SUGGESTIONS FOR THE USE OF WARBURG RESPIROMETERSIN PLANT PHYSIOLOGICAL INVESTIGATIONS'

JAMES W. BROW-N

(WITH TWO FIGURES)

Introduction

Although Warburg respirometers have been used extensively in animalphysiology for a number of years they have niot been employed by plantphysiologists as much as their accuraey and convenieniee seem to warrant.HARRINGTON (2) used a modification of the Warburg principle in his res-pirometers, but TANG (6) and MURLIN (5) apparently are the only workersto use the standard type of respirometer for studies of seed respiration.The Warburg respirometer permits simultaneous measurement of 02 con-sumption and CO2 production on small samples for short time intervals.This makes possible easy and rapid replicationi of determinations withoutexcessive use of material. By the use of respirometer vessels of suitable sizeand shape not only seeds, but small fruits, small seedlings, excised roots,buds, flowers and other plant parts may be placed in them.A careful study of the technique was made in colnnection with their use

in an investigation of the respirationi of acorns at various temperatures.It was found that several factors must be considered which are not includedin the technique as given in the papers describing their operation (4, 5, 6, 7).This paper suggests certaini modifications of the stanidard procedure togetherwith suggestions on the operation of both large anid small respirometer vessels.

Experimental methods

The usual procedure for obtainilnl coneurrenit measurements of 02 con-sumption and CO2 production is as follows. The time periods given arethose which the writer found necessary for the conditions of his investiga-tion and the letters refer to the sample record sheet at the end of the paper.For the most efficient operation of the respirometers it is desirable to deter-mine experimentally the time periods required for temperature adjustment.This can be done by placing the vessels in the bath under the conditions ofthe experiment and observing the time necessary to bring the manometricfluid to a constant level after the stopcocks of the maniometers are closed.

An acid, usually HCI, is placed in the side arm of the respirometer vesselwhich is then stoppered. An alkali, usually NaOH or KOH, is placed inthe bottom of' the vessel or in the central cup if onie is present, the support

1 No doubt some investigators have used most of the suggestions offered in thispaper, but since they have never been published in any collected form, they are presentedhere mainly for the consideration of those who are just beginning to use the method.

309

PLANT PHYSIOLOGY

for the respiring material is placed in position, the top of the vessel seatedand the whole unit attached to the manometer. The apparatus is set upona shaker so that the vessels are submerged in a constant temperature bath.After the vessel has come to the temperature of the bath (7 minutes) theapparatus is removed from the shaker, taken apart, the respiring materialplaced in the vessel upon the support, and the unit reassembled and re-turned to the shaker. The time a is recorded when the material is placedin the vessel, and this time is used as the start of the period of CO2 produc-tion. Three minutes later, after respirometer and material are at thetemperature of the bath, the manometric fluid is levelled to the 150-mm.mark in the arm connected to the vessel and the level in the open arm isrecorded, b, with the time c, and the stopcock is closed. The time c marksthe start of the period of 02 consumption. The apparatus is left on theshaker for the length of the experiment (30 minutes). The shaker wasoperated at 70 complete oscillations per minute with a throw of 7 cm.Whenever a reading is made care must be taken to have the manometers ina vertical position.

When the period of 02 consumption is ended, the time d is recorded andthe manometric fluid is again brought to the 150-mm. mark and the openarm reading e is made. The length of the 02 consumption period is thend - c, while the change in manometric graduations, b - e, represents theamount of 02 consumption. The apparatus is taken from the shaker imme-diately after recording d and e and turned in such a position as to cause theacid from the side arm to flow into the alkali in the vessel and liberate theCO2 which has been absorbed by the alkali. An excess of acid is necessarvand where very small amounts of the reagents are used it is advisable to tiltthe unit so that the alkali flows into the side arm and then allow the mixtureto return to the vessel to insure more thorough mixing of all of the reagents.The apparatus is then replaced on the shaker and after a period of threeminufes the level of the fluid in the closed arm of the manometer is broughtagain to the 150-mm. mark and a reading of the level in the open arm, f,is recorded with the time g. The time g marks the end of the period ofCO2 production, g-a, and the "apparent" amount (see corrections 1 and3) of CO2 produced is represented as the change in level of f - e of experi-mental respirometer minus f - e of the thermobarometer which is substitutedfor h in the following formula. The volume of the respiring material isdetermined and allowance for this volume is made in the formula. At theend of the experiment the dry weight of the material is determined if prac-ticable in order to-place the results on a dry weight basis.

