Plant, Cell and Enviromnent (1984) 7, 693-697

TECHNICAL REPORT

A new stem hygrometer, corrected for temperaturegradients and calibrated against the pressure bomb

M. A. DIXON & M. T. TYREE Botatiy Departtnent, University of Tot-onto, Tot-onto, Ontario, CanadaM5S lAl

Received 26 March 1984; accepted for publication 11 June 1984

Abstract. A simple stem hygrotneter for attachmentto a bared section of sapwood or a cross-sectionalcut end of a shoot is described. Two welded chromel-constantan thertnocouples inside the chatnber, onetouching the sample and the other in the chatnberair, allowed measurement of and eorrection for thetemperature gradient between the sample and thedewpoint tneasuring junction. The instrument wasattached to the cut end of an apical shoot of Thufaoccidentalis L. protuding from a Scholander-Hammelpressure bomb. Cut-end water potential (i/'hyg)'measured using the stem hygrometer, was comparedto xylem pressure potential (i/'j,,) while the latter wasmanipulated in the pressure bomb. After an initialequilibration time of 3-4 h, hygrotneter equilibriutnvalues were achieved within 1.5-4.0 min of changingi/ ,p in the pressure bot-nb. The half-titne (r,/.) forvapour pressure equilibration was 15-40s. Stabletemperature gradients between the sample anddewpoint measuring junction of 0.01-0.1 C weremeasured. Correcting 1//,,,, for the temperaturegradient resulted in excellent agreement with t//jp.

Key-words: sten-i hygrometer; equilibration time; temperaturegradients; pressure botnb.

Introduction

Although a t-elatively new teehnique in the field ofwater relations research, the thertnocouple hygro-tneter (Spanner, 1951) has etijoyed a wide audience.Recently, the attraction of continuous non-destruetive monitoring of plant tissue water potentialhas received attention and variations of this highlyadaptive technique have appeared in the literature(Neumann & Thurtle, 1972; Michel, 1977; McBurney& Costigan, 1982). Most in situ hygrometers havebeen designed for attachtnent to leaves, but thet-e ateadvantages in using stem hygrometers. Less signifi-cant energy balanee disruptions and ease of attach-ment of the stem hygrometer favour its use over thatof the leaf hygrotneter.

This paper describes a simple stem hygrotneter andits perfortnance t-elative to the Scholandet-Hatntnel

Correspondence: M, A, Dixon, University of Toronto, SnakeIsland Institute, PO Box 1, RR 2. Keswick, Ontario, CanadaL4P 3F.9,

pressure bomb (Seholander et al, 1965). Unlike itsuse in similar studies in the past (e.g. McBurney &Costigan, 1982), the stetn hygrometer is attached tothe end of a shoot protruding frotn the pressurebotnb. This reduces probletns concerning waterpotential gradients between the sites of hygrometerand pt-essure botnb tneasurements and facilitates amore precise comparison.

The thertnocouple hygrometer relies, for itssuccess, on the accurate determination of very smalldilTerential tetnperatures (0.01-0.5 °C), and also onthe assumption that the initial measuring junctionand satnple tetnperatures are identical. It has longbeen lealized that failure to achieve the latter is atnajor souree of error in hygrotnetry.

Michel (1979) addt-essed the problem oftemperature gradients in stem hygrometers butstopped short of actually tneasuring the error (i.e.the error-indueing gradient is that between thesatnple and the measuring junction, not thetneasuring junction and the reference or instrumenttetnperature). He relied on predicting the former bytneasuring the latter. In our experience, however, thevariability between instruments, as well as installa-tions, precluded reliable predictions. Calissendorf &Gardner (1972) deseribed a leaf psyehrotneter whichdid measure the temperature gradient between thesatnple and the tegion of the measuring junetion.However, the tesults they t-eported when using theirinstrument on corn leaves were unrealistically lowcotnpared to results frotn sitnilar studies (e.g.Neumann & Thurtle, 1972).

This paper describes an instrument similar, inprinciple, to that described by Calissendorf &Gardner (1972) in that it allows tneasuretnent oferror-inducing temperature gradients within thechatnber, but for use on stems rather than leaves.Furthertnore, the technique described allows con-curt-ent and directly related tneasuretnents of waterpotential via an independent method (pressurebomb).

