Plant Physiol. (1980) 66, 267-2710032-0889/80/66/0267/05/$00.00/0

Growth Responses and Adaptations of Fraxinus pennsylvanicaSeedlings to Flooding'

Received for publication November 7, 1979 and in revised form March 26, 1980

A. R. SENA GOMES2 AND T. T. KOZLOWSKIDepartment of Forestry, University of Wisconsin, Madison, Wisconsin 53706

ABSTRACT

Flooding induced several physiological and morphological changes inFraxinus pennsylvanica seedlngs, with stomatal closure among the earliestresponses. Subsequent changes included: reduction in dry weight incrementof roots, stems, and leaves; formation of hypertrophied lenticels andproduction of adventitious roots on submerged portions of the stem abovethe soil line; leaf necrosis; and leaf abscission. After 15 days of stomatalclosure as a result of flooding, stomata began to reopen progressively untilstomatal aperture was similar in flooded and unflooded plants. Adventitiousroots began to form at about the time stomatal reopening began. As moreadventitious roots formed, elongated, and branched, the stomata openedfurther. The formation of adventitious roots was an important adaptationfor flooding tolerance as shown by the high efficiency of adventitious rootsin absorption of water and in high correlation between the production ofadventitious roots and stomatal reopening.

Several investigators have emphasized that Fraxinus pennsyl-vanica Marsh. is highly tolerant of flooding. This species occurson alluvial soils along rivers and brooks. It is commonly found onland that is subject to periodic flooding and grows vigorously evenwhen flooded for much of the growing season (2, 7, 17).The high flooding tolerance of certain species of forest trees has

been attributed to one or more adaptive mechanisms, includingcompensation for poor aeration of the normal root system byproduction of AR. Often species that survive flooding best arethose that form AR3 at, or below, the water line (5, 6, 14, 20, 21).Although some investigators considered such roots to be merelysymptoms of flooding stress (7), others concluded that the capacityof some species to survive flooding depended wholly or partly onthe activity ofAR (1, 1 1). Clemens et al. (4) showed that floodingtolerance of three species was in the following order: Eucalyptusgrandis > Eucalyptus robusta > Eucalyptus saligna and that thisorder was correlated with capacity for production of AR.Our previous experiments (18) indicated that stomatal closure

of several species ofwoody plants was one of the earliest responsesto flooding. There is some evidence (13) that stomata of F.pennsylvanica may reopen after a critical period of flooding. Withthese considerations in mind, further experiments were conductedon growth responses of flooded F. pennsylvanica seedlings, effects

'This research was supported by: the College of Agricultural and LifeSciences, University of Wisconsin, Madison; CEPEC (Cocoa ResearchCenter), Bahia, Brazil; and EMBRAPA (Brazilian Research Institute),Brasilia, Brazil.

2Present address: CEPEC, Caixa Postal 7, Itabuna, Bahia, Brazil.3Abbreviations: AR: adventitious roots; r,: leaf diffusion resistance;

WAR: without adventitious roots.

of flooding on stomatal aperture, the importance of flooding-induced AR in absorption of water, and timing of AR productionin relation to stomatal aperture.

MATERIALS AND METHODS

Dry Weight Increment. F. pennsylvanica seedlings were grownfrom seed in the greenhouse in sand-loam (1:3) contained in 15.5-cm pots. Thirty plants of each of two different ages, 8 and 10weeks old, were randomly separated into three groups of 10 plantseach: (a) plants for initial determination of seedling dry weight,(b) unflooded control plants, and (c) plants to be flooded.

Initial dry weights were obtained separately for leaves, stems,and roots after drying at 70 C for 48 h. The plants of groups b andc were transferred to a growth chamber in which PAR was 226pE m2 s ' 20 cm above the pots; daylength was 16 h beginning at0700 h; day and night temperatures were 25 and 20 C, respectively;and RH was approximately 80%Yo.

After 11 days in the growth chamber, plants of group c (10 ofeach age class) were flooded by immersing the pots in tubs ofwater for 30 days. The water level was maintained at 2 cm abovethe soil line. Unflooded control plants were watered daily. Threedays after termination of flooding, both flooded and unfloodedplants were harvested. Roots, stems, and leaves of each plant wereseparated and their dry weights were determined.

