KETELLAPPER-ENDOGENOUS AND ENVIRONMENTAL PERIODS IN GROWTH

It has been concluded that peanut and tomatopossess an endogenous, time-measuring mechanism,which is slightly temperature dependent, and that foroptimal development the external period must be syn-chronized with the endogenous period of the plant.

ACKNOWLEDGEMENTSThe author wishes to acknowledge the advice of

Dr. James Bonner and Dr. Frits W. Went. He isindebted to Dr. Walton C. Gregory (North CarolinaState College) for providing the peanut seeds.

LITERATURE CITED1. BPNNING, E. 1931. Untersuchungen iiber die au-

tonomen tagesperiodischen Bewegungen der Prim-arblitter von Phaseolus multiflorus. Jahrb. wiss.Bot. 75: 439-480.

2. GARNER, W. W. and H. A. ALLARD 1931. Effectof abnormally long and short alterations of lightand darkness on growth and development of plants.Jour. Agr. Res. 42: 629-651.

3. GRIESEL, W. 0. and J. B. BIALE 1958. Respiratorytrends in perianth segments of Magnolia grandi-flora. Amer. Jour. Bot. 45: 660-663.

4. HIGHKIN, H. R. and J. B. HANSON 1954. Possibleinteraction between light-dark cycles and endo-genous daily rhythms on the growth of tomatoplants. Plant Physiol. 29: 301-302.

5. HILLMAN, W. S. 1956. Injury of tomato plants bycontinuous light and unfavorable photoperiodiccycles. Amer. Jour. Bot. 43: 89-96.

6. NANDA, K. K. and K. C. HAMNER 1959. Theeffect of temperature, auxins, antiauxins, and someother chemicals on the endogenous rhythm affectingphotoperiodic response of Biloxi soybean (Glycinemax. L.; Merr.). Planta 53: 53-68.

7. RICHTER, G. and A. PIRSON 1957. Enzyme vonHydrodictyon und ihre Beeinflussung durch Beleu-chtungsperiodik. Flora 144: 562-597.

8. VERKERK, K. 1955. Temperature, light, and thetomato. Mededel. Landbouwhogeschool Wagenin-gen 55: 176-224.

9. WENT, F. W. 1957. The Experimental Control ofPlant Growth. Chronica Botanica Co., Waltham.

RELATIONSHIP OF GROWTH AND DEVELOPMENT TO CHANGES IN SUGARS,AUXINS, AND GIBBERELLINS IN FRUIT OF SEEDED AND

SEEDLESS VARIETIES OF VITIS VINIFERA 1,2

B. G. COOMBE3DEPARTMENT OF VITICULTURE AND ENOLOGY, UNIVERSITY OF CALIFORNIA, DAVIS

There are large differences in the degree of seeddevelopment in different varieties of Vitis vinifera(European grape) ranging from nil in parthenocarpicvarieties to complete seed development from severalof the four ovules in the berries of certain othervarieties. A few varieties are "seedless" because oftheir stenospermocarpic fruit development, i.e., theirseeds begin to develop after fertilization but they abortbefore the development is complete. These differ-ences in seed development are reflected in growth ofthe berries. This study was made to relate the de-velopment and growth of the fruit of several varietiesto changes in their sugar, auxin, and gibberellin con-tents between anthesis and maturity.

The morphological changes in fruit of a few seededand seedless varities of grapes have been described(16, 20, 24). Fruit morphogenesis was measured in

Received July 23, 1959.2 Portion of a thesis submitted in partial satisfaction

of requirements for the Ph.D. degree in the GraduateDivision, University of California, Davis, under the direc-tion of Dr. R. J. Weaver.

8 Present address: Waite Agricultural Research Insti-tute, Adelaide, S. Australia.

a semi-quantitative way in the varieties used here tohelp interpret the other factors measured. Fruitgrowth studies of grapes are not extensive. Mosthave the disadvantage that they are based on fruitdiameter measurements (33). In this study freshweight, dry weight and volume of berries, and sugarconcentration in the juice were measured.

The most detailed work on auxins in grapes hasbeen done by Nitsch and coworkers. They demon-strated the presence of several types of auxins innearly mature grapes (14) and made a comparativestudy of auxin changes during the development ofberries of seeded and seedless Concord (16). Gustaf-son's comparison of auxins in seeded and seedlessgrapes was made on flowers before anthesis (6).Ferenczy (5) showed the presence of an auxin, pos-sibly IAA4, in the juice of ripe grapes. The measure-ment of gibberellin activity was attempted because of

4Abbreviations used are: IAA, 3-indoleacetic acid;IAN, 3-indoleacetonitrile; IAE, ethyl ester of IAA;4-CPA, 4-chlorophenoxyacetic acid; GA,, gibberellinA3 (synonymous with gibberellic acid), similarly for GA,and GA,,.

241

www.plantphysiol.orgon May 31, 2020 - Published by Downloaded from Copyright © 1960 American Society of Plant Biologists. All rights reserved.

PLANT PHYSIOLOGY

the recently appreciated importance of these com-pounds as plant growth hormones (22) and the pro-found effects of exogenous GA3 on the growth ofseedless grapes (28).

MATERIALS AND METHODS

Five varieties of grapes were chosen to give threetypes of seed development: A. Perfect (Muscat ofAlexandria and Emperor), B. Stenospermocarpic(Seedless Emperor and Sultanina) and C. Parth-enocarpic (Black Corinth). The two Emperor va-rieties mature later than the other varieties. Thedevelopment of young fruit of Muscat and Sultaninawas followed by microscopic examination of longi-tudinal and transverse sections (prepared in 1934 byProf. H. P. Olmo) of ovaries taken at intervals from10 days before to 23 days after anthesis. Measure-ments were made of dimensions and cell size withindifferent seed and fruit structures to estimate changesin the rate of cell division. For example, counts weremade of the number of cells along minimal distancesacross the pericarp; this was done in five positionson each of six berries at each stage. For Muscat,this number doubled between the third and tenth daysafter anthesis an(l during this time many new crosswalls were evident. These observations were supple-mented by dissection of preserved and lyophilized ber-ries. Few measurements were made of embryos be-cause they were difficult to dissect.