Materials and methods

A thermoeouple hygt-otneter ehatnber (Fig. I) wasmachined frotn the head of a 25 mm stainless-steel

693

694 M. DIXON & M. T. TYREE

Output wires

Copper-constantanthermocoupleClampChromel-constantanthermocouples

Calibration discholder

Figure I. Diagram (not to scale) of stainless-steel stem hygrometerand calibration disc holder. The clamp can be used to attach theinstrument to the side of stems.

bolt (AISI 304). It was fitted with two weldedchromel-constantan measuring junctions (wireavailable from Goodfellow Metals, Milton Rd,Cambridge, U.K.). One of the junctions was incontact with the sample surface while the other wasin the chamber air. A Weseor HR-33T microvolt-meter (Weseor, Inc., Logan, Utah, U.S.A.) was usedto measure the temperature difference between thetwo junctions as well as to provide the Peltier coolingcurrent for the junction in the chamber air. Asoldered eopper-eonstantan thertnocouple wasembedded in the chatnber body to measureinstrument temperature. The instrument wasdesigned to allow attachment to either the cut end ofa shoot or the side of a bared seetion of sapwood.

The hygrometer was calibrated using filter-paperdiscs soaked with standard sodiutn ehloride solutionsof molality 0.1-0.5 mol kg"* in increments of0.1 mol kg '. The water potential of each standardsolution at 25 °C was obtained from tables (Lang1967).

An apical shoot of Thufa oecidentalis, 0.6-0.8 mlong and 10-12 mm butt diameter, was enclosed in aSeholander-Hammel pressure bomb (Seholanderetal, 1965) with 30-60 mm of the butt-endprotuding. The stem hygrometer was attached to theeut end of the shoot (Fig. 2). Grease was applied atthe junction of wood and metal surfaces to ensure avapour seal. The assembly at the top of the bombwas then enclosed by a temperature-controlledehamber (Fig. 2). This consisted of a metal eylinder(105 mm diam. x l75mm high), completely linedwith 5-mtn-diameter copper tubing and covered with30-mm-thick Styrofoam (Styrofoam SM). The endsof the cylinder were sealed with eotton wool andStyrofoam. Air was circulated within the chatnber bya 6 V fan (Mieronel Ltd, Zurich, Switzerland). Fluidfrotn a constant-temperature bath (Exacal, ModelEX 100, Neslab Instruments, Portsmouth, NewHatnpshire, U.S.A.) was putnped through the eoppertube and maintained at 25 ±0.01 °C. Fluctuations inambient air tetnperature were minimized (+1.0°C)using an air conditioner.

The sample shoots were air-dehydrated overnight(10-12 h) before they were placed in the pressure

FanCopper tubingTin cylinder

Styrofoam

Stem hygrometer

Grease

Cotton wool

0 - ring

Rubber seal

Pressure bomb

Cedor shoot

Figure 2. Diagram (not to scale) of a cedar shoot clamped in apressure bomb with the stem hygrometer attached to the cut endand cnclo.sed in a temperature-controlled chamber.

bomb. Xylem pressure potential (1// ,,), as determinedat the start of each experitnent, was typically — 1.0 to— 2.0 MPa. Following i/',,, determination the bombpressure was lowered 0.1 MPa and held. Thehygrometer and temperatute control systetn werethen installed and allowed to achieve vapourpressure and tetnperature equilibt-ium (usually3-4 h). The botnb pressure was then lowered in0.2-0.4 MPa steps at a rate of 3.0 kPa s""' and heldfor at least 15 tnin at each step. Hygrometer readingsof water potential (i//|,yg) at the cut end were takenevery 5 min and compared to computed \\i^^ (presentbotnb pressure minus initial balance pressut-e).Pressure changes were reversed upon reaching thelowest i/'xp (i.e. bomb pressure = 0).

Vapour pressure equilibration

Analysis of vapour pressure equilibration time wasaehieved by operating the HR-33T in the Dew PointMode (Weseor Instruction Manual, 1982) andcontinuously recording the output on a strip ehartrecot-der while manipulating \\i^^ in the pressurebomb. The botnb pt-essure was raised or lowet-ed at arate of about 40.0 kPa s^ ' to effect a total change of0.8 MPa in each case. In the ease of these rapidpressure changes, tetnpei-ature changes inside thepressure bomb were inevitably eaused (Puritch &Turner, 1973) and the extent to which this affectedthe temperatut-e at the cut end of the shoot wasassessed as outlined in the following section.

