Plant Physiol. (1982) 70, 1475-14790032-0889/82/70/1475/05/$00.50/0

Properties of Peach Flower Buds Which Facilitate SupercoolingReceived for publication May 4, 1982 and in revised form August 4, 1982

EDWARD N. ASHWORTHUnited States Department ofAgriculture, Agricultural Research Service, Appalachian Fruit Research Station,Kearneysville, West Virginia 25430

ABSTRACT

Water in dormant peach (Pws persica IL.i Batsch. var. 'Harbrite')flower buds deep supercooled. Both supercooling and the freezing of waterwithin the bud axis and pnmordium as distinct components depended onthe viability of the bud axis tissue. The viability of the prmordium was notcritical. Supercooling was prevented by wounding buds with a dissectingneedle, indicating that bud structural features were important. Bud mor-phological features appeared to prevent the propagation of ice through thevascular tissue and into the primordium. In dormant buds, procambial cellshad not yet differentiated into xylem vessel elements. Xylem continuitybetween the bud primordium and adjacent tissues did not appear to beestablished until buds had deacclimated. It was concluded that structural,morphological, and physiological features of the bud facDiltated supercool-ng.

Dormant flower bud primordia of several species avoid freezinginjury by low temperature supercooling. Supercooling has beenobserved in the buds of azalea (7), blueberry (2), grape (11),forsythia (10), Cornusflorida (16), and several Prunus species (13,15). When these species were exposed to lethal low temperatures,death of the bud primordium was associated with the suddenfreezing of a fraction of supercooled water. It has been shown bydirect observations (4, 14) and thermal analysis (2, 3, 7, 8, 13, 14)that water in the bud primordium supercooled even though icecrystals were present in the adjacent bud axis and scales. Theseobservations have led to proposals that a physical or thermody-namic barrier must separate the primordium from the rest of theplant tissues. This hypothetical barrier must exist to prevent thedirect growth of ice into the primordium and the sublimation ofsupercooled water to ice in the adjacent tissues (8, 14).George et al. (7) observed that both living and dead azalea

florets supercooled. Since exotherms produced by the freezing ofdead flower primordia were identical with those of living primor-dia, they concluded that the ability to supercool was based onstructural features of the tissue. In contrast, dead blueberry floretsdid not supercool (2). The authors concluded that supercoolingwas a function of cell membrane continuity and compartmentali-zation. It was surprising that two members of the same family,blueberry and azalea, behaved differently.

Studies on peach buds led Quamme (14) to conclude that thenucleation barrier was organized at the tissue level. He proposedthat, during freezing, water moved out of the bud axis and frozein the scales. This created a dry region in the bud axis which mayhave prevented the spread of ice into the primordium. Althoughlow temperature supercooling has been observed in several species,the nature of the nucleation barrier, which enables water in theprimordium to supercool, is unknown.The purpose of this study was to further characterize the

physiological and structural features of the peach bud whichfacilitate supercooling.

MATERIALS AND METHODS

Peach (Prunus persica [L.] Batsch. var. 'Harbrite') buds wereobtained from dormant 9-year-old trees at the West VirginiaUniversity Farm at Kearneysville between October 1981 and April1982. One-year-old twigs were harvested, put into plastic bags,and placed in an insulated box containing either crushed ice orsnow. The tissue was harvested daily and kept on ice until it wasprocessed. Buds were excised from twigs, so that a small portionof twig was left attached. Primordia were isolated by carefullyremoving the bud scales and excising the primordium from thesubtending axis tissue.Thermal Analysis. The freezing of water within buds was char-

acterized using both TA' and DTA. TA was a modification of apreviously described technique (1). The temperature of individualbuds was measured using a 36-gauge copper-constantan thermo-couple. Temperature was monitored using a Fluke 2200B DataLogger (John Fluke Manufacturing Co., Everett, WA)2 interfacedto a computer. The computer determined the exotherm initiationtemperature. Tissue was cooled at approximately -5 to -7°C/hby placing it inside Dewar flasks which were, in turn, put into amanually controlled deep freeze. All manipulations of buds weremade outside the laboratory building to prevent warming andminimize deacclimation. The DTA technique was a modificationof that described by Quamme et al. (12). Excised flower buds wereplaced in small aluminum foil containers along with the junctionof a 36-gauge copper-constantan thermocouple. Freeze-dried tis-sue was used as a reference. Output of thermojunctions wasmonitored with a strip chart recorder (0.5 mv/full scale). Sampleswere placed in a Kjeldahl flask partially submerged in an alcoholbath. The bath was cooled using a cold finger, and bath temper-ature and cooling rate were controlled using a heating elementand temperature programmer. Samples were routinely cooled at5°C/h unless otherwise indicated.

