Influenceo fexterna lfactor so ngrowt han ddevelopmen t ofsugar-bee t (Beta vulgaris L.)-

CENTRALELANDBOUWCATALOGU S

0000006 0653 9 A.L. Smit

Department of Field Crops and Grassland Science, Agricultural University, Wageningen.

Influence of external factors on growth and development of sugar-beet (Beta vulgaris L.)

Centre for Agricultural Publishing and Documentation

Wageningen - 1983

'2~&é>\\.~l CIP-GEGEVENS

Smit, A.L.

Influence of external factors on growth and development of sugar-beet (Beta vulgaris L.) / A.L. Smit. - Wageningen : Pudoc. - Fig., tab. - (Agricultural research reports ; 914) Met samenvatting in het Nederlands. - Ook verschenen als proefschrift Wageningen. - Met lit. opg. ISBN 90-220-0812-6 Siso 630 UDC 631/632 Trefw. : landbouw.

ISBN 90 220 0812 6

The author graduated on 18 February 1983 as Doctor in de Landbouwwetenschap­ pen at the Agricultural University, Wageningen, the Netherlands, on a thesis with the same title and contents.

©Centre for Agricultural Publishing and Documentation (Pudoc), Wageningen, 1983.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, mechanical, photocopying, recor­ ding, or otherwise, without the prior written permission of the publisher Pudoc, P.O. Box, 4, 6700 AA Wageningen, Netherlands.

Printed in the Netherlands. Abstract

Smit,A.L . 1983.Influenc e of external factors on growthan d development ofsugar-bee t (Betavulgari s L.). Agric.Res .Rep . (Versl. landbk.Onderz. )914 ,ISB N9 022 0081 26 , (xii) +10 9p. ,2 8tables ,3 6figs . 103refs . Eng.an dDutc hsummaries . Also:Doctora lthesis ,Wageningen .

Several trials on the quantitative influence ofphotophase ,chillin g (vernalization) andhig htemperatur e (devernalization)o nboltin go fsugar-bee twer eanalyse do nth ebasi s of a simplephysiologica l model,i nwhic hboltin g is considered as the finaleven to f dynamic,momentar y and quantitative processes in theplant .Trial s inth efiel dan di n growth chambers examined factorsi nchillin gan di nligh tresponse .Th einhibitor yeffec t onboltin g ofhig htemperature san dth erol eo fphotophas ei nthi sproces swa sinvestiga ­ tedfo rsevera lperiod safte rvernalization . Growth and bolting seem tob e correlated,a splant s with justvisibl e boltingwer e usuallyheavier . A possible relationbetwee nboltin g resistancean dvigou rwa sinvestigated .Als oth ein ­ fluenceo fphotophas ean dcol dtreatmen to ngrowt hwa smeasure d ina trial . Someway s are showno f usinga climati cfacto rlik etemperatur et opredic tboltin gi n thefield .Finall ysom erecommendation s forsugar-bee tbreeder sar edraw nup .

Free descriptors:vernalization ,photoperiod ,bolting ,daylength ,temperature ,chilling , model,flowering ,generative ,devernalization ,regression ,optimization ,breeding . Contents

Abbreviations and symbols

1 Introduction 1 1.1 Boltingi nsugar-bee t crops 1 1.2 Influenceo fagricultura l practice 1 1.3 Breedingo fcultivar s resistant toboltin g 2 1.4 Scopeo fth epresen t study 3

2 Factors influencing bolting in sugar-beet 4 2.1 introduction 4 2.2 Temperature 4 2.3 Lightphas e 8 2.4 Interactionsbetwee n lightphas e andtemperatur e 9 2.5 Growth andboltin g 10

3 Modelsfor bolting 11 3.1 Modelsi nliteratur ebase do n(hypothetical )plan thormone s1 1 3.2 Amathematica l model forboltin gi nsugar-bee t 12

4 Materials andmethods 16 4.1 Plantmateria l 16 4.2 Field trials 16 4.3 Indoortrial s 17 4.3.1 Growing rooms 17 4.3.2 Rooms forcold 'treatmen t 18 4.3.3 Greenhouses 18 4.3.4 Small greenhouses 18 4.4 Lightcondition s 18 4.5 Raising andchillin go fplant s 18 4.6 Observations 20 4.7 Statistics 20

5 Results 21 5.1 The juvenile stage andvernalizatio n 21 5.1.1 Effecto fdaylengh tan d coveringwit hplasti c 21 5.1.1.1 Introduction 21 5.1.1.2 Materials andmethod s 22 5.1.1.3 Results 22 5.1.2 Effecto fgerminatio n temperature 24 5.1.2.1 Introduction 24 5.1.2.2 Materials and methods 24 5.1.2.3 Results 25 5.1.3 Effecto fplan t sizeo nvernalizatio n inth efiel d 26 5.1.3.1 Materials andmethod s 26 5.1.3.2 Results 26 5.1.4 Effecto fplan t sizeo ncol d treatment 28 5.1.4.1 Materials andmethod s 28 5.1.4.2 Results 28 5.1.5 Conclusions 30 5.2 Temperature ofcol d treatment 30 5.2.1 Materials andmethod s 30 5.2.2 Results 31 5.2.3 Conclusions 33 5.3 Light phase andvernalizatio n 33 5.3.1 Influence ofligh t phase after vernalization 33 5.3.1.1 Materials andmethod s 33 5.3.1.2 Results 33 5.3.2 Effecto fcultiva r 35 5.3.2.1 Introduction 35 5.3.2.2 Materials andmethod s 35 5.3.2.3 Results 35 5.4 Devernalization 37 5.4.1 Materials andmethod s 37 5.4.2 Results 38 5.5 Temperature andligh t phase immediately after vernalization3 9 5.5.1 Trial 1 39 5.5.1.1 Materials and methods 39 5.5.1.2 Results 40 5.5.2 Trial 2 43 5.5.2.1 Materials andmethod s 43 5.5.2.2 Results 44 5.6 Discussion 48 5.6.1 Introduction 48 5.6.2 Seedling stage 48 5.6.3 Vernalization 49 5.6.4 Post-vernalization 50 5.6.5 Physiological background ofth e model 51

Relation between growth and bolting of sugar-beet 55 6.1 Introduction 55 6.2 Influence ofcondition s ofgrowt ho nboltin g 56 6.2.1 Introduction 56 6.2.2 Materialsan dmethod s 56 6.2.3 Results 58 6.2.4 Plantweigh to fbolter san dnon-bolter s 60 6.2.5 Conclusions 61 6.3 Influenceo fcol d treatmentan dphotophas eo ngrowt h 63 6.3.1 Introduction 63 6.3.2 Materials andmethod s 64 6.3.3 Results 65 6.3.4 Discussion 77 6.4 Influenceo fselectio n towardsboltin g resistance 80 6.4.1 Introduction 80 6.4.2 Materials andmethod s 80 6.4.3 Results 80 6.4.4 Discussion 82

7 Temperature and bolting under field conditions 84 7.1 Introduction 84 7.2 Therelatio nbetwee nboltin gan dsowin g date 84 7.3 Aregressio n approach 85

7.4 Anoptimizatio n approach 92

8 Final remarks 97

Summary 100

Samenvatting 103

References 106 Abbreviations and symbols

* P<0. 1 ** P<0.05 *** P<0.01 **** P<0.001 ANOVA Analyses of variance d.05 Studentized range, Tukey, atth e5 % level F Hypothetical flower hormone 2 LA Leaf areape rplan t (cm) 2 -1 LAR Leaf area ratio (cm g ) -2 -1 NAR Netassimilatio n rate (gc m d ) r Coefficient of correlation 2 r Coefficient of determination RER Relative expansion rate (d ) RGR Relative growth rate (d ) 2 -1 SLA Specific leafare a (cm g ) W Dryweigh tpe rplan t (g) V Hypothetical substance thoughtt o accumulate during chilling 1 Introduction

1.1 BOLTING INSUGAR-BEE TCROP S

Usually sugarbee t (Beta vulgaris L.) behave s asa biennial . Inth e first year of growth, the plant produces a rosette of leaves and aroot .Afte r overwintering, inth esecon d yearo fgrowth ,dr ymatte r distribution becomes totally different. InMa yo rJune ,th e stem appears and thenth eplan tflow ­ ers. Inmos t years alloverwintere dplant sma y flower,althoug h incertai n years some plants do not.Als o in some years, someplant sru nt o seedi n their first season, thebee tproductio n year,an dd ono tbehav e asa bien ­ nial.Thes eplant s arecalle dbolters .Bolter s appear fromth ebeginnin go f Julyunti l theharves to fth ecro p inOctober . Ingeneral ,plant swhos e stem is visible in theNetherland s before 1Augus t arearbitraril y called early bolters andafte rtha tdat e latebolters . For several reasons bolters are not wanted in a rootcrop .Especiall y earlybolters ,whic hd ono thav etim et odevelo p anorma l root,reduc eyield . According toLongde ne t al. (1975), yield isreduce db y about1 % forever y 4 % of plants showing signs of bolting. There is also aneffec to nsuga r content of roots;ove r a 0-50 % range in the frequencyo fbolters ,a 1 % increase causes a 0.05 % decline in sugarconten t (Longdene t al., 1975). The values vary according to whether bolters are removed inth e fieldan d whether the remaining plants can compensate forth e gaps.Othe rdisadvan ­ tages are the woody texture of the root, especially ofearl ybolters ,s o thatfarmer shav edifficult y intoppin g theircro pproperl y and sugarfacto ­ ries have to sharpen theirknive smor eoften .Finall ybolter s giveris et o self-sown beet in following crops,especiall y ifearl ybolter s areno tre ­ moved and are allowed to shed viable seeds (VanSteyvoor t &Va nStallen , 1973).Whethe rbolter s appear inth esugar-bee tcro p depend onth e following factors: agricultural practice choice ofth ecultiva r - environmental conditions (temperature, daylength)

1.2 INFLUENCEO FAGRICULTURA L PRACTICE

The farmer can alter the risk of bolting by shifting the sowingdate . However, extension of the growing season by aiminga ta nearl yclosin go f the canopy has proved more advantageous foryiel d ofsuga rtha navoidanc e ofbolting . So inrecen tyears ,ther eha sbee n ashif ttoward s earlier sow­ ing dates,encourage d alsob y the improved quality of seed frommor e south­ erlyareas . Some other influences onboltin g havebee nreported . Nitrogen fertilizer and lowerplan tdensit y seem topromot ebolting . Itseem s asthoug h allfac ­ tors that enhance growthpromot e bolting.A s there are several indications to arelationshi p betweenboltin g and rapid growth, growers areunlikel y to curb bolting by changing cultivationmethod s: th emeasure s againstboltin g could reduceyield . Sot opreven texcessiv e bolting insugar-beet , cultivars resistant toboltin gmus tb e developed.

1.3 BREEDING OFCULTIVAR S RESISTANT TOBOLTIN G

Although the environment plays amajo r role in theboltin gbehaviou r of sugar-beet, there are largedifference s between cultivars.Th e inheritance of the bolting resistance isno t fullyunderstood . Dominant,recessiv e and polygenic control havebee n reported inliterature . Although great progress has been made in the past 20year s indevelopin g cultivars resistant to bolting, bolting has not disappeared. In certain years, itca n still reduceyield . Breeders, however,hav e to solve several problemsbefor e abolting-resis ­ tant cultivar can be obtained. Because sugar-beet cross-pollinates, the plants are rathervarie d incharacteristics , includingbolting .A self-pol ­ linating crop would show anal l ornothin g response,whic hmake s selection easier. Selection forboltin g resistance isnowaday s by sowing early inth esea ­ son (February,March) .Whe nbolter s arediscarded , the remainingplant swill , in general, have less tendency tobolt .Sowin g early about 500k mnort ho f thebeet-growin g area turned outt ob e evenmor esuccessful . The seed crop is atpresen t grown inmor e southerly countries like France and Italy,wher e climate during ripening isbette r foryiel d andqualit y of seed. An other advantage oftha t area for seedproductio n is that ripening seed onth eplan tdoe sno tvernalize ,wherea s inN.W . Europe,boltin g behav­ iour of the nextgeneratio n isaffecte d by conditions during seed ripening (O'Connor, 1970;Bosemark , 1970;Bornscheuer , 1972; Lexander, 1969). When sugar-beet is grown for seed, breeders need tob e certain of flowering in allplants ,otherwis e therewil l be ashif ttoward smor eboltin g inth eprog ­ eny.Her e lies also oneo fth e limits forth ebreeder s increatin g acultiva r resistant tobolting .A cultivar extremely resistant toboltin gwoul d bring problems inho wt oobtai n seed.Movin g seedproductio n tosouther n countries only aggravates this problem: duration of vernalizing temperature willb e shorter and also daylengthwil lb e shorter than innorther nEurope . Over anumbe r ofyears ,anothe r complication hasemerge d of seedproduc ­ tion in the south of Europe. The seed stock canb e contaminated bycross - pollination with annual wild beets like Beta maritima and B. patellaris, because the tetraploid pollinators usednowaday s inmoder nbreedin gproduc e less pollen and reach theirdail ypolle npea k later thandiploi dwil dbee t plants (Scott& Longden , 1970). However,breeder s cantes t for contamination with annualbeet sb y growing seed stocks inlon gday s above2 0 °C. This defect of hygiene during seed multiplication isoutsid e thescop e ofthi sreport .

1.4 SCOPEO FTH EPRESEN T STUDY

Several research workers have studied the relationship ofenvironmenta l factors to bolting, including the extensive researchb yChrobocze k (1934) onbeetroo t (redgarde nbeet )an d studiesb y Curth (1955;1960 ;1962 ; 1963). Bothworkers ,however ,worke d withoutgrowin g rooms,whic h areessentia l to quantitative measurements ofinfluence s oftemperatur e and lightphase (day- length) . Chapter 2describes ,base d ondat a from literature, the influence of light and temperature onboltin g andChapte r 5present s trials onthi smatte rwit h four genotypes of different susceptibility to bolting. Such information should helpbreeder s indevelopin gmor e specific selectionmethod s toreduc e susceptibility to bolting, and to improve adaptation to localconditions . Practical implications willprobabl y alsob e found inthi s chapter, asbreed ­ ers are interested in making their crosses in greenhouses toshorte nth e growthcycle . Chapter 6 analyses a relationbetwee ngrowt h and appearance ofbolters . Chapter 7describe s techniques torelat e aclimati c factor like temperature toproportio n ofbolters . In recent years, bolters have not been such a severe problem in the Netherlands. Insurroundin g countries,wher ebee tca nb e sownver y early in theyea ro n sandy soils,crop shav ebolte d severely, forcing farmers toaban ­ donsom e fields forbee tgrowing . Ifautum nsowin go fsugar-bee t should ever become possible, extremely resistant cultivarswil lhav e tob ebred , util­ izingquantitativ e information aboutth eboltin gprocess . 2 Factors influencing bolting in sugar-beet

2.1 INTRODUCTION

Many extensive reviews of flowering physiology are available (Chouard, I960; Lang, 1965;Nap p Zinn, 1961, 1973;Naylor , 1961;Purvis , 1961;Zee ­ vaart, 1976). So only data will be reviewed here that iscomparabl e with the situation inbeet . Sugar-beet requires cold for flowering: alon g interval of lowtempera ­ ture forvernalization .Vernalizatio n has twomeanings :first ,th ephysiolo ­ gical process inth eplant ; secondly, atreatmen t (ina col d room or inth e field)t ohaste n development.Vernalizatio n induces to flower.Th e term 'rip­ eness to flower' isuse d (NappZinn , 1973)becaus en oexterna l modification can be seen at the end ofth echillin gperio d and theplan tneed s another stimulus (daylength)befor e the apex starts todifferentiate .Sugar-beet , being a long-day plant,therefor e needs chilling first,the n atim ewit ha relatively longphotoperio d atsomewha thighe r temperature (inth e following the words light phase or photophase willb euse d fordaylengt h andphoto - period). Not only low temperature but also moderate orhig h temperature canin ­ fluencebolting .A relatively shorttim e athig h temperature cannullif y or moderate the effect of previous vernalizing temperatures.Thi s phenomenon iscalle d devernalization. Inth enorma l 2-yearcycl e of sugar-beet,al lprerequisite s for flowering are fulfilled inth e second year.B y overwintering, theplant s have undergone a long period ofvernalizin g temperatures and theplant s all startboltin g whendaylengt h gradually increases inMa y orJune . In the first season of growth, beets are fairlywel lprotecte d against undesired bolting. With a normal sowing date, longvernalizatio n doesno t usually occur.Moreove r some authors report ajuvenil ephase , inwhic hsugar - beet in the germination or early-leaf stages isno tver y sensitive tolo w temperature (Chroboczek, 1934). The influence ofenvironmenta l conditions willb e considered inmor ede ­ tail inth e following sections.

2.2 TEMPERATURE

Temperature influencesman yprocesse s inth eplant .Amon g the leastunder ­ stood is vernalization which induces the plant to flower.Althoug h widely used since Lyssenko introduced vernalization for germinatingwhea tseeds , themechanis m ofth eparadoxa l forcing effecto f lowtemperatur e isunknown . Chroboczek (1934)demonstrate d that only part ofth eplan tneed s tob e exposed to lowtemperature .Windin g arubbe r tubing around thecrow no fbeet ­ roots justbelo w thepetiole s andrunnin g coldwate r throughth ecoils ,mad e all plants flower,wherea s control plants did not flower atall .Coolin g the lower part of the root caused only 10% ofth eplant s to flower.Als o forothe r rosetteplants ,chillin g ofonl y theape xwa s sufficient toobtai n flowering (Curtis& Chang , 1930).However , the influence ofvernalizin g tem­ peratures is not restricted to apices alone,sinc eWellensie k (1964a)de ­ monstrated thepossibilit y ofvernalizin g leafcutting s of Lunaria biennis. Theearlie r concepttha tdividin g cellswer e aprerequisit e forvernalizatio n was almostabandone dbecause ,i nsom eplants ,vernalizatio n tookplac eunde r conditions thatpracticall y excluded celldivision . Instead itwa s thought possible that also cells preparing themselves fordivisio ncoul dperceiv e the vernalizing action oflo wtemperature .Likewis e treatmento froo tseg ­ ments of Cichorium intybus, which areknow n forrapi d regeneration, resulted in flowering, although itwa s lesseffectiv e thansee do rplan tvernaliza ­ tion (Wellensiek, 1964b). Asmitoti c cells seem topla y animportan t role inth evernalizatio n pro­ cess, some authors prefer tospea ko fa vernalize d conditiono fth eplant , rather than toassum e adiffusibl e substance accumulating during coldtreat ­ ment. Grafting trials totransfe r thevernalize d conditiono fth edono rt o the receptor almost always failed. If they were successful, photoperiod played an importantrole ,s oi tca nb e argued thatno tth e immediate product of vernalization is transmitted, but the flower-inducing end-product. Barendse (1964)conclude d from his trials with Cheiranthus allionii that the direct vernalization effect was immobile andwa s translocatedb y cell divisiononly . One way of studying vernalization is to measure organic substances in vernalized and unvernalized plants.Howeve r one cannot be sure whethera change inconten to fa substanc e isdu et overnalizatio n (the flower-inducing process), to 'coldmetabolism ' or to other differences betweenvernalize d andunvernalize d plants.Adequat e testsmus t include also formsno t requiring cold. Carbohydrates,especiall y sucrose,wer e supposed topla y an important role invernalization , since Gregory & Purvis (1938a,b)demonstrate d that vernalization of caryopses of Petkus winter rye failed in theabsenc eo f oxygen and sufficientsugars . Nowadays many authors try to relate thevernalizatio nproces swit hDN A andRN A synthesis.Accordin g toBesnard-Wibau t (1977)inductio n atlo wtem ­ perature specifically actso nth e axial cells ofth e shoot apexo f Arabidop- sis thaliana, where theDN A synthetic capacitywa s increased.Als o Shiomi & Hori (1973)observe d an increase inDN A synthesis invernalize dbarle y seed­ lings. Inwhea t embryos,Tateyam a et al. (1978)foun d anincrease d content of RNA and DNA during cold treatment compared to germination ata norma l temperature. Numerous reports mention an increased or decreased content of several growth substances invernalize d plants,especiall y the gibberellins areofte n linked with thevernalizatio n process. Insugar-beet ,Margar a (1967)showe d that application ofG A tounvernalize d plants could lead to stemelongation . Infull yvernalize d plants,G A advanced thedat e of flowering and increased the number of flowerbuds .Afte rvernalizatio n periods too short toobtai n total flowering,G A applicationwa s effective tocomplet e thevernalizatio n process (Margara, 1960;Gaskill , 1957). Although gibberellin couldb ein ­ volved in the vernalization process,Margar a (1960)di dno tdetec tdiffer ­ ences in gibberellin contentbetwee nvernalize d andunvernalize d plantsb y severalbiologica l tests.Sug e (1970)measure d adouble d contento fgibberel ­ lins in response to vernalization of radish seeds or seedlings.Thi smay , however,hav ebee n anindirec teffec to fvernalization .Accumulatio n ofgib ­ berellinsma y also occur as aconsequenc e ofvernalization ,whe n precursors of the gibberellins areproduce d during thevernalizatio n treatment,o ri t mighteve nb e adirec t consequence ofth e flower initiationproces s itself. Moreover,gibberellin s cannotb e considered asrea l flowering hormones,thei r main effect being stem elongation, even inunvernalize d plants undernon - inductive conditions. In Hyoscyamus ni ger, gibberellinparticipate s inth e mechanism of flowering only by its indirecteffec to n stem elongationan d does not act directly on flower formation itself, according to Mugnier (1977). In sugar-beet, content of gibberellin inapice s increased sharply justbefor e or aftervisibl e bolting (Lentone t al., 1975), suggesting that gibberellins are somehow involved in the expression of the vernalization stimulus. A different approach tovernalizatio n ist orevea l more ofth e kinetics of the process by studying different conditions before, during and after cold treatment. The effectiveness of a certain environmental conditiont o induce 'ripeness to flower' canb emeasure d asth enumbe r of leavesproduce d before flowering, as the proportion of plants flowering or bolting ora s the time from chilling to firstvisibl e symptoms of flowering. Insugar-beet , theproportio n ofplant s flowering isofte nuse dbu tca n depend alsoo ntem ­ perature and light phase (photoperiod)afte rvernalization . Ifther e isa n interactionbetwee neffectivenes so fcol d treatment andclimati c conditions after vernalization, the proportion ofplant s floweringwil l notb e agoo d measure ofth e effectivenesso fth e applied cold treatment.

The effect of cold treatment can depend on the presence of ajuvenil e stage, temperature, duration of cold treatment, and on temperature after vernalization. A juvenile stage,i nwhic h sugar-beet is lessresponsiv e to lowtempera ­ turewa s reportedb yChrobocze k (1934), who foundtha tth eyounge r theplant s atth ebeginnin g ofcol d treatment,th e lowerth eproportio n of seedstalks . He also suggested that low temperature might have no lesseffec t attha t stagebu ttha tsubsequen t devernalizing temperatureswer emor eeffectiv e in youngplants .Margar a (1960)foun dn o floweringwhe nplant swer evernalize d atth ecotyledonar y stage,no teve nunde r subsequent continuous illumination. However in his trial temperatures after vernalization were rather high (18-22 °C), so devernalization may have played arole .Gaskil l (1963)an d Curth (1955), the latter with Steckling beet,observe d thatag eo fplant s atth etim eo fcol d treatmentwa spositivel y correlated with theproportio n ofbolters . Inbeetroot ,Junge s (1959), alsoobserve d thatolde rplant s couldb ever ­ nalized more readily. In his trial, there could, however, havebee nsom e induction ofolde rplant sbefor evernalization , astemperatur e during raising wasrathe r low,10-1 5 °C.Th e samehold s forVos s (1936)fo rfodde rbeet . Wood & Scott (1975) sowed sugar-beet in autumn but encountered excessive bolting inth e following spring,excep t forplot s sownlat ei nautumn ,per ­ hapsbecaus e ofa juvenil estage . By contrast, Heide (1973) showed that beetrootswer e responsivet olo w temperature at any age,thoug hth e sensitivity tochillin g increasedsome ­ whatwit h age.Kloe n (1952)an dWiebosc h (1965)indee d found thateve n seed of sugar-beet couldb evernalized .Eve n inimmatur e seedo nth emothe rplant , vernalization seems possible (Lexander, 1969; O'Connor, 1970; Scott & Jaggard, 1978;Bosemark , 1970).

The literature does not agree on optimum temperature ofvernalization , ofwhic h severalhav ebee n reported. Curth (1960)state s 3 °Cfo r Steckling beets, Fife & Price (1953)6 °C,Bachman n etal . (1963)8 °C,Stou t (1946) 6-9 °C,Curt h (1962)normall y 4 °Can dwit h a 'photothermic' treatment(si ­ multaneous low temperature and long light phase) 8 °C and Lasa & Silvan (1976)wit h 'photothermal'treatmen t also8 °C .

The duration ofvernalizatio n influencesproportio n ofplant s flowering. Plants longer exposed to low temperature, morebol t and the firstbolter s appear sooner after cold treatment (Curth,1955 ;Heide ,1973 ;Wellensie k & Verkerk, 1950). Theminimu m durationo fth ecol dperio d depends oncultivar , since unsusceptible cultivars require farstronge r induction thansuscept ­ ibleones .

Chroboczek (1934) wa sprobabl y the firstresearc hworke rwh o didsystema ­ tic researchwit hbeetroo t onth e influence oftemperatur e after vernaliza­ tion.Hig h temperature (21-27 °C)tende d torevers e theeffect s ofpreviou s cold treatment. But this reversal could be counteractedb y extending cold treatment and also byextendin g the lightphas e upt ocontinuou s illumina­ tion.Heid e (1973)als omention s interactions betweendaylengt h anddeverna - lizing temperatures in beetroot: shorter light phase allowed reversal at lower temperature (18 °C or more) than did alonge r lightphas e (24 °Co r more). Apart from the influence of high temperature as such,duratio no f the warm period isdecisive :Curt h (1960)foun d complete reversalwit h2 5 days at3 0 °C.Curt h also found thatth eregio n ofth egrowin gpoin t isth e receptor oftemperature , as forvernalization . Incrop s likewhea t or rye,devernalizatio n occurs especially whenplant s are exposed to high temperature immediately aftervernalizatio n (Purvis & Gregory, 1952). After alon gcol d treatment orwhe nvernalizatio n is followed by arathe r shortperio d of intermediate temperature (12-15 °C),th eproces s ofvernalizatio n was assumed tob e fixed andhig htemperatur e could notexer t a devernalizing influence anymore.Thi s fixation isofte n called stabiliza­ tion. In Arabidopsis thaliana, however,devernalizatio n canoccu r irrespec­ tive ofth e lengtho fth evernalizatio nperio d (Napp Zinn, 1957). ButNap p Zinnshowe d that also in A. thaliana thedevernalizin g action ofhig htemper ­ ature strongly depends onth e timeelapse d afterth een d ofth e coldtreat ­ ment, immediately aftervernalizatio n being themos teffective . For beet, there is little evidencewhethe r astag eo rconditio n iseve r reached where vernalization is fixed or 'stabilized': in certain other plants, certain treatments result in fixation and reversal isn o longer pos­ sible.