*Whenever a series of determinations is to be made a control respirometer(thermobarometer, check, or blank) must be subjected to the same condi-tions, reagents,.and treatment as the series except that it contains no respir-

310

BROWN: WARBURG RESPIROMETERS

ing material. This provides a correction for any temperature and externalpressure changes during the run, and for any CO2 which may be dissolvedin the reagents before the experiment.

The following formula is used to calculate the volumes of gas absorbedand produced by the respiring material.

FV273+ (x=h +Vf(a)]- 2r~~~=h

Powhere:

x =mm.3 of gas at S. T. P.h = height change in manometric graduations, b - e.Vg=free volume of gas in the vessel and manometer to the level of

manometric fluid (total volume of apparatus less the volume of thesample, liquids and detachable supports placed in the vessel). SeeDIXON (1).

T = absolute temperature of the water bath surrounding the vessel.Vf = the volume of all liquids in which the measured gas might dissolve.a = the solubility of the gas being measured in the contained liquids at

the temperature T. BUNSEN's solubility table has been recom-mended (4), Vf (a) gives the volume of the dissolved gas and thisis added to the free gas to give the- total volume (this expressionwas omitted by the writer for reasons presented later in this paper).

PO = normal pressure in terms of manometric fluid (for BRODME'S solu-tion, 10,000 mm.).

In experiments where a constant volume of material is to be used atconstant temperature, the quantity within the brackets in the above formularemainis constant for a given vessel and can be assigned a value which maybe referred to as the vessel constant. The vessels used in the present work

FIG. 1. Modified Warburg respirometer having a volume of approximuately 90 ml.

311

PLANT PHYSIOLOGY

had a capacity of about 90 ml. (fig. 1) and vessel constants of from 7 to 9whereas-the majority of vessels described in the literature are of about15-ml. volume and have correspondingly smaller vessel constants of about1 to 3. Where the vessel constant of an apparatus is 1 the reading of achange in pressure on the manometer may be transferred directly to termsof volume; but as the vessel constant increases, the change in pressure asshown by the manometer is smaller by the reciprocal of the vessel constant.When a vessel constant becomes as large as 10, the accuracv with which thereading may be interpreted becomes 0.1 as great as if the vessel constantwere 1, or each manometric graduation would represent an internal changeof 10 mm.3, whereas with a vessel constant of 1 it represents an internalchange of 1 mm.3

The -use of the large vessels has made evident certain factors that appar-ently have frequently been overlooked in calculating the results of deter-minations made with the smaller vessels. Some of these factors are equallyimportant, however, for small and large vessels.

CORRECTION FOR INTERFERENCE OF 02 CONSUMPTION WITH CO2PRODUCTION READINGS

In previous investigations involving the siinultaneous determination of02 absorption and C02 production with the Warburg apparatus- the factapparently has been overlooked that these two processes are continuouslyconcurrent. The usual technique (4, 5, 6) does not measure both processesfor the same period of time.