Pretreatments.. Buds were exposed to various pretreatments tonote the effect on subsequent freezing characteristics. For the dryice and heat treatments, twig pieces were wrapped in aluminumfoil and placed on either dry ice or in an oven at 65°C for 30 min.Tissue was then returned to room temperature. Buds were excised,and freezing was characterized using DTA and TA.

In the second pretreatment experiment, twig pieces werewrapped in aluminum foil, placed in Dewar flasks, and put intoa manually controlled deep freeze. Tissue temperature was mon-itored with a 26-gauge copper-constantan thermocouple. Samples

'Abbreviations: TA, thermal analysis; DTA, differential thermal anal-ysis.

2Mention of a trademark, proprietary product, or vendor does notconstitute a warranty of the product by the United States Department ofAgriculture and does not imply its approval to the exclusion of otherproducts or vendors that may also be suitable.

1475 www.plantphysiol.orgon August 22, 2020 - Published by Downloaded from Copyright © 1982 American Society of Plant Biologists. All rights reserved.

Plant Physiol. Vol. 70, 1982

were cooled between 5 and 7°C/h. When samples reached thespecified temperature, the Dewar flasks were removed from thefreezer and transferred to an incubator at 3°C. The followingmorning, buds were excised, and freezing was characterized byTA.

In the wounding experiment, a 32-gauge dissecting needle was

inserted through the cut surface of the excised bud deep enoughto impale the primordium. Wounded buds were subsequentlyanalyzed using both TA and DTA.Bud viability was determined visually. Following treatments,

samples were incubated for 2 d in an aluminum foil envelope atroom temperature. Buds were then bisected longitudinally andexamined for tissue browning.Dye Uptake Experiments. One-year-old twig pieces 15 to 20 cm

long were harvested and brought to the laboratory. The cut baseswere placed into a 1% (w/v) azosulfamide solution and incubatedat room temperature. At various time intervals, buds were bisectedlongitudinally with a scalpel and the location of the red dye was

observed using a dissecting scope.Bud Anatomy. Excised buds were vacuum infiltrated and fixed

with formalin-acetic acid, dehydrated with an alcohol-tertiarybutyl alcohol series, and imbedded in paraffin (9). Ten-,um sectionswere cut using a rotary microtome. Sections were stained usingsafranin-fast green.

RESULTS

DTA of dormant peach buds detected two distinct exothermsbetween 0 and -30°C when cooled at 5°C/h (Fig. 1). The firstexotherm was broad and occurred between -3 and -7°C. Thesecond exotherm appeared as a sharp spike on the recorder. Thisexotherm generally appeared between -15 and -25°C and variedseasonally. DTA of dissected flower buds demonstrated that thefirst broad exotherm was associated with the freezing of water inthe bud axis, and the second exotherm was associated with thefreezing of water within the primordium (Fig. 1). Peach buds andisolated primordia thawed near 0°C. No endothermic events wereobserved between -15 and -25°C (data not presented).

Effect of Cooling Rate. The extent of supercooling was influ-enced by cooling rate. Buds cooled at 2°C/h exhibited two distinctexotherms. Both the broad high temperature exotherm and thesharp low temperature exotherm were observed. When buds werecooled at 20°C/h, the results were in sharp contrast, and lowtemperature supercooling was not observed (Table I). Instead,exotherms appeared as a single peak or as two fused peaks. Ifbudswere first cooled at 2°C/h to -10°C and then subsequently cooledat 20°C/h, the freezing pattern was the same as that observedwhen buds were cooled exclusively at 2°C/h. Water in isolated

i c

I

-30-10 -20

TEMPERATURE, C

FIG. 1. DTA of peach bud tissues. A, Intact bud; B, isolated primor-dium; and C, isolated bud axis. All samples were cooled at 5°C/h.