2.3 LIGHTPHAS E

Photoperiodism is a response toth ephas e andperio d of lightan ddark ­ ness. Incident radiant energy above acertai n threshold iso f secondary im­ portance. When sugar-beet is vernalized, it behaves as a long-dayplant : stem elongation and subsequent flowering isadvance d and accelerated inpro ­ portion to light phase. Response depends largely on duration ofpreviou s cold treatment.Usuall y there isn o response withoutcol d treatment. According to Schneider (1960), the apex of sugar-beet starts todiffer ­ entiate only after cold treatment. During cold treatment lasting 84days , he observed no changes in the structure of the growingpoint .Obviously , after chilling, a certain time with suitable lightphas e and good growing conditions were required for actual stem and flower formation. Curth (1960)measure d the influence ófdaylength . Insugar-bee t chilled for a given time,continuou s illuminationwa s slightly superior to aligh t phase of 21h . With shorter light phase,respons e decreased sharply until with 8h stems didno telongat e andn oplant s flowered. Also after the stem started elongating, a long photophase was still required. Margara (1960) reported a standstill of stem elongation and a delay in flowering after transfer from long to shortphotophase .Usuall y theplant s startt o forma rosette of leaves again andeve n the rootstart s toswel l again atth etop . Margara observed thattransfe rbac k to longphotophas e afterprotracte dsho i photophase no longer provoked flowering. This phenomenon is often called 'SD devernalization' but Irestric t the words vernalization and reversal (devernalization)t otemperature-dependen tprocesses . Research workers usually study lightphas e with incandescent tungsten lamps toexten dnatura l daylight.Fo r sugar-beet Curth (1960)foun d lowlu ­ minous fluxdensitie s (areic luminous flux)t ob e sufficient.Abov e 100-200 lx, anincreas e influ x fromtungste n lampsdi dno t advancebolting .Durin g that extension of daylight, such a flux allows negligible photosynthesis and avoids differences ingrowt hrate .

Curth (1960)als o studied the influence ofspectra l distribution ofsever ­ alligh tsource s used fordaylengt h extension.Mos teffectiv e forphotoperi - odical responses were sources with apea ko f luminous fluxdensit y inth e blue or orange-red regions. Sources with a peak in the greenregio nwer e lesseffective . Lane et al. (1965)compare d red light (wavelength 600-700nm )wit hfar - red light (700-770nm ) for daylength extension. In an annual sugar-beet strain, far-red extensionwa smor e effective toobtai n floweringplant s than red light. Inothe rplants ,however , likehenban e andpetunia ,a mixtur e of red and far-red was superior.O fth e luminous flux from incandescent lamps inth ewaveban d 600-770nm ,hal f isre d andhal f far-red. Inflora l initiation,plant sperceiv ephotoperio d by theleave s (Withrow et al., 1943;Naylor , 1961;Lang , 1952). Thepigmen tphytochrom e islikel y to be involved as the receptor of thedaylengt h stimulus,considerin g the red and far-red actionspectra . For phytochrome, there are two types ofreactions : induction-reversion reactions and 'high-irradiance' reactions. Involvement ofphytochrom e ina ligthmediate d response requires thata ninductio neffec tb y apuls e ofre d lightca nb e fully reversed by asubsequen tpuls e of far-red light.Wit h an exposure time of lesstha n 5min ,th e action spectrum for theinductio no f a light-mediated response shows a peak at wavelength 660n m and reflects theabsorptio n spectrum ofth ered-absorbin g form ofphytochrom e (P_). Thehigh-irradianc e reaction occurswit hprolonge d irradiation (e.g.day - lengthextension) . Its actionspectru m iscompletel y different from thein ­ duction spectrum: always apea k inth e far-red regiono fth espectru m (wave­ length 700-730nm ) and several peaks in the blue region of the spectrum (Schäfer, 1976). Curth (1960)an d Lanee t al. (1965)reporte dpeak s ofth e action spectra in theblu e and red to far-red regions forbeet , suggestive ofth ehigh-irradianc e reaction.

2.4 INTERACTIONS BETWEEN LIGHTPHAS EAN D TEMPERATURE

Themos tcommo n interactionbetwee n theeffect s ofdaylengt h andtemper ­ ature istha tunvernalize d plants fail torespon d to lightphase .Th e 'cri- tical photophase1 seems toshif twit h the degree ofvernalization . Forun - vernalized plants, the threshold is apparently sohig h thatth e stem does notelongat ewit h longphotophases . Inothe rwords ,th evernalize d condition seems to lower thecritica l photoperiod: sufficiently vernalized plantsd o not need an extremely long photophase tobolt .Fif e& Pric e (1953)showe d complete substitution ofvernalizatio n and lightphas e for sugar-beet.Whe n Steckling beets were vernalized at 4 °C for extremely long (100-300d) , plantsbolte d and flowered at2 1 °C,eve n incomplet edarkness . For beetroot, Heide (1973)reporte d bolting and flowering in continuous light,withou t any cold treatment attemperature s usually considered as 'ver­ nalizing',bu ttemperatur e inhi stria lwa s not above 18°C .

2.5 GROWTHAN DBOLTIN G

Several research workers havepointe d out aconnexio nbetwee n growth and bolting insugar-beet .Usuall y avigorou s early growth leads tomor ebolter s; inhibiting circumstances on the other hand reduce bolting. According to Röstel (1968)ther e is apositiv e correlationbetwee n theproportio n ofbol ­ tersan d soil fertility.Röste l alsoreporte d apositiv e influence ofirri ­ gation on bolting. He stated that bolting resistance wasmor eneede d when growth-stimulating cultural practices were successful. Many reports deal with the influence ofnutrien t supply,especiall y nitrogen,-onth e proportion of bolters (Mann, 1951;Gorodni i & Sereda, 1975;Hoekstra , 1960;Lüdecke , 1938; Schneider, 1960;Lysgaard , 1978). Also an influence hasbee nreporte d ofplan t density,whic h affectsin ­ dividual plantgrowth , (Jorritsma, 1978), lowerplan tdensitie s causingmor e plants to run to seed. Warne (1949)observe d ina nexperimen t thatborde r rowplant s had agreate r tendency tobolt .Dowke r &Jackso n (1975)observe d the samephenomen a incarrots . Good growing conditions clearly promote the tendency toru n to seed for thesespecies .

10 3 Models for bolting

Several models have been elaborated for the flower-promoting actiono f lowtemperatur e ando fsubsequen t lightphase .

3.1 MODELS INLITERATUR EBASE DO N (HYPOTHETICAL)PLAN THORMONE S

Lang& Melcher s (1947)an dPurvi s& Gregor y (1952)conceive d oftw oreac ­ tionswithi nth eplant :synthesi s of aflower-promotin g substanceB and des­ truction ofthi s substance (Figure1) . Reaction 1wa s thoughtt oprocee d evena tlo wtemperatures ,bu tth erat e ofReactio n 2i slo w atlo wtemperature s and increaseswit htemperatur emuc h more rapidly than forReactio n 1.A t lowtemperature , substance Bwoul dac ­ cumulate and at higher temperature destroyed (converted to C, orperhap s back to A). Atmoderat e temperature,B wa s thoughtt ob e converted byReac ­ tion3 to D, which was not destroyed by high temperature.S oafte r afe w days atthes emoderat e temperature reversalwa sn o longerpossible . Reaction 3migh tb e governedb y lightphase ,wherea sReactio n 1an d2 migh tb einde ­ pendent of daylength. Substance Bmigh tb e identical withth e hypothetical substance vernalin, considered tobuil d updurin gvernalizatio n (Melchers, 1939), andD migh tb eth e final flowering substance florigen. Inthi smode l 'vernalin' isa precurso r offlorigen . Anothermode l toexplai nexperimenta l resultswa s thato fNap pZin n (1957). He assumed themor ecomplicate d system of several 'labile'an d 'stable'stage s invernalization , someo fwhich ,however ,coul db ebypassed . As Chouard (1960) states: 'theseformulation s arehand y tomemoriz ean d they stimulate further investigations onhypothetica l substances,bu tthe y provide no more clarification than the authors' description of theirow n results.Furthe r theyma y require adjustment foreac hne wdiscover y andthe y also change for each plant type thatbehave s ina particula rwa y anddoe s not fitth eparticula r representation'.

B

Fig.1 .Reaction si ninductio no fflowerin g(Lan g& Melchers ,1947) .Reactio n1 ,synthesi s ofa flower-promotin gsubstanc eB ;Reactio n2 ,destructio no fth esubstance .

11 A major assumption of the models is build-up of a substance, usually call­ ed vernalin, during vernalization. This substance could be a precursor of florigen or could, together with light phase, regulate florigen formation. However there is no experimental evidence for 'vernalin', in contrast to the final flowering hormone(s).

3.2 A MATHEMATICAL MODEL FOR BOLTING IN SUGAR-BEET

To account for the experimental results for sugar-beet of Chapter 5, a different approach was necessary: as it is likely that the flowering process as a whole is a dynamic, continuous and quantitative process, a relational diagram was drawn with the conventions of Forrester (1961), like models used in simulation studies (Figure2) . The rectangles in the model represent a quantity of specific substances. Such a quantity is subject to change, the rates of which are indicated with valve symbols. Factors influencing the rate of these changes are drawn with dotted arrows. Flow of material (substance) is drawn with solid arrows. The model assumes two substances:

V: A substance resulting from vernalization. F: A final flowering substance. The term flowering substance needs explana­ tion as this publication reports only 'bolting'. In the trials, bolting was always followed by flowering. According to Heide (1973) the period between visible bolting and flowering is almost constant. Stem elongation and flower­ ing may, however, differ in physiological mechanism. An illustration was given by Curtis (1964), who showed in a grafting experiment that flowering could occur without previous stem elongation. Under normal circumstances, however, it may be assumed that conditions favouring the synthesis of sub­ stances involved in stem elongation also favour the synthesis of the flower­ ing substance.

The model distinguishes the following processes.

Process I (vernalization) is the synthesis of a substance V at a rate that is positively temperature-dependent but still proceeds at low temperature.

Process II (devernalization) is the breakdown of V at a rate that is also temperature-dependent. At low temperature, Process II is slower than Process I. So after a long time in the cold, a considerable amount of substance V is available, because Process II is practically still.

Process III is the synthesis of a final flowering substance F. After vernali­ zation, synthesis of F starts, if temperature is raised and light phase (pho- toperiod) is long. The rate of synthesis of F is (in this quantitative model) determined by three conditions.

12 PROPORTION BOLTING

Fig. 2. Relational diagram of a simple model for the bolting process in sugar-beet. Sub­ stance V is hypothetical and is involved in vernalization. Substance F is thought to be associated with one or more flowering hormones. Solid lines mean flow of substance. Dotted lines are transfer of information: e.g. dotted line e means that the amount of V, together with photophase (f) and temperature (g), regulates the rate of synthesis of F.

- The light phase influences the rate of synthesis of F, longer lightphase speeding up this rate (relation f in the model). - After longer cold treatments, bolters appear sooner and in larger number. So there must be a relation between the amount of V and the rate of synthe­ sis of F (Figure 2, relation e): the larger the amount of V, the higher the rate of synthesis of F. - Temperature must also influence synthesis of F, since most biochemical reactions are temperature-dependent, higher temperatures accelerating syn-

13 thesis and since synthesis also might uepend on the growth rate of the. plants, which is itself a temperature-dependent process. Rapid expansion of leaf area or leaf number accelerates reception of the light phase. Favour­ able conditions of growth have increased the number of bolting plants in some cases (relation g in the model).

The following mathematical eguation fulfills all three conditions and gives the momentary rate of the synthesis of F at any time:

dF/dt ~ kQ V *p (1) in which

F, V is substance content of F and V

t is time k is a temperature coefficient k a photophase coefficient

The temperature coefficient increases with temperature and the photophase coefficient with light phase.

The model simulates the following observations of other workers. Unvernalized plants do not bolt, although enough leaf area is produced to perceive a suitable light phase. Vernalized plants (with enough V) do not bolt with a short photophase, which does not allow synthesis of F. Under certain conditions higher temperatures advance bolting, for in­ stance after long cold treatment. Then plants start to bolt earlier under better growth conditions.

Rates introduced depend on state variables. For instance, rate of synthesis of F depends at any moment on the amount of V. As the amount of V is continu­ ously changing, not only during vernalization but also after vernalization, the direct relation of content of V with rate of synthesis of F implies that the rate of increase in F is also continuously changing. So a plot of the content of F against time can have different shapes according to temperatures after vernalization, duration of chilling and photophase. When conditions are favourable for bolting and in the right sequence, enough F is produced and bolting begins when a certain threshold content of F is reached. This threshold is probably subject to variation between plants, according to genotype. When F is produced at a low rate, with short photo- phases or after less cold treatment, the interval from appearance of the first bolter and the final one will be long. When F is produced very rapidly, however, the threshold will be passed quickly for all plants, so that they

14 will bolt in a few days. For example,wne n sugar-beet isgrow n forseed , plants vernalized during thewinte r haveproduce d aconsiderabl e amounto f V; inMa y orJun ewhe nth e lightphas e isadequat e and theraise d temperature has allowed formation ofsevera l leaves,synthesi s ofF isunlimited .

Although themode l isprobabl ymuc h toosimple ,i taccount s formos to f the observations reported inth e literature. Its framework canb e extended and modified to results reported in Chapter 5. The model canserv e asa n aid in quantitative interpretation ofthes e results.Althoug h thepropose d model isbase d onth eexperimenta l results,i ti spresente d already atthi s pointo fth emanuscrip tt omak e iteas y forth ereade rt oinstantl y compare theresult s ofth eexperiment s with themodel .

15 4 Materials and methods

4.1 PLANT MATERIAL

The cultivars used in growing rooms, greenhouses and in the field differed widely in susceptibility to bolting (Table 1).Becaus e sugar-beet is cross- pollinated, each cultivar, even a single cross, would include different geno­ types. Most trials were with four single crosses (G1-G4) supplied by the breeding firm D.J. van der Have B.V. (Kapelle). Commercial cultivars were sometimes included. For most indoor trials, each treatment consisted of 20 plants in two replicates of 10 placed randomly. Trials were of factorial design.

4.2 FIELD TRIALS

Triald fields were at three sites near Wageningen in the years 1975-1978: - Wageningen-Hoog on a coarse-sand soil with a low content of organic matter. - Wageningen (Haarweg, on the western outskirt of Wageningen) on a heavy-clay soil. - Achterberg (near Rhenen to the west of Wageningen) on a black peat soil. The experimental design was randomized blocks or split plots. In 1975 and 1976, daylength was extended in the field with incandescent light from bulbs of 100 W at a height of 1.5 m above the ground and with 1 bulb for every 2 5 m .A t plant height this resulted in an areic luminous flux of 50-110 lx,

according to distance from the bulb. For technical reasons, the assigned

subblocks could not all be lit at the same time. So, some were lit from sun-

Table 1. Cultivarsuse d inth eproject . Code Cultivar Type Bolting Breeder resistance

Gl P2272 singl s cross strong D.J. van der Have G2 35848-74 singl s cross rather stro ng D.J. van der Have G3 35872-74 singl s cross low D.J. van der Have G4 P6672 singl s cross very low D.J. van der Have G5 Donor comm. cultivar strong Hilleshög Fro G6 Kawepoly comm. cultivar rather low Kleinwanzleben G7 Monohil comm. cultivar rather stro ng Hilleshög Fro G8 Monokuhn comm. cultivar rather low Kuhn & Co G9 Polykuhn comm. cultivar rather low Kuhn & Co G10 Zwaanpoly comm. cultivar rather strong Zwaanesse Gil MK711 test cultivar low Kuhn & Co

16 Light phase(h) 21

daylength • extension

natural daylength N• civil twilight

natural daylength 52° N

April May June July Aug. Sept. Fig.3 .Daylengt hi nth eNetherland si nth eperio dApril-Septembe r(LI )an dligh tphas ei n fieldtrial sa tWageninge nextende dwit hincandescen tbulb s (L2). settil l 00h30, theres t from OlhOO till sunrise.A tWageningen , astronomical midnight is 00h45.Th e sequence was reversed everyweek .Th econtro lsub - blocks were protected from illumination by their distance from the light source (atleas t3 m forth eneares tplots )an db yplasti c shades.Th eday - length (lightphase )s oobtaine d isshow ni nFigur e 3.Fo rth eunli tplots , thegrap ho fdaylengt h +civi l twilightprobabl y reflects theeffectiv eday - length.

4.3 INDOOR TRIALS

The Department of Field Crops & Grassland Science of theAgricultura l University at Wageningen provided growing rooms for raising plants, cold treatment and subsequent treatments asfollows .

4.3.1 Growing rooms

Sixgrowin g rooms4. 5 mx 3. 2 mx 2. 2 m, regulated intemperature ,humid ­ ity and partial pressure of carbondioxide, wereused .Arei c radiantpowe r 2 in the waveband 400-10.000n m was 125W/ m and inth ewaveban d 400-700n m 80W/m 2 from 187 'TL' MF 140W/3 3 + 17 'TL'M 140W/3 3 fluorescent tubes, togetherwit h 18 incandescentbulb s ofelectrica lpowe r 150W and 18incan ­ descent bulbs of 100W . That areic radiant power in thevisibl e spectrum corresponded to an areic luminous flux of about2 0klx , according todat a of Gaastra (1959). In the middle ofthes e cellstemperatur e couldb e kept constant to within about0. 4 °C.Nea r thecel lwalls ,however , temperature could deviate about 1 °Cfro mth e desiredvalue .B ymovin g theplants, place d oncarts ,a tleas ttwic e awee k around thecells ,difference s due topositio n were reduced.Thes e cellswer euse dmainl y forraisin gplant s and fortreat ­ ments afterchilling .

17 4. 3. 2 Rooms for cold treatment

Four rooms were used for cold treatment, equipped like those used for raising plants, lit by 8 HPLR bulbs and 4 incandescent lamps of electrical power of 100 W, giving an areic radiant power in the waveband 400-10000ran 2 2 of 45 W/m and in the waveband 400-700 nm 23 W/m .Th e areic radiant power in the visible spectrum corresponded to an areic luminous flux of about 6.6 klx. Unless otherwise stated, cold treatment was at 3 °C. 4. 3. 3 Greenhouses

2 Three greenhouses with a floor area of 40 m each, were used. Temperature was controlled to within 0.3-0.5 °C at night and 2 °C with low radiation during the day and 4-5 °C or even more in summer with high solar radiation. In winter, natural radiation was supplemented by 32 HPLR bulbs, giving an 2 additional radiant power in the waveban;he d 400-10.000 nm of 45 W/m and in the 2 waveband 400-700 nm of 23 W/m

4. 3. 4 Small greenhouses

2 Three smaller greenhouses with a floor area of 30 m each were used, in which 20 HPLR bulbs could be installed if supplementary lighting was needed.

4.4 LIGHT CONDITIONS

The photosynthetic phase of the daily cycle was extended with normal in­ candescent bulbs of electrical power 100 W, giving an areic radiant power 2 in the waveband 400-10.000 nm of 5.3 W/m and in the waveband 400-700 nm of 2 0.88 W/m and an areic luminous flux of about 210 lx. The extension was di­ vided equally before and after the phase of high-intensity lighting. The light phase will be annotated below as 2 numeric values, the first the phase of high intensity and the second the phases of extension. For instance, 14 h + 4 h indicates a total photophase (TP) of 18 h, 14 h of high intensity and 4 h of extension. The greenhouses could be covered with blinds during the dark phase to exclude natural light or artificial light from adjacent green­ houses.

4.5 RAISING AND CHILLING OF PLANTS

Plants were grown in a mixture of equal volumes of sandy soil (pH of KCl extract 5.6; mass fraction of organic matter 3.8 %) and peat soil (pH of KCl extract 4.6; organic matter 62 %). The KCl extract of the mixture had a pH of 5.2. Sugar-beet was sown either directly in plastic square 2L pots (0.13 metre cube) or in plastic boxes 0.46 m x 0.31 m x 0.16 m and trans- planted into the square pots after chilling.Whe n seedwa s sown inboxes , heat-sterilized soil was used to prevent infectionb y soilpathogens .Te n rowswer e sowni neac hbo x and afteremergenc e thinned inth e rowt o adis ­ tance of 2c m between seedlings.Usuall y theplant swer e grown inth esi x raising rooms (Section4.3.1 )wit h aligh tphas e 14h + 0 h an d temperature 15 °Cunti l the firsttru e leaveswer e about2- 3 cm.Boxe s orpot swer e then transferred to the cold rooms.Whe n chilling or lightphas ewa s notunde r test, atemperatur e of3 °Can d aligh tphas e of1 4h +0 h wa sused . When the duration ofchillin g orplan tag ebefor e chillingwa s afacto r in the experiment, sowing andchillin gwer e started on suchdate stha tth e cold treatments couldb e finished onth e samedate ,s otha tal lplant swer e under identical conditions from thenon .Whe nth eplant swer e sown inboxes , theywer eusuall y kept awee k at 10 °Cafte r thecol d treatmentbefor etrans ­ planting into thesquar e 2Lpot s andtransferre d toeithe r growing roomso r greenhouses,wher e lightphas eo rtemperatur e werevaried , 15° Can d aphas e ofhigh-intensit y of 14h bein g thebasi c conditions. Immediately afterth e transfer, however, the plants were usually keptye tanothe rwee k at1 0° C to acclimatize after transplanting.Th eplant sreceive d 100m l ofa nutri ­ tion solution at least once a week according to growth stage (Table2) . Although the plants weregrow n insmal lpot s to save floor space andwoul d lose some nutrients by percolation, they grew vigorously in alltrials . Times ofchange s incondition s during growth areexpresse d as fara spossi ­ blewit h respectt oth edat eo f sowingunti l cold treatment andwit h respect toendin go fcol dtreatmen tthereafter .

Table2 . Compositiono fth enutrien tsolutio nfo rsugar-bee tgrow nindoors .

8 ml Ca (N03)2 (1mol/1 ) 8 ml KNO (2mol/1 ) 5 ml KH^PO, (1mol/1 ) 2 4 3 ml ofa solutio ncontaining :

2 gMnCl 2.4H20/l 3 gH3BO3/ I 0.5g ZnS O.7 H0/ 1

0.1g CuSO 4-5H20/l

0.1g Mo0 3/1 2 mlo fa solutio ncontainin g3 5gFeEDTA/ 1

5 mlMgS0 4( 1mol/1 ) 69 mlH O

100m l

19 4.6 OBSERVATIONS

In the field and indoors number ofbee tboltin gwa s recorded regularly, sometimes twice a week. A plant was considered abolte rwhe n theste m had elongated by at least 5cm .Whe nbatche s were harvested periodically, plants were separated in leaf blades, petioles andbee t root.Lea f areao fyoun g plantswa smeasure dwit h anelectroni c device and ofolde rplant sb ymatchin g the leaves againstphotocopie s of leaves ofknow n area.Yiel d ofdr ymatte r was estimated by drying at10 5 °C fora tleas t 18h .

4.7 STATISTICS

Datawer eprocesse d onth eDEC-1 0compute r of theAgricultura lUniversi ­ ty, either with Fortran programs or with SPSS (Statistical Package ofth e Social Sciences).Dat a onth eproportio n ofbolter s were transformated (arc- sine-method) for statistical analyses toobtai n amor e normal distribution (Bliss, 1937).

20 5 Results

5.1 THEJUVENIL E STAGEAN DVERNALIZATIO N

Thepossibilit y ofa juvenil e stage insugar-bee t iso fconsiderabl ere ­ levance toundesire d bolting.Figur e4 showsth e average course oftempera ­ ture during spring in the Netherlands. If temperatures below about 11° C vernalize,th eproces s cancontinu e forth ewhol e dayunti l about1 0April . By then,plant s inth eNetherland s areusuall y germinating or inth ecotyle - donary stage. The following trialswer edesigne d totes t forvernalizatio n inthes e early stages.

5. 1. 1 Effect of daylength and covering with plastic

5.1.1.1 Introduction

In 1976 atWageningen-Hoog , emergence was accelerated by covering the seed-bedwit h aperforate d polyethene sheeting. Ifa juvenil e stageexists , such a treatment should increase proportion of bolters because itwoul d shorten thisstage .Th e lightphas ewa s alsoextende dwit h artificial light (Section4.2) .

Temperature CO 26 f maximum temperature 22 18 14 / 10 -—. / _,•"" minimuf.m 6 ' /•"" temperature 2 0 -2 * MARCH APRIL MAY JUNE JULY AUG. Fig.4 .Averag ecours eo ftemperatur e (minimuman dmaximum )fro mMarc ht oSeptembe ri nth e Netherlands (datafo rD eBil tnea rUtrecht) .

21 5.1.1.2 Materials and methods

The experimental scheme (split plot) included the following factors:

Covering of the seed-bed: Ml: control M2: covered Light phase (main factor): LI: natural daylight L2: light phase extended to about 20 h (Figure 3) Two genotypes: Gl and G4 (Table 1) Sowing dates: Tl = 2 March T2 = 17 March T3 = 1 April T4 = 15 April Two replicates.

2 2 The trial consisted of 64 plots, each of area 30 m (net area 20 m )i n 2 2 the lit parts and 42 m (net 30 m )i n the unlit parts. Directly after so­ wing, the assigned plots were covered with the sheeting. As soon as seed­ lings emerged the sheeting was removed, in order to protect the seedlings from too high temperatures, which can easily occur under plastic. After -2 emergence, the plants were thinned to a stand density of 9-10 m (rows 50 cm apart). Fertilizers were applied on a normal basis.

5.1.1.3 Results

Measurements with a temperature recorder showed that on sunny days the plastic raised maximum temperatures at sowing depth to about 4 °C above those of controls. Germination took much less time (Table 3).O n 17 May, covered plants had grown to almost twice the plant mass and leaf area of the control plots (Table 4).Sowin g date and covering both had a signifi­ cant effect on mass and area (P < 0.001). Although it could be expected that these bigger plants would have been

Table3 . Effecto ndat eo femergenc eo fcoverin gth eseedbe dwit hperforate d transparantplasti c sheetingdurin ggermination .