Leaf Diffusion Resistance. We monitored r for plants used inthe previous experiment. Environmental conditions were as pre-viously described. For the 10 flooded and 10 unflooded plants ofeach age class, ri was determined for the abaxial surface of onefully expanded leaf per plant. Measurements of ri were initiated1 week after the plants were transferred to the growth chamber. ALambda porometer was used to monitor ri daily for 3 days beforeflooding, for 30 days during flooding, and for 3 days after floodingwas discontinued. During the 3 days before flooding, ri wasmeasured once daily at 0800 h. Thereafter, measurements weremade eight times daily at 2-h intervals beginning at 0800 h. Theporometer was calibrated in the growth chamber at 25 C by theprocedure described by Kanemasu et al. (12).

Effect of Adventitious Roots on Transpiration. Sixty F. penn-sylvanica seedlings were grown from seed as described above untilthey were 4 months old. Then 50 plants were immersed in tubs ofwater. The water level was maintained at 2 cm above the soil line.In addition water was replaced regularly at 2-week intervalsthroughout the flooding period. Environmental conditions were:minimum day and night temperatures, 25 and 22 C, respectively;daylength, 15 h (beginning at 0700 h) with incandescent lamps,150 w each; RH, 56 to 80%.After 42 days of flooding, approximately half of the flooded

plants had developed adventitious roots on the submerged part ofthe stem above the soil line. Six flooded plants with well-developedAR and six WAR were removed from the tubs and the soil wasallowed to drain for 3 days on a greenhouse bench. On thefollowing day, transpiration measurements were initiated for each

267

Dow

nloaded from https://academ

ic.oup.com/plphys/article/66/2/267/6077173 by guest on 02 January 2022

GOMES AND KOZLOWSKI

seedling by the gravimetric method (15, 16). Plastic bags were

placed around each pot and secured to the lower plant stem. Thisarrangement prevented direct water loss from the soil. The plantplus pot system was weighed daily at 1130 h for 5 days. Dailywater loss was replaced by adding water to the soil through a

watering tube.In both AR and WAR plants, r, was measured on three fully

expanded leaves per plant once only on the 5th day just beforethe daily transpiration rate was determined. Temperature in thegreenhouse was 27 C and RH was 46%.

After transpiration rates were determined in the greenhouse,AR and WAR plants were moved to a growth chamber foradditional measurements of transpiration rates. Conditions in thegrowth chamber were: day and niFht temperatures, 27 and 23 C,respectively; PAR, 283 ,IE m2 s ; daylength, 16 h beginning at0700 h; RH, 55 to 70%o. The plants were acclimated to growthchamber conditions for 3 days and were watered daily. Dailytranspiration rates then were determined gravimetrically begin-ning at 1130 h for 7 consecutive days. All AR roots of AR plantswere severed from the stem on the 3rd day. Two r1 measurementswere taken daily on the abaxial surface of four fully expandedleaves per plant, beginning at 0900 and 1400 h, for the first 6 days.No new leaves developed during the greenhouse or growth

chamber experiments. After daily transpirational losses were de-termined, the leaves were harvested and leaf surface areas were

determined for each seedling. Leaf areas were calculated byreproducing the leaf outlines on photosensitive paper and thencutting and weighing them. Leaf area was obtained by the ratio ofweight to area of the paper.

Absorption of Water by Normal and Adventitious Roots. Thenormal (nonadventitious) root system and each system of flooding-induced AR of each of three 4-month-old intact plants that hadbeen flooded for 70 days were inserted separately into potometersto measure their efficiency in absorption of water. A potometersystem consisted of two pieces of vertical glass tubing held by a

clamp attached to a support. The two glass units were connectedat the basal end by U-bent glass tubing 7 mm in diameter insertedthrough rubber stoppers. The original root system was placed ina large potometer (glass tube: 25 cm long, 2.2 cm in diameter)connected to a 50-ml burette. Smaller potometers (glass tube: 15cm long, 1.3 cm in diameter) connected to a 30-ml burette accom-odated each AR system that emerged from the stem. For eachseedling, three to four of the small potometers were used, depend-ing on the number ofAR. Potometers were calibrated by removingknown volumes of water from the glass tube and recordingcorresponding changes in water volume in the burette.The potometer plus plant systems were kept in a growth cham-

ber under the following environmental conditions: PAR, 305 ,uEm-2 s-'; daylength, 16 h beginning at 0700 h; temperatures, 21and 20 C during the day and night, respectively; and RH, near80%.The experimental plants were acclimated to growth chamber

conditions for 3 days. Beginning on the 4th day, absorption ofwater by original and AR was measured separately, twice daily at0800 and 1400 h, for 8 consecutive days. At the end of theexperiment, dry weights of the original roots and AR (< 1.00 mmin diameter) were determined separately.