Ten vines of each variety were selected in the uni-versity vineyard at Davis during the summer of 1958.In addition, ten vines each of Corinth and Sultaninawere trunk-girdled at the end of bloom an(d 10 daysafter bloom, respectively. Girdling consisted of re-moving a complete strip of bark (5 mm wide) downto the canibium after making two parallel knife cutsaround the trunk. From these seven lots of vines.flower or fruit samples were taken every 3 to 5 daysfrom the beginning of bloom until maturity. Faclhsample was a composite of several berries from manyclusters on all vines. In all, no more than a tenthof each vine's crop was removed. A sub-sample of200 berries (gra(dually decreasing to 50 berries asthey enlarged) was weighed, its volume determinedby water displacement, then dlrie(l at 900 C to meas-ure dry weight. Berries reaching mlaturity dried veryslowly. A secondI sub-samiiple was crushed and thepercentage of sugars in the free-run juice measuredwvith an Abbe refractomiieter. This instrument iscalibrated against sucrose solutions but it gives aclose approximation to the amount of total solublesolids, mainly glucose andl fructose, in grape juice.Excluding sampling error, the error in measuringfresh weight was low (+0.2 %) but that of volumewas high ('-1 to 2 % ). To help follow the changesin volume, values for dlensity (weight/volume) ofeach variety were graphed and a smoothed curvedrawn (the density remained between 1.00 and 1.01g/ml until the beginning of sugaring, when it in-creased gradually to 1.08-1.10 g/ml); volumes wererecalculated from thie densities read from these curves

and from the fresh weights. The size of the samplingerror can be gauged by the distance of points fromthe weight curves shown for Muscat (fig 3). Othervarieties behaved similarly so their fresh weightcurves are omitted and only smoothed volume curvesare shown.

On 16 of the 25 sampling occasions, about oneliter of berries was frozen shortly after collecting,held at -12° C and later lyophilized. Samples col-lected 17 days after anthesis, and later, failed to dryduring 2 days lyophilization unless the skins of theberries were cracked; this was done by striking theplastic bag of frozen berries against a bench. A fewsamples collected between 17 and 25 days after anthesisspoiled because they were not dIried completely.

EXTRACTION AND MEASUREMENT OF AUXINS.Preliminary comparisons of extraction and purifica-tion methods led to adopting a procedure similar tothat used by Nitsch (13). A weighed sample of ly-ophilized berries (1-1.5 g dry wt) was ground andextracted with 50 ml methanol at 40 C for 3 to 4 hours.The filtrate was dried under an airstream and the resi-due extracted four times with acetonitrile and hexane(1: 1) in a separatory funnel. After shaking, theclear yellow acetonitrile solution was run off andI di-vided into two or three equal parts (in case repeatswere neede(d), then dried under an airstream andstore(l at -12° C. Three hours before bioassay, ameasured volume of sucrose-buffer solution (15) wasadded to the residue to give a (lesiredl concentration ofgrape extract. After 1 hour, this was filtered and0.5 ml aliquots of the filtrate (or (lilutions of it) -wereuse(l for the bioassay.

Auxin was measured by the method of Nitsch andNitsch (15) using straight growtlh of sections ofAvena first internodes. Their method was modifie(Ias follows: seed was grown in vermiculite for 60hours at 260 C: only two sections were placed in eachtube containing 0.5 ml sucrose-buffer-auxin solutionand each treatment was replicated four times (26);the tubes were rotated at 0.2 rpm in a roller tubeapparatus. A control without IAA and five solutionscontaining TAA at concentrations between 0.001 and0.1 Ag per 0.5 ml were run with every test.

For each extract, three dilutions Nere assayed (oltwo in some repeated runs). Solutions from somiieextracts with high auxin concentration had to bediluted to as little as 5 mg of berries (dry wft) p)er0.5 ml to fall within the test's sensitivity. Thestrength of solutions from extracts with low auxinconcentration was increased to 24 mig per 0.5 ml.Some strong solutions of extract weere toxic to theinternode sections causing tlheml to become flaccidIand reducing their elongation. Bentley (1) has alsodescribed this. The trouble from phvtotoxicitv -asconfined to extracts of berries sampled later than14 days after anthesis and ceased when sugaring be-gan. Usually only slight (lilutions were needIe(I to

remove phytotoxicity and reveal auxin activity. Tiieach run, from 8 to 12 extracts were compare(I at two

or three dilutions of each. Bach samiple of lvophlil-

242

www.plantphysiol.orgon May 31, 2020 - Published by Downloaded from Copyright © 1960 American Society of Plant Biologists. All rights reserved.

COOMBE-CHANGES IN SUGARS, AUXINS, AND GIBBERELLINS

ized grapes was extracted and assayed two or threetinmes.

The dilutions of grape extracts showed a steeperdosage-response curve than IAA. The average ratioof the slope of the grape extract response curve tothe IAA curve was about two. This difference couldbe due to the presence of inhibitors or of auxins otherthan IAA. Nitsch and Nitsch (14) have shown thatgrapes probably contain IAN and IAE as well asIAA. They later (15) showed that the standardcurves for IAN and IAE are steeper than that forIAA. Because of this difference in slope the IAAstandard curve could not be used to compare results,so the following arbitrary procedure was used: Anew standard curve was prepared, double the slopeof the IAA curve and intersecting the latter at itsmidpoint, viz. 0.01 yg TAA per 0.5 ml (this reducedhigh values and increasedl low values). From thiscurve, internode elongation was corrected to "/lg-equivalents of auxin". These were adjusted to "'g-equivalents of auxin per gram dry wt. of grapes"and these are called "auxin units".