Table I. Effect of Cooling Rate on the Freezing of Water within DormantPeach Buds

The freezing of water within peach buds was monitored using DTA.Excised buds were cooled at 2, 20, or at 2°C/h to -10°C and subsequentlyat -20°C/h.

X Exotherm Initiation Tempera-ture

Cooling RateFirst Second

exotherm exotherm

OC

2°C/h -3.5 -18.920°C/h -4.2 -5.320C/h to -10, then 20°C/h -3.5 -20.8

Table II. Effect of Heat and Freeze Treatments on the Ability of IntactBuds and Isolated Primordia to Supercool

Pretreatment Intact Buds Isolated Primordia

LTexo (oC)aControl 18.9b -18.2650C, 30 min -6.7 -14.7Dry ice, 30 min -6.4 -16.6a Mean low temperature exotherm initiation temperature.b Results are the mean of at least five replicates per treatment.

Table III. Effect of Temperature on the Subsequent Ability of Buds toSupercool

Twigs were cooled between 5 and 7°C/h to various temperatures,thawed, and stored overnight at 3°C. The following day, buds were excisedand the extent of supercooling was determined by thermal analysis. Therewas a minimum of 10 replicates per treatment.

Pretreatment x Exotherm

Exposure Primordia Bud axis Initiationtemperature viability viability Temperature

oc oC3 + + -18.9

-10 + + -19.4-25 - + -21.4-40 - - -6.2

primordia was observed to supercool when cooled at either 2 or20°C/h (data not presented).

Relationship of Bud Viability to Supercooing. Buds which werekilled by either a 65°C heat treatment or by freezing with dry icedid not low temperature supercool (Table II). Water in heat- andfreeze-killed buds no longer froze as two distinct components.Instead, exotherms appeared as a single peak or as two fusedpeaks. Heat treatment lowered the initiation temperature of thehigh temperature exotherm slightly (data not presented). In con-trast, primordia isolated from treated buds supercooled followingeither heat treatment or freezing on dry ice (Table II).The effect of freezing on supercooling was examined further by

cooling budwood between 5 and 7°C/h to -10, -25, or -40°C.Primordia were less hardy than the bud axis tissue (Table III).The -25°C pretreatment was lethal to peach primordia but not tobud axes. When exposed to -40°C, both the primordia and theaxes were killed. Primordia of buds frozen to either -10 or -25°Cwould subsequently supercool when refrozen (Table III). Super-cooling was not dependent on the viability of the primordia,inasmuch as a -25°C pretreatment killed the primordium but didnot prevent water in the dead primordium from subsequentlysupercooling. Exposure to -40(C was lethal to both the primor-

1476 ASHWORTH

www.plantphysiol.orgon August 22, 2020 - Published by Downloaded from Copyright © 1982 American Society of Plant Biologists. All rights reserved.

SUPERCOOLING IN PEACH FLOWER BUDS

-300 -10 -20TEMPERATURE, C

FIG. 2. Effect of wounding on the freezing of peach buds. A 32-gaugedissecting needle was inserted into the cut surface of an excised bud deepenough to enter the primordium. Buds were cooled at 50C/h and freezingwas monitored using DTA. A, Control; B, wounded bud.

FIG. 3. Movement of dye into peach buds. The cut ends of excised twigpieces were placed in 1% azosulfamide and incubated 7 d at room

temperature. Buds were bisected longitudinally and examined (x 45).

dium and the bud axis and prevented supercooling.Effect of Wounding. Low temperature supercooling was pre-

vented if buds were wounded with a dissecting needle. DTA ofwounded buds revealed only a high temperature exotherm (Fig.2). This exotherm appeared either as a single peak, a large peakwith a shoulder, or two fused peaks. Water in a wounded bud was

not observed to freeze as two distinct components.Dye Uptake Experiments. The water-soluble dye azosulfamide

FIG. 4. Vascular tissues of dormant peach flower bud. Tissue was

harvested and fixed in December. Procambium cells were elongated andcontained densely stained cytoplasm, and the cell walls lacked secondarythickenings (X 1000).

was translocated through the apoplast. Dye movement was pri-marily through the xylem, but movement through the intercellularspaces also occurred. Dye was observed in the axis of dormantbuds after only 1 d of incubation. Although the bud axis was

readily stained, dye was not observed in the primordia of dormantbuds, even after 1 week (Fig. 3). Dye was excluded from the budprimordium, even ifbuds were heat killed or exposed to metabolicinhibitors. Neither 100 mm sodium azide, 100 mg/ml cyclohexi-mide, nor 100 mg/ml actinomycin D enabled dye to enter theprimordium (data not presented).Dye was observed in the primordia of deacclimated buds.