Sowing date Date of emergence Time of Re rmination (d) (month-•day) (month-day) control mulched control mulched

03-02 04-09 03-27 38 25 03-17 04-15 04-04 29 18 04-01 04-19 04-12 18 11 04-15 04-29 04-23 14 8

22 Table 4. Leaf area and mass of dry matter per plant for covered and open plots on 17 May 1976. Average for the genotypes Gl and G4.

2 Sowingdat e Leaf area (c( m,) Masso fdr ymatte r (g) (month-day) open covered open covered

03-02 27.7 52.1 0.28 0.52 03-17 20.8 36.2 0.22 0.39 04-01 15.0 22.6 0.15 0.24 04-15 3.1 7.2 0.04 0.08

vernalized more effectively, the opposite was true (Table 5): the covered plots bolted much less (P < 0.05). Perhaps covered plants were devernalized in the few days after emergence and before removal of the plastic mulch, especially for the third sowing for which emergence started on a Friday but the cover was not removed until Monday. With the sunny weather of that week­ end air around the plants would have been very hot. Such an explanation would assume either that vernalization proceeded during germination or that the seeds were already more or less vernalized when sown. Alternatively vernalization could not proceed to such an extent as in the controls during germination because of the higher soil temperature. That explanation requires also that vernalization was already proceeding during germination. Extension of the light phase to about 20 h induced more bolting (P < 0.025, Table 5) and may thus be a useful tool for breeders to test bolting behaviour of their plant material or to be used in more specific selection methods. In 1976, the effect of sowing date was marked (P < 0.001), especial­ ly in the cultivar unsusceptible to bolting, presumably because of the very high (devernalizing) temperatures during spring and summer. Cultivar Gl and G4 showed large difference in tendency to bolt (P < 0.001).

Table 5. Influence of covering, light phase and sowing date on proportion of bolters on 23 August 1976 for 2 cultivars.

Genotype Cover Natural photophase Light phase extended to 20 h sowing date (month-day) sowing date (month-day) 03-02 03-17 04-01 04-15 03-02 03-17 04-01 04-15

Gl 6.8 0.3 0.0 0.0 12.9 3.2 0.0 0.0 Gl 6.6 0.0 0.0 0.0 8.8 0.3 0.3 0.0

G4 88.9 77.0 43.8 23.0 94.6 88.6 71.3 50.9 G4 89.0 66.0 21.8 17.7 100.0 86.1 58.4 41.5

23 5.1.2 Effect of germination temperature

5.1.2.1 Introduction

In1977 ,temperatur e of soil atgerminatio n depthwa s kept down bycover ­ ing field plots withplate s ofpolystyren e insulating material 2c mthick . Thiswa sver yeffectiv e inkeepin g the soiltemperatur e low atsowin gdepth .

5.1.2.2 Materials andmethod s

The experimental field atWageningen-Hoo g was laid out in afactoria l block designwit hplot s 4m x2. 5 m including the following factors:

Three sowing dates:3 an d 15Marc h and 5April ;onl y cultivar Gl (unsus­ ceptible tobolting )wa ssown . Some plots were covered with 3 polystyrene plates (1.22m x 2.4 4 mx 0.02 m) as soon as possible after sowing. Immediately after emergence of the firstplant s ina plot ,th eplate swer e removed. Sowing ofread ygerminate d seed. Seedwa s soaked for2 h inrunnin gwate r at2 5 °C,an d allowed togerminat e for2 0h betwee n two layers of filterpa ­ per ata temperatur e of20/3 0 °C (day/nightcycle )i n an incubator.A tsow ­ ing, almost all seedswer evisibl y germinated. Fourreplicates .

All combinations were included, sotha tth e trial consisted of4 8plots . Temperature was recorded with aFlat-Ba t 24-points recorder incovere d and

Maximumdail ytemperatur e atsowin gdept h(°C )

20

16 y »DO: control plots ( ) ^S-.^ "~ .£-o' ; A»»: plots with polystyrene cover ( ) 12

4 -

1 oU - MARCH APRIL MAY Fig. 5.Maximu m soiltemperature sa tsowin gdept ha sinfluence db ycoverin g (filled symbols) withpolystyren eplate si n1977 . AA,sow n3 March ;DB ,sow n1 5March ,0* ,sow n5 April .

24 Table6 . Influenceo fcoverin gth eseedbe dwit hpolystyren eplate sdurin g germinationo nth edat eo femergence ,arei cnumbe ro fplant s (standdensity ) andproportio no fbolter so n4 Octobe r 1977fo rread ygerminate d andcontro l seed.

Dateo f Ready Dateo f Areic Proportion (month-day) germinated emergence number ofbolter s(% ) (month-day) (m^) sowing covering removal

03-03 control 03-23 9.3 6.5 03-03 control + 03-20 8.1 4.6 Mean (8.7) (5.5)

03-03 03-07 03-22 03-27 3.9 13.8 03-03 03-07 03-22 + 03-25 4.1 13.7 Mean (4.0) (13.8)

03-15 control 04-10 10.3 1.7 03-15 control + 04-04 10.0 1.7 Mean (10.1) (1.7)

03-15 03-15 04-18 04-30 5.0 8.2 03-15 03-15 04-06 + 04-20 3.7 5.6 Mean (4.3) (6.9)

04-05 control 04-30 11.1 0.3 04-05 control + 04-30 10.3 0.0 Mean (10.7) (0.1)

04-05 04-06 05-02 05-04 5.9 0.9 04-05 04-06 05-02 + 05-04 5.2 2.0 Mean (5.6) (1.4)

Significance of the effects of: Sowing date (P< 0 1%) (P< 0.1%) Imbibition etc. (P< 3 9%) ( n.s.) Polystyrene plates (P< 0 1%) (P< 0.1%) open plots at sowing depth. After complete emergence, plants were thinned _2 to obtain, if possible, a stand density of 10 m

5.1.2.3 Results

Withpolystyren e plates,th e seed-bedwa s 4-6 °Ccooler ,somewha tdiffer ­ ent foreac h sowing date (Figure 5). However,th eplate s alsohindere dger ­ mination, possibly because ofsaturatio no fth e seed-bedwit h condensation water from theplate s incombinatio nwit h the lower temperature duringger - -2 mination.Onl y anirregula r standwit h adensit y ofabou t4 m was achieved inth ecovere d plots. Covering delayed emergence andconsiderabl y increasedboltin g (Table6) . The effect on bolting couldb edu eeithe r togerminatio n temperature ort o stand density, since Jorritsma (1978)reporte d 2.4,3. 1 and 3.5 %bolter s _2 for respective stand densities of12.6 ,6. 9 and 3.4m .Th eeffect s Iob ­ served were, however, morepronounce d thanthos e reportedb y Jorritsma.S o 25 veryprobably ,boltin g isenhance d mainly by the lower germinationtempera ­ ture inth ecovere dplots .Covere d plots of the second sowing gave evenmor e bolters than the control plots of the first sowing, asmigh tb e expected from thedifferen t temperatures (Figure 5).Sowin g ofread y germinated seed had no significant influence on bolting. Despite the disturbing effecto f differentplan tdensities ,thi s experiment provided another indicationthat , already during germination, temperature maypla y an important role inachiev ­ ing avernalize d condition.

5.1.3 Effect of plant size on vernalization in the field

5.1.3.1 Materials andmethod s

In 1976, single cross G4 was transplanted into the field atdifferen t stages. The trial (block design with 4 replicates and withplot s ofare a 2 20m )consiste d ofth e followingtreatments . A. Sownnormall y outdoors on 18Marc h andemergin g on2 1April . B. Sown in paper pots (Nippon Tensai Seito Kabushiki Kaisha)o fdiamete r 1.9 cm and a height 13c m on 12Apri l ina greenhous e with respective day and night temperatures of 18 and 15 °C.O n 21April ,plant swer e aslarg e as those sown inth e field on 18Marc h (TreatmentA )an dwer e transplanted toth e field. C. Sown in 16Marc h and transplanted outdoors on 18March , being fortw o days inth e greenhouse atda y andnigh ttemperature s of 18 and 15°C . D. Sown in the greenhouse inpape rpot s on 25Februar y and transplanted to the field on 18Marc h when two true leaves had developed. In the week after transplanting, severe frost damage occurred and transplanting was repeated (with spareplants )a t2 5March . E. Sown inth e greenhouse inpape rpot s on1 2Februar y and transplanted out­ doors on 18Marc h in the 4-leaf stage.Als o inthi s treatment,becaus e ofth e frostdamage ,a ne wbatc hwa s transplanted on2 5March .

The plants were transplanted or, for direct sowing, were thinned toa stand density of 12m .

5.1.3.2 Results

Although on 21April , outdoor-sown (TreatmentA ) and later indoor-sown plants (TreatmentB )wer e atth e same stage ofgrowth , theoutdoor-sow n beet were heavier on 1June , probably because of an initial inhibitionb yth e paperpot .A tth een d ofth e season,plant s ofTreatmen t B,withou t a 'cold' germination period, had bolted significantly less than the outdoor sown plants of TreatmentA (Table 7).Temperatur e during the spring of 1976 is shown in Figure 6. The timing of thedifferen t treatments ispresente d at thebotto m of thisgraph .

26 Table7 . Effecto ftransplantin g atsevera lstage so fgrowt ho nth e masso fdr ymatte rpe rplan to n1 Jun e197 6an dboltin go n2 2September .

Treatment Site Sowing Stage Masso f dry Proportiono f date on2 1Apri l matter (g) bolters(% )

A Outdoors 03-18 cotyledonary 4.2 47.3 6 Indoors 04-12 cotyledonary 1.3 3.0 C Indoors 03-16 2-4lea f 2.1 57.0 D Indoors 02-25 4-leaf 10.3 26.4 E Indoors 02-12 6-leaf 14.7 69.5

Significanceo ftreatmen teffec t (P< 0.01 ) (P< 0.01 )

Again results indicate that vernalization proceeds very early in growth, considering the difference in bolting between Treatment A and B. Beet sown indoors in March and transplanted outdoors two days later bolted more than beet sown directly outdoors (Treatment C and A).Bee t sown early February and planted out late in March (Treatment E) bolted severely, indicating easy vernalization of that older plants, but those sown later in February and planted out at the same time (Treatment D) bolted less than treatment C. Proper comparisons with Treatment A, B and C cannot, however, be made be­ cause of the delayed planting of treatments D and E.

TemperatureC O 30r- .'Maximum 25 ,•Averag e Treatment A \ 20- emergence / \ I ; V

15 oMinimum Treatment A sowing A/A--/ / 10 ... y , S'

,o--° 'J I *" March April May Treatment • A -*B

IHIIIIIIIIItllllllllllllll lllllllllllllllllllllllllllllllllllllll

• = Field iiiiiiii = Greenhouse Fig. 6. Course of maximum, minimum and average temperature in spring 1976.Timin g of the treatments A, B, C,D and E is indicated below the graph. Treatment A is field sown.Treat ­ ment B, C and D: plants were raised in the greenhouse at 18/15 °C forvariou s durations and afterwards transplanted to the field.

27 5.1.4 Effect of plant size on cold treatment

5.1.4.1 Materials and methods

In the raising rooms, plants of cultivars Gl, G2 and G3 (Table 1) were raised to different ages before cold treatment. They were sown in 2L pots and raised at 20 °C with a light phase of 14 h during raising and cold treat­ ment. Sowing dates were so organized that plants were of 5 ages at the be­ ginning of cold treatment and that both cold treatments of 49 and 28 days at 4 °C were completed simultaneously: 51 = Immediately chilled. Sown and immediately transferred to the cold rooms. 52 = Chilled after 2 days. Sown 2 days before transfer to the cold treatment. 53 = Chilled after 14 days. 54 = Chilled after 28 days. 55 = Chilled after 42 days. After cold treatment, the plants were transferred to a greenhouse where the temperature was kept at 10 °C for 1 week and then raised to 15 °C, which could be rigidly maintained. Only towards the end of the trial (90 days after cold treatment) did the sun cause temperature to rise to a maximum of 18 °C during some days. In the greenhouse, the total light phase was 14 h + 10 h = 24 h.

5.1.4.2 Results

At the beginning and end of cold treatment, number of leaves per plant was counted for each treatment (Table 8).Despit e the low temperature, treat­ ments SI and S2 emerged during the cold treatment. Germination may have been stimulated by use of sprinkling water of 10-15 °C. At the end of cold treat­ ment for 49 days, S2 seedlings already had stretched cotyledons, whereas those of SI were just visible. Although plants were genetically as uniform as possible, Figure 7 indi-

Table8 . Numbero ftru eleave spe rplan t (meanfo r3 cultivars )fo ru p to4 2day sa t2 0° Cbefor ecol dtreatmen tfo r2 8o r4 9day sa t4 °C .

Stage Numbero fleave sfo rplant s of treatment chilled for2 8d chilled for4 9d

timeo fraisin ga t2 0° C timeo f raisinga t 20° C

0 2 14 28 42 0 2 14 28 42

Starto f chilling 2 9 12 2 6 11

Endo fc h illing * * 2 10 14 3 10 15

noemergenc evisibl e cotyledonswer evisibl e

28 Proportion bolting(% ) 100 100

80 100 120 80 100 120

100 100 r

20-~^ 120 60 80 100 120 Timeafte r cold treatment (d) Fig. 7. Bolting as influenced by plant age at begin of cold treatment. A. Cultivars Gl and G2 with cold treatment of 28 days. B. Cultivars Gl and G2 with cold treatment of 49 days. C. Cultivar G3 with cold treatment of 28 days. D. Cultivar G3 with cold treatment of 49 days. Age at start of cold treatment 0, 0 days; A, 2 days; V, 14 days; •, 28 days; », 42 days. cates the great variability in bolting date. Longer cold treatment and (in other trials) longer photophases seemed to reduce that variability, as all plants then bolted within a shorter time interval. The results do not indicate how much of the variability was genotypic or phenotypic. In some treatments, bolting and non-bolting plants can be distinguished, but the difference could be caused by rather small genotypic or phenotypic differences. Perhaps the non-bolting plants were near the threshold for bolting. In Figure 7A and 7B, the mean is taken for the two cultivars, as no dif­ ferences in reaction were observed. Against expectation, the plants chilled at a younger stage showed the earliest bolters, 30 days after cold treatment (Figure 7A). At the final count however, those treatments resulted in signi-

29 ficantly fewer bolters. Inplant s chilled for 7weeks ,th e differences are less pronounced (Figure 7B), althougheve nther eth eplant s chilled atth e youngest stages started bolting first. Themor e susceptible cultivar G3reacte d differently (Figures 7C and7D) . The quicker onseto fboltin gwit h directchillin gwa s similar to all culti- varsbu t allplant s ofcv .G 3ha dbolte d with that treatmentwithi n 60days , before theothe r treatments.Wit hplant s chilled atincreasin g stages,a de ­ pression in sensitivity to cold seems tob e followed by an increase.Als o against expectation,plant s chilled after germination for 2day sbolte d less thanthos e directly chilled. Longer cold treatment leads tomor ebolter s and afaste r onset ofbolting , especially inth emor e resistantgenotypes .

5.1.5 Conclusions

Whether youngerplant s canb evernalize d ispertinen t tocommercia l beet growing and to breeders wanting to vernalize plants ina controlle d room. Ifyoun g plants canb evernalized , space,labou r andtim eca nb e saved. The results (Section 5.1) indicate that a true juvenile stage doesno t exist in sugar-beet, although chilling during germination induces somewhat lessboltin g than during later stages ofgrowth .

5.2 TEMPERATURE OF COLD TREATMENT

5.2.1 Materials and methods

Several values for the optimum temperature of vernalization have been reported inliteratur e (Section 2.2).Fou r cultivars were therefore chilled to four temperatures ina nindoo r trial.T o separate effects of lightphas e from those of low temperature, arathe r shortphotophas e of 14h wa smain ­ tained during raising and chilling.Al l four single crosses Gl,G2 , G3an d G4wer e directly sown in2 Lpot s and kept in aroo m at2 0 °Cunti l twotru e leaveswer e formed after 14days . Theplant s were then transferred to4 col d roomswit h temperatures of3 ,7 ,1 1 and 15 °Crespectivel y for 55days . After chilling, allplant s were transferred toa greenhous e wheretempera ­ turewa s kept at1 2 °C forth e first8 days .Th e temperature was thenmain ­ tained at1 5 °Can d the lightphas ewa s altered to 14h + 10h =2 4h .Tem ­ perature was reasonably well regulated, rising to 18 °Co nonl y 7day si n the interval 15-40day s afterchilling . Unchilled plants of cultivars G3 andG4 ,sow n 3week s before theen do f the cold treatment, raised at 20 °C and atota l lightphas e of 14h ,wer e transferred onth e same day to thegreenhous e asth e chilledplants .

30 Theexperimen tca nb e summarized asfollows .

Factors 1) Genotype: Gl , G2, G3, G4 2) Cold treatment: VI; unvernalized (onlyG 3 andG4 ) V2; 55day s 3) Temperature ofcol dtreatment : Tl = 3° C T2= 7° C T3 =1 1° C T4= 1 5° C

5.2.2 Results

As a consequence of the different temperatures ofchilling , thenumbe r of leavespe rplan tdiffere d atth een d ofth evernalizatio n treatment(Ta ­ ble9) . Theunsusceptibl e cultivarsG l andG 2hav e reacted identically totemper ­ atureso fchillin g (Figures8 Ae n 8B). The lowerth e temperature,th e sooner plants started to bolt. Even the unsusceptible cultivars chilled at1 5° C started bolting about 100day s later. If the temperature had beenhighe r afterchillin g (inthi stria l also 15 °C),ther emigh thav ebee n fewbolters , ifany . The susceptible cultivar G4 responded in the same way to temperature of chilling: the lowertemperatur ewa smor eeffectiv e (Figure 8D). Only inth e cultivar G3 did thecol d treatment at7 °Cresul t inslightl ymor eboltin g thantha ta t3 °C (Figure 8C). Unchilledplant so fcultiva rG 3differe dcon ­ siderably from those chilled at 15 °Cbu tth edifferenc ewa sver y small in cultivarG4 .

5.2.3 Conclusions

A tentative conclusion inth eterm so fth emode l inSectio n 3.2 istha t eitherProces s 1proceed s faster atlowe r temperature orth edifferenc ebe -

Table9 . Numbero fleave spe rplan ta tth e endo fth ecol dperio d(mea no fth efou r cultivars)i nrelatio nt otemperatur ean d durationo fcol dtreatment . Unchilledplant sha d3 leaves .

Temperature Durationo f the (°C) cold period (d)

31 55

3 4 3 7 4 6 11 9 12 15 10 14

31 Proportion bolting (%) 100 00 O-O-O— O- O —O —D —•— Q A/ikV-A-A-A-A/^-A' . B 80 t I /

60

40 -

20 A /a-d ,/ ,v-w-v/ i „ 20 40 60 80 100 120 140 160 20 40 60 80 100 120 140 160

100

80

60 h

40

20

0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 Time after cold treatment (d) Time after cold treatment/raising (d) Fig.8 .Boltin ga sinfluence db ytemperatur eo fcol dtreatmen t0 ,3 °C ;A ,7 °C ;V ,1 1°C ; •,1 5°C ;• ,unchilled . A.Cultiva rGl ;b .Cultiva rG2 ;C .Cultiva rG3 ;D .Cultiva rG4 . tween the rates ofProcesse s 1an d 2 isgreates t attemperature s as lowa s 3 °C. But even at 15 °CV accumulates continuously, though very slowly. Although the trial suggests chilling at the lowest temperature toobtai n earliest flowering, this mightb ewron g (especially for susceptible culti- vars), since there is asecon d temperature-dependent step inth emodel ,th e synthesis of F. Synthesis of F during chilling wasprobabl y slow,becaus e of the short photophase and, in sometreatments ,th e low temperature.Ha d the lightphas e during chillingbee n 24h , theoutcom e would havebee ncom ­ pletely different.Fo r example, inunchille d plants ofcultiva r G4,boltin g started in continuous light 55day s after raising,wherea splant s chilled at 3 °C with aligh tphas e of 14h for 55day s required already 55+ 2 4= 80day s (Figure 8D).So , for several cultivars vernalization couldb emos t effective at intermediate temperatures like 8-10 °Cwit h longphotophases , becauseV coul d accumulate and also synthesis of Fcoul d start.Th e optimum temperature for vernalization was found indeed higher if the lightphas e during chilling was longer (Section 2.2).I na tria l reported below (Section 5.4), however, lightphas e during chilling at3 °Cha d no significant effect on subsequentbolting ,wherea s in another (unpublished)tria l at8 °C,ligh t

32 phase had a slight influence. Thus the action of lightphas e seemst ob e limitedb y lowertemperatures ,a sha sbee n assumed inth emodel .

5.3 LIGHTPHAS EAN D VERNALIZATION

5.3.2 Influence of light phase after vernalization

5.3.1.1 Materials andmethod s

The response of bolting toligh tphas e afterdifferen t durationo fcol d treatment was estimated in the four single crosses (G1-G4). Plants were raised in plastic boxes (Section4.5 ) in agreenhous e at1 5 °C,th e first true leaves appearing after14-1 5 days.The ywer ethe ntransferre d toa col d room at3 °Cwit h aligh tphas eo f 14h for0 , 14,2 8an d4 2days .Afterward s theywer etransferre d to agreenhous e at1 5 °C,stil lwit h lightphas e 14h , and after 4day s transplanted into 2L pots. The plants were thendivide d between 3 greenhouses atth esam etemperatur e butafte r another 3day sth e lightphas ewa seithe runchange d or supplemented for4 o r 10h . Temperature inth egreenhouse s could notb e regulated asaccuratel y as inpreviou strials . Onsunn y days, thetemperatur e rose4- 5 °Cabov eth e intended 15 °C,thoug h to almost the samedegre e inth ethre egreenhouses .However ,i nth egreen ­ housewit h aphotophas e of 14h , atechnica l failure caused aris et o2 8° C fora fewhour s2 3day s after coldtreatment .Th econsequen tdevernalizatio n wasprobabl y notto odisastrou s butcoul dno tb e assessed quantitatively. A comprehension ofth eexperimenta l treatments:

- Genotypes: Gl, G2, G3 andG 4 Durationo fth ecol dperiod : VI, 0week s V2, 2week s V3, 4 weeks V4, 6 weeks ase: PI, 14+ 0= 14 h P2, 14+ 4= 18 h P3, 14+ 1 0= 24 h

5.3.1.2 Results

To somedegree , inductionb y cold and subsequent lightphas ewer einter ­ changeable in a susceptible cultivar (Figure 9B).Wit h continuous light, shorter chilling retarded appearance of the firstbolter s andne wbolter s appeared slower,bu t even in unchilled plants 80% of theplant s finally bolted. With a subsequent light phase of 18h the effecto fchillin gwa s morepronounce d buteve no funchille d plants,5 % bolted .Wit hth e shortest

33 Proportion bolting (%) UU 1001-

80 _ A • 60 / . • m" • • 40 J 20 •—• 1 /

J , / 1 : i 20 40 60 80 100 40 60 80 100 Timeafte rchillin g(d ) Fig.9 . Bolting without chilling (•) orafte rchillin g for1 4(X) ,2 8(A )o r4 2day s (•) and with subsequent light phase of1 4(-.-.-) ,1 8( )o r2 4h ( ).A .Cultiva rG2 ; B.Cultiva rG4 .

light phase of 14 h, bolters appeared only after chilling for 42 or 28 days. However, for the more resistant cultivar (Figure 9A),n o beet bolted with a light phase of 14 h and even under continuous light the plants only bolted if they had been chilled for 28 to 42 days. A certain proportion of bolters, for instance 50 %, after a certain time after cold treatment can be obtained in different ways: longer chilling and shorter light phase or shorter chill­ ing and longer light phase (Figure 10A and 10B).Lo w temperature and subse­ quent long light phase are complementary.

Proportion bolting (%) f-100

41 27 13 0 Time of chilling (d) Fig. 10.Proportio no fbolter safte r 100day sa sa functio no fduratio no fchillin g ando f subsequent lightphase . A.Culti-i ' G2;B .Cultiva rG4 .

34 Theothe r cultivarsG 3 andG l reacted similarly tocultivar sG 4an d G2. Interpretation interm so fth emode l suggests thatdurin gvernalization , only Process 1proceeds .Th e light phase during vernalization (14h )wa s not appropriate for Process 3; also because ofth e lowtemperatur e (3°C ) probably no F wouldb e synthesized. Aftervernalization , different amounts of V would bepresen t inth eplant s according toth eduratio no fchilling . With temperature raised to1 5 °Cunde rdifferen t lightphases, difference s insynthesi s ofF depen d onth e amounto fV ando nligh tphas e inthi strial . InFigur e 9B,th ecombinatio nwit h 6week s chilled and then 14h lightphas e obviously results inth e samepatter n ofsynthesi s ofF a sunchille d and2 4h photophase, as is possiblewhe nassumin g arelatio nbetwee n synthesiso fF and temperature, lightphas e and amounto fV asi nSectio n 3.2

5.3.2 Effect of cultivar

5.3.2.1 Introduction

Ina simila r trial toth epreviou s one and ina fiel d trial,al l 10cul ­ tivars of Table 1wer e tested to findou twhethe r the foursingl e crosses were ina simila r rangeo f susceptibility tocommercia l cultivars ofsugar - beetuse d inth eNetherlands .

5.3.2.2 Materials andmethod s

AtAchterberg , thete ncultivar swer e sowno n2 an d2 9Marc h and 12Apri l 2 1978 inplot s of 15m in 4 replicates. After emergence,th eplant swer e _2 thinned to a stand density of 10m .Indoors , the cultivarswer e raised for 6day s at 20 °C,the n 5day s at1 5 °Cwit h aligh tphas e of1 4h and3 days at1 0 °C.Afterward s theywer echille d foreithe r 31o r4 9day s at3 °C, switched to 10 °C for 3days , transplanted and transferred to greenhouses with the following lightphases :

PI 14h +0 h for5 6days , afterwards 14h + 4 h = 1 8h . P2 14h + 4 = 18h unti l theen do fth etrial . P3 14h + 1 0h = 2 4h unti l theen d ofth etrial .

Afterchilling , temperaturewa s kepta t1 3 °C forth e first4 day s andsub ­ sequently at1 5 °C.