Temperature gradients

Stable temperature gradients within the hygrotneterchamber between the sample surface and thedewpoint measuring junetion were aehieved using thetemperature-controlled chamber. The value of the

TEMPERATURE-CORRECTED STEM HYGROMETER 695

gradient was determined from the differential outputof the two welded thet-tnocouples. In the temperaturerange of interest, the tempet-ature coefficient ofcht-otnel-constantan thertnocouples is 0.061 mV/°C(Otnega Engineering, Inc., Temperature MeasuringHandbook, 1980, p. A-11).

Transient temperature gradients within theehatnber, which were apparent only during t-apidbotnb pt-essut-e changes, were also measured in theabove tnanner and continuously recorded duringrapid pressut-e changes.

A correction to i/'i.y,, resulting frotn thetemperatut-e gradient tneasured at the time of eachreading was applied according to the fortnula:

w

whet-eip^ = corrected water potential (Pa),ijj^ = et-roneous water potential (Pa),R = gas constant (J mol"' K~'),T = tetnperatut-e (K); subscripted

surface temp.,1 dC*

's' = sample

(MPa)

-2

-2

Figure 3. The relationship between computed i//,p (bomb pressureti-iinus initial balance pressure) and ij/^,, as measured by the stemhygrometer before correction for temperature gtadients. Only ftveexperiments are shown to avoid the confusion of coincident data

C* = satutated water vapour concentration attemp. T,

K = partial tnolar volutne of water (m^ tnoP'),AT = tetnperature difference between stem surface

and dewpoint measuring junction.

The correction factor {kRTJVJ evaluates to 7.77MPa K"' at 25"C.

Results

Following initial equilibration, 1/ ,, , as measured bythe stetn hygrotneter, was always lower(0.2-1.0 MPa) than 1//,,,, before correction for themeasured tetnperature gradient had been taken intoaccount (Fig. 3). As long as the tetnpet-ature gradientremained constant, the relationship between ,,yg andt/ ,,,, was linear. Deviations from linearity withinexpetitnents usually reflected tninor fluctuations inthe tetnperature gt-adient. Measuretnent of thegt-adient and subsequent correction of i/ yg resultedin the tclationship shown in Fig. 4.

The half-time (z,/, = time to reach 50% equilibrium)for vapour pressut-e equilibt-ation was usually lessthan 60 s and seemed to vary mainly according to thelength of stetn ptotruding from the pressure bomb(the longer the stetn, the longer the half-titne). Figure5 shows t-eplicate determitiations of equilibrationtitne on the satne shoot with 35 mtn protruding. Thepoints rept-esent the natural logarithtn of theabsolute value of the difference between apparenti//hyg and eotnputed i/ p after the pressure change hadbeen cotnpleted.

The time course for temperature equilibrationfollowing rapid pressure changes in the bomb isshown in Fig. 6. Generally, the half-time for

(MPa)

- 2

y

expipoittt.s.

- ' - 2

Figure 4. The relationship between computed i ,p (bomb pressureminus initial balance pressure) and (/f yg as measured by the stemhygrometer after eorrection for temperature gradients. The dashedline represents a 1 : 1 correlation.

696 M, DIXON & M. T. TYREE

Or

- 3

30

Time (s)

6 0

Figure 5. The absolute value of the difference between apparenti/i,,,!, and eotnputed \j/^^ plotted as a nalural log-arithm againsi limefor a shoot with 35 mm protruding from the pressure bomb. Thesereplicates depict r-apid water potential changes in both the positiveand negative directions. The average slope is —0.036 +0.003 s" ' .The half-time for equilibration is 19.3 s.

temperature equilibration was roughly equal to thatfor vapour pressure equilibration. The magnitude ofthe temperature change at the cut end of the shootwas influenced by the rate of pressure increase ordecrease as well as the length of stem protrudingfrom the bomb.

Discussion

Our method of concurrently measuring stem waterpotential using two independent techniques showedexcellent agreement between the two whencorrections for tetnperature gradients were t-nade. Itis noteworthy that these gradients persisted eventhough the stem hygrometer and a significant portionof the stem satnple were housed in a tetnperature-controlled chamber. This was probably the con-sequence of heat conduction through the rathermassive stem from regions outside the temperature-controlled chamber. It is surprising that temperature-corrected hygrometers/psychrometers have not beenused more widely. In most chambers designed totake leaf discs or leaf strips the sample is usuallycompletely surrounded by a metallic thermal mass;thus substantial gradients of temperature betweenthe leaf surface and thermocouple junction are lesslikely than in our systems. However, a temperature-corrected hygrometer might be advantageous whentemperature gradients due to the heat of respirationare thought to be substantial (Barrs, 1964).Tetnperature corrections are clearly needed in stemhygrometers and may be advisable for in situ leafhygrometers, especially when used under fieldconditions.

