainly explained, either as
artifacts arising from induced cavitation within conduits, or
from encountering immature xylem initials, or from the
localized release of water from the sieve tubes as a result of
phloem unloading. The major misconception has been that
the entire xylem system is tightly integrated. Xylem conduits
do not deŠelop as a single integrated system ; furthermore
lateral and radial hydraulic resistances can be considerable,
restricting easy migration of sap from one conduit to
another. Thus, from a cellular perspective, localized atypical
pressures can develop, mainly in the outermost xylem,
adjacent to the sieve tubes.
The claims that the organic molecular contents of xylem
conduits can contribute in some mysterious way to xylem
transport (Zimmermann et al., 1994 a) are readily explicable.
Observations using the pressure probe seem to be the
consequence of wounding immature xylem elements, when
they are still replete with solutes as a necessary precondition for vacuolation prior to lignification. This must
generate initially positive, then later negative, pressures as
the vessels mature. The explanations provided here raise
new challenges for those who object to the cohesion
hypothesis. If results from the pressure probe in single cells
are to be properly reconciled with cell-population methods,
such as the pressure bomb, a comprehensive awareness of
plant functioning is essential. Measurements documenting
the sequence of pressure changes in developing xylem
initials are needed urgently. Lateral fluxes of Mu$ nch water
transport also need to be measured accurately within stems
of many plants. The liberation of Mu$ nch water also
explains how xylem conduction could be restored in the
outermost conduits under appropriate circumstances.
The present challenges to the cohesion theory may soon
be regarded as a ‘ storm in a teacup ’. It seems appropriate
to paraphrase G. Santayana’s maxim : ‘ Those who fail to
remember the lessons of plant physiology, anatomy and
development are condemned to repeat them ’. Nevertheless
the present controversy may be beneficial if it leads to a
wider appreciation of the holistic way in which plants
operate, especially in utilizing both xylem and phloem to
transport water over long distances. The conventional view,
that sap ascent is induced in plants because substantial
tensions, established by the evaporation of water, are
transmitted via cohesional forces throughout the
vasculature, has yet to be seriously challenged.
A C K N O W L E D G E M E N TS
This work received no external funding. I am grateful to Dr
D. Woodcock, University of Hawaii for information ; R.
406
Milburn—Sap Ascent in Vascular Plants
Kenny, C. Brennan, and V. G. Williamson for assistance
with analyses and graphics ; Emeritus Professor G. R.
Morris for Russian translation ; Drs W. Allaway and S.
Tyerman kindly provided criticism.
LITERATURE CITED
Balling M, Zimmermann U, Buchner KH, Lange OL. 1988. Direct
measurement of negative pressure in artificial biological systems.
Naturwissenschaften 75 : 409–411.
Balling M. Zimmermann U. 1990. Comparative measurements of the
xylem pressure of Nicotiana plants by means of the pressure bomb
and pressure probe. Planta 182 : 325–338.
Bennet JP, Anderssen FG, Milad Y. 1927. Methods of obtaining
tracheal sap from woody plants. The New Phytologist 26 : 316–325.
Bollard EG. 1953. The use of tracheal sap in the study of apple-tree
nutrition. Journal of Experimental Botany 4 : 363–368.
Bose JC. 1923. The physiology of the ascent of sap. London : Longmans
Green and Co.
Briggs LJ. 1950. Limiting negative pressure of water. Journal of Applied
Physics 21 : 721–722.
Bu$ sgen M, Mu$ nch, E. 1929. The structure and life of forest trees. 3rd
edn. Revised by E. Mu$ nch ; English Translation by T. Thompson.
London : Chapman and Hall.
Canny MJ. 1990. What becomes of the transpiration stream ? The New
Phytologist 114 : 341–368.
Canny MJ. 1995. A new theory for the ascent of sap–cohesion
supported by tissue pressure. Annals of Botany 75 : 343–357.
Crombie DS, Hipkins MF, Milburn, JA. 1985. Gas penetration of pit
membranes in the xylem of Rhododendron as the cause of
acoustically detectable sap cavitation. Australian Journal of Plant
Physiology 12 : 445–453.