5.3.2.3 Results

Theresult s (Table 10)wer e comparablewit hthos e inth epreviou strial . The proportion of bolters forth e firstsowin g date inth e fieldtria l are shown at the bottomo fth etabl e for2 dates ,3 Augus t (earlybolters )an d

35 Table 10. Proportion of bolters for the cultivars G1-G10 (defined in Table 1) in a greenhouse at 5 intervals (30, 44, 74, 102, 125 days) after chilling and in a field trial (on 2 dates). Treatment VI, chilled for 31 days; V2, chilled for 49 days; PI, light phase of 14 h for 56 days and later 14 h + 4 h; P2, 14 h + 4 h; P3, continuous light (14 h + 10 h).

Time Treat- Prop ortiono f bolters(% ) Spearman's rwit h *-.- i— it-.- — after chil ment piUJJU IL i.UII UUlLlllg ling(d ) Gl G2 G3 G4 G5 G6 G7 G8 G9 G10 inth e field trialon : 08-03 09-21

30 V1P1 0 0 0 0 0 0 0 0 0 0 - - V2P1 0 0 0 0 0 0 0 0 0 0 - V1P2 0 0 0 55 0 0 0 0 0 0 0.52*1 0.52* V2P2 10 0 25 84 0 20 5 10 20 5 0.39 0.52* V1P3 0 0 0 100 0 18 5 10 20 0 0.26 0.46* V2P3 60 55 90 95 30 60 50 75 70 75 0.48* 0.48*

44 V1P1 0 0 0 5 0 0 0 0 0 0 0.52* 0.52* V2P1 0 0 0 50 0 0 0 0 0 0 0.52* 0.52* V1P2 0 0 0 90 0 0 0 10 0 6 0.44 0.44 V2P2 55 25 75 100 5 35 5 31 45 25 0.31 0.27 V1P3 20 5 45 100 5 59 25 44 30 5 0.58** 0.64** V2P3 65 90 100 100 65 80 60 80 95 85 0.52* 0.35*

74 V1P1 0 0 0 45 0 0 0 0 5 0 0.16 0.39 V2P1 0 10 5 85 0 5 0 5 0 5 0.77*** 0.25 V1P2 5 0 10 95 0 0 0 10 10 11 0.22 0.39 V2P2 70 55 100 100 20 55 10 48 60 30 0.32 0.17 V1P3 45 20 70 100 10 59 45 50 58 35 0.50* 0.59** V2P3 70 100 100 100 75 85 75 85 100 90 0.51* 0.30

102 V1P1 0 0 10 75 0 5 0 5 5 0 0.66** 0.72** V2P1 10 20 60 90 5 40 45 30 20 20 0.73*** 0.72** V1P2 5 5 30 100 5 10 10 25 25 11 0.54* 0.76*** V2P2 85 80 100 100 55 85 35 58 70 50 0.47* 0.19 V1P3 50 30 85 100 25 69 60 78 79 40 0.49* 0.68** V2P3 70 100 100 100 90 85 90 90 100 100 0.35 0.30

125 V1P1 10 0 25 90 0 10 0 15 5 5 0.53* 0.51* V2P1 20 35 70 90 20 50 55 65 45 35 0.77*** 0.84**** V1P2 15 20 35 100 20 20 25 40 45 23 0.47* 0.78*** V2P2 95 90 100 100 80 90 65 73 80 75 0.36 0.08 V1P3 55 40 90 100 35 74 65 84 84 60 0.55** 0.73*** V2P3 95 100 100 100 90 90 95 90 100 100 0.10 0.02 Pro portiono fboltin g (%) in field trial

Date Gl G2 G3 G4 G5 G6 G7 GS G9 G10

08-03 1.4 4 5 29.1 85.5 4.1 4.3 4. 1 4. 8 3.1 4.0 09-21 2.1 9 842. 1 89.5 13.5 10.8 15. 4 15. 7 11.2 10.5

1.* ,P <0. 10; ** P< 0 .05; ***,p <0.01 ) **, P <0. 001

21 September (total bolters). Single crosses Gl and G2 were less susceptible and G3 and G4 were more and far more susceptible, respectively, than commercial cultivars.

For breeders using regulated growing conditions, these conditions should be such that genotypic differences in bolting behaviour, as observed in the

36 field,coul d bereproduced . Ifth edifference s arelarg eenough ,ther ei sn o problem.However ,smal lbu tsignifican t differences inboltin g are difficult todistinguis h indoors.Althoug h itwa sno tth epurpos e ofthes etrial san d thenumbe ro fplant swa ssmall ,th eproble mmigh tb esolve da sindicate d in thelas ttw ocolumn so fTabl e 10.T ocompar e ranking orderso fth egenotype s inth e field andgreenhouse ,non-parametri c Spearmancorrelatio n coefficients were calculated for theproportio n ofbolter s inth e field at2 date swit h those ineac htreatmen t atth e 5interval s after coldtreatment . Chilling for 49day s andcontinuou s lightdi dno trevea ldifference s in bolting found in the field.Chillin g for4 9day s and aligh tphas e of1 4h and later 18h , which isclose rt onatura l circumstances,gav ebette rcor ­ relation.Curt h& Fiirste (1960)state ,however ,tha tcontinuou s illumination improves the accordancebetwee n field andgreenhous e boltingtrials . Cultivars could differ inrat eo fdevernalizatio n (Process 2), sensitivi­ ty to light phase (Process 3), cold requirement (Process 1)an d thevalu e ofth ethreshol d forboltin g (Process 4). Becauseo fth einterdependenc e of theseprocesses ,i twoul db edifficul tt odistinguis h these characteristics forth ecultivars . Ifgenotype smainl y differed inth erat eo fdevernaliza ­ tion, the method could be improved by ahig h temperature fora shor ttim e afterchillin g instead ofa constan ttemperatur e of1 5 °C.If ,however ,th e difference were in sensitivity to lightphase ,temperatur e after chilling should not be toohigh . Instead the lightphas e shouldb etha twhic hgive s best indicationo fsensitivit y towards lightphase . Iti sunlikel y thatdif ­ ferences inboltin g aredu et oa singl echaracteristic . Sot otes t forbol ­ ting resistance the conditions should imitate thoseo fearl y spring (with the same sequence of lowtemperatur e andhig htemperature , and arathe r short lightphase . If,however ,th ebreede rwoul d liket o select fora particula r characteristic, thecondition s canb echose n accordingly. Probably every climatic region where beet isgrow nha s specificcondi ­ tions for bolting. Dutchexperienc e istha tplan tmateria l fromelsewhere , for instance Poland and NorthAmerica , tends to bolt more thanth eDutc h commercial cultivars. Temperature conditions in spring inthos e countries are quite different andma y allowmor edevernalization ,o r lessvernaliza ­ tion,o rboth .

5. 4 DEVERNALIZATION

5.4.1 Materials and methods

To find the effect of high temperature after vernalization, the four singlecrosse s (Table 1)wer evernalize d andthe nkep ta ttw oconstan ttem ­ peratures 15 and 25 °C.The y were raised at 15 °C in plastic boxes ina greenhouse for1 4day swit hligh tphas e 14h andthe nkep tcol d at3 °Cfo r 0, 14,2 8 and 42 days. Besides chilling, theeffec to fligh tphas e during

37 cold treatment was tested with 14 + 0 = 14o r 14+ 1 0= 2 4h .Afte r cold treatment, the temperature was raised to 10 °C,an dth eplant s weretrans ­ planted into 2Lpot s 2day s later.Th eplant swer e divided between 2green ­ houses with a light phase of 14h + 10h = 2 4h and, incontras tt oothe r trials, immediately kepta t1 5 °C.On ewee k aftertransplanting , thetempera ­ ture inon egreenhous e was raised to2 5 °Can d theothe r maintained at1 5°C . Temperature could not be regulated very rigidly inth ecoole rgreenhouse . On several sunny days, temperature rose to 18-20 °C.Th ewarme r house became infested with aphids and mites, despite precautions, and thatpar t ofth e trial had tob e terminated earlier thanplanned .Th eexperimenta l treatments canb e summarized asfollows :

1.Singl e crosses:Gl ,G2 , G3 andG 4 2.Duratio n ofcol dtreatment : VI; 0week s V2; 2week s V3; 4week s V4; 6week s 3.Photophas e during coldtreatment : PI; 14h + 0h = 14h P2; 14h + 1 0h = 2 4h 4.Temperatur eafte r coldtreatment : Tl; 15 °Can dphotophas e 14+ 10h =2 4h T2; 25 °Can dphotophas e 14+ 1 0h = 2 4h

5.4.2 Results

Lightphas e duringvernalizatio n had no significant influence onbolting , possibly because ofth e lowtemperatur e duringvernalization . With atempera ­ ture of 15 °C after vernalization, the 4cultivar sbolte d inth e samewa y as in the trials on lightphas e (section 5.3).A t2 5 °C,however , onlyth e most susceptible cultivarbolted . So apparently devernalization inth e less susceptible cultivarswa s sufficient topreven t all theplant s frombolting . Bolting of cultivar G4 was reduced byhig h temperature after all times of cold treatment, although the difference between 15 °C and 25 °C wasmor e pronounced after shorter chilling or no chilling (Figure 11).Unchille d plants did notbol t at2 5 °C,bu tha d almost allbolte dwithi n 90day s after raising at1 5°C . According toth etim eo fchilling , different amounts ofV woul db epresen t inth eplants .Plant s chilled for4 2 dayswoul d bolt first:afte rvernaliza ­ tion, Fwoul d berapidl y synthesized at1 5 °C, alsobecaus e the amounto fV was not destroyed at that temperature andpositivel y influences thissyn ­ thesis. Shorterchillin gwoul d result inles sV and reduce,bu t still allow, synthesis ofF .Plant s would reach thethreshol d forboltin g later.

38 O 20 40 60 80 100 Timeafte rchillin g(d )

Fig. 11.Influenc eo ftemperatur e ( ,1 5°C ; ,2 5°C )afte rcol dtreatmen t fordif ­ ferent times (•, 0 d;A ,1 4d ;T ,2 8d ;• ,4 2d )o nboltin go fcultiva rG4 .Averag edat a for lightphase s of 14an d2 4h durin gchilling .Unchille dplant sdi dno tbol ta tal la t 25°C .

At 25 °C, V would break down. If so, the smaller amount of V would direct­ ly influence the rate of synthesis of F in proportion to that amount. During breakdown of V, F would still be synthesized, though at progressively lower rates until V was depleted. Obviously, the amount of F produced until that stage allowed bolting in only 40 and 80 % of the plants chilled for 28 and 42 days respectively. After shorter chilling, less V will be available, which means that syn­ thesis of F will stop earlier, the amount of F may then not be sufficient to allow any of the plants to bolt. In bolting-resistant genotypes, deple­ tion of V might take place in such a short time that insufficient amounts of F can be synthesized, even at that long daylength.

5.5 TEMPERATURE AND LIGHT PHASE IMMEDIATELY AFTER VERNALIZATION

In the next two trials, high temperature was applied for only a short time, to avoid treatments that induced no bolting.

5.5 . 1 Trial 1

5.5.1.1 Materials and methods

Seeds of cultivars G1-G4 were sown in plastic boxes, temperature being kept at 25 °C until emergence was complete and then lowered to 15 °C. After 13 days, cold treatment started. After 52 days at 3 °C, the plants were transplanted into 2L pots and transferred to 4 growing rooms at 15 °C. After

39 Table 11. Temperaturean d lightphas efro m thetim eo fendin go fcol dtreatmen t inTria l1 .

Treatment Time interva 1 (d)

17-28 29-41 41-end (Period I) (Period II) (Period III)

1 Temp. (°C) 15 15 15 Phase (h) 18 18 18

2 Temp. (°C) 15 15 15 Phase (h) 24 24 24

3 Temp. (°C) 25 15 15 Phase (h) 18 18 18

4 Temp. (°C) 25 15 15 Phase (h) 24 24 24

5 Temp. (°C) 15 25 15 Phase (h) 24 24 24

6 Temp. (°C) 15 25 15 Phase (h) 18 24 24

8 days, the light phase was increased from 14+ 0 = 1 4h to 14+ 4 = 18h and, after 17days , theplant swer e subjected to 6combination s oftempera ­ ture and lighting (Table11) . Besides the chilled plants,unchille d plants ofcultivar s G3 andG 4wer e included inth etrial .The ywer e sown 16day sbefor e theen d ofchillin g in the samewa y asth echille dplant s and allowed togerminat e at2 5 °C.Thre e days before chilling ofth e otherplant s ended, theywer e cooled to 11°C . Theywer e transplanted atth e same time aschille dplant s and furthertreate d identically.

5.5.1.2 Results

In Gl and G3,hig h temperature inth e early interval (Period I)reduce d bolting considerably (Treatment 3), evenwit h continuous light (Treatment4 , Figure 12A and 12B).I n unchilled plants of the most susceptible cv. G4 (Figure 12C)wit h continuous lightboltin gwa s delayed rather thanreduced . Ata photophas e of 18h ,n oboltin g tookplace . Inth evernalize d plants of G4, devernalization seemed tohav e less influence,probabl y the applied cold treatmento f5 2day sha d induced this susceptible genotype somuc h thatothe r factors didno thav emuc h influence anymor e (Figure12D) . With continuous light inth eunsusceptibl e cultivar Gl,warmt hwa smor e effective in preventing bolting ifi twa s earlier (Treatment4 )tha n ifi t was later (Treatment 5, Figure 13A).However , a later interval of warmth alsoreduce dbolting , ifth e lightphas e inth eearlie r intervalwa s shorter

40 Prop ortion bolting (%) 00 tOOr

80 A 1 60 1 40 - I

20 1 n-oo-DO-o frA-O-ûi ff o-oooo i —1 [>g-»oiOO-0-S*^wl W MP^^W- 20 40 60" "80 100 120 20 40 60 80 100 120 i_ Period: I m Time after chilling(d) Period: I Time after chilling(d)

100 100 r

80 80

60 60 temperature/photophase in Period: 1 n HI 40 40 15/18 15/18 15/18 15/24 15/24 15/24 25/18 15/18 15/18 20 20 25/24 15/24 15/24

-a'_ _i i I 0 20 40 6(5" 80 100 120 40 60 80 100 120 i Period: I m Time after raising(d) Period: I IL IE Time after chilling(d) Fig. 12.Boltin g inTreatmen t 1(0) ,2 (A) ,3 (V)an d4 (D )o fTria l 1(Tabl e11) . A.Cultiva rGl ;B .Cultiva rG3 ;C .Cultiva rG4 ,unchilled ;D .Cultiva r G4,chilled .

(Treatment5 an d6) . In the more susceptible genotype G3 (Figure 13B),late rwarmt hdi dno t reduce bolting at all (Treatment 5), except if the early lightphas ewa s short (Treatment6) . Unchilled plants of the most susceptible cultivar (Figure 13C)reacte d towardswarmt h like chilled plants of themoderatel y susceptible cultivar (Figure 13B).However , devernalization shifted theboltin g toa late rdat e and did not prevent ultimate bolting of all plants.Probabl ythi s effect was associated with abette r 'vernalizability'tha nG 3unde rth ecoo lcon ­ ditions of Period III.Afte r sometim eunde rth econdition s oftha tperio d (15 °C,2 4h tota l lightphase )th eplant sca nb e considered asvernalized . Bolting is then rapidly complete.Fo rthi scultiva r intervalso fhig htem ­ perature retard thisdat eo fbein gvernalized .

41 Proportion bolting(% ) 100 uu

80 . B h-'v-'v-v-w-w*- w

60

40 ^^ooW^S 20

i -JTJ I 1 1 1 0 20 40 60 80 100 period: I u period-. J u Time after chilling(d) Time after chilling(d)

uo temperature/photophase in period: I H UI o 15/24 15/24 15/24 80 c A 25/24 15/24 15/24 v 15/24 25/24 15/24 60 • 15/18 25/24 15/24

40 -

20

1 - 100 ». period: I H UI Time after raising (d) Fig. 13.Boltin gi nTreatment s 2(0) ,4 (A) ,5 (V)an d6 (•) ofTria l 1(Tabl e11) . A.Cultiva rGl ;B .Cultiva rG3 ;C .Cultiva r G4,unchilled .

An attempt will be made now to examine the results of this experiment in a more quantitative way, with respect to the presumed processes occurring in the plants. The question arises how the different effects of high tem­ perature in Period I or II can be explained. (Treatments 2 or 4 and 5, all three treatments having a constant photophase of 24 h throughout the trial. If in Period I temperature is raised to 25 °C, this has consequences for the destruction rate of V, but also for the rate of synthesis of F. Both processes will be enhanced by high temperature. However, since the rate of synthesis of F is also directly dependent on the amount of V, high tempera­ ture has a positive influence on synthesis of F in a direct way (via tem­ perature dependence of the true synthesis of F) and an indirect (negative) way (via the amount of V in the plants). When at the end of Period IV is

42 depleted totally andno tenoug hV canb eproduce d anymor e in followingpe ­ riods, no further synthesis of F takes place in succeedingperiods .Thi s explains the curves forTreatmen t4 :no tenoug hF ha sbee nproduce d atth e endo fPerio d It oallo wcomplet eboltin g incultivar sG l and G3. When the devernalizing temperatures act upon the plants in the second period,th e amounto fF produce d inth epreviou sperio d governsho w farbol ­ ting is reduced. For genotypeG 3thi s amountwa sclearl y sufficientt oin ­ duce bolting. Probably alsoi nthi s genotype,V isdestroye d inth esecon d period, but this will have less effect,becaus e enoughF ha s alreadybee n synthesized inth epreviou sperio d tosurpas s theboltin g threshold for8 0 % of the plants. In such away , one can alsoexplai nwh yhig h temperatures some time after vernalization canstil l reducebolting , aslon ga s shorter photophasesprevai l inth epreviou speriod . ComparingTreatmen t 5wit h 6an drememberin g thatProces s 3proceed s ata much slowerrat e inTreatmen t 6, oneca nse eth econsequence s forth e amount of F at theen do fPerio d I.I nth eplant s ofTreatmen t 6, the destruction by high temperature inPerio d IIdoe shav e aneffect ,becaus e the levelo f already synthesized F isstil l too low forman y ofth eplants .I nthi streat ­ ment, synthesis of additional F will alsob e reducedbecaus e ofth e lower amounto fV .Th e finalamoun to fF isclearl y thento o lowt o allowboltin g toa leve lcomparabl e toTreatment s 5o r 2. So stabilization of thevernalize d condition insugar-beet , ifany ,de ­ pendsno to ntime ,bu tespeciall y onphotophase .

5.5 . 2 Trial 2

5.5.2.1 Materials andmethod s

Plants ofcv .G l andG 2wer e raised asi nTria l 1bu ta t2 0 °Cinitially . Sixday s after sowingth etemperatur ewa s lowered to 15 °C for8 days ,the n plants were chilled to 10 °C for 4 days, 6 °Cfo r2 day s and 3 °C for5 4 days. The temperature was then raised to 10 °C; after 2days , theplant s were transplanted andtransferre d togrowin g rooms;an d after another 8day s the temperature was raised to 15 °Can dchillin g ended.Th ecol d induction was thusprobabl y stronger thani nth epreviou s trial.Afte r 12day s (photo- phase 14h +0 h = 14 h), treatments fortim e intervals of8 day sbega n (in contrastt o1 2day s inth epreviou s trial)(Tabl e 12).A sbefore ,unchille d plants ofG l andG 2wer e included inTreatment s 2,4 and6 ,bu tnon eo fth e plants bolted during the experimental period. Unchilled plants ofG 4wer e present in all treatments.Afte r the firstinterval ,al lplant sexcept.o f Treatments 7an d8 wer e subjectt oth e sametemperatur e of 15 °Can dt ocon ­ tinuous light. Treatments 7 and 8 measured theeffec to fa late r interval ofwarmt h afterdifferen t lightphase s inth e firstinterval .

43 Table 12. Temperaturean dligh tphas efro mth e timeo fendin go fcol dtreatmen ti nTria l2 .

Treatment Timeinterva l (d)

12-19 20-27 28-end (Period I) (PeriodII ) (Period III)

1 Temp. :°c) 10 15 15 Phase :io 14 24 24 2 Temp. :°o 10 15 15 Phase :h) 24 24 24

3 Temp. °c) 15 15 15 Phase th) 14 24 24 4 Temp. :°o 15 15 15 Phase :h) 24 24 24

5 Temp. °C) 25 15 15 Phase h) 14 24 24 6 Temp. :°o 25 15 15 Phase :h) 24 24 24 7 Temp. :°o 15 25 15 Phase rh) 14 24 24

8 Temp. °C) 15 25 15 Phase rh) 24 24 24

5.5.2.2 Results

As expected, the lower the temperature inth e first interval,th emor e and the sooner bolting appeared (Figure 14A, Treatment 1, 3 and 5).Hig h temperatures later (Treatment 7)ha d far less effect thanearlie r (Treatment 5). Bycontrast ,wit hcontinuou s lightdurin g the firstinterval ,th ehighe r thetemperatur e inPerio d I,th esoone rbolter s appeared (Figure 14B).Wit h continuous light in Period I, high temperature (Figure 14B, Treatment 6) didno treduc e earlyboltin g although later alowe r finalvalu ewa s reached than with the lower temperature treatments.Tha tpictur e was common toth e twounsusceptibl e cultivars (Figures 14A,B , C,D) . The unchilled plants of G4 in thesequenc e ofligh tphase s 14h ,2 4h , 24h (Figure 14E)behave d like chilled plants ofG l and G2. Incontinuou s light (Figure 14F) the patternwa s somewhat different,i ntha tTreatmen t4 andno ton eo fth ehig h temperature treatments gave (asi nFigure s 14B,14D ) theearlies tbolters .

An interpretation in terms of the model would assume thatwhe nplant s were subjected to a photophase sequence of 14h ,2 4h ,2 4h (Figures 14A, C), synthesis ofF i nPerio d Iwoul db e lowbecaus e ofth e rather shortday -

44 Proportion bolting (%) 100|- ÎEMP./PHOTOPHASE 100 IN PERIOD: /' J IL JIL / -i=i :/' • '10/14 15/24 15/24 80 •" »15/14 15/24 15/24 M..^ A »X »25/14 15/24 15/24 •15/14 25/24 15/24 60 60 IEMP./PHOIOPHASE IN PERIOD: I II III 40 40 • 10/24 15/24 15/24 * 15/24 15/24 15/24 * 25/24 15/24 15/24 • 15/24 25/24 15/24 20 20

_l_ _l 20 40 60 80 100 120 20 40 60 80 100 120 Time after chilling (d) 1 1 ' II ' Iff"' I II III Proportion bolting(% )

1001" TEMP./PHOIOPHASE 100 INPERIOO: /' 'T*-*-*-*-*-* ./" • 10/14 15/24 15/24 / . - ._./ 80 80 » 15/14 15/24 15/24 / ._./ ~ » 25/14 15/24 15/24 / ^'«-/

60 60

JI_ JIL 40 40 • 10/24 15/24 15/24 « 15/24 15/24 15/24 * 25/24 15/24 15/24 20 20 • 15/24 25/24 15/24

40 60 80 100 120 20 40 60 80 100 120 _^ Time after chilling (d) > Time after chilling (d) I II III I II III 100 loor .X—X—X—X—X—X—X- sf 80 / y ~-'7 .y/ ' ^A~»-A'

60

TEMP./PHOTOPHASE IN PERIOD: 40- I II III 10/24 15/24 15/24 -x- 15/24 15/24 15/24 20 -»- 25/24 15/24 15/24 -•- 15/24 25/24 15/24

_L 20 40 60 80 100 120 0 20 40 60 80 100 120 Time after raising(d) Time after raising (d) i n ...... II I III Fig. 14.Boltin gi nth ecultivar sGl ,G 2an dG 4a sinfluence db ytemperatur ean dligh tphas e afterchillin g (Glan dG2 )an dafte rraisin g(G4) . A.Cultiva rGl ,Treatment s 1(•) ,3 (X) , 5 (A)an d7 (•) B.Cultiva rGl ,Treatment s2 (•) ,4 (X), 6 (A)an d8(• ) C.Cultiva rG2 ,Treatment s 1(•) ,3 (X), 5 (A)an d(•7 ) D.Cultiva rG2 ,Treatment s2 (•) ,4 (X), 6 (A)an d8 (•) E.Cultiva rG4 ,unchilled ;Treatment s 1(•) ,3 (X) ,5 (A ) and 7 (•) F.Cultiva rG4 ,unchilled ;Treatment s2 (•),4 (X),6 (A ) and 8 (•) Treatmentsar edefine d inTabl e12 . 45 Proportion bolting (%) 100 h 100 o-o .o-o—o—o—o-o—o o-o—o-o— O-O 80 80 - B

60 60

TEMP./PHOIOPHASE 1EMR/PH0T0PHASE 40 40 IK PERIOD: - \ 1 IN PERIOD: I II III 1 I II III 10/K 15/24 15/24 15/14 15/24 15/24 20 -o- 10/24 15/24 15/24 20 - 1/ -°- 15/24 15/24 15/24 1 I i i 20 40 60 80 100 120 2Jl0 410 61 0 80 100 120

I II III I II III

100 100

80 - 80 - D „/ o-_0-0-9—Ol.^ / / / 60 60 - / / o / ._._/TEMR/PH0TOPHASE / IEMP./PHOIOPHASE 40 - / IN PERIOD: 40 - / IN PERIOD: / I II III I/ , I II III /• 25/14 15/24 15/24 15/14 25/24 15/24 20 -o- 25/24 15/24 15/24 20 -o- 15/24 25/24 15/24 / - I o / ./L 1 1 I I 1 20 40 60 80 100 120 2I-0 40 60 80 100 120 —>—i »• Time afterchilling(d) I II III I II III a Fig. 15.Th einfluenc eo fphotophas e inPerio d I(lA h (0)an d24 h (0)o nth eproportio no f bolterso fgenotyp eG la tth evariou s temperatures inth esam ePerio d 10° C (A),1 5° C(B) , 25° C (C)an d 15° C (+25° Ci nPerio d II)(D) .

length in that period. The amount of V, at the beginning of Period II; which is influenced by temperature in Period I, will then greatly influence final yield of F. When, however, photophase in Period I is 24 h (Figures 14B,D) , synthesis of F and destruction of V will take place concomittantly in this period. Considering Treatment 6 (25, 15, 15 °C),th e only conclusion can be that at a high temperature V is in fact broken down at a certain rate, but in the mean time, before V is depleted completely, rapid synthesis of F can take place (in contrast to Treatment 5) because both (high) temperature and photophase favour this synthesis. Obviously the result is that Treatment 6 is the first all treatments to reach the threshold level. In Figures 15 and 16 the main features are given of the response of synthesis of F to tempera­ ture and V-level. In these graphs the 2 photoperiodic treatments in Period I

46 Proportion bolting (%) 100 t- 100 r- o—o-o—o—o—oVó—o -o—o—o-o—o--0 o'°" / / o' 80 80 - B / , ./* / - ./ 60 60 -

TEMP./PHOTOPHASE 40 TEMP/PHOIOPHASE 40 IN PERIOD: IN PERIOD: I II III f I II III -.- 10/14 15/24 15/24 15/14 15/24 15/24 20 - 20 -o- 10/24 15/24 15/24 i 1 - 15/24 15/24 15/24 / 7' ~ _l_ _L 1 rf^J L i 1 i 80 100 120 20 40 60 80 100 120 I . i.. i I II III I II III

100 00 - o-o-o—o—o j-O-O o—o—o—o—o—o— o 80 n 80 ^—y y / / 60 60 / / J " / / 40 - ; 40 IEMP./PHOTOPHASE / . s INPERIOD : / y TEMP/PHOTOPHASE " / / I II III 0 / IN PERIOD; — 15/14 25/24 15/24 20 / / I II III 20 -o- 15/24 25/24 15/24 y/ — 25/14 15/24 15/24 / S* -o- 25/24 15/24 15/24 ,-—^i i i i i J A , i i i -J2 0 40 60 80 100 120 20 40 60 80 100 120 i . i„ I —• Time after chilling I II III I II III Fig. 16. Like Fig. 15, now for genotype G2.

are drawn for each temperature sequence.Whe n in Period Itemperatur e is low (10 °C) (Figures 15A, 16A)littl edifferenc e inreactio nbetwee n 14an d 24h is observed. Obviously inbot h treatments synthesis ofF isalmos ta t the same level, despite the difference inphotophase :th elo wtemperatur e reduces synthesis ofF i nthi speriod .Whe ntemperatur e inPerio d Ii shighe r (Figure 15B, 16B)th eeffec to fth edifferenc e inphotophas e inPerio d Ibe ­ comes more pronounced, which is in agreementwit h the assumed temperature dependence of synthesis ofF .A t stillhighe r temperatures inPerio d I(Fi ­ gure 15C, 15D, 16C, 16D)no tonl ytemperatur e dependence ofsynthesi s ofF isinvolved , alsoth edependenc e onth e amounto fV accounts forth ediffer ­ ence in the ultimate amounts ofF synthesize d andth ecorrespondin g number ofboltin gplants . Inunchille dplant s ofG 4a somewha tdifferen t situationwil loccur .On e

47 may assumetha ta tda yO n oV ispresent .The nhig h temperature inPerio dI will only delay V-synthesis,becaus e inthi s genotype itma yb e assumed that afterwards at 15 °C V-synthesis will still take place. As canb e seeni n Figure 14F, 15 °C is optimal forrapi dbolting ,possibl ybecaus ebot hsyn ­ thesis ofF an dV procee d atthi s temperature,whil e at 10 °Cthoug hsynthe ­ sis ofV ma yb e faster,synthesi s ofF is slower.Highe r temperatures cause these plants, in contrastt ochille d plants ofG l andG2 ,t obol t lateri n continuous light. Thedifferenc ema yb eexplaine d by assuming thatth e amounto fV isstil l rather low atth ebeginnin g ofth ehig htemperatur e interval,wherea s inG l and G2, thehig h amounto fV atth e starto f 25 °Calread y allows synthesis ofF righ t fromth ebeginnin g and so inducesboltin g more rapidly.