During most of a determination all C02 produced is absorbed by alkaliin the bottom of the vessels so any change of volume is caused by consump-tion of 02 by the respiring material. At the end of this period the manom-eter is read and the acid in the side arm mixed with the alkali to release theabsorbed C02. After several minutes the manometer is read again and allchange in pressure is attributed to the C02 produced bv respiration and thatalready in the reagents. Oxygen consumption, however, has continuedduring this period so in reality the change in pressure represents that pro-duced by C02 minus the 02 absorbed during the period required for mixingthe reagents and again bringing the temperature of the vessels into equi-librium with that of the bath. It is therefore quite evident that in order toobtain the actual volume of C02 produced the volume of 02 absorbed duringthe period of mixing and temperature adjustment must be added to theapparent volume of C02 and the thermobarometer correction applied. This

002)is particularly important where a low R. Q. (respiratory quotient = 2)

exists and also where the time interval is relatively long. Under ordinaryconditions the time interval required for mixing reagents and returning the

312

BROWN: WARBURG RESPIROMETERS

vessels to the temperature of the bath is 3 to 5 minutes which is 10 to. 15per cent. of the time for which 02 consumption is being measured. Thesituation is represented diagramatically in figure 2. The manometer read-

A B c D-10

-20

-30

-40E0~~~~~~~~~~~~~~~~!Z -50

(D-60

-70

2-80

0 5 10 IS 20 25 30 35 40MINUTES

FIG. 2. Suggested procedure compared with usual procedure for obtaining a truereading of CO2 production.

BH-Progress of 02 consumption during experiment. Measured at H in 26.5 minutes(BC) by line CH.

H-Point at which reagents were mixed to liberate CO2. Used by most workers as basefor obtaining CO2 readings.

E-Point at which CO2 production is measured by difference (vertical or EF) betweenbasic reading at H and finial reading at E.

EF-CO, production for 41 minutes measured by generally accepted method.HG-Extrapolation of BH showing estimated progress of 02 consumption while waiting

for reading of CO2 production at E.G-Point 02 consumption is estimated to have reached at time of reading E.

FG-Additionial 0, consumption while waiting to make reading E. Should be addedto EF to make correct CO2 reading.

G-Base which should be used instead of H for correct CO2 readings.JG-Length of time CO2 production is measured. Same as AD. Should be used as base

line (zero point) for measuring CO2 produced during the length of the experiment.JE-Progress of CO2 production. When used in conjunction with JG it is possible to

make readings of CO2 at any point during experiment.BG-Progress of 02 consumption which imiakes possible a determination of 02 consumed

at any time during the experiment.

ing for 02 consumed (fig. 2) in 26.5 minutes is 91 (CH) or approximately141 for 41 minutes. By the usual method the manometer reading for CO2production is 51 (EF) for 41 minutes but bv the suggested procedure be-comes 66.3 for the same period. Disregarding the solubilitv of these gasesfor reasons presented later, the R. Q.'s differ as follows:

313

PLANT PHYSIOLOGY

141 = 0.36 R. Q. by the usual procedure

141 = 0.47 R. Q. by the suggested procedure

CORRECTION FOR SOLUBILITY OF GASES

A fault of the usual technique exists in the method of correcting for thesolubility of gases in the reagents. Theoretically, the solubility of 02 andCO2 in the reagents should be considered, but in most experiments it is anegligible factor and correction for solubility by the usual method mayactually decrease the accuracy of the measurements. The usual proceduresuggests the use of Bunsen's solubility tables (3) but these tables representdeterminations using distilled water as the solvent and the gas at 760 mm.of pressure. Neither of these conditions holds for the respiration experi-ments, for the final solution in the respirometer is a salt solution in whichgases are less soluble than in pure water and the concentration of gases,especially CO2 is low. In the writer's experiments the concentration ofCO2 in the vessels did not exceed three times its concentration in the nor-mal atmosphere and its pressure would be only about 0.76 mm. or 0.1 percelnt. of that called for by standard solubility tables. The alkali used bythe writer for absorption of CO2 was 1.5 ml. of 2N NaOH and the acidused to liberate the CO2 was 0.6 ml. of 6N HCI. Actually there was muchmore CO2 in the alkali before it was placed in the vessels than in the saltsolution at the end of a determination, hence none of that produced inrespiration could be lost by absorption in the reagents.

Both reagents have been exposed to the 02 of the air hence are in equi-librium with the concentration of 02 existing in the vessels at the beginningof the experiment. Since only a very slight change in 02 pressure occursduring an ordinary determination, solubility of 02 may also be safelydisregarded.