EXTRACTION AND MEASUREMENT OF GIBBEREL-LINS. The terms "gibberellin" and "gibberellin ac-tivity" are used to include known gibberellins andother gibberellin-like compounds which have theproperty of causing elongation of seedlings of thedwarf mutants of maize used in the bioassay. Meas-urements of gibberellin activity were made on allsamples which were assayed for auxin content, and.in addition, flowers at anthesis were assayed afterseparation into ovaries and stamens. The followingextraction procedure was used (Phinney, personalcommunication): a lyophilized sample was weighed(1-8 g dry wt of ovaries, less for stamens), groundwith a pestle and mortar, extracted with 30 to 50 mlacetone for 3 hours at room temperature, then filter-ed; the filtrate was dried under an airstream, extractedwith ethyl acetate in the presence of excess anhydroussodium sulfate to remove water, then filtered; thefiltrate was dried under an airstream and stored at-12° C until an hour or two before its bioassay whena measured volume (3-10 ml) of water plus Tween20 (polyoxyethylene sorbitan monolaurate, 0.05 %v/v) was added. After filtering. 0.1 ml aliquots wereapplied to each test plant.

The amount of gibberellin in these extracts wasmeasured by Phinney's dwarf maize bioassay (11:Phinney, personal communication). The dwarf mu-tants d-1 and d-5 were used for these assays (seedkindly supplied by Prof. B. 0. Phinney); d-1 wasused for the first test of all samples, then, when re-testing active samples, comparisons were made withd-5. Unknown extracts, a control solution and stand-ard solutions containing from 0.001 to 0.2 ILg GA. per0.1 ml were combined in randomized complete blockdesigns with 4 to 8 replicates. Seven days aftertreatment the lengths of the first and second leafslheaths were measured in mm and added (=D ).The average of the values for control seedlings (DO)w,as subtracted from the individual values (DO) for

control and treated seedlings to give a value for elon-gation (E. = D. - Do). Results were comparedby analysis of variance after transformation to log(E, + 10). This transformation gave a straight linerelation to concentration of GA3 on a log scale be-tween 0.001 and 0.2 Ag GA3 per plant and it in-creased the sensitivity of comparisons at low concen-trations.

This study of the changes in the levels of auxinsand gibberellins during the development of grapesshould be considered as exploratory. Only crude ex-tracts were made and, for gibberellin assays, dilutioncurves of these extracts were not made. More ex-periments are needed using berries sampled at morefrequent intervals, extracted by several methods, puri-fied by chromatography and tested with more thanone bioassay method.

RESULTS AND DISCUSSIONFRUIT DEVELOPMENT. The gross changes in seed

and fruit structures during the development of Muscatare shown in figure 1. An estimate of the rate of celldivision in different structures of the fruit of Muscatand Sultanina is shown in figure 2.

The development of a 1Iuscat berry is probablytypical of all seeded grapes apart from varietal differ-ences in dimensions. It can be considered in twocycles:

Cycle I (0-45 days after anthesis): The growthrates of most parts of the seed and fruit rise to a maxi-mum and then fall. The ratio of seed width and peri-carp width remains nearly constant, i.e., both grow atan approximately equal rate. Growth is due to bothcell dlivision and cell enlargement. The radial diam-eters of cells of the outer integument and nucellusincrease about three-fold and those of the pericarpand septum about six-fold. These four structurescomprise the bulk of the seed and fruit until the endo-sperm takes the place of the nucellus. Meristematicactivity in the outer integument is high; the rate ofcell division increases to a maximum at 20 to 25 daysthen dleclines to nil at 45 days. The three layers ofcells of the inner integument divide anticlinallv toaccommodate the large size and shape changes of theouter integument. Cell division in the nucellus pro-ceeds at a rate comparable to that in the outer integu-ment but it declines earlier as the endosperm nucleiproliferate. Mferistematic activity in the endospermreaches a peak at 35 days. Only a few divisions oc-cur in the zygote until 25 days after anthesis. Mostcell divisions in the pericarp occur between the fifthand tenth days after anthesis (divisions occur also 5-10 davs before anthesis). The skin, with four toeight layers of cells. is distinguishable at 16 days andis well clefined by 23 days. Meristematic activitv inthe sentum continues throughout cycle I. By 45days. the seeds are nearly full-sized, but the berrv ish-ard and green and only a third of its mature volume.

Cycle II (45 days - 110 days after anthesis):Berry volume triples and the berry softens and ma-tures. Pericarp width increases two to three times

243

www.plantphysiol.orgon May 31, 2020 - Published by Downloaded from Copyright © 1960 American Society of Plant Biologists. All rights reserved.

PLANT PHYSIOLOGY

and all of this increase is due to cell enlargement inthe pericarp and septum; there are no cell divisions.Presumably, the embryo continues to grow.

Sultanina berries show similar changes to Muscatexcept that seed developnment is imperfect. Pearson(20) has described the excessive growth of the tipsof the inner integument which can be seen at anthesis.Olmo's sections show that these abnormalities are evi-dent 11 days before anthesis. The outer integumentand nucellus commence growth witlh rapid( cell (livi-sions. The rate of cell division in these parts reachesa maximum 15 days after anthesis and stops by about30 days. Endospermii nuclei commllence (livisions butusually, after 20 days, en(lospermii an(d nucellus cellsstart to degenerate. In the mature berry, the steno-spernmic see(l is a slender, soft structure. In one ma-

ture see(l an undegenerate nucellus was foundl.Black Corinth berries grow partlhenocarpically.

DAY S AFTER ANTHESIS

16 23

2 SULTANINASEED

v I

OUTER

TESTAINNER

NUCELLUS

ENDOSPERM _

EMBRYO

FRUIT

ER ICARP

SEPTUM

20 40 0 20 40 60

DAYS AFTER ANTHESIS

FIG. 1. Diagrams of transverse sections of berries of

Muscat of Alexandria at varying times between anthesis

and maturity.FIG. 2. Diagram of changes in rate of cell division in

different seed and fruit structures of Sultanina and Muscat.

Compiled from sections prepared byr Prof. H. P. Olmo.

Diagram design after Wright (34).