5.6 DISCUSSION

5.6 . 1 Introduction

This section summarizes the main effects before,durin g and afterver ­ nalization, and discusses theirphysiolog y inth e lighto fth emodel .

5.6.2 Seedling stage

Isther e aminima l size sugar-beetplant smus t reachbefor e vernalization cantak eplace ?Th etrial s suggested notru e juvenile stage,thoug hvernal ­ ization seemed lesseffectiv e inver yyoun gplants .Thi s loweredeffective - nes isles spronounce d thanstate db yMargar a (1960; 1968)wh o didno tsuc ­ ceed inbringin g sugar-beetplant s to flower,whe nvernalize d inth ecotyle - donary stage.Als o Junges (1959)an dChrobocze k (1934)detecte d such ajuve ­ nile stage in red gardenbeet .Bot h authors,however , could not avoidhig h temperatures aftervernalization . Especially the earlyvernalize d plants can also perhaps be devernalized more easily. Heide (1973)o n theothe r hand showed results in beetroot that were similar to the effects described in Section 5.1. Despite a somewhat lower vernalizability in the germination stage, a great part of the vernalization process in practical sugar-beet growing should take place before emergence, as judged by the temperature course inmos tyears . Thedepresse dvernalizabilit y ofgerminatin g seedsma yb edu et o shortage ofcarbohydrates .Thes e are,accordin g toPurvi s (1944), essential forver ­ nalization.Figur e 7A,B ,C ,D showed thatpre-germinate d seeds kept for2 days at2 0 °Cwer e slightly inferior inboltin g thanunpre-germinate d seeds. Especially thispre-germinatio n ata rathe rhig h temperature mayhav eresul ­ ted in a lowcarbohydrat e content,wit h consequences forth e finalpropor ­ tion of bolters inthi streatment .Als oWellensie k (1964b),wh o vernalized roots of Cichorium intybus, foundnegativ e effects ofpreviou s exposure to

48 20 °C,an d thoughti tmigh thav ediminishe d thesubstratu m forth evernali ­ zationprocess .Highki n (1956)report s aretardin g effecto fa pre-treatmen t at 20 °C or 26 °C for up to 5day s before theoptimu m coldtreatmen t for peas. This treatment resulted in aprogressiv e loss of the ability tob e vernalized. Besides causing ashortag e ofcarbohydrate ,hig htemperature sbefor ever ­ nalization could also have an effect similar todevernalization , assuming thatplant s have an 'initial degree of vernalization'. Raising atrathe r warm temperatures determinesthen :

losso fth e initial degree ofvernalizatio n increase invernalizabilit y (becauseo fth e larger plants)

Intha twa y iti smor eunderstandabl e thatth e firstappearin gbolter swer e observed intreatment svernalize d atth eyounges t stage (Section5.1.4 )(th e plants kept their initial "degreeo fvernalization" )bu tth ehighes t final proportion ofbolter swer e foundwhe nvernalize d aslarge rplan t sizes(be ­ causethe ycoul db evernalize d easier).

5. 6. 3 Vernalization

Vernalization (accumulation ofV inth emodel )wa s faster,th e lowerth e temperature (Figure 8), event o3 °C.Ther ewa sn o interactionbetwee n cul- tivars andvernalizatio ntemperature . In this trial a long-day influence during vernalizationwa sprecluded . A longerphotophas e duringvernalizatio nwoul dhav e favoured especially the treatments at 7 °C and 11 °C,becaus e according toth emode l suchtempera ­ tures allow both vernalization and synthesiso fF .Thi sexplain s the shift towards higher optimum temperatures of vernalization, as found by Curth (1962)whe n a photothermal treatment was applied (simultaneous action of low temperature and long photophase). Gaskill (1952)mention s an optimum temperature ofvernalizatio n of9 °C,bu t in facth eto o applied a2 4h pho ­ tophase duringchilling .Recentl y Lasa& Silvan (1976)reporte d 8 °Ca sop ­ timum, also with aphotophas e of 24h . Itexplain s further alsowh yHeid e (1973) found that a long photophase duringvernalizatio nprevente d subse­ quentdevernalizatio n tosom eextent . Thequestio n ariseswhethe r aconstan ttemperatur e duringvernalization , as is usual maintained intrials ,i sth emos teffectiv eprocedure . Inbol ­ ting-resistant cultivars, itmigh tb eusefu l firstt ocreat e ahig h amount of V at 3 °C (the duration of this period couldb e adjusted according to the susceptibility of the cultivar) and to 'convert' itcautiousl y inth e secondpar to fth ecol d treatment,a tsomewha thighe r temperatures andwit h long days, to the more stable substance F (without destroyingV b y ato o high temperature).Eve ndail y fluctuations intemperatur emigh tb emor eef -

49 ficient.Wellensie k (1979)reporte d thatvernalizatio n of Silene armeria L. during only thedar kperio dwa s evenmor e effective thanvernalizatio n dur­ ingth ewhol eday . Without cold-treatment and at temperatures as high as 15 °C butunde r continuous light,boltin g is bound to start after a longperiod , also in the most bolting-resistant single-cross.Obviousl y somevernalizatio n pro­ ceeds at these hightemperatures ,bu t atsuc h alo w rate thatboltin g only starts at extremely longphotophase s andno tbefor e 6month s after sowing. Vernalization does not seem tob e aproces s restricted toth e lowtempera ­ ture, as is usually thought, but proceeds slowly,eve n at 15 °C. Itmigh t be a common process also in plants without a quantitative or qualitative cold requirement.Heid e (1973)als omention s that inbeetroot ,thoug hgener ­ ally more liable to bolt than sugar-beet, a constant temperature ofeve n 18 °Cunde r continuous illumination triggered flowering.

5.6.4 Post-vernalization

According to the proposed model, the synthesis of the final flowering substance inth epost-vernalizatio n period isdependen t onsevera l factors, which also interact.Thi smake s italmos t impossible todiscus s thediffer ­ ent processes like devernalization, stabilization and photoperiodical in­ fluences separately.Whe n only length ofcol d treatment and lightphas e after vernalization werevaried , these two factorswer emor e or lessinterchange ­ able (Figures 9 and 10).Accordin g to Fife & Price (1953), these factors can even be interchanged completely, asextremel y long chillingperiod s of 100-300 days couldprovok e flowering incomplet e darkness.Th e reverse seems tob epossibl e also (unchilled plants flowering under continuous light,e.g . Figures 8an d 9). According to theresult s shown inSectio n 5.5, temperature conditions immediately aftervernalizatio n were crucial for final proportion ofboltin gplants .Hig h temperatures immediately after cold treatmentstrong ­ ly inhibit flowering.Ye tther e isa n important interactionwit h photophase as in Trial 1 (Section 5.5.1). Also after aperio d ofneutra l (stabilizing inth e literature)temperatures ,devernalizatio n remainspossible ,provide d theplant s received ashorte r lightphas e inth eneutra lperiod . Forpracti ­ cal sugar-beet growing, thiswoul dmea ntha tespeciall y inearl y springwit h a daylength ofonl y 14-15h as inth eNetherlands ,day swit h ahig hmaximu m temperature would thusb ehighl y effective inpreventin g bolting. If thermostable end-products of the vernalizationproces s doexist ,a s proposed by Napp Zinn (1957) for Arabidopsis and Devay et al. (1976)fo r winterwheat ,thei r rate ofsynthesi s should depend notonl y on temperature but also on photophase. For sugar-beet,withi n thequantitativ e andmomen ­ tary approach of themodel ,ther e seems tob en onecessit y to assume other intermediate substances thanth e final flowering substance. Oneo fth epropertie s ofth epresente d model isth epositiv e temperature

50 dependence of photoperiodic action,whic hmean s thatphotophas e iso fles s orn o influence atver y lowtemperatures . Inpractice ,thi swoul dmea ntha t photophase duringvernalizatio nwa s ofmino r importancebecaus e ofth eusu ­ ally low temperatures in this process.Whe nvernalize d at3 °Cth eplant s showedn o significantdifferenc ebetwee n lightphase s of1 4an d 24h durin g cold treatment (Section 5.4). Voss (1940)als oreporte d thatphotophas edu ­ ring chilling had no influence onbolting .Heid e (1973)o nth eothe rhand , mentions that in beetroot,alread y at5 ° Ca 2 4h photophas eprevente dde - vernalization slightly in the post-vernalization period. Hencephotophas e plays amajo r role atmoderat e temperatures aftervernalizatio n (Figure 9B). Inth edescribe d trials thedifferenc ebetwee n 18an d2 4h seemst ob emor e pronounced thanindicate db y Curth (1960).

5. 6. 5 Physiological background of the model

It isworthwhil e torecal lsom erecen tphysiologica l results,tha tsup ­ portth evalidit y ofth emodel .I tmust ,however ,b eborn e inmin d thatman y remarks orassumption s haveno tbee nteste d experimentally directly and are only suggestions. Why does the model assume2 substance s tob e active? Could not lowtem ­ perature and alon gphotophas e influence the synthesiso fonl yon esubstan ­ ce, because photophase and temperature aremor e orles scomplementary ?Ar ­ guments against such aconcep t areth edifferen t siteso fperceptio n (apex and leaves) and theexistenc e of amajo r interactionbetwee nvernalizatio n and photophase. Vernalization plusphotophas e aremuc hmor eeffectiv e than vernalization or a long photophase alone.Th e facttha t long lightphase s aftervernalizatio nhav e farmor e effecttha nbefor evernalizatio n suggests that cold treatment influences subsequent reaction towardsphotophase .Al ­ though iti ssuggeste d inth epropose dmode l thatV is aspecifi c substance thataccumulate s duringvernalization ,ther e ishardl y anyevidenc e for such a substance,a sman y grafting experiments have failed. Schneider (1960)foun d that partial plant vernalization (budvernalization )i nsugar-bee tdi dno t lead to flowering in unvernalized buds, which showsth e untransferability ofvernalization . Therefore it has beenthough ttha tvernalization , rathertha nproducin g a substance, brings forth a certain condition, which is transmissableb y cell division only (Barendse, 1964). Ifth eplan t isi na differen tcondi ­ tion after vernalization, what would be theeffect ?On epossibilit y ist o assume that the apex region has obtained anincrease d sensitivity towards flowering hormone(s). In this view the growingpoin twoul d react (bydif ­ ferentiation into stem and flowerbuds )onl y aftervernalizatio n tosubstan ­ ces exported by leaves in long-day conditions.Thi s idea is,however ,un ­ likely sinceth egraftin g experiments ofCurti s& Hornse y (1964)an dMargar a (1960)showe d thatunvernalize d growingpoint s couldb ebrough t intoflower ,

51 provided they were grafted on aflowerin gplant ,whic h indicates thatals o unvernalized apices canb ebrough t into differentiation. Instead iti smor e likely that vernalization is a process necessary for subsequent synthesis of the final flowering hormone in leaves grown out of such a vernalized growingpoint .Th erat eo fthi s synthesis,moreover , seems todepen d onth e length of the photophase.Th e existence oftransmissabl e hormones exported by leaves is,i ncontras tt oth evernalizatio n substance,beyon d doubt (Lang, 1965). Soth e amounto fV inth emode l could quantitatively represent acertai n condition ofth eplants ,rathe r than asubstance ,a sha sbee nsuggeste dun ­ til now, mainly for simplicity. What could be the nature of such aplan t condition? Wellensiek (1977) discussed the principles of flower formation ingenera l and distinguished between the following groups ofgenes .

Flower-forming genes:a collectiv e name foral l genes influencing flower formation in someway . Flower-hormone-forming genes: those genestha t aremainl y active inth e leavesb yproductio n of floralhormone . Floral genes: genes which act inth e apex anddetermine , forinstance , shape andcolou r ofth e flowers and shapeo f the inflorescence.

Wellensiek discussed the possibility that certain groups of genes are blocked, repressed or inhibited: 'thisblockin g is immobile, inth e senceo f not-translocated, and itma y occur indifferen t quantities,i ndifferen tin ­ tensities, atdifferen t levels'.Vegetativ e plantspresumabl y produce leaves in which the genestha tproduc e flowerhormon e areinactiv eb y ablocking . This blocking can be prevented, however, by aperio d of lowtemperature . Leaves growing out of avernalize d growingpoint ,carr y the genes forming flower hormone ina mor e orles s deblocked state.Th e longer thecol dtreat ­ ment lasts the more the flower-hormone-forming genes of the subsequently appearing leaves are deblocked. The deblocked state ismaintaine d by cell division, unless reblocking takesplac e athig h temperature,whic hcorres ­ ponds to the devernalizing action of hightemperatures . Inagreemen twit h this isth e conclusion ofCurt h (1960)tha tth e site ofperceptio n ofdever ­ nalizing temperatures also lies inth eapex . Assuming thatblockin g ofsuc h specific genes canoccu r atdifferen tin ­ tensities, it would not lead to anelementar y change inth emode l ifsub ­ stanceV inth emode lwer e replaced by avariabl eV indicating the intensity ofdeblockin g (offlower-hormone-formin g genes). Should this intensity refer toth ecellula r orth eplan t level?Lookin g ata cellula r level,th e flower- hormone-formin g genes themselves mightthe nb eblocke d indifferen tintensi ­ ties. If it would turnou ttha t only two states ofblockin g exist,blocke d andunblocked , the intensity ofdeblockin g mightb e considered atplan tlev ­ el,wher e itmigh t represent the ratiobetwee n thenumbe r ofblocke d andun -

52 Rate of increase of substance F dF/dt

NON-BLOCKED

HALF "-BLOCKED

TOTALLY BLOCKED

24 Photophase(h ) Fig.17 .Th e hypothetical relationbetwee n the intensity ofblockin go fflower-hormone - forminggenes ,photophas ean dth erat eo fsynthesi so fth ehypothetica lflowe rhormone(s) .

blocked,cells . One may assumetha tth erati o ofunblocke d toblocke d cells would shift ifvernalizatio nwer e longer.Wellensie k (1964a)als omentione d competitionbetwee n thenumbe r ofvernalize d andunvernalize dcells . Without the occurrence ofdevernalizin gtemperatures ,leave s grown from a vernalized growing point can then export final flowering substances to the apex,wher ethe ypla y arol e inth edifferentiatio n into stemo r flower buds.Whethe r thesedeblocke d leaveswoul d exportthes e final flowerpromo ­ ting substances orno t is,howeve rhighl y dependento nphotophase .Th e fol­ lowing scheme could elucidateth erelatio nbetwee nblockin g andphotophas e (Figure 17).Suc h a relationcoul d account forth ecomplementar y actiono f low temperature and photophase. As can be seen,photophas e determines the rate of synthesis of F in dependence on the intensity ofblockin g ofth e flower-hormone-forminggenes . Themode l alsoassume stha ttru esynthesi so fF i sdependen to ntempera ­ ture. Salisbury (1963,p . 163)describe d the 'photoperiodical'proces sa s being temperature sensitive.No tonl yth e synthesis ofth e flowerhormone , butals osensitivit y toth ehormon emigh tb e temperature-dependent. Stoddart et al. (1978) for example found inlettuc ehypocotyl s ashar pdecreas e in therespons e togibberelli c acid attemperature s lowertha n1 3 °C.Thi spo ­ sitive temperature-dependence is usually masked in plantswit h acol dre ­ quirement. In these plants,hig h temperatures have atwofol d effectafte r vernalization:mor e rapid initial synthesis ofF an denhancemen to fth ere - blocking ofth e flower-hormone-forminggene s(devernalization) . The model further assumestha tth e flowerhormon eF accumulate s untila certain level is reached, after which differentiation ofth e apexstarts . The validity of this assumption issupporte db y Zeevaart (1976)wh ointer -

53 preted grafting experiments in Perilla by King & Zeevaart (1973)t omea n thatth e stimulus hast oaccumulat e to athreshol d inth e apexbefor e flower formation can take place. The flower hormone will acti nth eape xwher ea possible deblocking of the floral geneswil l takeplace ,resultin g indif ­ ferentiation into stem and flowerbuds . It remains to be seen whether only one substance isinvolved . Bernier (1976)offer s aplausibl e picture, showing the existence of atleas ttw o components in the floral hormone in Sinapis alba, oneo fthe mbein gcyto - kinin. In sugar-beet, itcoul db e imagined that oneo fth e components isa gibberellin-likesubstanc emainl y active inste m elongation. More attention should be given to the consideration that flowering is the final resulto f astrictl y quantitative process.Thi smigh tconnec tva ­ rious response types (e.gS Dplants ,L Dplants ,wit h orwithou t acol dre ­ quirement)t oth e samebasi cprinciples .

54 6 Relation between growth and bolting of sugar-beet

6. 1 INTRODUCTION

External factors that encouraged growth also enhancedbolting , forin ­ stance low plant density (Jorritsma, 1978;Warne , 1949), heavy nitrogen dressing (Gorodnii & Sereda, 1975;Mann , 1951;Hoekstra , 1960;Ludecke , 1938; Lysgaard, 1978)an d irrigation (Röstel, 1968). Generalizing, onecoul d assumetha talmos tan yexterna l factor,tha tsti ­ mulates growth after vernalization could lead tomor ebolters .Thi swoul d also implytha tplant s ina fiel d cropwit h locallybette r growing conditions (e.g. more nitrogen, more space or early emergence)ru na greate r risko f bolting.Suc ha relatio ncoul dwel l account forth egreate rplan tweigh to f boltingplants ,a sfoun db y Lysgaard& Hol m (1962)wh o showed thatlat ebol ­ ters had considerably higher root and topweight s thanvegetativ eplants . The smaller rootweigh ta tth e finalharvest ,observe d forearl ybolter sn o doubtha sbee ncause db y thechange d dry-matter distribution afteronse to f stem elongation. One could imagine that early bolters also showa highe r plant weight, when measured immediately after visible stem formation. In the following somedat awil lb egive no nthi ssubject . More rapid growth ofboltin gplant sbefor evisibl e stemelongatio nnee d notresul tonl y frombette r external growing conditionsbu tals o fromquit e different factors. Inth e literature,som eevidenc e canb e found thatplant s thathav e advancedmor e towards flowering show an increased growth.Behaegh e (1975)observe d growthdifference sbetwee nvernalize d andunvernalize d Lolium perenne and Dactylis glomerata. Vernalized plants showed anincrease dto p growth,thei r specific leafare awa s considerably greater andne tphotosyn ­ thesispe runi tlea f areawa shigher .Davie s (1971)reporte d similar results inperennia l ryegrass.Relativ e growthrate s insward s ofvernalize d plants were 50% greatertha ni nsward s ofcomparabl e unvernalized material.Davie s attributed this effect toa change d distribution ofth eproduct s ofphoto ­ synthesis andt odifference s inth erat eo flosse so fdea dmatter .June se t al. (1975)reporte d thatvernalizatio n ofa winte rwhea tcultiva r increased thephotosyntheti crat eb y 16% , incontras tt o asprin gwhea twher en osuc h increasewa smeasured . Such observations could indicate that the advanced stateo f flowerin ­ duction itself or other circumstances leading to flowering also promote growth of the plants.Suc ha relatio nwoul d also lead tohighe rweight so f boltingplants ,assumin g thatespeciall y thoseplant smor e sensitive tolo w

55 temperaturean dphotophas e andtherefor e further advanced to flowering would show amor e vigorous growth. Soon ema y suspecttha t selection forboltin g resistance by breeders could have a negative effect on vigour. Selecting for vigorous growth would, inversely, increase bolting susceptibility. If such arelatio n should exist,breeder s of sugar-beet arecaugh t ina viciou s circle. For carrots,wit h asimila r relationbetwee nboltin g and growth,Dowke r & Jackson (1975) suggested that a checkwa s necessarywhethe r a selection forreduce d bolting doesno t lead to anundesirabl e reduction ingrowt h rate in the selected lines.A ver y clear examplewa s givenb y Parlevliet (1967) in spinach cultivars. Parlevlietdetermine d thegrowt h rateo f fivegroup s of spinach cultivars, differing in earliness. Earliness inspinac h canb e taken to be similar toboltin g susceptibility insugar-beet .H e foundtha t after equal growingperiods ,th e later the cultivar (morebolting-resistant) , the lower itsyield .Parlevlie t also reported that selection for lateboltin g in apopulatio n of mainly fast-growing plants almost inevitably leadst o slower growth. If,however , selection for fast-growingplant swa s carried outi na populatio no fmainl y late-bolting plants,th epopulatio nwil lbecom e earlier. However, the literature provides no clear evidence that such arelatio n alsohold s forsugar-beet .Thi s chapter reports sometrial s onth e influence of external growth stimulating factors, such as additional nutrients and irrigation.

6.2 INFLUENCE OFCONDITION S OFGROWT HO NBOLTIN G

6.2.1 Introduction

Considering the positive effect ofextr aminera l nitrogen onbolting ,a firstquestio nwa swhethe r this effectwa s associated withth e sluggish res­ ponse to the vernalizing actiono flo wtemperatur e during theearl y stages of growth. Nitrogen couldpla y arol eb y accelerating progress towards the stage that ismor e sensitive tocold .However , differences ingrowt h caused by nitrogen usually appear after the periods ofvernalizin g temperatures. Therefore the influence ofnitroge n onboltin g mightb emainl y active after vernalization.

6.2.2 Materials and methods

In 1976 and 1977 trials were done tostud y therelatio nbetwee n growth and bolting. In 1976 a trial of factorial design (splitplot )wa s setou t atWageninge n (Haarweg)i nthre e replicates,bein g sowno n2 5Februar y with cultivar Glwit h aStanha yprecisio ndrill ,wit h rows 50c m apartan d seeds 5 cm apart in the row. After emergence on8 Apri l theplant swer e thinned

56 —2 to a stand density of 10m .I nth etria l the following factorswer ein ­ vestigated: 1 Irrigation (main factor): R0= contro l Rl =irrigate d Thecro pwa s regularly irrigated from 5Ma yt o 1July . Intotal ,th eir ­ rigatedplot s received 145m m ofwater .Afte r 1July , allplot swer eir ­ rigated asnecessary .

2 Rateo fN : Nl = 25kg/h a N2 =10 0kg/h a N3 =17 5kg/h a N4= 25 0kg/h a Nl+ T = 25+ 7 5= 10 0kg/h a (T= secon d timeo f application) N2 +T =10 0+ 7 5= 175kg/ha .

On 15Marc h the whole field was dressedwit h 250k go f4 3% superphos ­ phate and 500k g of K-40.O n 31March , before emergence, nitrogenwa s applied at 25 (Nl) and 100kg/h a (N2,N 3 and N4). Someplot sgive n 100 kg/ha were given a further 75 or 150kg/h aafte remergenc e on2 9Apri l (N3an d N4). Forth etop-dressin g treatments,N l+ T andN 2+ T ,th e first dressing of 25 and 100kg/ha , respectively, was giveno n3 1March ,to ­ gether with the N1-N4 treatments, the remainder of 75kg/h a was given on 15June .