If the vessel constants are corrected for solubility by using the standardsolubility tables they are increased. Thus, the product of the vessel con-stant and the manometric reading is also increased. It happens that theextent of the increase of CO2 production corrected on the basis of thesolubility tables in the usual procedure approaches the increase obtained bythe suggested procedure (fig. 2) and the final volume of CO2 found by bothmethods may be nearly equal in some cases. The 02 reading, however, isinvariably higher when the solubility tables are used in the usual procedure.The attempt to obtain more accurate results by correcting for solubility ofgases by these tables, therefore, actually decreases their accuracy. Forthese reasons the writer did not use the expression Vf (a) in the formulapresented earlier.

314

BROWN: WARBURG RESPIROMETERS

CORRECTION FOR DIFFERENCE BETWEEN VOLUMES OF THERMOBAROMETERAND OTHER VESSELS

In making corrections for pressure changes in the vessels as indicatedby the behavior of the thermobarometer certain additional corrections arenecessary.

During the period in which 02 absorption is measured there is no changein the mass and volume of gas contained in the thermobarometer eventhough it may be subjected to temperature and pressure changes. Thissituation holds for any empty respirometer. The manometric changecaused by external conditions for a period of time will be equal in allrespirometers regardless of size and content. This equality is evident whenthe respirometers are run empty but it still exists when they contain respir-ing material. The thermobarometer correction for pressure and tempera-ture changes during the period of 02 consumption can therefore be applieddireetly to the determination of changes in 02 volume of all respirometers.As the temperature decreases, or the external pressure increases, the mano-metric reading of the respirometers will decrease; and as the temperature in-creases or the external pressure decreases the manometric reading of therespirometers will increase. Thus, the thermobarometer correction may beapplied by either subtracting the decrease of the thermobarometer readingfrom, or adding the increase of the thermobarometer reading to, the readingof each of the experimental respirometers.

The correction for CO2 contained in the reagents cannot be applieddirectly to the other respirometers, however, unless they all contain equalvolumes of gas, because the introduction of equal volumes of gas into sys-tems of different volumes will produce different changes in pressure. If,for example, the total volume of one system is 500 ml. and the volume ofanother system is 100 ml., an addition of 10 ml. of gas to each of the systemswill cause the pressures of those systems to change. The extent of these

changes will be 10 and 10500 100'

10 x 760 = 15.2 mm. increase in pressure or 775.2 mm. total pressure.

100° x 760 = 76.00 mm. increase in pressure or 836.00 mm. total pressure.

Thus it is clear that the pressure change in the 500-ml. system is lessthan the pressure change in the 100-ml. system when equal volumes of gasare introduced into them.

Table I shows the relation between changes of manometric readings ofthe thermobarometer and the corresponding changes of each of the indi-vidual respirometers. This table is based on the values obtained by using

315

PLANT PHYSIOLOGY

TABLE IMANOMETRIC READINGS FOR EQUAL QUANTITIES OF GAS INTRODUCED INTO RESPIROMETER

SYSTEMS OF DIFFERENT VOLUMES

CONTROLRESPIROM- RESPIROM- RESPIROM- RESPIROM- RESPIROM-

(THERMO- ETER 1 ETER 2 ETER 3 ETER 4BAROMETER

Volume in mm.3 ...... 90512 94686 90475 88945 99885

Change in mano- r 30 28.68 30.00 30.53 27.18metric read- q 60 57.36 60.00 61.06 54.36ing in mm. l 90 86.04 90.00 91.59 81.54

the volume of the control respirometer as the numerator and the volume ofthe respirometer in question as the denominator.

As has been shown, the larger volume would be affected least by an addi-90512tion of a unit of gas, thus: -8= 0.906, showing that a change of 1 mm. of99885

the control manometer would be analogous to a change of 0.906 of the ma-nometer of respirometer 4. Of course, the larger the volume of gas whichis introduced into the systems becomes, the more conspicuous this differencebecomes, until a change of 90 mm. of the control manometer would indicatea change of only 81.54 mm. in the manometer of respirometer 4.