This is due to defective embryo sac fori iationl andnot to defects in any other tissue of the ovule (20).In most berries, the ovule does not (levelop beyondanthesis; if it does, it develops a lhard testa and theberry enlarges. The stimulus of pollination and gird-ling are needed for good set (17); this is (lefined byStout (24) as stimulative parthenocarpy.

Seedless Emperor berries are stenospermocarpic.Their seed development is very incomplete andl varia-ble ranging fronm none at all to some larger than thestenospermic seeds of Sultanina. No hard testas were

found. There is a correspondingly large variation inberry size because the growth of each berry is propor-

tional to the amount of seed growth within. Mostberries measure fromii 9 by 11 to 12 by 15 mm and, inthese, seeds rarely lengthen to more than 1.5 mm

(c.f. Sultanina seeds which average 3 mm in lengthandl Muscat seeds, 6-7 mm). A few berries enlargeto 16 by 20 mm ain(l some are as small as 5 by 5 nmml.Olmio (18) found from one to four small rudimentaryseeds in about 70 percent of the berries and. in the rest,no ovule dlevelopment occurred after anthesis. Con-sidering the slight seed development, it is remarkablethat Seedless Emperor berries grow to about 1.5 mlvolume compared with 0.25 ml in Corinth.

FRUIT GROW'TH AND CHANGES IN SUGAR. Thechanges in berry volume and the rate of increase involumle from anthesis to maturity are shown for thefive varieties in the top graphs of figures 3 to 7.Simlilarly, changes in sugar concentration in the juiceare shown in the middle graph. The changes in berryvolunme aln(d sugar concentration brought about bygir(lling of Sultanina an(l Corinth are shown by theirrate curves.

All varieties show a double-sigmoidI growtth curxeand( hence a two-humped rate curve. These humpscorrespond to cycles I and II mentioned above; simi-larly shaped growth curves of many fleshy fruits havebeen commonly described in three periods or stages(see review by Nitsclh, 12). The late-maturing varie-ties, Emperor and Seedless Emperor. shlow this fea-ture clearly. It is barely perceptible in Corinth be-cause the first growth cycle is small. Winkler andlWilliaams (33) concluded that their measurementsof berry (liameter of Sultanina did not show a double-sigmoi(l growth curve (although their graph sug-

gests it); the volume measurements reported here

show it clearly.Sugar concentration in the berries of MIuscat.

Sultanina, and Corinth rises suddenly, almost on a

certain day, and the increase in sugar content reachesits maximum rate within ten (lays. In the later-maturing Enmperors. sugaring begins slowly for 5 to10 days, then rises to a rapid rate. The maximum

rate of sugaring for Corinth is nearly (loul)le the maxi-mum rate for the other varieties, probably because of

its small size. The increase in sugar concentration of

the juice of grapes is remarkable (it changes from4-2() percent in 30 or 40 days).

A comparisoln of the sugar and growth rate curves

reveals the striking coincidence of the commencemenit

244

www.plantphysiol.orgon May 31, 2020 - Published by Downloaded from Copyright © 1960 American Society of Plant Biologists. All rights reserved.

COOMBE-CHANGES IN SUGARS, AUXINS, AND GIBBERELLINS

MUSCAT OF ALEXANDRIA

20 40 60 80

DAYS AFTER ANTHESIS100

4

>.

0

ow

4

0-

x

4c

EMPEROR

DAYS AFTER ANTHESIS

SEEDLESS EMPEROR

2

z

X2'I

_________II

,1~~~~~I -- ,4

06 ?.,o20 40 6o0 0DAYS AFTER ANTHESIS

rSt run

cond run

run

100

n2z

ZI

4

FIGS. 3, 4, 5, 6. Changes in berry volume, juice sugar

Muscat, Emperor, Seedless Emperor and Sultanina fromare expressed as cumulative and rate curves; for girdledordinate scales differ for each variety. The term "auxin

VOLUME

0 20 40 o0 *oDAYS AFTER ANTHESIS

100

<9C'

120

concentration, and auxin concentration in berries ofanthesis to maturity. Volume and sugar measurementsSultanina, only the rate curves are shown. Note thatunits" is defined in the text.

245

SUGARSRate

I

II

7

6

5

z

3z

x 24 II

AUXINS Z-K run

*---o Second run\ I\7X

3 2-

2 Di

aw

12

n

-o

I I I ,~ I , ,,r

AUX INS , First run

2 - -e_ Second run -

0 20 40 60 s0 100 2O

SULTANINA 5

-_ F)rrt run

Seco,id 'r'unThSimd ruh

O( I

J, I I I I . . . . . - .

oilU . . . . . . . .

-

I

I I YI MQAUXINS a-

0-

0-- rir-

www.plantphysiol.orgon May 31, 2020 - Published by Downloaded from Copyright © 1960 American Society of Plant Biologists. All rights reserved.

PLANT PHYSIOLOGY

of sugaring and( the second( growth cycle. Corinthshows a slight (liscrepancy but, having smallest ber-ries, this has the greatest percentage error. Bothsugaring an(l growth rate curves increase sharply and,in each variety, the maximiiumii rate of sugaring isreache(d a few (lays before the maximium growth rate.It is possible that the second growth cycle of grapeberries could be cause(d by the influx of sugar intothe berry. The changes in auxiins andCI gibberellins(discusse(l below) and the morphological changes donot suggest other nmechanislmis to explain why the cellsof the pericarp an(l septum enlarge as they do duringthis stage. From 80 to 90 percent of the total solublesolids in grape juice are D-glucose and D-fructose inabout e(qual amounts: a 20 ,( solution of glucose hasa theoretical osmotic pressure of about 97 atmospheres.Thomas, Ranson, and Richardson (25, p. 121) quotean osmotic pressure of 40 atmosplheres measured ingrapes. It is suggested that as sugar moves into thecells of the flesh, water also moves in to adjust diffu-sion pressure deficits. The berries contain some aux-ins, probably in sufficient amounts to permit stretch-ing of the cell walls. The literature on changes insugar and fruit volume in the developing figure (3)andl apricot (4) suggests that this hypothesis shouldbe examinedl in studies on the dlevelopment of othersugary, fleshy fruits.