Experiments conducted in March and April demonstrated thatdye was translocated into the primordia within 3 d. Dye was

observed in the base and walls of the primordium, the base of thepistil, and within the anther filaments.

Anatomical Observations. When examined in midwinter, dor-mant peach primordia have differentiated so that the stamens andpistils were easily recognized. An epidermal layer separated theprimordium from the bud scales. The top of the primordium was

enclosed by overlapping sepals and petals, and the base fused tothe bud axis. Parenchyma cells within the primordium were

characteristic of those observed in meristematic tissues (5). Cellswere small and tightly packed and had angular walls. Since thecells were tightly packed, little intercellular space was observedwithin the primordium. In contrast, the parenchyma cells in thebud axis were larger, more spherical, and the tissue containedmore intercellular space. However, no sharp delineation in cellsize at the zone where the primordium attached to the bud axis

L)xwx

0

B

A

1477

a A A

www.plantphysiol.orgon August 22, 2020 - Published by Downloaded from Copyright © 1982 American Society of Plant Biologists. All rights reserved.

Plant Physiol. Vol. 70, 1982

k.: .

FIG. 5. Vascular tissues of developing peach flower bud. Tissue was

harvested and fixed in mid-March during the bud swell period. Xylemvessel elements void of cytoplasm and with secondary cell wall thickeningswere observed (X 1000).

was observed. Also, no morphological or anatomical barriersseparating the primordium from the bud axis were observed.The vascular system in the dormant and developing flower bud

was observed using a combination of longitudinal and cross-

sections. Discrete bundles of vascular tissue were observed con-

necting the bud axis to the primordium. These bundles were

distributed throughout the bud primordium. They were observedin the developing sepals and petals and ran the length of thefilaments to the developing anthers. The cells in the vascularbundles of dormant buds were identified as procambial cells.These cells were elongated and contained densely stained cyto-plasm and no secondary wall thickenings were observed (Fig. 4).Xylem vessel elements were observed later in the spring in buds

that had visibly swollen. Vessel elements were easily distinguishedfrom procambium inasmuch as they lacked cytoplasm and hadcell walls with secondary thickenings (Fig. 5). The number ofxylem vessel elements increased as floral development proceeded.

DISCUSSION

Water in dormant peach buds froze as two distinct componentswhen cooled at 50C/h. Freezing occurred in the bud axis andscales a few degrees below0WC. Water in the primordium super-

cooled and froze between -15 and -250C. Quamme (13) firstdemonstrated that water in peach primordia supercooled. Heobserved that the melting and freezing points differed greatly andthe freezing point could be raised if the tissue water was inoculatedwith ice (14). He also demonstrated that the temperature at whichwater froze in the primordium was closely correlated to thetemperature at which this tissue was killed. Several lines of

evidence indicated that once ice formation was initiated in theprimordium, the entire tissue froze as a unit: (a) exotherms asso-ciated with the freezing of water in the primordium appeared assharp spikes on a DTA plot; (b) although considerable variabilityas to the freezing point of individual primordia existed, water ina single primordium froze over a very narrow range; and (c)injured primordia were never observed. Dormant primordia wereeither dead or alive following freezing treatment.

Intact peach buds killed by either a heat or freeze treatment didnot low temperature supercool. Water was no longer observed tofreeze as two separate components. It was concluded that heat andfreeze treatments destroyed some form of barrier, which in theliving bud prevents direct growth of ice into the primordium.Presumably, heat and freeze treatments disrupted cellular features,but had little effect on bud structure. Heat and freeze treatmentshad no effect on supercooling of isolated primordia. Furtherexperimentation demonstrated that pretreatment freezes whichkilled the primordium, but did not kill the hardier bud axis, wereineffective at preventing supercooling. Only treatments whichkilled both the primordium and the bud axis would preventsubsequent supercooling. It was concluded that the viability of theprimordium was not critical. Instead, a viable bud axis was aprerequisite for water in the primordium to supercool. This pre-sented an interesting situation where the properties of one tissueinfluenced the freezing of water in another.