RESULTS

Morphological Changes. By the 5th day of flooding, hypertro-phied lenticels and aerenchyma tissue were observed on thesubmerged portion of the stem above the soil line. After 15 daysof flooding, a few thick, white AR had grown through thesehypertrophied lenticels. After 30 days of flooding, seedlings ofboth age classes had developed additional extensively branchedAR on the submerged portion of the stem above the soil line.Similar AR developed in 4-month-old plants that were flooded

for 42 or 70 days (Figs. 1 and 2; Table I). Flooding also inducedsome marginal leaf necrosis and caused abscission of a few leaves.The original roots of flooded plants darkened to an almost blackcolor, were sparsely branched, and had many dead tips.Dry Weight Increment. Flooding significantly reduced dry

weight increment of whole plants and various plant parts of the 8-and 10-week-old seedlings (Table II). However, the rate of dryweight increase was reduced more in the younger plants. In both

FIG. 1. Hypertrophied lenticels on submerged stem portions of 4-month-old F. pennsylvanica plants after 7 days of flooding. The arrow

indicates the height to which the stems were flooded.

FIG. 2. AR and hypertrophied lenticels on submerged stem portions of4-month-old F. pennsylvanica plants after 30 days of flooding. The arrowindicates the height to which the stems were flooded.

Table I. Weights ofNormal Root Systems and ofARThe dry weights of plants were determined on submerged portions of

stems after 70 days of flooding of 4-month-old F. pennsylvanica seedlings.% of Root Sys-

temNormal Rootsa AR on Stema

Normal ARRoots

Dry weight (g) 2.0088 + 0.3472 0.3494 ± 0.0550 85 15Dry weight of

roots up to 1mm diameter(g) 0.6295 ± 0.1789 0.1702 ± 0.0003 79 21a Values are ±SEM.

268 Plant Physiol. Vol. 66, 1980

Dow

nloaded from https://academ

ic.oup.com/plphys/article/66/2/267/6077173 by guest on 02 January 2022

FLOODING OF FRAXINUS PENNSYLVANICA

ages of plants, dry weight increment was reduced most in roots.Hence, root to shoot ratios of both age classes were also greatlyreduced by flooding.

After 4-month-old seedlings were flooded for 70 days, the ARthat developed on the submerged portion of the stem comprised15% of the dry weight of the total root system and 21% of the dryweight of the roots up to 1 mm in diameter (Table I).Leaf Diffusion Resistance. During the 3 days before flooding,

r1 at 0800 h was low and similar for both flooded and unfloodedseedlings of both age classes. Thereafter, in unflooded plants, r,values remained consistently low and did not exceed 10 s cm-'during the 37 days of the study (Fig. 3).

Flooding greatly increased r1 within 1 day in 8-week-old plantsand within 2 days in 10-week-old plants. The ri values of floodedplants of the younger age class increased rapidly from near 5 scm-l to more than 15 s cm-' (1200-h readings) within 24 h. Highdaily ri maxima of 27 to 35 s cm-' were recorded from the 7th tothe 15th day of flooding. Beginning on the 15th day of flooding,r, values began to decline progressively until, by the 30th day offlooding, they approximated those of unflooded plants.

Diurnal changes in ri occurred consistently in both flooded andunflooded plants, regardless of age class. In flooded 8-week-oldplants, the diurnal change was marked and, on some days, r1

fluctuated by as much as 20 s cm-'. After 15 days of flooding thediurnal amplitude in ri decreased appreciably. In 10-week-oldflooded seedlings, diurnal variation in r1 was only about 7 s cm-'.In unflooded plants, diurnal changes in ri were much lower,usually about 2 to 4 s cm-'. In flooded plants, ri was lowest at0800 and 1600 h and highest between 1200 and 1400 h. Bycomparison, in unflooded plants the lowest values occurred atnoon.The r1 values were significantly different for flooded and un-

flooded plants, except for a few days in the older seedlings. Dailyr, values of the younger seedlings were significantly different forflooded and unflooded plants (0.01 level) from the 2nd to the 21stday of the experiment.