Nitsch (12) an(d others have suggested that thelowvered fruit growvth rate midway between anthesisand(l maturity may be due to competition between thefruit and the embryo which is developing during thisperio(l. This hypothesis cannot be applied to seedlessgrapes because there is no embryo development.

CHANGES IN AUxINS DURING FRUIT DEVELOP-M[ENT. Changes in the level of auxin1s are shown inthe bottolmi graphs of figules 3 to 7. In all varietiesexcept See(dless Emiiperor, auxins rise fromii a lowlevel just after anthesis to a hiiglh lexvel (luring thefirst few weeks. In the twN-o seedel varieties, Muscatan(d Emperor, this level is higher and maintained fora longer period thian in the seecdless varieties Sultan-ina and Corinth. NTitsch et al (16) have shown asinmilar difference betwveen seededl ancd seedless Con-cordl and have related the changes to (lifferences inmeristemiiatic activity in the seedl. The same could besai(l for M\luscat an(l Sultanina: mleristematic activityin the see(l of Muscat is lmlaintained about 10 dayslonger than Sultanina and, similarly, auxin does not(lecline to a lo\ level until 35 dai's after anthesis inAMuscat compare(l with 20 days after in Sultanina.Tn Emperor, which has big seeds, auxin does not reacha low level until 40 (lays after anthesis. There is nor1ise in auxin (luring cycle T in Seedless Emperor.Corinth provides an exception to this correlation be-tween mleristematic activity an(l auxin productionbecause it lhas a brief but significant fluslh of auxin at11 (lays after anthesis wvhen its ovule is static: noexplanation can be given for this.

In all varieties except Corinth there is a second,smaller rise in auxin concentration reaching a maxi-nmum at about the tinme that the secondl growth cycle

andl sugaring begin. It is (lifficult to explain, onmorphological grounds, why auxin should rise as itdoes at this stage. Wright (34) was able to correlatea second rise in auxin level of Ribes nigrum fruitwith meristematic activity in the embryo. This can-not be done in seedless grapes because there is noembryo development. Perhaps the rise in auxin justbefore sugaring, if real, has some causal relationshipN-ith sugaring. Sonme related facts are that exogenousauxins affect the concentration of sugars in planttissues (7) and also some workers consider that auxinacts as a trigger in affecting plant mletabolism (19).It is interesting that (lefoliation did not delay theonset of the second growth cycle (33). The mechan-ism whereby sugars accunmulate in grapes and( otherfruits, apparently against a concentration gra(lient,remains one of the puzzles of plant physiology. Thekey to sugaring is probably held( by the berries them-selves.

\Vith the increase in berry size (luring cycle II,the auxin concentration, on a dry weight basis, de-clines to a low level. Recalculated to the amount ofauxins per berry, the sanme general changes are ap-parent except that, at about 20 days after the begin-ning of cycle IT, auxins show yet another rise: thisis probably too late to enter into causal considera-tions of growth during this cycle. Changes in thedry weights of berries of the five varieties al-e shownin table I so that auxin (and gibberellin) figures maybe converted to amounts per berry.

CHANGES IN GIBBERELLINS DURING FRUIT DE-VELOPMENT. Results of the gibberellin bioassays areshown in figure 8 and table II. Duplicates showedhigh variability. It must be emphasized that the ab-sence of significant activity in these tests dloes notmean that gibberellins were absent. Better purifica-tion might have reveale(l smaller anmoulnts or removedmasking agents.

The results show that the ovaries of the seededlvarieties Mfuscat and Emperor, have n1o significantactivity at any stage tested but that their stanmens haveactivity at anthesis. On the other hanid the ovariesof the three seedless varieties, especially SeedlessEmperor, show activity during the first 14 days afteranthesis. but not subse(luently. Also, like seeded va-rieties, their stamens show activity at anthesis. The

TABLE IDRY \WFIGHTS OF BERRIES OF 5 GRAPE VARIETIES AT

CERTAIN INTERVALS AFTER ANTHESIS

VTARIETY

MuscatEmperorSeedless

EmperorSultaninaCorinth

DRY WETGHT PER 'BERRY (MG)10 30 50 70 90 MATURE*

DAYS AFTER ANTHESIS

6 78 170 450 860 1,3003 77 137 150 250 800

343

30355

41 45 14580 270 470' 1 36 57

35050060

* Dry weight estimated by slight extrapolationl.

246

www.plantphysiol.orgon May 31, 2020 - Published by Downloaded from Copyright © 1960 American Society of Plant Biologists. All rights reserved.

COOMBE-CHANGES IN SUGARS, AUXINS, AND GIBBERELLINS

BLACK CORINTH 81b 4

4 I.

si-

I:

a: I0! {a5

go 3zW2

il ,_

4 "

3

<o 2

O

AT ANTHE S

MUSCATj

EMPEROR

SEEDLESS EMPERORl

_ K + _i

SULTANINA 3

BLACK CORINTH

Jxo~~~~~~~:-oi . , , I, ,,,0 20 40 60 80 100 -1 5 0 5 10 I5 20 25 30 45 80 75

DAYS AFTER ANTHESIS DAYS AFTER ANTHESIS

FIG. 7. Changes in berry volume, juice sugar concentration, and auxin concentration in berries of Black Corinthfrom anthesis to maturity. Volume and sugar measurements are expressed as cumulative and rate curves; forgirdled Corinth, only the rate curves are shown. The term "auxin units" is defined in the text.

FIG. 8. The gibberellin activity of extracts of grape stamens and ovaries of different age and variety as meas-ured by the d-1 maize bioassay; note the different ordinate scale for Sultanina.

results suggest also that there may be activity in theovaries of seedless varieties at anthesis. A fourthseedless variety (Olmo's hybrid No. Q25-6) was

tested in June, 1959, and found to have significantgibberellin activity in its young berries, although lessthan Seedless Emperor.