Cruiziat P. 1995. Water relations in trees : some aspects of the present
state of the art. 3rd International Congress on ‘ The Tree ’. 11–16
Sep. 1995, Montpellier. Montpellier, France : Pub. Sauramps. (in
press).
Dixon, HH. 1914. Transpiration and the ascent of sap in plants. London :
Macmillan.
Dixon HH, Joly J. 1895. On the ascent of sap. Philosophical
Transactions of the Royal Society (London) 57B : 563.
Henderson SJ, Speedy, RJ. 1987. Temperature of maximum density in
water at negative pressure. Journal of Physical Chemistry 91 :
3062–3068.
Holbrook MN, Burns MJ, Field CB. 1995. Negative xylem pressures in
plants : a test of the balancing pressure technique. Science. 270 :
1193–1194.
Magnani F, Borghetti M. 1995. Interpretation of seasonal changes of
xylem embolism and plant hydraulic resistance in Fagus sylŠatica.
Plant, Cell and EnŠironment 18 : 689–696.
Milburn JA. 1966. The conduction of sap. I. Water conduction and
cavitation in water stressed leaves. Planta 65 : 34–42.
Milburn JA. 1970. Cavitation and osmotic potentials of Sordaria
ascospores. The New Phytologist 69 : 133–141.
Milburn JA. 1972. Phloem transport in Ricinus. Pesticide Science 3 :
653–665.
Milburn JA. 1973 a. Cavitation in Ricinus by acoustic detection :
Induction in excised leaves by various factors. Planta 110 :
253–265.
Milburn JA. 1973 b. Cavitation studies on whole Ricinus plants by
acoustic detection. Planta 112 : 333–342.
Milburn JA. 1975. Mechanisms of phloem transport Chapter 14
Pressure flow. In : Zimmermann MH, Milburn JA, eds.
Encyclopedia of plant physiology New Series. New York : SpringerVerlag, 1 : 328–353.
Milburn JA. 1979. Water flow in plants. (Integrated Themes in Biology
Series). London : Longman Group, 1–255.
Milburn JA. 1991. Cavitation and emboli, development and
significance. In : Raghavendra AS, ed. Physiology of trees. New
York : John Wiley & Sons, 163–173.
Milburn JA. 1993. Cavitation. A review : past, present and future. In :
Borghetti, M. Grace JA, Rashi A, eds. Water transport in plants
under climatic stress. Proceedings of an International Workshop
(May–June 1990), Vallombrosa, Firenze, Italy. Cambridge :
Cambridge University Press, 14–26.
Milburn JA, Johnson RPC. 1966. The conduction of sap. II. Detection
of vibrations produced by sap cavitation in Ricinus xylem. Planta
69 : 43–52.
Milburn JA, Kallarackal J. 1991. Sap exudation. In : Raghavendra AS,
ed. Physiology of trees. New York : John Wiley & Sons, 383–400.
Milburn JA, McLaughlin ME. 1974. Studies of cavitation in isolated
vascular bundles and whole leaves of Plantago major L. The New
Phytologist 73 : 861–871.
Molotovsky GK. 1934. Opyty po proverke oteorii Miunkha. Botanic
Journal of the U.S.S.R. 19 : 225–230.
Mu$ nch E. 1930. Die Stoffbewegungen in der Pflanze. Jena : Fischer.
Passioura JB. 1991. An impasse in plant water relations ? Botanica Acta
104 : 405–411.
Pate JS, Davidson NJ, Kuo J. Milburn, JA. 1990. Water relations of the
root hemiparasite Olax phyllanthi (Labill) R.Br. (Olacaceae) and
its multiple hosts. Oecologia 84 : 186–193.
Pate JS, Peoples MB, Van Bel AJ, Kuo J, Atkins C. 1985. Diurnal
water balance of the cowpea fruit. Plant Physiology 77 : 148–156.
Pockman WT, Sperry JS. O’Leary JW. 1995. Sustained and significant
negative pressure in xylem. Nature 378 : 715–716.