3 N source.Th enitroge nwa s supplied aseithe r

51= Ca(N03)2 or

52= (NH4)2S04 To avoid any differences in pH, an additional dressing of 0.61 kg of acid-bindingmateria lwa s given foreac hkilogra m ofammoniu m sulphate.

The trial consisted of7 2 fieldplots ,eac ho f6 m x7 m .Eac hplo twa s divided intotw oparts ,on ehal fuse d forperiodica l harvests ofvegetativ e plants on the followingdates :1 9May ,2 4June ,2 1July ,2 5Augus tan d1 4 September.O nthes edate s sampleso f10 ,20 ,20 ,2 0an d4 0plants ,respect ­ ively,wer e taken from eachplot . Inth eothe rhal fo feac hplo twit h about15 0plants, boltin gplant swer e harvested almost weekly on the following dates:1 4June ,2 4June ,1 July , 9 July, 15July , 22July , 29July , 5August , 12August , 25Augus t and1 4 September. On each side of the plot, 2 rows were lefta sborde r rows,t o avoid interferencesbetwee nplots . Forth e same reasonplant swer e leftun - harvested between successive gapsresultin g fromharvesting .

57 Also in the following year (1977) dry weights of bolting plants were com­ pared with those of vegetative plants. For this purpose the two single- crosses Gl and G2 were sown on 16 March at Achterberg. For each cultivar four plots of 9 m x 14 m were sown, which were thinned on 25 and 26 May to _2 a stand of 10 m .Durin g the growing season 25 vegetative plants per plot were harvested at regular intervals together with bolting plants, which had started bolting since the previous harvest. From both groups of plants, dry weights were calculated.

6.2.3 Results

Effects on plant growth were as follows. Due to the low temperature in the month of March 1976, field emergence did not take place before 8 April. During this period, vernalization of the germinating seeds was probably strong. After emergence, however, an unusually dry and warm summer ensued which no doubt reduced the potential number of bolting plants by devernali- zation. Especially in the first week of May, when the plants had reached the two-leaf stage, there were several days with high maximum temperatures. Irrigation until July encouraged growth early in the season (Table13) . Afterwards the irrigated plots did not maintain their lead and ultimately there was no significant influence of irrigation. Especially with little N, growth was even retarded, perhaps because of poor rooting in the irrigated plots, which would have had influence in the following dry and warm summer.

Table13 . Masso fdr ymatte rpe rplan ta tsuccessiv e harvestdate sa sinfluence db yirrigation ,for man d rateo fnitroge ndressing .

Treatment Mass (g)on :(month-day )

5-19 6-24 7-21 8-25 9-14

Irrigation Unirrigated 0.47 38.4 105.5 193.8 206.8 Irrigated 39.2 108.6 183.9 198.9 effect °;55 - - - Formo fN Ca (NO) 0.57 42.3 109.8 190.5 202.2 (NH4)J0 4 0.48 35.4 104.3 187.2 203.5 effect

Rateo fN (kg/ha) 25 0.44 32.3 100.2 185.1 192.9 100 0.55 38.3 103.3 180.2 209.9 175 0.60 44.1 112.5 191.1 209.6 250 0.57 47.5 115.7 204.3 215.1 25+ 7 5 0.45 30.3 103.6 181.0 198.7 100+ 7 5 0.54 40.5 107.2 191.4 199.7 effect

1)* ,P < 0.10 ;** ,P < 0.05 ;*** ,P < 0.01 ;**** ,P _< 0.00 1 Table 14. Masso fdrie dlea fpe rplan t (g)a tsuccessiv eharves t dates,a sinfluence db yirrigatio nan drat eo fnitrogen .

Rate of N (kj 5/ha) Mass (g/pl .) o n

Unirrigated 5-191 6-24 7-21 8-25 9-14

25 0.45 22.5 45.6 60.6 51.5 100 0.46 23.2 40.2 64.1 61.2 175 0.54 24.7 47.9 75.0 65.2 250 0.48 26.4 55.0 81.0 70.8 25 + 75 0.41 22.6 50.3 66.2 58.5 100+ 75 0.47 26.2 49.1 67.7 66.8

Irrigated 25 0.44 16.7 38.5 58.7 48.0 100 0.64 24.7 50.1 61.6 59.1 175 0.66 32.5 57.2 67.8 65.7 250 0.66 37.4 55.1 77.8 67.5 25 + 75 0.49 16.6 44.6 63.6 53.9 100+ 75 0.61 27.2 50.7 69.6 62.9 Significance of interaction *** """" **** - *

1.Fo rth eharves to n1 9Ma ytota lplant srathe rtha nleave s weretaken .

Ammonium sulphate initially retarded growth, as would be expected. After 21July , however, total plantweigh tdi dno tdiffe r significantly, whereas leafweigh tremaine d differentthroughou t the season.Especiall y inth e first part of the season,nitroge ndressin g and irrigation interacted onmas so f leaves (Table 14).Mor e nitrogenha dmor e effect inth e irrigated thanun - irrigated plots, as would be expected inth edr y springo f1976 .Late ri n the season, growth increased with top-dressing butno tt oth eexten t (for mass ofleaf )o fearl ydressin g at10 0o r 175kg/ha , the sametota lamount . In summary, irrigation up to 1Jul y especially had apositiv e influence in the firstpar t of the season, except for the Nltreatments ,wher e growth was reduced. Boltingwa s influenced inthi stria lb y irrigation althoughth e difference did not reach statistical significance until 14Septembe r (Figure 18).Th e form of nitrogen had no significant effect (Figure 19). Rate ofnitroge n had asignifican tpositiv e effecto nbolting ,excep to n 14June ,9 Jul yan d 15August . It had little effect in unirrigated plots (Figure 20A)bu tha d more pronounced effect on irrigated plots (Figure20B) .To pdressing sen ­ hanced bolting, so that improved growth aftervernalizatio n is responsible forboltin g rather thanearl y stimulationo fgrowt h andconsequen t increased responsiveness to low temperature. The same was suggestedb y agreenhous e trial on the effect of nitrogen. Inthi stria lmor eN aftervernalizatio n increased theproportio n ofplant sbolting . Final proportion of bolterswa s thuscorrelate d to (leaf)growth ,espe ­ cially growth inJun e and July seemed crucial forbolting . For leafmas so n 21July ,correlatio n toproportio n ofbolter swa s 0.59.

59 Proportion bolting (%) 4r irrigated \

/ unirrigated / -+ •' ,+' +' ./ '/

/

_1 I U I 1_ H 24 1 9 15 22 29 5 12 25 14 June I July | Aug. I Sept. date Fig. 18.Boltin ga sinfluence db yregula rirrigation .

Proportion bolting (%) 4r

Ca(N03),

_^ (NHt)2SOt

jr ?7 s/•V

// X- 1 I 1 1 1 L, I 1 L K 24 1 9 15 22 23512 25 | 14 | 12 June I July I Auç Sept. date Fig. 19.Influenc eo ftw onitroge nsource s (NH) S O andCa(N O) onbolting .

6.2.4 Plant weights of bolters and non-bolters

The mass ratio of incipiantly bolting plants to vegetative plants on the same date for each treatment (Figure 21) was usually more than 1. The weight of vegetative plants at a date between two harvests was estimated with a

60 Proportion bolting (%) 4 Unirrigated

H 24 1 9 15 21 291 5 12 June I July I Fig.20 .Influenc eo frat e(arei cmass )o fnitroge ndressin gbolting . A.Unirrigate dplots .B .Irrigate dplots .Label so fline sar erate so fnitroge ni nkg/h a eithera sbas edressin gu pt o25 0kg/h ao rwit htop-dressin go f7 5kg/h a(indicate db y+) .

regression equation (weight againsttime )fo reac ho fth e2 4treatment s in 2 this trial. Coefficient ofdeterminatio n (r )wa sbetwee n0.94 0an d 0.996, witha naverag e of0.977 .Calculatin g theratio s foreac htreatmen tseparate ­ lyavoide dbia s from treatments,i nwhic hboltin g aswel l asgrowt hwa sin ­ creased, which would lead to an overestimate of the differences inplan t weightsbetwee nboltin g andvegetativ eplants .I ngeneral ,bolter swer eheav ­ ier than vegetative plants in the corresponding treatments.Tha t relation held forearl yo r latebolters .Th edifferenc ewa s duet o leafmor etha nt o root.

Thenumbe r ofharveste d boltingplant s foreac hgenotyp e inth e 1977tria l can be found inTabl e 15,whic h also showsth e significance ofth ediffer ­ ences inplan tweight .Boltin gplant s turned outt ohav e adistinc t lead in plantweigh tthroughou t the season,a si nth e 1976tria l (Figure22A ,B) .

6.2. 5 Conclusions

The results presented inthi ssectio n show thatplant s thatru nt osee d have alarge rplan tweigh tthroughou tth egrowin g season,i fharveste d when their dry matter distribution is still comparablewit hvegetativ eplants . The irrigation and fertilization experiment showed,however ,tha tstimula -

61 T2.9 2.6r MASS RATIO

2.4

2.2

2.0

1.6

1.4

1.2

1.0 ..«•-

0.8

0.6

0.4-

0.2 =media n

0.0LA ,_i _ 6-24 7-1 7-9 7-15 7-22 7-29 8-5 8-12 8-259-14 date

Fig. 21. Mass ratio of dry matter per beet plant with just visible bolting and vegetative plants at the same date. The median on a date is indicated by .

tion of growth by external factors enhanced bolting, so this larger plant weight might be largely explained by better local circumstances of growth (e.g. plant space) which would favour growth and enhance undesirable bolt­ ing. The mechanism of how growth stimulation influences bolting remains un­ certain. The explanation might be that in the faster growing plants the de- vernalizing temperatures have a smaller influence because the threshold for bolting is reached earlier in the season. The period in which devernalizing temperatures can have their effect is then shortened.

62 Table15 . Numbero fbolter san dnon-bolter sharveste d ondifferen tdate si nth e 1977tria l

Cultiva r Type Harvest date (month -year) 06-16 06-27 07-06 07-13 07-26 08-09 08-22 09-14

Gl bolters 5 6 26 8 22 16 16 14 Gl non-bolters 101 100 100 100 100 101 102 100

G2 bolters 2 4 39 37 67 53 58 42 G2 non-bolters 100 100 100 100 100 100 101 100 Significanceo f differencei nplan t weight betweenbolter s andnon-bolter s - *** "fc*T - *** •** **

6.3 INFLUENCE OF COLD TREATMENT AND PHOTOPHASE ON GROWTH

6.3.1 Introduction

The observed larger plant weights of bolting plants could also be ex­ plained by assuming that especially those plants more responsive to low tem­ perature or light phase reacted with an improved growth. To study the in­ fluence of vernalization and photophase on growth, sugar-beet plants were subjected to various conditions in growth chambers, with extreme differences in time of chilling and in photophase.

Masspe rplan t(g ) 2601- 240- 240 + bolters 220 non bolters 220 200 200 180 180 B 160 160 140 :/ 140 120 120 - 100 100 80 80 :/ 60 f 60h y 40 V 40 20 20 y v ^ June July August ' Sept. June July August Sept. date Fig.22 .Mas s of dry matter perplan t ofjus t visibly boltingbee tplant san d (attha t moment)vegetativ eplants . A.Cultiva rGl ;B .Cultiva rG2 .

63 6.3.2 Materials and methods

The influence was investigated of two factors ongrowth ,eac h inthre e 2 levels (3 factorial).

1 Timeo fchilling : VI, unchilled V2, 27day s at3 ° C V3, 55day s at3 ° C

2 Photophase aftervernalization : PI, 9h+ 0h= 9h P2, 9h+ 6h=15h P3, 9 h + 15 h = 24 h

Allpossibl e 9combination swer e included inth etrial .

Raising theplants .Seed s ofcultiva rG lwer e sown inpape rpot s andwer e placed in a glasshouse at2 5 °C for 8days , afterwhic h emergencewa scom ­ plete.Th e temperature was then lowered to 15 °Cunti l two leaves developed threeweek s after sowing. Vernalization.Plant swer e chilled ingrowt h chambers (Section4.3.1 )a t a temperature of3 °C, totalphotophas e 15h + 0h , rel.humidit y 0.7.Ver ­ nalization took 0,2 7an d 55day s forth ethre eexperimenta l groups.Sowin g dateswer e soorganize d thatth ecol d treatments ended on the sameday . Post-vernalization. Aftervernalizatio n theplant swer e transplanted in­ towhit eplasti c 6Lpots , andkep t ata temperatur e of 10 °C for three days. Before transplanting 2x 10plant spe rcol dtreatmen twer e sampled andmas s of dry matter and leaf area perplan twer emeasured .Next ,th eplant s for eachvernalizatio n groupwer e splitu p into three subgroups,eac h henceforth receiving one of the following photophases: 9h+0h=9h, 9h+6h= 15h , and 9h + 1 5h = 2 4h a t a temperature of 15 °Can d arel .humidit y of about 0.80.Th ephotophas e was extended inth ewa y described in Section 4.4.Th e 9treatment s consisted of 18plants . From these 18plants , 16plant s wereharveste d onthre e successive harvest dates: threeweek s aftervernalizatio n (4plants) , 5week s aftervernaliza ­ tion (4 plants) and finally 8week s after vernalization (8plants) . The plants were taken at random from each group. To excludevariatio nwithi n thegrowt h chambers,th eplant swer eplace d oncart s andwer e switched about twice per week. Differencesbetwee n thegrowt h roomswer eprevente d asfa r aspossibl eb y exchanging theplant s atregula r intervalsbetwee n thecham ­ bers, thereby, ofcourse ,als o changing the assigned photophase.Thi sresis ­ tant genotype bolted inth eV3P 3 treatment after the 3rdharves t (1plant , whichwa sno tharvested) .A t the4t hharvest ,3 plant s showed visible bolt-

64 Table 16A. Mass of dry matter per plant and leaf area per plant at t = 0 for the three cold treatments. Treatment

VI V2 V3 d.051

Dry matter (mg) 13.7 31.4 92.3 20.0 Leaf area (cm ) 1.52 2.42 6.12 1.98

1. d.05 = studentized range, (Tukey) at the 5 %- level. VI, unchilled ; V2, 27 days at 3 °C; V3, 55 days at 3 °C

ing. As these plants were then still comparable with vegetative plants, they were included to obtain an orthogonal scheme, which facilitated the statis­ tical procedures. At each harvest date, the following measures of dry matter were calculated: beet-root per plant, petioles per plant and leafblades per plant. Also leaf area per plant was recorded with either an electronic de­ vice at the first 2 harvests or later by scaling to photocopies of leafblades of known area.

6.3. 3 Results

Atth een do fcol d treatment (t= 0 )an dbefor e submittingth eplant st o differentphotophases ,dr ymatte ran dlea fare awer e firstmeasured . Table 16A shows that longerchillin g resulted inmor edr ymatte ran dlarge r leaf area, because plant growth didno tsto pentirel ya t3 °C .Th edifference s indr ymatte ra tt = 0 betwee nth echillin g treatmentsma ymak ea late rcom ­ parison spurious.Fo rexample ,eve nwhe nassumin gexactl yth esam e (exponen­ tial)growt h forth ethre e chilling treatments,difference si ndr ymatte r would become even greatera tfollowin gdates .T oovercom e thisproble man d stillt oinvestigat eth epossibl epositiv e effecto fcol d treatmento nsub ­ sequentgrowth ,a procedur ewa sdevise d forcomparison swithi neac hvernal ­ ization level,i.e .betwee nth ethre e lightphases .Suc hcomparison sca nb e madea teac hmomen t aftervernalizatio nwithou tobjection ,a sth e(average ) startpositio na tt = 0 i sth esam efo rthos etreatments . InTable s 16B,C an dD ,th efirs tthre e linespresen tth emai neffec to f chilling (averaged over thethre e lightphases) .Th efourt h line indicates thesignificanc eo fth evernalizatio neffec tan dals ogive sth eStudentize d Range (Tukey)a t5 % ,fo rcomparison sbetwee ntw omeans .Th enex t four lines indicate themai n effect ofligh t inth esam eway .Th efollowin g lineso f thetable s showth emea nfo reac ho fth enin e combinations.Th esignificanc e of aninteractio n canb erea di nth elas tline ,togethe rwit hth eStudent ­ izedRang e forcomparison sbetwee nan ytw omeans .

65 Table 16B. Harvest 3we e ks (t= 3) after end of chi Hing.

Treat­ Leaf Petio­ Beet Total dry Leaf 2 2 Petiole Number ment blade les root matter area T -h T -^ fraction of (g) (g) (g) (g) (cm ) (cm g ) (cm g ) of sprout leaves (%)

VI 0.21 0.07 16 0.30 73 349 243 25.1 7.3 V2 0.32 0.10 46 0.47 94 296 200 24.8 7.4 V3 0.57 0.21 86 0.86 164 299 193 27.2 9.2 0.04 0.14 1.1 d.05 0.09 24 26 30 17 _

PI 0.41 0.09 49 0.55 106 265 200 17.7 7.8 P2 0.40 0.14 58 0.60 123 323 216 26.4 8.3 P3 0.30 0.15 43 0.49 103 356 221 33.0 7.8 d.05 0.09 0.04 - - - 30 17 3.4 ~

V1P1 0.24 0.05 22 0.30 69 288 224 16.2 7.0 V1P2 0.19 0.07 14 0.28 68 351 244 26.4 7.8 V1P3 0.20 0.10 13 0.31 83 408 262 32.7 7.0

V2P1 0.31 0.07 38 0.41 79 256 191 18.1 7.0 V2P2 0.44 0.13 68 0.64 129 295 203 23.4 7.8 V2P3 0.22 0.11 33 0.37 76 337 206 32.8 7.5

V3P1 0.68 0.16 86 0.92 170 250 184 18.6 9.5 V3P2 0.56 0.23 91 0.88 174 322 201 29.5 9.3 V3P3 0.47 0.23 81 0.78 149 324 195 33.4 8.8 d.05 0.21 - - - 60 ------VI,unchilled ;V2 ,2 7days ;V3 ,5 5day sa t3 ° C PI= tota lligh tphas e9 h ;P 2= 1 5h ;P 3= 2 4h d.05= Studentize d rangea tth e5 % leve lfo rcomparison sbetwee nmeans . SLA= specifi c leafare a LAR= lea fare aratio .

Influence of photophase Light phase had great influence, even though all plants received the same amount of photosynthetically active radiation. (The light was varied with incandescent bulbs of low radiant flux density, which had negligible effect on photosynthesis).Ye t the different photophases had a pronounced influence on dry matter production. The crop at t = 3 (3 weeks after chilling. Table 16B) showed that sprout growth was especially influen­ ced. Lengthening the light phase from 9 h (PI) to 15 h (P2) increased total 2 leaf area, though not yet significantly, and Specific Leaf Area (SLA = cm leaf area per g dry matter of leaf laminae). There was also a marked influence of light on petiole lenght and on pe­ tiole dry matter. With longer light the mass fraction of dry matter of pe­ tiole to total sprout was significantly larger. A still longer light phase (24 h) did not further increase dry matter production. On the contrary, a slightly inhibitory effect could be observed with respect to 9 h. The plants under continuous light developed narrower and yellower leaves. Petiole growth was however not inhibited to such an extent, so that the proportion of petiole dry matter increased even further.

66 Table16C . Harvest5 week s (t= 5 )afte rchilling .

Treat­ Leaf Petio­ Beet Totaldr y Leaf Petiole Number ment blade les root matter area (cmg ) (cmg ) fraction of (g) (g) (g) (g) (cm2) ofsprou t leaves (%)

VI 1.42 0.59 0.39 2.40 378 265 158 28.9 10.2 V2 1.87 0.86 0.38 3.11 475 254 152 31.4 10.2 V3 2.58 1.34 0.49 4.41 621 242 141 34.3 11.2 d.05 0.61 0.38 - 1.02 158 19 18 3.5 1.3 **** *•*** - **** *** ** •k *** *

PI 2.03 0.71 0.51 3.25 434 220 141 25.4 10.8 P2 2.27 1.13 0.44 3.85 636 281 165 32.4 10.9 P3 1.58 0.95 0.30 2.82 404 259 145 36.9 9.8 d.05 0.61 0.38 0.20 1.02 158 19 18 3.5 1.3 ** •ÄrÄr "n * *** **** *** **** «:

V1P1 1.31 0.41 0.20 1.92 311 244 166 24.1 10.0 V1P2 1.65 0.71 0.59 2.95 480 292 159 30.4 10.8 V1P3 1.32 0.64 0.37 2.33 342 259 148 32.3 9.8

V2P1 1.81 0.61 .55 2.96 376 208 127 25.4 10.5 V2P2 2.33 1.11 .33 3.77 673 290 180 31.6 10.8 V2P3 1.47 .87 .26 2.60 376 263 147 37.2 9.3

V3P1 2.97 1.11 .79 4.88 616 208 129 26.6 11.8 V3P2 2.84 1.57 .40 4.81 754 261 154 35.3 11.3 V3P3 1.94 1.34 .26 3.54 494 256 140 41.1 10.5 d.05 - - .47 - - - 42 - - - - *** - - - ** - -

At the third harvest (Table 16C),a similar phenomenon was observed. Lengtheningphotophas e from 9t o1 5h increased leafare ape rplan tfro m43 4 2 to63 6c m .Th e increasewa sno tdu et oa n increase innumbe r ofleaves ,bu t to a higher massic areao fleave s (SLA).Th ephotophas e induced fasterex ­ pansion of leaf area, resulted in an increase in drymatte ro fal lplan t parts at the finalharves t forth eP 2plant s (Table 16D).A ttha ttim eth e P3 plants had practically caught up with theP 2group ,s otha ta positiv e influence of this photophase onroo tdr ymatte rwa s observed.Th e negative influence on laminae growth however remained, buttogethe rwit h theposi ­ tiveeffec to npetiol e growth,P 3plant swer e significantly heavier thanP I plants.O fth ethre ephotophases ,1 5h wa s optimal fordr ymatte r production perplant .A n increase of4 2% wa s observed: from 13.91g t o19.7 6g !

Influence of cold treatment Throughout the trial the chilled plants (V2 and V3)wer e heavier and had larger leaf area thanunchille d plants.Th e tables do not indicate how much ofth e increase shouldb e attributed toa positive effecto fchillin g onsubsequen t growth andho wmuc h toth enatura l consequence ofth edifferen t starto fth ethre etreatments .T oovercom e this difficulty and toinvestigat e thechillin geffec ti nmor e detail,th e foll­ owingprocedur ewa s developed.

67 Table16D . Harvest8 week s (t= 8 )afte rchilling .

Treat­ Leaf Petio­ Beet Total dry Leaf Petiole Number ment blade les root matter area (cm g ) (cm g ) fraction of (g) (g) (g) (g) (cm ) of sprout leaves (%)

VI 5.61 3.56 2.90 12.07 1156 209.3 97.2 39.2 13.6 V2 6.71 4.74 4.77 16.22 1315 198.6 82.8 41.1 15.5 V3 8.23 6.48 6.04 20.74 1569 192.4 77.2 43.6 16.3 d.05 1.26 1.12 0.93 2.64 233 15.6 7.4 4.2 1.6 •îrk-lcîc iVVwWfc •irk-kic ArrtWr •k-k •&"iWc>V **

PI 7.05 4.02 2.84 13.91 1387 198.6 101.2 36.3 18.3 P2 7.90 5.89 5.98 19.76 1510 194.4 78.6 42.4 13.6 P3 5.60 4.87 4.89 15.36 1143 207.3 77.3 45.2 13.5 d.05 1.26 1.12 0.93 2.64 233 - 7.4 4.2 1.6 •JWWVÏV Vw'w'C I'wWwV **** Vw'wV - ;'wWrtr -W-V* ï'wWwV

V1P1 5.46 2.91 1.98 10.35 1147 210.4 111.5 34.8 16.6 V1P2 6.84 4.46 3.65 14.95 1403 211.1 94.0 40.7 12.9 V1P3 4.54 3.30 3.08 10.91 919 206.4 86.1 42.0 11.3

V2P1 7.23 4.07 2.86 14.15 1396 193.7 97.8 36.5 18.9 V2P2 7.74 6.00 6.80 20.54 1480 191.3 72.9 42.6 13.4 V2P3 5.16 4.15 4.66 13.97 1069 210.6 77.6 44.2 14.1

V3P1 8.47 5.08 3.68 17.22 1619 191.6 94.4 37.7 19.3 V3P2 9.11 7.21 7.48 23.81 1647 180.7 68.8 43.9 14.5 V3P3 7.11 7.14 6.95 21.20 1441 205.0 68.4 49.3 15.1 d.05 - - 2.16 ------

Standard deviationso fplan tdr ymatte r andlea f area increasedwit htime . To render thevariabilit y more homogeneous with time,value st otota ldr y matter perplan t (W)an dlea f areaspe rplan t (LA)wer e transformed tona ­ tural logarithms (Hunt& Parsons , 1974). Foreac ho fth e9 V Pcombinations ,a 2nd-degre epolynomia lwa sthe n fit­ tedt oth etransforme d data, accordingt oth efollowin g regressionmodel :

2 ln(x..7(t))=a. +ß.. t+ y..t +e..,(t ) (2) x i]k '' l H2j ' ij ijk [e (t)= v(0,o)] ijk in which In (x... (t))refer s toth enatura l logarithm ofeithe rmas s (W) ijk or leaf area (LA)o fplant s submitted toth eit hchillin g treatment (i= 1 , 2, 3)an dafterward s growing underth ej't hligh tregim e (j= 1 ,2 ,3) . The subscript k refers toth eordina l numbero fth eharveste d plant (k =1 ,2 , 3, 4o r k =1 ,2 ....8)withi na ViP jcombinatio n attim e t aftervernali ­ zation (t= 0 ,3 ,5 ,8 weeks) .Th emode l required thatal lobservation s with the same degree ofvernalizatio n were fittedwit hequation swit hth esam e constanta. .Suc ha regressio nmode lwa schose nbecaus e allplant swit hth e samecol d treatmentV iha dha dth esam e treatmentunti l t - 0.Hun t& Parson s

68 Table17 . Analysiso fvarianc e formas so fdr ymatte ran dlea fare ape rplant .

Plant Sourceo f Sumo f Degreeso f Mean F P parameter variation squares freedom square

40.02 1 Dry c.f' 2 77.73 1 matter 3 137.10 1 (W) 254.85 (3)

linear 425.26 9 47.25 593.01 <0.001 quadratic 8.14 9 0.91 11.35 <0.01 error 10.28 129 0.08

total 698.53 150

Leafare a 1 1683.98 1 (LA) cf. 2 1803.47 1 3 2016.05 1

5503.50 (3)

linear 288.95 9 32.11 473.18 <0.001 quadratic 30.90 9 3.43 50.60 <0.001 error 8.75 129 0.07

total 5832.09 150

1. cf. correctionfactor .