In some cases the temperature gradient did not

-2-0 -

-3-0

-4 -0

30

Time (s)

60

Figure 6. The absolute value of the temperature dillerence betweensample and dewpoint measuring junction plotted as a naturallogarithm against time for a shoot with 35 mm protruding fromthe pressure bomb. The.se replicates depict temperature changesresulting from rapid bomb pressure changes in both the positiveand negative directions. The average slope is -0.026 +0.001 s" ' .The half-time for equilibration is 26.6 s.

account for all the apparent disagreement betweenthe two tnethods. For example, in Fig. 4, note thatthe relationship between 1// ^ , and computed tp^^tended to deviate from the 1 : 1 correlation aseotnputed i/',,, increased in one of the trials.Subsequently, it was found that excess water used tosoak filter paper inside the botnb had accumulated ina pool on the bottom. The increasing bomb pressuremay have caused water to infiltrate the xylemresulting in an error in eotnputed i/ ,,,; hence thegradual deviation as ij/^^ increased.

The stem hygrotneter showed rapid vapourpressure and temperature equilibration; half-titncsnever exceeded 60 s (Figs 5 and 6). These favourablecharacteristics are highly desirable for reliable use ofthe instrutnent.

Comparison of equilibration characteristicsbetween our instrutnent and others reported in theliterature is diflicult since a standard for comparisonis rarely used. Boyer (1972) t-eported an equilibrationhalf-time of 4 min when using an isopiestictechnique (Boyer & Knipling, 1965) on sunflowerleaves. This half-titne is determined by the combinedhalf-titnes for tissue and chamber equilibtation.Equilibration times, reported in the literature, varyfrom scvet-al minutes to several hours (Millar, 1974;Campbell & Catnpbell, 1974). A variety of factorsinfluence equilibration time, not least of which is theehatnber material. Chambers with rubber seals oroxidized or dirty metal surfaces will displayequilibration characteristics dominated by those ofthe ehatnber material (Dixon & Grace, 1982).Another factor influencing equilibration is thecuticular resistance of the sample (in the case of leaftissue). This can be lessened by removal of cuticularwaxes (Neumann & Thurtle, 1972) or abrasion of theleaf surface (Brown & Tanner, 1981). Our stem

TEMPERATURE-CORRECTED STEM HYGROMETER 697

hygrometer exhibited a very fast equilibration timebecause near optimum chamber-wall materials wereused (Dixon & Grace, 1982), rubber seals wereeliminated, and cuticular barriers to water evapora-tion from the plant surface did not exist. Also, thetisstie in equilibrium with the stem hygrometer has ahigh hydraulic conductivity and low capacitancerelative to leaf tissue. Therefore, water potentialequilibration within the stem tissue is quite fastfollowing an increase or decrease in bomb pressure,(If the bomb pressure does not exceed the balancepressure then very little water redistribution isneeded inside the shoot to bring about a change inxylem water potential,)

Transient electrical zero offsets are anothersignificant source of error in the use of thermocouplepsychroineters/hygrometers. It was found that thezero offset of the HR-33T microvoltmeter wasaffected by proximity to other electrical equiptnentand to human bodies. Poorly earthed equipment andproximity of a,c, mains cables to the hygrometeroutput leads was found to cause significant zerooffsets and errors in l|/^,y^, Shielding the wires did noteliminate but, in fact, accentuated the problem insome cases. Spatial isolation from other electricalequipment of the microvoltmeter and wires within atleast a 2 m radius minimized these errors. Strictadherence to experimental protocol enhanced thereproducibility of data.

We conclude from our tests and calibrations thatour version of the stem hygrometer is a reliable andaccurate instrument for measuring in situ stem waterpotentials. Of course, certain precautions must beobserved, especially when employing this techniquein the field.

Avoiding exposure of the installation to directradiation is helpful in reducing the magnitude ofthermal gradients. Cotton wool and Styrofoam areappropriate insulating materials, A final reflectivelayer of aluminium foil is recommended as well.Elaborate temperature control, although sometimespossible in the field, is difficult to achieve. Sufficientinsulation and careful measurements of chambertetnperature should produce reliable results whencorrected for temperature.