Priestley JH Scott LI. 1955. An introduction to botany. 3rd Edition.
London : Longmans Green and Co., 351–353.
Renner O. 1915. Theoretisches und Experimentelles zur
Ko$ hasiontheorie der Wasserbewegungen. Jahrbuch fuX r
Wissenschaftliche Botanik 56 : 617–667.
Richter H, Kikuta S. 1995. Xylem water transport : evidence consistent
with the cohesion theory. 3rd International Congress on ‘ The Tree ’.
11–16 Sep. 1995 Montpellier. Montpellier, France : Pub. Sauramps,
(in press).
Ritman KT, Milburn JA. 1990. The acoustic detection of cavitation in
fern sporangia. Journal of Experimental Botany 41 : 1157–1160.
Salleo S, Hinckley TM, Kikuta SB, Lo Gullo MA, Weilgony P, Yoon TM, Richter, H. 1992. A method for inducing xylem emboli in situ :
experiments with a field-grown tree. Plant, Cell and EnŠironment.
15 : 491–497.
Salleo S, Lo Gullo MA, De Paoli D, Zippo M. 1995. Refilling of
cavitated xylem conduits in young plants of Laurrus nobilis L. A
possible mechanism. IUFRO World Congress, IUFRO, Tampere,
Finland, 6–12.
Scholander PF. 1968. How mangroves desalinate seawater. Physiologia
Plantarum 21 : 251–261.
Scholander P, Hammel HT, Bradstreet ED Hemmingsen EA. 1965. Sap
pressures in vascular plants. Science, 148 : 339–346.
Scholander PF, Bradstreet ED, Hammel HT, Hemmingsen EA. 1966.
Sap concentrations in halophytes and some other plants. Plant
Physiology 41 : 529–532.
Scholander PF, Hammel HT, Hemmingsen E, Garey G. 1962. Salt
balance in mangroves. Plant Physiology 37 : 722–729.
Smith AM. 1994. Xylem transport and the negative pressures
sustainable by water. Annals of Botany 74 : 647–651.
Sperry JS, Saliendra NZ. 1994. Intra- and inter-plant variation in
xylem cavitation in Betula occidentalis. Plant, Cell and EnŠironment
17 : 1333–1241.
Steudle E. 1993. Pressure probe techniques : basic principles and
application to studies of water and solute relations at cell, tissue
and organ level. In : Smith JAC, Griffiths H, eds. Water deficits :
plant responses from cell to community. Oxford : Bios Scientific
Publishers, 5–36
Ursprung A, Blum G. 1919. Zur Kenntnis der Saugkraft. IV Hedera
helix. Abgeschnittenes Blatt. Berichte der Deutschen Botanischen
Gesellschaft 37 : 453–462.
Tyree MT, Sperry J. 1989. Vulnerability of xylem to cavitation and
embolism. Annual ReŠiew of Plant Physiology 40 : 19–38.
Zheng Q, Durben DJ, Wolf GH, Angell CA. 1991. Liquids at large
negative pressures : water at the homogeneous nucleation limit.
Science 254 : 829–833.
Zimmermann MH. 1983. Xylem structure and the ascent of sap. Berlin :
Springer Verlag.
Zimmermann U, Meinzer F, Bentrup FW. 1995. How does water ascend
in tall trees and other vascular plants ? Annals of Botany 76 :
545–551.
Milburn—Sap Ascent in Vascular Plants
Zimmermann U, Zhu JJ, Meinzer F, Goldstein G, Schneider H,
Zimmermann G, Bankert R, Thurmer R, Melcher P, Webb D,
Haase A. 1994 a. High molecular weight organic compounds in the
xylem sap of mangroves : implications for long-distance water
transport. Botanica Acta 107 : 218–229.
407
Zimmermann U, Zhu JJ, Meinzer F, Goldstein G, Schneider H,
Zimmermann G, Bankert R, Thurmer R, Melcher P, Webb D,
Haase A. 1994 b. Xylem water transport : is the available evidence
consistent with the cohesion theory ? Plant, Cell and EnŠironment.
17 : 1169–1181.