(1974)an dNicholl s& Calde r (1973)hav ewarne d againstoverfitting ,meanin g thatquadrati c orcubi c terms shouldno tb e included inth eregressio nmode l if insignificant. The ANOVA tables forW and LA (Table 17)sho wtha tals o thequadrati c term issignifican t forbot hplan tparameters .S oa 2nd-degre e polynomial was fittedt oth etransforme ddata . The estimated regression coefficients forW andL A arepresente d inTa ­ bles 18 and 19.The y allowestimate s tob emad e ofW andL Athroughou tth e trial without any awkward interpolation. Figures 23an d2 4sho wth eresul ­ tingcurve s fordr ymas s and area foreac ho fth ecombinations . Relativegrowt h rates (RGR)ca neasil yb e computed,as :

RGR(t)= 1/ Wx dW/d t =d[l n (W(t))]/dt (3) inwhic h InW(t )represent s theestimat e ofth e logarithm ofmas s attim et weeks aftervernalization . Differentiation of Equation 2allow sth eRG R tob eexpresse d asa func ­ tiono ftime :

RGR(t)= ß + 2yt (week-1) (4)

Figures 25A , B, C shows the calculated course of therelativ e growth

69 Table 18. Regression coefficients formas stota l drymatte rpe rplan ta sa functio no ftime : 2 In(W.. . (t) y JJ)= a .+ ß. .t + Ev. .t ijk - -i M - *ij_-

Treatment Regression coefficients

chilling photophase a. ß Y ( l ü ii V (Pl) VI PI -4.52 1.275 -0.0526 VI P2 -A.5 2 1.309 -0.0505 VI P3 -4.52 1.338 -0.0599

V2 PI -3.68 1.116 -0.0409 V2 P2 -3.68 1.232 -0.0497 V2 P3 -3.68 1.029 -0.0299

V3 PI -2.55 0.987 -0.0390 V3 P2 -2.55 0.920 -0.0256 V3 P3 -2.55 0.831 -0.0163

Forexplanatio no ftreatmen t codesse eTabl e16B .

rate for each ofth e 9combinations .Fo reac h ofth e three chillingtreat ­ ments, P2 plants had ahighe r RGR thanth ecorrespondin g PIplant s during thegreate rpar to fth etrial .Continuou s light (P3)reduce d theRG R inun - chilledplants . Chilled P3 plants initially grew slower toobu t later grew faster thanP I andP 2plants . Figure 25 also shows thatth eRG R decreasedwit h advancing stage ofde ­ velopment,hamperin g direct comparison ofrelativ e growth ratesbetwee ndif ­ ferent chilling treatments,becaus e ofth e differentplan tweight s attim e t =0 .Rathe r thant ocompar e theRG R ata give nmomen to ftime ,a compari ­ son atth e samegrowt h stagewoul d be abette r approach.Althoug h ratherar -

Table19 . Regressioncoefficient s forlea fare a perplan ta sa functio no ftime .

Treatment Regression coefficien t

chill ing photophase a. ß Y (V l M M VI PI 0.44 1.471 -0.0814 VI P2 0.44 1.544 -0.0868 VI P3 0.44 1.585 -0.0993

V2 PI 0.85 1.386 -0.0743 V2 P2 0.85 1.653 -0.1062 V2 P3 0.85 1.392 -0.0784

V3 PI 1.86 1.303 -0.0769 V3 P2 1.86 1.347 -0.0822 V3 P3 1.86 1.228 -0.0692

Forexplanatio no ftreatmen tcode sse eTabl e16B .

70 Total mass per plant (g) 20

1r 2

1r 3

0LV> 3 4 5 6 7 8 Timefro men do fchillin g (weeks) Fig.23 .Mas so ftota ldr ymatte rpe rplan ta sa functio no ftim ewit hdifferen tperiod so f chilling (VI,unchilled ;V2 ,2 7day s and V3,5 5day schilled )an ddifferen tphotophase s afterchillin g (PI,9 h ;P2 ,1 5h an dP3 ,2 4 h).

71 Leaf area per plant (cm2) 20O0r

1500

1000

1500 V, :chille d 55 days

1000

500

0L* 3 4 5 6 7 8 Time from end of chilling (weeks) Fig.24 .Lea fare ape rpLan ta sa functio no ftim ei nth evariou s combinationso fchiliin g andphotophas e (asi nFigur e23) .

72 RGR (week"1) 1.2 1.2 1.2

1.0 1.0- 1.0

.8 .8

.6 .6

.4 .4

.2 V, :unchilled V2: chilled 27 days .2 " V3:chille d 55 days

J i_ i 0 2 4 6 8 Time after chilling (weeks) Fig. 25. Relative growth rates (RGR) after cold treatment as influenced by length of the chilling period and subsequent photophase. bitrarily, in the following procedure, total dry weight per plant is chosen as a yardstick for development.

At time t after vernalization (Equation 2), mass of dry matter Wwil l be

.2 In (W) = a + ßt + vt (5) or (a-ln (W)) + ßt + yt2 = 0 (6)

A commontyp eo fequation .

-ß+ 2 -4(g -i n (W))Y = h (7) 2Y canb econverte d to

ß + 2yt =V l ß" -4( a -ln(W)) v (8)

Comparison with Equation 4 shows that the left term represents relative growth rates which can be calculated from the right sidewhe nplan tmas s equals W. Relative growthrat e isthe na functio no f a, ß, Y andW , rather thana functio n ofß , Y andt a si nEquatio n4 :

RGR(W)= Vß 2 -4(o t- I n (W))Y (week ) (9)

This equation allows comparison between cold treatments at a givenplan t mass rather than at a certain moment, thereby preventing differences in growthparameter s arising fromdifference s indevelopment . InFigur e 26,th eRG R isplotte d againstplan tmass .Th enatura l decrease

73 RGR (week"1) 1.0 h 1.0 1.0 - Photophase Photophase :1 5 h Photoph nse :2 4 h .8 .8

.6 .6 ==—v- 3r3 V P t V,2Rr2 .4 V,P2 4 ^V P

.2 .2

i i I I 0 12 16 0 4 8 12 16 Mass per plant (g) Fig. 26. Relative growth rates as function of mass of dry matter per plant for various combinations of chilling and photophase (as inFigur e23) .

in relative growth rate with advancing stage was less marked in chilled plants. At aplan tmas s of 16g ,V 3 and/orV 2plant s grew fasterwit h each light phase. Immediately after chilling, the contrary seems tob e true,perhap s because cold-treatment plants hadmor e acclimatization problems inresump ­ tiono fgrowt h thanunchille dplants . In the same way as relative growth rate based onplan tmass , agrowt h parameter canb ebase d on leaf area,th erelativ e expansion rate (RER). Normally the formuleis :

RER (t)= l/LA(t ) dLA/dt (week -1, (10)

To make RER a function of leaf area rather than time,th e same procedure canb e followed as for theRGR :

RER(LA) =r 4(a -I n(LA)) y (ID inwhic h a, ßan d yar eth e regression coefficients for theestimat e oflea f area per plant. In Figure 27,RE R isplotte d against leaf areape rplant . Ateac h leaf area,P 2 leaves expanded faster than PI. Those of P3, however, expanded slower in all chilling treatments. InFigur e 28,th e effecto fa previous cold treatment onexpansio n of leaf areaca nb e read foreac h light phase, againa ta particula r leafarea .Th e impression isobtaine d thatcol d treatment affects this ratetoo .Especiall y theplant swit h 9an d 24h light phaseha d agraduall y increasing advantage with increasing coldperiod . Yet another growthparameter ,ne t assimilation rate (NAR), caneasil yb e computed with thecalculate d regression coefficients:

74 Relative expansion rate (week ) 1.0 h

V, :unchille d

V3 : chilled 55 days

^V3P2 .4 - VV3P,

.2 -

_l_ _L _l_ _L 0 200 400 600 800 1000 Leafare ape rplan t (cm2) Fig.27 .Influenc e ofphotophas eo nrelativ erat eo fincreas ei nlea fare a (asa functio n oflea farea )afte rchillin g for0 ,2 7an d5 5days .

75 Relative expansion rate (week" 1) 1.0 -

.8 Photophase :9 h

2 -

0

1.0

Photophase :1 5 h

V9P7

-V,P, B

0 j i_

1.0

Photophase : 24h

0 200 400 600 800 1000 Leafare ape rplan t (cm2) Fig.28 .Influenc eo ftim eo fchillin g (VI,0 ;V2 ,2 7an dV3 ,5 5days )o nth erelativ erat e of increase in leaf area (asa functiono flea farea )wit hphotophase so f9 h (A), 15h (B)an d2 4h (C )afte rchilling .

76 NAR(t) = l/LA(t) dW/dt (12)

l/LA(t)W(t ) 1/W(t) dW/dt (13)

W(t)/LA(t) RGR(t) (week-1) (14)

Figure 29 showp assimilation rate for the various treatments as a func­ tion of time. Until at least 5 weeks after the end of chilling, the higher growth of P2 treatments did not result from higher assimilation. After that date, assimilation of P2 and P3 plants increased markedly. It might be simplistic to attribute this to daylenght as such. For example, it could be caused by a higher senescence rate of leaves. Table 16C and 16D indeed show that the number of leaves of the P2 and P3 treatments was much reduced after the third harvest. This might point towards a suddenly in­ creased loss of older leaves by the long-day plants. Such a phenomenon could account for the higher NAR, as the remaining leaf area seems then to be more efficient. Though often considered representative of photosynthetic capacity, one of the main drawbacks of using a growth parameter like NAR, is that it does not take into account the quality of the leaf area. The efficiency of the P2 and P3 plants after the harvest at the fifth week might therefore be overestimated.

6.3.4 Discussion

Both factors known to have a strong promoting influence on bolting also stimulated growth. Especially the influence of light was pronounced. The results were similar to those of Milford & Lenton (1976), who concluded that

NAR (mg .cm" week"2 1) 10r 10 10

V3P3 8 8 V3P2 V2P3

6 V2P2 6

V2R 4h A V3P,

2 V!: unchille d V2:chille d 27day s 2 V3: chilled 55 days

0 6 8 0 2 4 6 8 Time from end of chilling (weeks) Fig.29 .Ne tassimilatio nrat eafte rchillin gfo rth evariou scombination so fvernalizatio n andphotophase .

77 the increased growth at the longer photoperiod (light phases of 12h an d 12h + 4 h )wa s effectuated by achang e inlea f arearati o and specific leaf area andno tb y an increase inth ephotosyntheti c activity ofth e leafsur ­ face. Until the 5thweek , Figure 29 supports that view. As has alreadybee n mentioned, the rapid increase of the NAR after the 5thwee k forP 2 andP 3 plantsma yb eunreliable . The observed negative effect of continuous light was also observed in spinach by van Oorschot (1960), who found anoptimu m curve for drymatte r production. An optimum for fresh and dry matter was with aphotophas e of 21h . The extension of the light phase with incandescent lampsbrough t about morphogenetic effects,whic h resemble the effects whengibberelli n isap ­ plied to sugar-beet plants.Especiall y themarke d effecto npetiol e length and leaf shapewa sver y similar.Th e same effecto f longphotophas e andgib ­ berellinwa s observed inspinac hb y Zeevaart (1971), who suggested thatlon g dayspromot e ahighe r rate ofgibberelli nbiosynthesi s and increased sensi­ tivity togibberelli n tocaus e theobserve d growth responses.Probabl y this isa phytochrome-mediate dresponse , inducedb y redo r farre d lightemitte d by the incandescentlamps . Milford & Lenton (1976)mentione d that alsogrowt h inth e fieldma yb e influenced by the spectral radiant energy ofnatura l daylight.Accordin g to Smith (1975,p . 151),a shift towards far red canoccu rwithi n the canopy ofth esugar-bee t crop.Th e field observation thatpetiol e length andweigh t increased at high plant densities might alsob e connected with the change inth e spectrum insuc h acanopy . This morphogenetic behaviour (faster expansion of the leaf area)migh t help to achieve aclose d canopy earlier inth e season.Therefore ,possibl e genotypic differences in reaction towardsphotophas e or differences inre ­ action to particular wavelengths should perhaps receivemor e attentionb y breeders (orgrower s of indoorvegetable s to save energy). Alsopreviou s cold treatmentha d apositiv e effecto ngrowt h of theplants . RGR decreased ata slowe r rate inchille dplant s than inunchille d onesan d leaf areaexpande d faster. Are these effects,i fcause db y thecol d treatment, related tovernaliza ­ tion (the flower-inducing process)? Ordoe sth e lowtemperatur e cause some changes independent ofvernalization ? Behaeghe (1978), forgrasses ,consid ­ ered thepositiv e effecto f low temperature ongrowt h completely independent ofvernalizatio n and socoine d thenam e 'hibernation' (Figure30) . Yet, form y trial and thato fBehaegh e stimulation of growth shouldper ­ haps not be ascribed mainly to the differences in previous treatmentbu t rather as asid e effecto f lowtemperature .Fo r example,durin g coldtreat ­ ment, rootdevelopmen tmigh tb e relatively favoured,whic h could have apo ­ sitive influence inth e followingperiod .

78 COLDTREATMEN T

HIBERNATION VERNALIZATION

GROWTHSTIMULATIO N FLOWERINDUCTIO N

Fig.30 .Hypothetica lrelatio nbetwee ncol dtreatment ,growt hstimulatio nan dflowe rin ­ duction(afte rBehaeghe ,1978) .

Considering thepositiv e effecto nproductivit y resulting fromphotophas e and from lowtemperatur e onemus t askals owhethe r there isa definit erela ­ tionwit hth eusua l largerplan tweigh to fjus tvisibl yboltin gplants .Jus t as Behaeghe (1978)believe d thatth egrowth-stimulatin g effecto flo wtem ­ perature couldb e independent ofth e flower-inducing effect,th e samecoul d apply forphotophase .Figur e 31represent s sucha relation . Arguments for such anindependen t actiono fphotophas e ongrowt h ando n flower inductionare :

Only chilled plants respond to photophase in flowering, whereas both chilled and unchilled plants respond toth emorphogeneti c actiono fphoto - phase. - Thephotophas e causing themos trapi d flowering seemst ob e 24h (Chapter 2; Curth, 1960), whereas forgrowt h ashorte rphotophas ewa s optimum. Despite thesediscrepancies ,ther ewer e similarities.Gibberelli nma yb e oneo fth ecomponent s ofth e floweringhormon e orma ypla y some role inth e stem-formationprocess .Th eenhance dbiosynthesi s ofgibberelli nwit h longer

PHOTOPHASE

PHOTOMORPHOGENESISy .PHOTOPERIODISM

GROWTHSTIMULATIO N FLOWERINDUCTIO N

FLOWERING

Fig.31 .Hypothetica lrelatio nbetwee nphotophase ,flowe rinductio nan dgrowt hstimulation .

79 photophases (Zeevaart, 1971) could have twoeffects ,on e on stem formation and one on increasing leafexpansion .Thes eprocesse s may have adifferen t optimum concentration ofgibberellin .

6.4 INFLUENCE OF SELECTION TOWARDSBOLTIN G RESISTANCE

6.4.1 Introduction

Inth e schemes describing the influence ofphotophas e and coldtreatment , anarro wwa s drawnwit h aquestio nmark .(Figure s 30 and 31), since evidence was required whether plants induced to flower (or more liable to flower) had an increased growth rate. Ifso ,selectio nb ybreeder s to improve bol­ ting resistance could mean alos s ingrowt h rate.T o answer thisquestion , the plant breeders vande rHav eB.V . supplied several genotypes inwhic ha possible negative influence on productivity of such a selection couldb e tested.

6. 4. 2 Materials and methods

Most modern cultivars are triploids (3n) obtained by crossesbetwee na tetraploid (4n) and a diploid (2n) monogerm malesterile (HoMs) genotype. Selection towards bolting resistance in three tetraploids (Tl,T 2 andT 3) was carried out byva nde rHav e in197 5b y sowing early. From thenon-bol ­ tingplants ,see dwa s grown inth eyea r 1976:TS1 ,TS 2 andTS3 . Crosseswer e madebetwee n theselecte d tetraploids andon e diploid monogerm male-sterile genotype and alsobetwee n theorigina l populations and the same diploidge ­ notype.Th e resulting 6triploi d single crosseswer e sown out,togethe rwit h the 6 tetraploid genotypes (Table 20) at 5 sites in 1977:a tWageningen , Frederika Polder and ZimmermanPolder , being sown on 20,2 7an d 21April , respectively, and in Germany atCoverde n and Sollingen on 21Marc h and2 6 April, respectively. Except at Wageningen, the trials were supervised by vande rHave ,a spar t ofthei rvarieta ltrials .

6.4.3 Results

Inth eearly-sow n trial atCoverden ,man ybee tbolted . Comparison ofSe ­ lection 1wit h 2,3 wit h4 , and soon ,showe d that selection hadbee neffec ­ tivebot h fortetraploid s and triploids (Table20) . To investigate the influence of selection on productivity, periodical harvestswer e carried outa tWageningen .Tabl e 21 shows therelevan t figures of the first and second harvest.Th e data ofth e firstharves t indeed sug­ gest a slight reduction indr ymatte r and leaf areape rplant .Analysi s of variance showed,however , thatthi swa sno t significant. (P <0.143 )an d (P < 0.151), respectively. For the second harvesto n2 9 June especially root

80 Table 20. Influence of a selection for bolting resistance on the final proportion of bolters in a field trial at Coverden.

No. Genotype Seed Ploidy Selection Proportionbol ­ production for bolting tinga tCoverden , year res istance 11Octobe r

1 Tl 1071 4 n _ 10.8 2 TS1 1976 4 n + 1.7 3 Tl*MoMs 1975 3 n - 5.1 4 TSl*MoMs 1976 3 n + 0.3

5 T2 1974 4 n - 36.5 6 TS2 1976 4 n + 5.4 7 T2*MoMs 1975 3 n - 13.0 8 TS2*MoMs 1976 3 n + 6.0

9 T3 1973 4 n - 15.1 10 TS3 1976 4 n + 3.0 11 T3*MoMs 1976 3 n - 6.3 12 TS3*MoMs 1976 3 n + 0.8

weights tended to be less for the more bolting-resistant genotypes (P < 0.079). In succeeding harvests of the trial, the effect disappeared, however. At the final harvest, there was no significant difference in plant dry mat­ ter between selected and unselected genotypes. For the other four trials, only the final harvest was available. Because bolting in the Coverden trial prevents proper comparison for productivity between selected and unselected genotypes, those data were excluded. The mean for the three other trials are shown in Table 22. The selection effect

Table 21. Influence of a selection for bolting resistance on mass of total dry matters (g) and leaf area (cm ) per plant at two harvest dates.

Genotype Harvestdat e

27Ma y 29Jun e

drymatte r leafare a sprout beetroo t leaf drymatte r drymatte r area

Tl 0.25 33.1 32.0 8.8 3556 TS1 0.27 33.6 32.4 8.5 3668 Tl*MoMs 0.27 32.6 32.3 10.3 3587 TSl*MoMs 0.25 31.5 31.8 9.7 3553

T2 0.25 32.7 29.9 8.6 3240 TS2 0.25 31.6 30.3 8.4 3450 T2*MoMs 0.26 33.3 32.9 10.5 3590 TS2*MoMs 0.24 30.1 32.8 9.8 3546

T3 0.28 33.1 29.6 8.7 3402 TS3 0.24 29.8 29.2 8.8 3211 T3*MoM s 0.29 35.6 34.1 11.5 3440 TS3*MoMs 0.27 34.3 33.7 11.0 3744

81 Table 22. Influence ofa selectio n forboltin g resistance onfina l root yield andsuga r content.

Genotype Average of three trials (withe ut Coverden):

root yield sugar con tent white sugar (t/ha) (g/kg) (g/kg)

Tl 51.97 173.9 150.7 TS1 52.03 175.1 151.8 Tl*MoMs 54.53 179.3 154.9 TSl*MoMs 55.27 178.8 154.9

T2 45.97 189.1 165.9 TS2 46.10 189.7 166.4 T2*MoMs 51.83 186.6 162.7 TS2-MoMs 49.80 189.1 166.4

T3 53.37 177.6 151.4 TS3 52.03 178.8 152.9 T3-MoMs 57.13 181.8 157.1 TS3*MoMs 54.83 182.4 157.6

was insignificant (P < 0.19). Surprising was that the selected genotypes showed aconsistentl y higher contento f sugar ando fwhit e sugar (P <0.017 ) and (P< 0.015 ).

6.4.4 Discussion

Thepossibl e negative effect onproductivit y ofpreviou s selection against bolting could not be detected. Only the second harvest inth eWageninge n trial showed less rootbu t ata lo w significance.Th e influence of selection might, however, be confounded with differences in seed age (Table20) . Further the question ariseswhethe r these trialswer e sufficiently discrimi­ native todetec tpossibl e small effects of selection.Th e question alsoar ­ iseswhethe r thereha dbee nonl y aselectio n againstboltin g or that,unwit ­ tingly, also other selection criteria hadplaye d arole .O fth e non-bolting plants in the selection year, only themos tvigorou splant smigh tperhap s have been chosen forproductio n of the improved tetraploids.Suc h aselec ­ tion would counteract possible negative effects of selection only against bolting. For the positive effect on sugar content,n o explanation isavailable . According to the breeder, selection was only against bolting andno tfo r sugarcontent . Lysgaard (1978) investigated theeffec t of selection againstboltin g on yield in fodder-sugar-beet.Hi s trials,i nwhic h none ofth eplant sbolted , showed that selection againstboltin g did not reduce drymatte rproductivi ­ ty. Intw o ofth ecultivar seve n anincreas ewa s found.Als oYusubo v (1977) stated thateliminatio n ofboltin gbiotype s from tetraploid sugar-beetpopu ­ lations didno treduc eyield .

82 So there is nostron grelatio no fboltin g resistancewit hproductivity , in contrast to spinach where selection inevitably leads toreduce d growth (Parlevliet, 1967). In spinach, however, the difference betweenearl yan d late cultivars might be larger than insugar-bee tcultivars ,whic h already possess an acceptable resistance toboltin g (biennial character). Ingeno ­ types with more extreme differences in bolting, a relation mightcom et o light.

83 7 Temperature and bolting under field conditions

7.1 INTRODUCTION

In certain years,boltin g in sugar-beet occurs to such anexten t that yield is reduced. Inth eNetherland s theyear s 1972 and 1973wer e knowna s such "bolteryears" .Ofte n iti sno tknow n why suchyear sdeviate . Itmigh t be due to early sowing,whic h lengthens the subsequentperio d of low temperature. Further the temperature in corresponding periods between theyear smigh tb ebelo w average.Furthermor e insom eyears ,th e absenceo f high (devernalizing)temperature smigh t increase theris ko fbolting . This chapter correlates the course of temperature after sowing toth e observed finalpercentag e ofbolter s inth e field.Dat awer ekindl y supplied by the 'Instituut voor Rationele Suikerproduktie' (1RS)a tBerge n opZoo m and alsob y the 'Rijksinstituutvoo r Rassenonderzoek' (RIVRO)a tWageningen . These institutes arrange annual trials on sowing date for several cultivars atsite s throughout the country. Sowing and emergence dates,an d finalpro ­ portion ofbolter swer e recorded. Data were from trials with cvs.Monohi l and Polykuhn. For the period 1966-1976, 53 sowing dates (further indicated with cases)wer e available for Monohil and 60sowin g dates forPolykuhn .Th e finalpercentag e ofbol ­ ters was transformed to arcsines: angle (%) = arcsine (4%/Ï0 0) . In the trials, true 'annual'plant s (caused by 'contamination' with annual beet types)wer eno tinclude d inth e finalproportio no fbolters .

7.2 THERELATIO NBETWEE NBOLTIN GAN D SOWINGDAT E

A first impression of the data is given by Figure 32,whic h showsth e transformed percentages inth edifferen tyear s forcv .Monohi l and suggests rather good relation with the sowing date,a sexpected .A 2nd-degreepoly ­ nomial against time from 1Marc h to the sowing date had acoefficien t of 2 determination (r )o f 0.508. Figure 33 shows the same angles againstdat e 2 ofemergence .Th e relationwa spoorer : r - 0.282. Incv .Polykuhn ,th e res­ pective coefficients were 0.620an d 0.412.Th ebette r relationwit h sowing date than date ofemergenc e supports remarksmad e in Section 5.1 aboutth e absence of a true juvenile phase. Ifvernalizatio n took place only after emergence,th erelatio nwit h dateo femergenc ewoul db e atleas ta sgood . Figure 32 already allows satisfactory prediction of the proportion of bolters (r= 0.7!). However, thecours eo ftemperatur e after sowingo rafte r

84 ! of bolters

a =1966 r' =50.8% à. =1967 * = 1968 • = 1969 v = 1970 « = 1971 •=1972 • = 1973 + = 1974 0=1975 • =197 6

4 May Sowingdat e Fig. 32.Relatio nbetwee n sowing date and theproportio n ofbolter s (transformed into angles)unde rfiel dcondition sfo rcultiva rMonohil .

emergence for each casema ygiv e astil lbette r relation,whic hwoul d also be more causal than the indirect relation with the sowing date. Certain years showed asystemati c deviation from theproportio n ofbolter spredicte d (Figure 32).I n1969 ,1972 ,an d 1973,mor ebolter s appeared,wherea s in1971 , 1974an d 1976 fewerbolter s developed thanwoul db eexpecte d from the sowing date. An abnormal course oftemperatur ewa sprobabl y responsible forthos e deviations.Althoug h 1976ha d acol d spring,probabl y theproportio n ofbol ­ terswa sreduce db y thehig htemperature s inlat e spring and insummer .

7.3 AREGRESSIO N APPROACH

How can the relationtemperatur e andproportio n ofbolter sb e analysed? Severalchoice shav et ob emade .

Inwha t period should thetemperatur e be considered: afteremergenc eo r after sowing, andho w long should thisperio dbe ? Should daily,weekl y ormonthl y temperaturesb eused ? Which temperatures shouldb eused ,maxima ,minim a or average temperatures?