It may be significant that the difference in gib-berellin activity between seedless and seeded berriesis correlated with their responsiveness to exogenous

GA3. Without exception, the berries of seedlessgrape varieties have been found to enlarge followingtreatment with GA3 at, or soon after, anthesis, e.g.,

Corinth and Sultanina (28), Black Monukka andBeauty Seedless (Weaver, private communication),Concord Seedless (23), Seedless Emperor and Olmo'shybrid No. Q25-6 (the author, unpublished data).On the other hand, the berries of seeded grape varie-ties will not enlarge appreciably after treatment withGA3 (23, 28). These results suggest that gibberellinsmay be involved in the growth mechanism during cycleI of seedless berries and that the level of naturallv-occurring gibberellins may be an important factorlimiting their growth. Gibberellins may not be in-volved in the growth mechanism of seeded grapes

or they are destroyed. More information is neededbefore these questions can be decided.

The interactions between naturally-occurringauxins and gibberellins of seedless grapes are inter-esting. Seedless Emperor has low auxin content dur-ing cycle I yet it grows five times bigger than Corinth.Perhaps this is explained by the high gibberellin ac-tivity in its young berries. This idea is supportedby the fact that Corinth berries, when dipped in GA3solution (500 ppm) at anthesis (28), enlarge to aboutthe same size as, and ellipsoidal shape of, SeedlessEmperor berries even though their ovules remainminute. Thus it is possible that natural levels of gib-berellins may be the most important factor limitingtheir size. Girdling and 4-CPA are not as potent asGA3 in enlarging Corinth berries.

Comparisons of the activity of grape extracts inelongating d-1 and d-5 maize seedlings are shownin table IT. It can be seen that, in most cases, boththe young berries of seedless varieties and mature sta-mens of seeded and seedless varieties induce greaterelongation in d-5 than in d-1. Obviously, it wouldhave been better to use d-5 to examine change's indeveloping grapes rather than d-1. A similar differ-

w-gI a I 1-S: I -," -

a 0 - --*--O a -_

247

www.plantphysiol.orgon May 31, 2020 - Published by Downloaded from Copyright © 1960 American Society of Plant Biologists. All rights reserved.

PLANT PHYSIOLOGY

TABLE IICOMPARISON OF ACTIVITY OF GRAPE EXTRACTS IN

ELONGATING DWARF-1 AND DWARF-5MAIZE SEEDLINGS f

ELONGATION OF GA3 EQuIvs./EXTRACTr 1 ST AND 2ND 0.1 ML

LEAF SHEATHS EXTRACTVARIETY PART MM G X 10 9

d-1 d-5 d-1 d-5

Gold(seeded) Stamens 2.0 11.6** 0 1.5

Muscat Stamens 11.0** 34.9*** 2.2 50Seedless Stamens 14.3*** 22.7*** 4 8.3

Emperor Ovaries++ 9.5** 12.9** 1.6 1.8Sultanina Stamens 2.8 21.6*** 0 9

Ovaries 3.6 18.0*** 0 7Pedicels 3.2 4.8 0 1

Coritnth Flowers 1.6 6.9* 0 3

f 0.1 ml of extract (colntainiing from 1-11 mg dry wtof the original grape sample) was applied to each maizeseedling. Four or five replicates of pairs of dwarfs wereused. Elongations which were significantly greater thancontrol in the analyses of variance are marked by 1, 2, or3 asterisks when the probabilities that they were signifi-cantly greater than controls were better than 20: 1, 100: 1and 1000: 1, respectively; insignificant values are callednil.

+f Sampled 7 days after anthesis; all other parts inthis table were taken at anthesis.

ence in extracts of bean seeds has been found byPhinney and Neely (21) and WVest and Murashige(30). They purified two crystalline compounds fronmbean seed. Bean factor I behaved identically withCA1. Bean factor II was ten tinmes as active on(lwarfs -2, -3, -5 and an-1 compared with its activityon d-1. It also differed in its chromatographic be-haviour, infra-red spectrum and chemical propertiesfrom GA3, GA1, and GA.). Perhaps the activity ingrapes is due in part to the presence of bean factorII or other unknown gibberellin-like compounds.

EFFECTS OF GIRDLING ON BERRY GROWTH AND ONCONTENT OF SUGARS, AuXINS, AND GIBBERELLINS.It is common viticultural practice to girdle the trunksor canes of Corinth vines at anthesis to increase fruitset and size and to girdle Sultanina 10 days after an-thesis to produce large berries for table use (8). Itseffects on berry weight and volume (figs 6 and 7)are apparent within a few days of treatment and thesedifferences continue to increase until maturity. Thusgirdling affects both growth cycles. The secondgrowth cycle and sugaring began concurrently a fewdays earlier in girdled vines of both varieties than inungirdled. This supports the hypothesis that sugarinflux may cause the second growth cycle. The rateof sugar increase in berries on ungirdled vines even-tually exceeds that on girdled vines probably becauseof the bigger crop on the latter.

The effects of girdling on auxin content of theberries are shown in table ITI. The differences as-sociated with girdling are very small and are far ex-ceeded by variations betweeln tests. The results sug-

gest that the concentration of auxins on a dry weightbasis is unchanged. If this is true, the amount perunit volume or fresh weight would also be equal aspercentage dry weight and densities are equal ingirdled and ungirdled berries at these growth stages.However, the amount per berry would be greaterby a factor of 1.5 to 2 in girdled berries because oftheir greater size. Measurements of gibberellin ac-tivity in these samples are too few and too variableto be worth presenting. In four instances, there didnot seem to be any large difference on a dry weightbasis.

There are several considerations wvhich place im-portance on the events taking place in the berries ofSultanina during the period from 10 to 15 days afteranthesis: the effect of girdling and berry thinning inincreasing berry size is maximal (31, 32) ; treatmentwith 4-CPA and GA3 causes greatest enlargementof berries when applied during this period (27, 28);gibberellin activity can be detected in the berries (fig8) and auxin concentration is high (fig 6). The factthat the rate of cell division in the pericarp is alsomaximal (fig 2) suggests that girdling an(d othertreatments may increase berry size by increasing thenumber of cells. The possibility that cell size is alsoincrease(l should not be excluded.