Transpiration Rates. Transpiration rates per unit of leaf areawere significantly higher for 4-month-old seedlings with welldeveloped AR than for those that did not develop adventitiousroots (WAR) during 42 days of flooding (Figs. 4 and 5). In ARplants, transpiration rates varied from 0.75 to 0.84 g dm 2 dar-I;in WAR plants, they varied from 0.40 to 0.55 g dm-2 day in

both greenhouse and growth chamber experiments.Leaf diffusion resistance (ri) of AR plants was consistently

lower than in WAR plants at both 0900 and 1400 h (Fig. 5).Although ri readings in the greenhouse were taken once only, theywere also higher in WAR plants (e.g. AR = 4.7, WAR = 9.8 scm-2), suggesting that stomata were more open in plants with AR.

Leaf surface areas and leaf dry weights ofAR and WAR plantsdid not differ significantly. There were no significant differencesfor r, values taken at 0900 and 1400 h within the same treatmentin the growth chamber experiment. Transpiration decreasedslightly for 2 days after the AR were severed from the stem (Fig.

5), but the decrease was not statistically significant.Absorption of Water. After 4-month-old seedlings were flooded

for 70 days, both normal and AR contributed to total wateruptake. The rate of absorption by AR was slightly higher thanabsorption by the original roots for the first 4 days and slightlylower thereafter (Fig. 6). However, the differences in rates were

not statistically different for the two types of roots.

DISCUSSION

Flooding of F. pennsylvanica seedlings induced several physio-logical disturbances, including: stomatal closure; inhibition ofgrowth of leaves, stems, and roots; formation of hypertrophiedlenticels with aerenchyma tissue on submerged portions of thestem; production of functional AR; and some leaf necrosis andabscission..Flooding induced stomatal closure. Stomata began to close

within I or 2 days after plants were flooded and they continued toclose progressively during the next 7 days. For the following 7days, the degree of stomatal closure remained about the same.Large diurnal variations in stomatal aperture were superimposedon this pattern.

Stomatal reopening began after about 15 days of flooding andcontinued during the next 2 weeks. After 30 days of flooding, thestomata had reopened sufficiently so that leaf resistance was onlyslightly higher in flooded than in unflooded plants. These obser-vations are consistent with those of Kozlowski and Pallardy (13).Induction of stomatal closure of flooding, followed by stomatalopening after a critical period of flooding, has also been reportedfor Populus deltoides (19).

In the present study, some stomata may have been permanentlyinjured by flooding. When stomata began to reopen after 15 daysof flooding, leaf resistance of flooded plants remained slightly butconsistently higher than in unflooded plants for the next 14 daysof flooding and also throughout the 4-day postflooding period.Kozlowski and Pallardy (13) found that, after appreciable reopen-ing of stomata in previously flooded F. pennsylvanica seedlings,leaf resistance remained slightly higher than in unflooded plantsfor at least 17 days, further suggesting that some stomata wereinjured by flooding and failed to open after flooding was discon-tinued.The mechanism of rapid stomatal closure by flooding has not

been explained but appears to be more complex than merely theresult of leaf dehydration. In both short- and long-term experi-ments, flooding rapidly induced stomatal closure in several speciesof woody plants but did not decrease water potential of leaves(18). Furthermore, Populus deltoides plants remained turgidthroughout 28 days of flooding. Decreased absorption of waterdue to increased root resistance was offset by stomatal closure,thereby maintaining leaf turgor (19). Because flooding rapidlyclosed stomata of F. pennsylvanica plants without inducing waterstress in leaves, we speculated that a high CO2 content in substo-matal cavities of flooded plants and/or production ofABA by leaf

Table II. Effects of Floodingfor 30 Days on Dry Weight Increment of Leaves, Stems, and Roots of F. pennsylvanica Seedlings of Two Age Classes

Dry Weight IncrementRoot/Age Treatment % of %ofoot

Leaves % f Stems % f Roots % of Ttl % of ShootControl Control Control Total Control

weeks ratio8 Control 2.47 0.14 2.84 ± 0.04 2.19 ± 0.20 7.51 ± 0.28 0.41

Flooded 0.98 ± 0.05a 39.6 0.94 ± 0.09a 33.0 0.36 ± 0.03a 16.4 2.29 ± 0.16a 30.4 0.1910 Control 3.95 ± 0.31 3.31 ± 0.29 3.76 ± 0.29 11.03 ± 2.43 0.53

Flooded 2.39 ± 0.34a 60.5 2.60 ± 0.28h 78.5 1.21 ± 0.15a 32.1 6.22 ± 0.74a 56.3 0.26aValues significantly different from control (P = 0.01).b Not significant.