In the tests reported here the hygrometer wasattached to the cross-sectional cut end of the satnple.This was achieved using a variation of the clampingdevice depicted in Fig, 1, This tnode of attachtnent isnot necessary and is, in fact, not recommended inmany cases (e,g, resinous species, such as some pines,and species which have a significant proportion ofliving cells mingled with their xyletn conduits).Contamination of the chamber with symplasm orresin will adversely affect readings. The alternative isattachment to the side of the stem. Carefulpreparation of the sample by retnoving the bark,phloem and cambium to expose an appropriate sizedarea of the sapwood is required. After the area hasbeen cleaned and wiped dry the instrutnent can be

attached and the remaining exposed area sealed withsilicone grease. As long as the sapwood is notwounded there should be no problems. Obviouslythis technique is limited to those species which allowaccess to a portion of undamaged xyletn.

Acknowledgments

This work was carried out on Indian reserve land; wethank the Chippewas of Georgina and HaroldMcCue, in particular, for permission to work onSnake Island, This work was supported by grant no,A6919 from the Natural Sciences and EngineeringResearch Council of Canada,

References

Barrs, H.D. (1964) Heat of respiration as a po,ssible source oferror in the estimation by psychrotnelric methods of waterpotential in plant tissue. Nature. 203, 1136-1137.

Boyer, J.S. (1972) Use of isopiestic technique in thermocouplepsychrometry. III. Applieation to plants. In Psyetnomctry inWater Relations (eds R.W. Brown & B.P. van Haveren),pp. 220-223. Utah Agricultural Experitnent Station, Utah StateUniversity, Logan.

Boyer, J.S. & Knipling, E.B. (1965) Isopiestic technique formeasuring leaf water potentials with a thermocouplepsychrometer. Proeeedings of tlie National Aeademy of SeieneesU.S.A.. 54, 1044-1051.

Brown, P.W. & Tanner, C.B. (1981) Alfalfa water potentialmeasurement: a comparison of the pressure chamber and leafdewpoint hygrometers. Crop Seienee. 21, 240-244.

Calissetidorf, C. & Gardner, W.H. (1972) A temperaturecompensated leaf p.sychtotneter for in situ measurements ofwater potetitial. In P.syetuometry in Water Relations (eds R.W.Brown & B.P. van Haveren), pp. 224-228. Utah AgriculturalExperitnent Station, Utah State University, Logan.

Catnpbell, G.S. & Catnpbell, M.D. (1974) Evaluation of athermocouple hygrometer for measuring leaf water potentialin situ. Agroni)my Journat, 66, 24-27.

Dixon, M.A. & Grace, J. (1982) Water uptake by some ehatnbermaterials. Ptant. Cett and Environment. 5, 323-327.

Lang, A.R.G. (1967) Ostnotic eoellieients and water potentials of.sodium chloride solutions from 0 to 40 X . Au.stratian Journat ofCtiemistry. 20, 2017-2023.

McBurney, T. & Costigan, P.A. (1982) Measurement of stemwater potential of young plants using a hygrometer attached tothe stem. Journat of Experimentat Botany, 33, 426-431.

Michel, B.E. (1977) A tniniature stem thertnocouple hygrometer.Plant Phy.siotogy. 60, 645-647.

Michel, B.E. (1979) Correction of thermal gradient errors in stemthertnocouple hygrometers. Ptant Physiology. 63, 221-224.

Millar, B.D. (1974) Itnproved thermocouple psychrometer forthe tneasuretnent of plant and soil water potential. IILEquilibration. Journat of E.xperimentat Botany. 25, 1070-1084.

Neutnann, H.H. & Thurtle, G.W. (1972) A Peltier cooledthermocotiplc dewpoint hygrometer for in situ measurements ofwater potential. In P.syehrometry in Water Retations (eds R.W.Brown & B.P. van Haveren), pp. 103-112. Utah AgriculturalExperitnent Slalion. Utah Slate University, Logan.

Putilch, G.S. & Turner, J.A. (1973) Effects of pressure increaseand release on tetnperature within a pressure chatnber used toestitnate plant waler potential. Jinirnat of Experimentat Botany.24, 342-348.

Scholander, P.F., Hamntel, H.T., Bradstreel, E.D. &Hemtning,sen, E.A. (1965) Sap pressure in vaseular plants.Seienee. 148, 334-346.

Spanner, D.C. (1951) The Pcllicr cITecl and ils use in themeasurement of suction pressure. Jimrnat of ExperimentatBotany, 2, 145-168.