As the volumes of the vessels become smaller the differences between thevolumes of the control and the experimental respirometers must becomeproportionately smaller if the ratio between the volumes is to be kept near 1.Should the control respirometer and an experimental respirometer haverespective total volumes of 9 and 10 ml., the correction factor for the experi-

9mental respirometer would be- = 0.9, or practically the same as the correc-10 I

tion factor obtained in the preceding paragraph where the difference involume of the two systems is 10 ml. while in the present instance thedifference is but 1 ml.

It is important to note that a complete table of the type suggested bytable I cannot be constructed unless a constant volume of material is to beused. In such an instance the volume of the material and reagents shouldbe deducted from the total volume of the systems of the experimental respi-rometers. The volume of the reagents contained in the thermobarometershould also be deducted from its total volume.

Suppose, during the experiment, the control respirometer (10-ml. vol-ume) showed a manometric reading of 30 for CO2 and 1 ml. of the reagentsand respirometer no. 1 (12-ml. volume) showed a manometric reading of

316

BROWN: WARBURG RESPIROMETERS

80 for the C02 produced by 1 ml. of respiring material plus the amountcontained in 1 ml. of the reagents. The usual procedure has been to sub-tract 30 from 80 for a reading of the CO2 produced by the respiring mate-rial. The fallacy in this is shown by the following calculations.

The vessel constants of the systems are computed for 00 C. using theformula which has been given.

273(10-1) x27

10 = 0.9 for the control respirometer and

[12- (1+1)] 273

10 = 1 for respirometer no. 1.80- 30 = 50 manometric graduations of respirometer no. 1.50 x 1 (vessel constant of no. 1) = 50.0 mm.3 C02 by the usual method.A change of 30 graduations of the control, however, would be equal to

0.9 x 30 graduations of no. 1, therefore:80 - (0.9 x 30) = 53 manometric graduations of no. 153 x 1 = 53.0 mm.3 C02 by suggested procedure

or if one prefers to deal with actual volumes of gas instead of manometricgraduations the conversion for the same instance will be:

(80 x 1) - (30 x 0.9) = 53.0 mm.3 of C02.The difference shows

(53 50) x 100 = 5.66 per cent. too little C02 measured by the usual procedure.53In an experiment using a control respirometer of 92 ml. and an experi-

mental respirometer of 112 ml., suppose the control showed a manometricreading of 30 while the manometer of experimental respirometer no. 2,which contained 10 ml. of respiring material plus 2 ml. of reagents, read80 for the production of C02 at 0° C.

Then:

(92-2)x27310 = 9 the vessel constant for the control,

and

[112 - (10 + 2) ] x273x27310 - 10, the vessel constant for no. 2.

The conversion factor in this instance is the same as in the previous

experiment for90

= 0.9, therefore:

0.9 x 30 = 2780 - 27 = 53 manometric graduations of no. 253 x 10 = 530 mm.3 of CO2 by the suggested procedure

317

PLANT PHYSIOLOGY

and80 - 30 = 50 manometric graduations of no. 250 x 10 = 500 mm.3 of C02 by the usual procedure.

The difference shows

530500 x 100 = 5.66 per cent. too little C02 measured by the usual procedure.In this second instance the percentage difference is the same but the

amount of gas concerned is ten times as much as in the first instance. Itmay be repeated that the smaller the vessel constant of a system, the greateris the accuracy of manometric readings for that system. While the effect ofthe difference in volume between the control and experimental respirometersmay be considered negligible in some studies (when the vessels are of 10 to15 ml. capacity), in investigations where larger (100-ml.) respirometers areused the effect of this difference in volume would necessarily be consideredin order to obtain the greatest possible accuracy for the experiment.