The effects of girdling on berry size of Sultaninacan be duplicated by treatment with low concentra-tions of 4-CPA and GA,, so the mechanisnm of theaction of girdling may be due to its effect on naturally-occurring growth regulators. The behavior of seed-ed varieties supports this idea and emphasizes the

TABLE IIIAlUXIN CONCENTRATION IN BERRIES OF VARIOUS AGES

FROMI GIRDLED AND UNGIRDLED CORTNTHAND SULTANINA VTINES*

**

EXTRACTVARIETY DAYS AFTER

ANTHESISTESTNo.

AUXIN UNITS EXPRESSEDON DRY \VT BASIS

U NGIRDLED GIRDLED

Corinth 14 A 0.37 0.36B 0.42 0.34C 0.43 0.56

32 A 0.30 0.32B 0.33 0.38C 0.51 0.37

49 A 0.36 0.30Sultanina 17 A 1.08 1.02

B 1.59 1.9224 A 0.51 0.54

B 0.72 0.6128 A 0.45 0.55

B 0.45 0.5155 A 0.20 0.19

B 0.16 0.1575 A 0.25 0.19

* Equal dry weights of girdled and ungirdled berrieswere compared in each bioassay.

** The term "auxin units" is described in the text.

.248

www.plantphysiol.orgon May 31, 2020 - Published by Downloaded from Copyright © 1960 American Society of Plant Biologists. All rights reserved.

COOMBE-CHANGES IN SUGARS, AUNINS, AND GIBBERELLINS

probably important role of gibberellin; their berriesare devoid of (or low in) gibberellin and they do notenlarge much after girdling or spraying with 4-CPAor GA3. However, girdling affects the concentrationof many compounds, e.g., it increases the level ofcarbohydrates (29) and decreases the level of nitrate-nitrogen in petioles opposite the grape clusters (9).More information is needed before a conclusion canbe made on the mechanism of the girdling effect.

GENERAL CONCLUSIONSCorrelations between growth rate, auxins and

meristematic activity have been demonstrated in sev-eral fruits and other plant organs. There are ex-amples in fruits, however, where auxin increase canbe correlated with increased meristematic activity inthe seed (especially the endosperm) but this activitycannot always be correlated with increased growthrate in the fruit, e.g. apple (10) and figure 2. Fail-ure to show a correlation may be due to failure tomeasure the substances responsible for growth or toa different mechanism of growth.

In the varieties of grapes studied here, only theseeded varieties, Muscat and Emperor, show a cor-relation between meristematic activity in the seed,level of auxins and the first cycle of berry growth.In Sultanina, the auxin level parallels the rate of celldivision in the seed but the rate of increase in berryvolume is maintained at a high level 10 days longer.The situation in Seedless Emperor is not as clear be-cause no detailed study of seed development was madebut, in any case, auxin content remained low whilethe berries grew. In Corinth there was a sharp peakin auxin content 11 days after anthesis when therewas no seed development at all. It seems more likelythat the growth of these seedless grapes is controlledby gibberellin levels rather than auxins or that gib-berellins act together with auxins.

The second cycle in berry growth does not coincidewith a rise and fall of either auxins or gibberellins.As suggested above, a more acceptable hypothesis toexplain this growth cycle is that the movement ofsugars into the berries causes an osmotic attractionof water.

SUMMARYMeasurements were made of growth and develop-

ment of certain parts of-the fruit and seed of two seed-ed and three seedless varieties of Vitis vinifera(European grape) from anthesis to maturity. Thesewere compared with changes in the sugar concen-tration of the juice and the levels of auxins and gib-berellins in partially-purified extracts of berries.

The fruit of all varieties showed a double-sigmoidgrowth curve. In the two seeded varieties, the firstgrowth cycle was paralleled by a rise and fall in meri-stematic activity in the seeds and in the auxin contentof the berries. In the seedless berries, the firstgrowth cycle was greater than would be expectedfrom the berry's auxin content and the seed's meri-

stematic activity. This may be explained by the dis-covery of gibberellin activity in the young berries ofthe three seedless varieties. No gibberellin activitywas found in seeded berries at any time. These find-ings, plus the fact that berries of seedless varietiesenlarge considerably after treatment with gibberellin,auxin, and by girdling, whereas seeded varieties donot, suggest that gibberellins are important hormonesin the fruit of these seedless varieties. Mature sta-mens of all varieties showed gibberellin activity.

The second growth cycle could not be correlatedwith morphological, auxin, or gibberellin changesbut could be related to the influx of sugars into theberries. It is suggested that sugaring causes thisgrowth cycle by an osmotic attraction of water.A second rise in auxin level was found to reach

a maximum in four varieties at the beginning of'sugaring. These events may be related but no ex-planation can be given for this phenomenon. Trunkgirdling of two seedless varieties caused no measurablechange in auxin or gibberellin concentration expressedon dry weight, fresh weight, or volume bases, butlevels per berry increased because of the large increasein berry volume after girdling.

ACKNOWLEDGEMENTSThe author wishes to record his sincere thanks to

Drs. R. J. Weaver, H. P. Olmo, E. C. Maxie. andL. K. Mann for their guidance and encouragementduring these investigations.

LITERATURE CITED1. BENTLEY, JOYCE A. 1958. The naturally-occurring

auxins and inhibitors. Ann. Rev. Plant Physiol.9: 47-80.

2. CRANE, J. C., MURIEL V. BRADLEY, and L. C. LUCK-WILL 1959. Auxins in parthenocarpic and non-parthenocarpic figs. Jour. Hort. Sci. 34: 142-153.

3. CRANE, J. C. and J. G. BROWN 1950. Growth ofthe fig fruit, Ficuts carica var. Mission. Proc. Am.Soc. Hort. Sci. 56: 93-97.

4. CRANE, J. C., E. D. DE KAzos, and J. G. BROWN1956. The effect of 2,4,5-trichlorophenoxyaceticacid on growth, moisture, and sugar content ofapricot fruits. Proc. Amer. Soc. Hort. Sci. 68:105-112.