Conflicting results have appeared as to the importance of budviability on the subsequent ability to supercool. Dead azaleaflorets were observed to supercool, and exotherms obtained whenthese tissues were frozen resembled those of live florets. Theauthors concluded that the ability to supercool resided in structuralfeatures of the bud (7). In blueberry, dead buds were not observedto supercool, and it was proposed that a continuous membranewas a prerequisite for supercooling. Interestingly, both of thesesituations were observed in peach. Buds which were frozen on dryice or heat killed were subsequently unable to supercool. However,buds exposed to -250C so that the primordium was dead, but thebud axis tissue remained alive, still supercooled. The results inpeach may clarify the conflicting results observed in other species.

Cooling rate has no effect on the extent of supercooling of waterdroplets (6). Likewise, cooling rate had little effect on the super-cooling of isolated primordia. However, it had a marked effect onthe freezing characteristics of intact buds. When buds were cooledat 20'C/h, water in the primordium would no longer supercool.However, if buds were cooled slowly to -10OC and then subse-quently cooled at 20'C/h, supercooling was observed. It seemedthat the rate of cooling was critical between 0 and -10IC.Although the rate of cooling had no effect on the temperature

at which water froze in the bud axis and scales, it may haveaffected where the water froze.Examinations of frozen peach budsclearly demonstrated that a redistribution of water had occurred.Large ice crystals were observed within the bud scales. This wasfirst documented by Dorsey (4) and later by Quamme (14).Quamme (14) proposed that, during freezing, water migrated tothe bud scales and froze. A dry region would be created betweenthe bud primordium and the ice in the bud scales and axis. Itseems plausible that rapid cooling of peach buds between 0 and-10'C prevented, or partially prevented, the redistribution ofwater. Instead, water froze in the bud axis before it could migrateto the bud scales.Based on the available data, it is proposed that ice formation is

first initiated in the bud axis and scales. Freezing begins extracel-lularly and water migrates to regions in the bud scales. Once thewater in the bud axis and scales has migrated to preferred sitesand frozen, the water in the primordium can supercool. Theredistribution of water appears to be a prerequisite for deepsupercooling. If either the bud axis cells are not viable or thecooling rate is too rapid, water in the bud will freeze as a single

1478 ASHWORTH

www.plantphysiol.orgon August 22, 2020 - Published by Downloaded from Copyright © 1982 American Society of Plant Biologists. All rights reserved.

SUPERCOOLING IN PEACH FLOWER BUDS

component and supercooling will not occur. Presumably, underthese conditions, the redistribution of water is prevented and icecrystals will propagate into the primordium and nucleate thetissue. Quamme (14) proposed that the redistribution of waterwithin the bud axis would leave a dry region between the ice inthe axis and scales and the liquid water within the primordium. Ihave no direct evidence of a dry region, but its existence isconsistent with these data.

Structural features of the bud were also critical. Wounding ofbuds with a dissecting needle prevented supercooling. Presumably,this structural disruption enabled ice forming in the bud axis tonucleate water in the primordium and prevent supercooling. How-ever, no structural barrier was apparent at the light microscopelevel. It has been proposed that the small cell size in flower budprimordia would be advantageous for low temperature supercool-ing. Cells in the primordia were observed to be smaller than thosein the bud axis tissue. Also, cells in the primordia were denselypacked and little intercellular space was observed. Despite theseanatomical differences, there was not a marked distinction in cellsize and intercellular spacing in the zone where the bud axis andthe primordium join.Water in xylem vessels of peach and other woody plants freezes

above -10°C. Therefore, it was predicted that xylem vessels couldnot form a continuous network connecting the primordium to theremainder of the plant. If a continuous network did exist, iceforming in the xylem vessels would nucleate the supercooled waterin the primordium. Dye uptake experiments demonstrated that, infact, the xylem was not continuous into the primordium. Dye wasnot observed to move up twigs and into the bud primordium ofdormant buds. We have observed 10 additional Prunus specieswhich low temperature supercool and found that in each instancedye was not taken up by the primordium (E. N. Ashworth and D.J. Rowse, unpublished data). Direct anatomical observations in-dicated that the vascular tissues in the bud axis and primordiumhad not yet differentiated. These vascular traces consisted ofprocambium cells. These cells had deeply stained cytoplasm andlacked secondary wall thickenings. Later in the spring, maturexylem vessel elements were observed. Differentiation of the pro-cambium proceeded during blossom development. The appear-ance of mature xylem vessels and the uptake of dye into theprimordium appeared to parallel the loss of hardiness. However,present evidence was insufficient to establish causality.