SummaryA study of the operation of Warburg respirometers has been made in

connection with an investigation of the oxygen consumption and carbon

SAMPLE RECORD SHEET

EXPERIMENTAL RESPIROMETER THERMOBAROMETER OR CONTROL

Dry wt. of sample ingrams 10.4476

Volume of sample and/or reagents in ml. .... 17.6 2.1

Volume of system inexperiment in mm.3 87810 -17600= 70210 90512 -2100= 88412

TIME READING TIME READING

Set in bath ................. 9: 55 9: 55Seeds placed in vessel,

start CO ................. 10: 02 (a)Stopcock closed, start

02 .*----------------................. 10: 05 (c) 150.5 (b) 10: 05 (c) 149 (b)End 0,, dump HC1 ......... 10: 35 (d) 123 (e) 10: 35 (d) 151 (e)End C0, end of mixing 10:38 (g) 167 (f) 10: 38 (g) 181.5 (f)

EXPERIMENTAL CONTROL CORRECTIONRESPIROMETER FOR EXPERIMENTAL CONTROLGRADUATIONS RESPIROMETER

Change,02 ................. 150.5-123 = 27.5 2 151-149= 2Corrected 02 ................. 27.5 + 2 = 29.5

Change, CO2 ................. 167 - 123 = 44 80210 x 30.5 = 38.40 181.5 - 151 30.5

Plus 02 consumed in 3minutes ................. (44 - 38.40) + 2.95

Corrected CO2 ................. 8.55

318

BROWN: WARBURG RESPIROMETERS

CALCULATIONS

VESSEL CONSTANT (00 GC.)X MANOMETER x TIME - TOTAL GAS IN MM.3 (30 miN.)

87810- 17600 = 7.021 8.55 30 = 0.8333 50.02 mm.3 CO2

7.021 29.5 207.12 mm.3 02

Total CO2, 50.02Total 0o3R. Q. or 20712=0.2415

TOTAL GAS SAMPLE IN G. CORECTION = MM." OF GAS PER GM. DRY WT. PER HOUR

50.02 10.4476 2 = 9.58 mm.3 CO2

207.12 10.4476 2 = 39.64 mm.3 03

dioxide production of acorns. It was found that the accuracy of the methodcould be increased by modifying the usual technique in three ways. A cor-rection should be made for the absorption of oxygen during the intervalbetween the final oxygen reading and the final carbon dioxide reading.Theoretically, a correction should be considered for the solubility of thesegases in the. reagents before and after mixing, but since reliable solubilityvalues are frequently not available for these reagents more accurate resultsmay be obtained by omitting a correction for solubility. Differences involume of respirometers prevent the direct application of a control correc-tion for carbon dioxide in the reagents. A sample set of recordings andcalculations are presented.

The writer wishes to express sincere appreciation to Dr. P. J. KRAMER ofthe Department of Botany and to Dr. F. G. HALL of the Department ofZoology of Duke University for material assistance in the preparation ofthis paper.

DUKE FORESTDUKE UNIVERSITY

DURHAM, NORTH CAROLINA

LITERATURE CITED

1. DIXON, M. Manometric methods as applied to the measurement of cellrespiration and other processes. Cambridge Univ. Press. 1934.

2. HARRINGTON, G. T. Respiration of apple seeds. Jour. Agr. Res. 23:117-130. 1923.

3. HODGMAN, C. D. Handbook of chemistry and physics. 21st edition.Chemical Rubber Publishing Company, Cleveland, Ohio. 1936-1937.

319

320 PLANT PHYSIOLOGY

4. Loomis, W. E., and SHULL, C. A. Methods in plant physiology. McGraw-Hill Book Company, Inc., New York and London. 1937.

5. MURLIN, J. R. The conversion of fat to carbohydrate in the germinatingcastor bean. I. The respiratory metabolism. Jour. Gen. Physiol.17: 283-301. 1933.

6. TANG, P. S. On the respiratory quotient of Lupinus albus as a functionof temperature. Jour. Gen. Physiol. 15: 561-569. 1932.

7. WARBURG, 0. tber den Stoffwechsel der Tumoren. Berlin, JuliusSpringer. 1926.