5. FERENCZY, L. 1957. Examination of ether extract-able growth substances in grape and watermelonwith paper chromatography. Phyton (BuenosAires) 9: 47-52.

6. GuSTAFSON, F. G. 1939. The cause of naturalparthenocarpy. Amer. Jour. Bot. 26: 135-138.

7. HORSFALL, J. G. and A. E. DI'MOND 1957. Inter-actions of tissue sugar, growth substances. anddisease susceptibility. Z. Pflanzenkrankh. u. Pflan-zenschutz. 64: 415-421.

8. JACOB, H. E. 1928. Some responses of the seedlessvarieties of Vitis vinifera to girdling. Proc. Amer.Soc. Hort. Sci. 25: 223-229.

9. KISSLER, J. J. 1957. Nitrate fluctuations and petiolesampling technique with grapevines. M.S. Thesis,University of California, Davis.

249'

www.plantphysiol.orgon May 31, 2020 - Published by Downloaded from Copyright © 1960 American Society of Plant Biologists. All rights reserved.

PLANT PHYSIOLOGY

10. LUCKWILL, L. C. 1948. The hormone content ofthe seed in relation to endosperm development andfruit drop in the apple. Jour. Hort. Sci. 24: 32-44.

1 1. NEELY, P. M. and B. 0. PHINNEY 1957. Use ofmutant dwarf-1 of maize as a quantitative bioassayfor gibberellin activity. Plant Physiol. 32 suppl.:xxxi.

12. NITSCH, J. P. 1953. The physiology of fruitgrowth. Ann. Rev. Plant Physiol. 4: 199-236.

13. NITSCH, J. P. 1955. Free auxins and free trypto-phane in the strawberry. Plant Physiol. 30: 33-39.

14. NTTSCH, J. P. and COLETTE NITSCH 1955. Theseparation of natural plant growth substances bypaper chromatography. Beitr. Biol. Pflanz. 31:387-408.

15. NITSCH, J. P. anid COLETTE NITSCH 1956. Studieson the growth of coleoptile and first internode sec-

tionls. A new, sensitive, straight-growth test forauxins. Plant Physiol. 31: 94-111.

16. NITSCH, J. P., COLETTF NTTSCH, CHARLOTTE S.PRATT, and N. J. SHAUTIIS 1957. Auxins in theConcord and Concord Seedless grapes in relation toberry development and drop. Plant Physiol. 32suppl.: xx.

17. OLMIO, H. P. 1937. Pollination and the setting offruit in the Black Corinth grape. Proc. Amer. Soc.Hort. Sci. 34: 402-404.

18. OLMIO, H. P. 1940. Somatic mutation in the vini-fera grape. III. The Seedless Emperor. J.Heredity 31: 211-213.

19. OVERBEEK, J. VAN- 1959. Auxins. Bot. Rev. 25:269-350.

20. PEARSON, HELEN M. 1932. Parthenocarpy andseed abortion in Vitis viniifera. Proc. Amer. Soc.Hort. Sci. 29: 169-175.

21. PHINNEY, B. 0. and P. M. NEELY 1958. Differen-tial biological properties of gibberellin-like factorsisolated from beans and peas. Plant Physiol. 33suppl.: xxxviii.

22. PHINNEY, B. O., C. A. WEST, MARY RITZEL, andP. M. NEELY 1957. Evidence for gibberellin-like

substances from flowering plants. Proc. Nat.Acad. Sci., U. S. 43: 398-404.

23. SHAULIS, N. J. 1959. Gibberellin trials for NewYork grapes. Farm Res. 25: 11.

24. STOUT, A. B. 1936. Seedlessness in grapes. N.Y.Agr. Exp. Stn. Tech. Bull. No. 238: 1-68.

25. THOMAS, MEIRION, S. L. RANSON, and J. A. RrcH-ARDSON 1956. Plant Physiology. (4th Edit.)692 pp. Philosophical Library, New York

26. WALKER, D. R., C. H. HENDERSHorr, and G. W.SNEDECOR 1958. A statistical evaluation of agrowth substance bioassay method using extractsof dormant peach buds. Plant Physiol. 33: 162-166.

27. WEAVER, R. J. 1953. Furthur studies on effectsof 4-chlorophenoxyacetic acid on development ofThompson Seedless and Black Corinth grapes.Proc. Amer. Soc. Hort. Sci. 61: 135-143.

28. WEAVER, R. J. and S. B. MCCUNE 1959. Responseof certain varieties of Vitis vinifera grapes to gib-berellic acid. Hilgardia 28: 297-350.

29. WEAVER, R. J. and S. B. MCCUNE 1959. Girdling:its relation to carbohydrate nutrition and develop-ment of Thompson seedless, Red Malaga, and Ribiergrapes. Hilgardia 28: 421-456.

30. \VEST, C. A. and KATE H. MURASHIGE 1958. Theisolation of gibberellin A1 from beans and thechemical properties of other gibberellin-like factorsfrom beans and peas. Plant Physiol. 33 suppl.:xxxviii.

31. W"INKI,ER, A. J. 1930. Berrv thinning of grapes.Cal. Agr. Exp. Stn. Bull. No. 492: 1-22.

32. WINKLER, A. J. 1953. Producing table grapes ofbetter quality. Blue Anchor. 30: 28-31.

33. WINKLER. A. J. and WV. 0. WILLIAMS 1936. Ff-fect of seed development on the growtth of grapes.Proc. Amer. Soc. Hort. Sci. 33: 430-434.

34. WRIGHT. S. T. C. 1956. Studies of fruit develop-ment in relation to plant hormones. III. Auxinsin relation to fruit morphogenesis and fruit drop inthe black currant, RiThes n4grqj,. jour. Hort. Sci.31: 196-211.

250

www.plantphysiol.orgon May 31, 2020 - Published by Downloaded from Copyright © 1960 American Society of Plant Biologists. All rights reserved.