Plant Physiol. Vol. 66, 1980 269

Dow

nloaded from https://academ

ic.oup.com/plphys/article/66/2/267/6077173 by guest on 02 January 2022

GOMES AND KOZLOWSKI Plant Physiol. Vol. 66, 1980

Start ot Lenticels Adventitious End of

flooding roots flooding

35,

25..

Untlooded

°-

DAYS BEFORE OR AFTER FLOODING

FIG. 3. Effect of flooding on daily changes in leaf diffusion resistance (r,) of 8-week-old F. pennsylvanica seedlings.

0.9

E

z

0Q-

co

z

cc

0.8F

0.7k

0.6

0.51-

1 2 3

DAYS4 5

FIG. 4. Transpiration rates and standard errors, after 42 days of flood-ing, of 4-month-old F. pennsylvanica seedlings that had developed AR on

submerged portions of stems and plants that had not developed adventi-tious roots. Greenhouse experiment.

cells may have caused the stomata to close. Further experimentsare needed to determine the mechanism of stomatal closure fol-lowing flooding.The formation of AR appeared to be an important adaptation

for flooding tolerance as shown by two lines of evidence: (a) ARwere very efficient in absorption of water; and (b) production ofAR was correlated with stomatal reopening in flooded plants.The high efficiency of AR in absorption of water was demon-

strated by the much higher transpiration rates in plants with ARthan in those without them. The potometer experiments showedthat water uptake of flooded plants on a leaf area basis wasapproximately 80 to 90% higher in plants with AR than in thoselacking such roots.The observation that severing of AR from the stem did not

greatly influence transpiration rates might appear to deemphasizethe importance of AR in water absorption of flooded plants.Clarkson and Sanderson (3) showed that root pruning had littleeffect on the amount of water transpired by barley plants becausewater flux through the remaining roots was increased. Compen-

12.0

.10.0E

0.;2 8.0

6.0

4.0

, 0.8

E

0.7

la

z

0

Co

z

cc

vP. 1

0.6

0.5

0.4

1 2 3 4 5 6 7

DAYSFIG. 5. Leaf diffusion resistance (r,) and transpiration rates and stand-

ard errors, after 42 days of flooding, of 4-month-old F. pennsylvanicaseedlings that had developed AR on submerged portions of stems andplants that had not developed adventitious roots. Growth chamber exper-

iment.

satory water absorption by the remaining roots also apparentlyoccurred in the present study because transpiration was reducedslightly for 2 days after the AR were excised and then recovered.The pattern of stomatal reopening, beginning approximately 15

days after flooding, was closely correlated with production andgrowth of AR. Stomata, which had closed rapidly in response toflooding, began to reopen when AR first appeared and theycontinued to open over a 2-week period during which more ARformed, elongated, and branched, thereby progressively increasingthe effective absorbing surface for water and minerals. Whenstomatal aperture was compared for flooded AR and WAR seed-lings, stomata were more open in the former group.

270

Plants withadventitious roots

i

l_-9AA.M._I ~~~~~Plants without

adventitious rootsPlants with

_ adventitiousroots .2 P.M.

adventitiouadvetitousooot

I

_adventitious rootsitius oot

0.4~

I

Dow

nloaded from https://academ

ic.oup.com/plphys/article/66/2/267/6077173 by guest on 02 January 2022

FLOODING OF FRAXINUS PENNSYLVANICA

pe?

re(

ica

flath

ca

waU'flcan

tolof

75.0 LLITLKATUKL Ul'IILU

1. ARMSTRONG GW 1968 Oxygen diffusion from the roots ofwoody species. PhysiolPlant 21: 539-543

65.0 .2. BROADFOOT WH, HL WILLISTON 1973 Flooding effects on southern forests. JFor 71: 584-587

3. CLARKSON DT, J SANDERSON 1970 Relationship between the anatomy of cereal55.0 1 roots and the absorption of nutrients and water. Agric Res Counc (GB)