Several investigators have postulated that a barrier exists be-tween the supercooled water in the bud primordium and the ice

crystals in the bud axis, scales, and other tissues (7, 8, 14). Thisbarrier would prevent the nucleation of the supercooled water inthe primordium. Various types of barriers have been proposed (2,7, 8, 14). In peach, the hypothetical barrier appears to be acomposite of several features which enable the primordium tosupercool. These include cellular features of the bud axis tissue,structural features which allow for redistribution of water and theisolation of water in the primordium, and morphological featureswhich check the development of vascular elements.

Acknowledgments-The author wishes to thank Dianne Rowse and LindsayBillmyer for technical assistance, Barbara Rowland and Dixie Hammer for assistancein manuscript preparation, and Dr. Todd Cooke of the University of Maryland forhelpful discussions on plant anatomy and morphology.

LITERATURE CITED

1. ASHWORTH EN, GW LIGHTNER, DJ RowSE 1981 Evaluation of apricot flowerbud hardiness using a computer assisted method of thermal analysis. Hort-Science 16: 754-756

2. BIERMANN J, C STUSHNOFF, MJ BuRKE 1979 Differential thermal analysis andfreezing injury in cold hardy blueberry flower buds. J Am Soc Hort Sci 104:444-449

3. BuRKE MJ, C STUSHNOFF 1979 Frost hardiness: A discussion of the possiblemolecular causes of injury with particular reference to deep supercooling ofwater. In H MusselL RC Stapels, eds, Stress Physiology in Crop Plants. JohnWiley & Sons, New York, pp 199-226

4. DoRsEY MJ 1934 Ice formation in the fruit bud of peach. Proc Am Soc Hort Sci31: 22-27

5. ESAU K 1965 Plant Anatomy. John Wiley & Sons, New York6. FLErcmHR NH 1970 The Chemical Physics of Ice. Cambridge University Press,

Cambridge, England7. GEORGE MF, MJ BuRKE, CJ WEISER 1974 Supercooling in overwintering azalea

flower buds. Plant Physiol 54: 29-358. GEORGE MF, MJ BuRKE 1977 Supercooling in overwintering azalea flower buds.

Plant Physiol 59: 326-3289. JoHANsEN DA 1940 Plant Microtechnique. McGraw HilL New York10. Nus JL, JL WEIGLE, JJ SCHRADLE 1981 Superimposed amplified exotherm

differential thermal analysis system. HortScience 16: 753-75411. PIERQUET P, C STUSHNOFF, MJ BuRKE 1977 Low temperature exotherms in stem

and bud tissues of Vitis riparia Michx. J Am Soc Hort Sci 102: 54-5512. QuAsmM HA, C STUSHNOFF, CJ WEISER 1972 The relationship of exotherms to

cold injury in apple stem tissue. J Am Soc Hort Sci 97: 608-61313. QuAmME HA 1974 An exothermic process involved in freezing injury to flower

buds of several Prunus species. J Am Soc Hort Sci 99: 315-31714. QuAmwa HA 1978 Mechanism of supercooling in overwintering peach flower

buds. J Am Soc Hort Sci 103: 57-6115. RAJASHEKAR C, MJ BuRKE 1978 The occurrence of deep undercooling in the

genera Pyrus, Prunus, and Rosa: A preliminary report. In PH Li, A Sakai, eds,Plant Cold Hardiness and Freezing Stress. Academic Press, New York, pp213-225

16. SAKAI A 1979 Deep supercooling in winter flower buds of Cornus florida L.HortScience 14: 69-70

1479

www.plantphysiol.orgon August 22, 2020 - Published by Downloaded from Copyright © 1982 American Society of Plant Biologists. All rights reserved.