Letcombe Lab Annu Rep 16-25.4. CLEMENS J, AM KIRK, PD MILLS 1978 The resistance to waterlogging of three

45.0 Normal roots Eucalyptus species, effect of flooding and on ethylene-releasing growth sub-stance on E. robusta, E. grandis, and E. saligna Oecologia (Berl) 34: 5-131

5. GILL CJ 1970 The flooding tolerance of woody species. A review. For Abstr 31:671-688

6. GILL CJ 1975 The ecological significance of adventitious rooting as a response toflooding in woody species, with special reference toAlnusglutinosa (L.). GaertnFlora 164: 85-97

25.0 ~~~~~~~~~~~~~~~~7.HALL TF, GE SMITH 1955 Effect of flooding on woody plants. West sandyAdventitious roots dewatering project, Kentucky Reservoir. J For 53: 281-285

8. HOOK DD, OG LANGDON, J STUBBS, CL BROWN 1970 Effect of water regimes15.0 on the survival, growth, and morphology of Tupelo seedlings. For Sci 16: 304-

3119. HOOK DD. P KORMANIK 1971 Inductive flood tolerance in swamp tupelo (NJyssa

5.o1 sylvatica var. biflora (Walt) Sarg.). J Exp Bot 22: 78-891 2 3 4 5 6 7 8 10. HOOK DD, RH WETMORE 1972 Aeration in trees. Bot Gaz 133: 443-454DAYS 11. HOSNER JF, SG BoYCE 1962 Tolerance to water saturated soil of various

FIG. 6. Rates of water uptake by original and AR of 4-month-old F. bottomland hardwoods. For Sci 8: 180-186nnsylvanica seedlngs after 70 days of flooding. 12. KANEMASU ET, GW THURTELL, CG TANNER 1969 Design, calibration, and field

use of a stomatal diffusion porometer. Plant Physiol 44: 881-88513. KOZLOWSKI TT, SG PALLARDY 1979 Stomatal responses of Fraxinuspennsylvan-

ica seedlings during and after flooding. Physiol Plant 46: 155-158The mechanism by which activity of AR induced stomatal 14. KRAMER PJ 1951 Causes of injury to plants resulting from flooding of the soil.Opening in flooded plants has not been clarified. The physiolog- Plant Physiol 26: 722-736li role of AR in flooding tolerance of some species involves 15. KRAMER PJ 1969 Plant and Soil Water Relationships: A Modern Synthesis.1roleofAmfloodgtolerace of soe spec mvolves McGraw-Hill, New Yorkore than merely increasing the water absorbing surface. In 16. KRAMER PJ, TT KoZLOWSKI 1979 Physiology ofWoody Plants. Academic Press.)oded Nyssa sylvatica var. biflora plants AR not only increased New Yorke water absorbing surface but were additionally beneficial be 17. LouCKS WL, RA KEEN 1973 Submersion tolerance of selected seedling trees. J

For 71: 496497use they occurred in the upper flood water where the 02 content 18. PEREIRA JS. TT KoZLOWSKI 1977 Variations among woody angiosperms inis higher and toxic compounds were lower than in the soil. response to flooding. Physiol Plant 41: 184-192derflooding,the AR oxidized the rhizosphere, whereas un- 19. REGEHR DL, FA BAZZAZ, WR BOGGESS 1975 Photosynthesis, transpiration, and

rider flooding, the AR oxldlzed the rhlzosphere whereas un- leaf conductance of Populus deltoides in relation to flooding and drought.oded roots did not. The combined adaptations of increased Photosynthetica 9: 52-61aerobic respiration, oxidation of the rhizosphere, and high 20. ROWE RN, DV BEARDSELL 1973 Waterlogging of fruit trees. Hortic Abstr 44:leraneOf02bytheew ARappeaed toaccont fo toleance 534-548lerance of CO2 by the new AR appeared to account for tolerance 21. YELENOSKY G 1974 Tolerance of trees to deficiencies of soil aeration. Proc 40thflooding by this species (8-10). Int Shade Tree Conf: 127-147.

t

E

0o:0oo

Plant Physiol. Vol. 66, 1980 271I fTr D A Irl TDn Th''I

-j ---- x- - -1.

Dow

nloaded from https://academ

ic.oup.com/plphys/article/66/2/267/6077173 by guest on 02 January 2022