First itwa s decided tous e only thetemperatur e datao fth eMeteorologi -

85 Angleo fbolter s 16

a=196 6 ^=1967 *=1968 •=196 9 v=1970 »=1971 •=1972 •=1973 +=1974 0=197 5 •=1976

20 30 10 20 30 April | May Dateo femergenc e Fig. 33.Relatio n between date of emergencean dth efina lproportio no fbolter s (angles) underfiel d conditions forcultiva rMonohil .

cal StationD eBil tnea rUtrecht .O fcours e themos t accurate procedure would havebee nt ous eth e temperature ofth eneares t station toeac ho fth e trials. Inspection ofth eboltin g data,however , revealed noconsisten tre ­ gional deviation. There aren o extreme temperature differences betweenth e differentpart so fth eNetherlands . As the final proportion ofbolter s isrelate d toth enumbe r of 'vernal­ izing' days after sowing oremergence ,on emus tdecid ewha t should becon ­ sidered as a 'vernalizing'day .Therefore , foreac hcas e (observation,sow ­ ing date) the dayswer ecounte dwit h atemperatur e ofu p to 0,2 ,4 , 6,8 , 10, 12,1 4an d 16 °C,givin g 9number spe r case.Thes e countswer emad e for the period between sowing date and 1July , forminima ,maxim a and average daily temperatures. With each of the counts,a simpl e correlationcoeffi ­ cientwa s computed with the angle (arcsine)o fth e finalproportio n ofbol ­ ters (Table 23).Fo rbot h cultivars thecloses tcorrelatio n (0.76 and 0.83) was with number of days with a maximum temperature upt o 12 °C.Th e same procedurewa s followed fornumbe r ofday s from emergence rather than sowing (Table 24). On average, the correlation coefficients were smaller, ason e would expect by comparison between Figures 32 and 33.Th e relevantperio d should thereforebegi n atsowing .

86 Table 23. Simple correlation between the angle of bolters and the number of days from sowing date until 1 July with a (minimum, average or maximum) temperature 40, 2, 4 16 °C.

Variety Temperature Tempe rature (°C)

\<0 42 <<4 «6 v<8 410 s<12 .<14 N<16°C

Monohil minimum 0.52 0.50 0.54 0.64 0.67 0.67 0.69 0.68 0.68 n=53 average 0.06 0.14 0.37 0.49 0.75 0.73 0.73 0.68 0.70 n=53 maximum - - 0.02 0.09 0.43 0.72 0.76 0.76 0.74 n=53

Polykuhn minimum 0.53 0.68 0.66 0.76 0.75 0.71 0.72 0.72 0.72 n=60 average 0.02 0.03 0.23 0.63 0.78 0.79 0.78 0.76 0.73 n=60 maximum - - 0.06 0.06 0.35 0.76 0.83 0.79 0.81 n=60

-: correlation coefficient cannot be computed.

Daily maxima gave the best correlation with bolting. Daily minima are probably less indicative of a really vernalizing day. In the climate of the Netherlands, days in May or June with a low minimum temperature can be ab­ normally warm during daylight. Further studies revealed that also for the relation with a devernalizing day the best relation was found when daily maxima were used. Daily maxima were therefore used. The following question had to be answered now: in what period should tem­ perature be considered? A choice had to be made between two options:

a fixed period from sowing, e.g. 8 weeks after sowing a period from sowing to a fixed date, e.g. from sowing until 1 July

A fixed period has the advantage that time is equal for each case but the disadvantage that an early sowing date gives quite a different interval from a later sowing date and results in a different photophase. An interval to a fixed date means that at least the last part of the in­ terval is equal for each case. The beginning will differ for each sowing

Table 24. Simple correlation between the angle of bolters and the number of days from emergence until 1 July with a (minimum, average or maximum) temperature 40, 2, 4 16 °C.

Variety Temperature Temperature (°C)

40 42 4U S6 4& N<10

Monohil minimum 0.25 0.240.3 20.5 0 0.580.5 7 0.580.5 5 0.55 n=53 average - - 0.240.1 5 0.600.5 80.5 90.6 00.5 9 n=53 maximum - - - - 0.390.5 00.5 00.6 20.5 6 n=53

Polykuhn minimum 0.480.5 3 0.57 0.64 0.68 0.64 0.64 0.61 0.61 n=60 average 0.23 0.54 0.63 0.63 0.63 0.70 0.64 n=60 maximum - - 0.29 0.60 0.61 0.66 0.69 n=60

-: correlation coefficient cannotb ecomputed .

87 date,bu ttemperatur e will thenb e low andprobabl y photophase isthe nles s important (Section 5). A fixed dateprove d more succesful, according toto - 2. tal r inth e regressionmodels . Now the following regression model is proposed. For each observation, the following counts weremad e (and tested asvariâtes )fro m sowing tosi x fixeddates :1 5May , 1June ,1 5June ,1 July , 15Jul y and 1August :

n. =numbe r ofday swit h amaximu m temperature^ 12° C n_ =numbe r ofday swit h amaximu m temperature >1 2 and- S1 6° C n,= numbe r ofday swit h amaximu m temperature >1 6 and- S2 0° C n. =numbe r ofday swit h amaximu m temperature >2 0 and« 24° C n.= numbe r ofday swit h amaximu m temperature >2 4an d^ 28° C n,= numbe r ofday swit h amaximu m temperature >2 8° C

To find the relation with the process ofvernalization , aquadrati crela ­ tionship withn was assumed, therefore thevariabl e x andx _ were included inth eregressio nmodel :

x2 = n\

Further it is reasonable to assume that the effect of a devernalizing day will depend on the degree ofvernalizatio n inth eprecedin gperiod .A devernalizing daywil l notreduc e thenumbe r ofbolter s after alat e sowing date to the same extent than after anearl y sowing,becaus e thenumbe r of potential bolters is already lowerwit h late sowing.Therefor e aninterac ­ tion is assumedwit h thenumbe r of 'vernalizing' days. The following dependentvariable swer e computed (foreac h case):

X3 - B2 ni X4 = n3 ni X5 = n4 ni X6 = n5 ni n X = n 7 e • i

The following expressionwa s chosen asth e complete regressionmodel :

p..= ax..+ bx7 + ex. + dx„ +ex .+ fxfi + gx„ + C +e . . (15) where p . .wa s angle ofbolter s inth eyea r j after asowin g date i.

The bolting resistance of the cultivars was assumed toremai n thesam e from year to year, although this mayno tb e true.Breeder s are constantly selecting in their families and lines. It could only be hoped that this wouldno tb eto odisturbin g afactor .Regressio n analysiswa swit hth ecom ­ puterpackag e SPSS (SPSSmanual , 1975). Stepwise inclusionwa s combined with hierarchical inclusion.Th evariable sx . andx , (vernalization)wer e entered together in the firststep ,th evariable s x,t ox „wer e entered ina step ­ wise inclusion,provide d theyme tth e statistical criteria.Th e statistical criteria forthes evariable swer e theF value ,an d aparamete r calledtoler ­ ance.Th etoleranc eo fa n independentvariabl ebein g considered for inclusion isth eproportio n of itsvarianc eno texplaine d byth e independentvariable s already inth eregressio nequation .Th etoleranc e (avariabl ebetwee n0 an d 1)wa s seta t0.3 ,whic hmean s that3 0% ofth evarianc e ofa potentia lin ­ dependent variable is unexplained bypredictor s already entered.Th emode l as awhol e has the advantage of a limited number ofregressio nvariable s (atmos t 7).Especiall y ina smal lpopulatio n alarg enumbe r ofregressio n variables leadst ospuriou sresults .Hanu s& Aimille r (1978)hav e discussed thisproble m predicting cerealyield s inGerman y fromweathe r data.A further advantage of the model is that it is based onphysiologica l processes in theplants . To assess inwha tperio d temperature exerts the largest influence onbol ­ ting, the regression was analysed for 6 intervals up to 1August ,a tth e most. Table 25show sth eproportio n ofvarianc e explained byeac hvariabl ein - eluded inth eregressio nequation .Fo r allvariables ,tota l r (coefficient ofdetermination )i spresente d inth e lastcolumn .Th econtributio n totota l 2 r can be read foreac h ofth evariable s inth emodel ,an d itgive sthere -

Table25 . Attributedproportio no fvarianc et oeac ho fth eregressio n variables indat ao fth eyear s1966-1976 .

Timefro m 'Cold' Intera ctions Total sowingdat e 2 until <:12° C («12 °C)2 SUM 12- 16- 20- 24- >28° C - (Xj) (x2) 16° C 20° C 24° C 28° C (x7) (x3) (x4) (x5) (x6) Monohil 15Ma y 59.1 1.1 60.2 - 5.3 3.0 - 8.0 76.5% 1Jun e 61.3 1.6 62.9 - - 4.3 1.6 6.3 75.1% 15Jun e 59.3 1.2 60.5 - - - 15.7 1.7 77.9% 1Jul y 58.3 1.3 59.6 - - - 12.6 5.9 78.1% 15Jul y 58.3 1.3 59.6 - - - 17.3 4.0 80.8% 1Augus t 58.3 1.3 59.6 - - - 12.5 2.6 74.7% Polykuhn 15Ma y 74.5 9.7 84.2 - 3.6 4.0 1.2 1.6 94.5% 1Jun e 69.6 11.4 81.0 - - 9.6 - 1.0 91.6% 15Jun e 69.3 11.5 80.8 - - 7.2 - 1.0 89.0% 1Jul y 68.4 11.7 80.1 - - - 5.1 - 85.2% 15Jul y 68.4 11.7 80.1 - - - 4.5 0.4 85.1% 1Augus t 68.4 11.7 80.1 - - - 2.1 0.6 82.8% -:regressio nvariabl e isno tinclude d inth eequation .

89 Table26 . Regressioncoefficient s included inth eequatio no nth ebasi s ofstatistica lparameters .Year s1966-1976 .

Time Regression coefficients for variables C Xl X2 X3 X4 X5 X6 X7 Monohil 15Ma y 0.399E-1* 0.779E-2 0.203E-1 -0 142E-1 -0.730E-1 0.18 1Jun e 0.391E+0 0.279E-2 -0 189E-1 -0.197E-1 -0.426E-1 0.94 15Jun e 0.224E+0 0.489E-2 -0.277E-1 -0.155E-1 1.41 1Jul y 0.140E+0 0.878E-2 -0.203E-1 -0.122E-1 2.16 15Jul y 0.226E+0 0.806E-2 -0.209E-1 -0.612E-2 1.95 1Aug . 0.250E+0 0.683E-2 -0.148E-1 -0.483E-2 1.58

Polykuhn 15Ma y -0.190E+0 0.208E-1 0.432E-1 -0 429E-1 0.294E-1 -0.908E-1 0.98 1Jun e 0.443E+0 0.124E-1 -0 474E-1 -0.723E-1 0.43 15Jun e 0.453E+0 0.116E-1 -0 300E-1 -0.415E-1 0.38 1Jul y 0.696E-1 0.207E-1 -0.355E-1 1.31 15Jul y 0.798E-1 0.217E-1 -0.253E-1 -0.851E-2 1.26 1Aug . 0.306E-1 0.224E-1 -0.148E-1 -0.967E-2 1.14

.:variabl eno tinclude d inth eequation . ~:th evalue sar epresente d in 'E-format';th enumbe rfollowin g theE i sth epowe ro f10 . 0.123E-2 = 0.00123 0.123E+2 =12.3 0.123E+0 = 0.123 fore an impression of the influence on bolting of the temperature range corresponding with the variables. Table 25 shows that besides the variables x and x_ (which are always included and represent the low temperature ef­ fect), some of the higher temperature variables were included in the equa­ tions. These variables represent devernalization, because they reduce the number of bolters. For Monohil, increasing the interval from sowing only 2 slightly increased total r .Fo r Polykuhn, the largest value (the best ex­ planation) was obtained when only the temperatures in the short interval from sowing until 15 May were considered. 2 There were other differences between the cultivars. Total r was larger 2 for Polykuhn and the variables x and x_ contributed much to total r (about 2 80 %). For Monohil, total r was somewhat less, x..an d x? contributed only 60 % and in this more resistant cultivar, high temperature variables seem largely to determine final proportion of bolters. Table 26 shows the corres­ ponding regression coefficients for the variables included. The coefficient was positive for x. and x and, of course, negative for high temperature variables. 2 The observation that total r for Polykuhn decreased with a longer inter­ val (until 1 August) indicates that later in the season temperature does not have much influence on bolting, as observed in Section 5.5. For the resistant cultivar Monohil, not only high temperatures seemed to have more influence but also the interval in which these temperatures were effective seems to 2 be longer. Total r increased if the interval were extended to 15 July. Table 26 shows that upon extending the period in which temperature was considered from 15 May to 1 August variable x shifted from -0.07 to -0.0048 90 Table27 . Attributedproportio no fexplaine dvarianc et oeac ho fth e regressionvariables .Dat afro mth eyear s 1967-1975 (without196 6an d 1976).

Timefro m 'Cold' Interactions Total sowingdat e 2 until <12° C (<12 °C)2 SUM 12- 16- 20- 24- >28° C - (Xj) (x2) 16° C 20° C 24° C <28° C (x7) (x3) (x4) (x5) (x6) Monohil 15Ma y 64.7 2.0 66.7 - 7.8 4.8 - 2.8 82.0% 1Jun e 65.4 2.1 67.5 - - 7.3 1.4 3.2 79.5% 15Jun e 64.5 1.8 66.3 - - 2.7 9.7 - 78.6% 1Jul y 63.1 2.0 65.1 - - - 13.1 - 78.2% 15Jul y 63.1 2.0 65.1 - - - 14.5 - 79.6% 1Augus t 63.1 2.0 65.1 - - - 8.3 - 73.4% Polykuhn 15Ma y 73.5 9.8 83.3 - 3.7 5.7 1.4 - 94.1% 1Jun e 69.4 11.1 80.5 - - 11.3 0.4 - 92.2% 15Jun e 70.3 10.0 80.3 - - 9.0 0.5 2.9 92.7% 1Jul y 69.1 10.5 79.6 - - - 9.0 1.5 90.1% 15Jul y 69.1 10.5 79.6 - - - 6.5 3.1 89.2% 1Augus t 69.1 10.5 79.6 - - - 2.5 3.0 85.2% -:regressio nvariabl eno tinclude di nth eequation .

for Monohil and from -0.0908t o -0.00967 forPolykuhn , inlin ewit h anob ­ servation inTria l 1an d2 (Section5.5 )tha twar m days later aftervernali ­ zationha d less influence onth e finalproportio n ofbolters . To investigate theeffec to fdiscardin g someyear s fromth e datacollec ­ tion, the regression procedures were repeated without the years 1966an d 1976.Th e results arepresente d inTabl e 27an d 28.Som edifference s canb e observed between Tables 25 and 26.Variable s x_wa s less ofteninclude d in theequations .Discardin g ayea r like 1976 altered the significance ofthi s variable. This illustrates the importance of having a sufficiently large and representative population inorde r topreven t irrelevantvariable s from being included by coincidence or important variables frombein g leftout . InTabl e 27 (incontras tt oTabl e 25), thelarges tr (Monohil)wa s fortem ­ perature fromsowin gunti l 15May .However ,thi sequatio ndoe sno tsee mver y reliable, considering the positive regression coefficients forx andx_ . Days with amaximu m temperature above2 8 °Cwoul dhardl y increaseth epro ­ portion of bolting plants. So, it isstil lpossibl e thathig h temperature has astronge r influence,whic h remainsactiv e fora longe rperiod , inbol ­ ting-resistantcultivars . o The largerproportio n oflo wtemperature s inr doesno tnecessaril ymea n that Polykuhn would bemor e sensitive tolo wtemperature . Ifdevernalizin g temperatures less influenced the proportion of bolters in this cultivar, bolting would mainly be determined by lowtemperature s andwoul db e easier 2 topredic t (higherr )tha n ifboltin gwer e dependento ntw oprocesses . Similar studies on the relation of temperature toboltin gwer emad eb y

91 Table28 . Regressioncoefficient sinclude d inth eequatio no nth ebasi s ofstatistica lparameters .Year s1967-1975 .

Time Regression coefficients forvariable s

X, X3 X4 Monohil 15Ma y -0.528E+1 0.112E-1 0.213E-1 -0.220E-1 125E+0 2.26 1Jul y 0.465E+0 0.271E-2 -0.296E-1 -0.178E-1 113E+0 0.64 15Jun e 0.496E+0 0.119E-3 -0.104E-1 -0.252E-1 0.13 1Jul y 0.944E-1 0.103E-1 -0.213E-1 2.47 15Jul y 0.194E+0 0.895E-2 -0.221E-1 2.07 1Aug . 0.219E+0 0.729E-2 -0.142E-1 1.55

Polykuhn 15Ma y -0.267E+0 0.229E-1 0.410E-1 -0 468E- 1 0.291E-1 1.92 1 June 0.454E+0 0.136E-1 -0 535E- 1 -0.141E-1 0.57 15 June 0.323E+0 0.153E-1 -0 251E- 1 -0.139E-1 -0.112E+0 1.06 1 July -0.380E+0 0.320E-1 -0.324E-1 -0.592E-1 4.89 15 July -0.321E+0 0.327E-1 -0.229E-1 -0.424E-1 4.49 1 Aug. -0.161E+0 0.312E-1 -0.181E-1 -0.464E-1 3.34

.: variab ] e not incl uded in the regression equation

Wood & Scott (1975). They, however, studied temperatures in the period 4-6 weeks after emergence, in contrast to the foregoing approach with tempera­ ture from sowing. Also Lasa (1977), in correlation studies in Spain found it was best to start the interval at emergence or even 30 days after emer­ gence. Spanish conditions are, however, quite different from the Netherlands, as the crop is sown in dry soil and the seed does not germinate until the soil is wetted. According to Lasa, emergence is much delayed in some years, which explains why the best relation is found with temperatures from emer­ gence rather than sowing. Lasa (1977) also stated that devernalizing tem­ peratures did not have a great role in the field. However, with the approach of the regression model (inclusion of higher temperatures with an interac­ tion term), high temperatures still seem to reduce the proportion of bolters in the field (at least for Dutch conditions). A further difference is that Lasa used average and minimum temperatures instead of maximum temperatures, which gave best correlation in my study.

7.4 AN OPTIMIZATION APPROACH

The regression method of Section 7.3 has shown that proportion of bolting can be predicted from a few weather data, especially low and high tempera­ tures. Some objections can be raised to this method:

In most of the resulting models, only days with a maximum temperature above 28 °C influence the final proportion of bolters. It is reasonable, however, to assume that days with a somewhat lower maximum temperature also have an effect, though less.

92 Relative (m+A) rate of change of substance V

Fig.34 .Hypothetica l relationbetwee n relative rate of change of substance Van dtem ­ perature.

Because of the small sample, some spurious correlations wereprobabl y pickedup .

Adifferen t approachwoul db e toregar d theproces s of floweringmor e in terms of the model (Figure 2). The final amount of F synthesized in the plants will, of course, show a perfectrelatio nwit hth e finalproportio n ofbolters .Presumabl y the amounto fV ata specifi c date also shows agoo d relation with the proportion of bolting plants.Suppos eth erelativ e rate of increase and decrease ofsubstanc eV asa functio n ofdail y temperature is known. Probably a function like the one in Figure 34woul d result.A t lower temperatures,V issynthesize d and athighe r temperaturesV isdeple ­ ted.Th e following function isuse d todescrib e therelativ e rateo f increase anddecreas e insubstanc eV :

m + \ f (T) =- A . + (16) {^)2 *1 . in which T = daily maximum temperature, K =lowe r asymptoticvalu e ofth e daily rate of change, mmaxima l dailyrat eo fchang e at T= ( jan da devia ­ tionroun d T= u . Usewa smad ebelo wo fdat a forMonohi l (p., j =1 ,2 , 52,53) . Itwa s assumed thata tth e sowing dateth e amounto fV wa s 1 (inarbitrar y units)an d thatth e amounto fV on 15Jul ywoul d show agoo d (linear)rela ­ tionwit h the final angleo fbolters .Fo reac hcase ,p. ,th erat eo fchang e i days after sowingwil lbe :

m + k Fij =-*+ T . . + 1. (17)

93 The amounto fV on 15Jul y (nday s after sowing)fo r acas ep .wil l thenbe :

Vj =l(r l7.+ 1. ) (r2j.+ 1. ) (ri;j +1 ) {rnj + 1) (18) where V. i s final content of V on 15Jul y andr .•rat e ofchang e onda yi after sowing forcas e j. To predict the final angle of bolters p., the following relationwit h the contento fV . was assumed: J

b . = aV . (19) J J With the least-square method, the following function Fwa s defined:

j'=53 2 F= I ( Pj. - Pj) (20) J'=l

Iti sa functio n ofth e temperature course after sowing ineac hcas e ando f theparameter s k, m, a, \i and a. Itwa sminimize dwit h thenon-linea r opti­ misation package OPTPAC3, (vanKilsdonk , Philips, Eindhoven, 1977)usin g thezero-orde r method ofHook e &Jeeves .Fo r thosevalue s ofth eparameter s that minimized function F, predictions ofth eproportio n ofbolter s should begood . To simplify the problem and to reduce thenumbe r ofparameters , itwa s assumed that p = 5 "C. Although the problem required alarg e core-memory and a long computer run, aminimu m of Fwa s found atth e followingvalue s ofth eparameters :

A = 0.052 m= 0.062 a =10. 5 °C a = 4.7

Figure 35 gives the corresponding function ofth edail y relative change inth econten to fV inth eplants .Wit h such arelatio n topredic tp. ,76.2 % ofth evarianc e inangle s ofbolter s couldb e accounted for.Figur e 36show s the prediction for theyear s 1976,197 7 and 1978 asa function oftempera ­ ture after sowing. The year 1976 cannot be considered atru eprediction , because the data of 1976wer epar to fth e data themode lwa sbase d on.Th e curves for 1977 and 1978 are on the other hand truepredictions ,becaus e these observations are additional data,mad e availableb y the two research institutes ata mor e recentdate .Th enumbe r of additional observationswas , however, too smallt odecid ewhethe r the approachwa svalid .Th emetho dwa s probably less dependent on occasional deviations in temperature, because

94 -0.04 Fig.35 .Relatio nbetwee nrat eo fchang eo fsubstanc eV an dtemperatur ewit hoptimization .

every daybetwee n sowing and 15Jul ywa s included inth ecalculations .Thi s is in contrast with the regression method where in somemodel s only days with amaximu m temperature -$12an d >28 °C.wer euse d topredic t the final 2 angle of bolters.Despit e this more realistic approach, total r wasno t higher thani nth e regressionmethod .However ,temperatur e ofever yda y from sowing date was used.Thi s ismuc hmor e 'difficult' forpredictio n thani n theregressio nmethod ,whic h easily could givecoefficient s of determination over 90% , fora large rnumbe r ofx-variable san dwithou t restricted inclu­ sions. This optimization method will certainly require further research and perhaps amodifie d physiological basis.It sprinciples ,however ,see m tob e useful and could alsob e applied forothe rtemperature-dependen t processes likegrowth , forexample .A n optimized temperature function instead oftem -

An gle of bolters 32 Anejl e of bolters An gleof bolters i 28" 28 28 A x 24 1976 1977 \ 1978 24 24 " \ \ 20 - 20 20 \ \ 16 16 16 \ 12 s \ 12 s 12 \ \ *\" \ 8 - \ 8 8 \ s« s X 4 4 ~ x ** 4

X n 1 1 1 • 1 0 0 1 I I I 1 1 I—«- '.. -\ l MARCH APRIL MAY MARCH APRIL MAV MARCH APRIL MAV Sowing date Sowing date Sowing date Fig.36 .Predictio no fboltin gangle sfo rth eyear s1976 ,197 7an d1978 .Th edat afo r197 7 and197 8wer eno tpar to fth edat ao nwhic hth emode lwa sbased .

95 perature sums, time-temperature product ofgrowt h (in °C.d)migh tb e abet ­ ter physiological basis forprediction .A disadvantage istha t iti sstil l not adynami c approach.Equatio n 18t ocalculat e the final amounto fV shows that inthi smetho d awar m day inJun ewil l have the same effecto nth epre ­ dictedproportio n ofboltin g than awar m day inMay . Inth edynami c approach of the model (Figure 2), however, this certainly makes difference. Early destruction ofV hinders F synthesismor e thandestructio n later inth esea ­ son (Trials 1an d2) . A simulation approach isthe n the solution.Th e estimation ofth e relevant parameters is for themoment ,however , impossible asonl y the final result (bolting)i sknow n andno tth e intermediate steps inth eproces s ofboltin g and flowering.

96 8 Final remarks

Let us first consider bolting in sugar-beet from the pointo fvie wo f breeders.T oreduc eboltin g inthei rplan tmaterial ,breeder s as arul e sow early and discard boltingplants .I ncertai nyears ,however , sucha selec ­ tion ishardl ypossible ,becaus e onecanno tso wearl yo rbecaus eth eweathe r isno tcol d aftersowing .Th edegre e of flower inductionwil l thereforevar y fromyea rt oyear .No tonl yth etota l strengtho fth einductio n canb edif ­ ferent between years, also the pattern of flower-inducing factors mayb e quite different in someyear s fromothers .I ncertai nyears ,a col d spring is followed by aperio d of hightemperatures ;i nothers ,a shorte rperio d of vernalizing temperatures is followed by aperio d of 'neutral'tempera ­ tures. So the nature of the selection isdependen t onth eweather .T opu t it in physiological terms: some years, readily devernalizingplant s will remain inth epopulation ; inothe ryears ,plant swil lb e retained thathav e a high cold requirement. Itwoul d be interesting to findou twhethe r such differences exist andwhethe rmor e specific selection ispossible . Which componentso fth e floweringproces s arerelevan t forbreeder swhe n selecting fora boltin g resistantcultivar ? Iti spossibl e that differences inboltin g resistance ofcultivar s isdu et odifference sin :

juvenility cold requirement maximum temperature forvernalizatio n orminimu m temperature fordever - nalization sensitivity tolon gday s sensitivity todevernalization .

Ifpossible ,whic hcomponen t shouldb echose nb ybreeder s toachiev eboltin g resistance? My trials indicated thatboltin g resistance isdu et o several components,becaus eo fth eobserve d interactionbetwee nth esevera l factors involved. However, itma y stillb epossibl e andusefu l ina selectio nprogram mt o emphasize a single component of the bolting process. If so,photophas e should be a selecting tool. A smaller sensitivity to longphotophas ewil l have a direct effect,becaus e the synthesis of flowering substanceswil l slowdow n and the finalproportio n ofbolter swil lb e reduced. Itwil l also have an indirect effect: a lowered synthesis ofF (interm s ofth emodel ) wouldpostpon e reaching thethreshol d levelo f flowering toa late rdat ei n

97 the season.Thi s delay,wil l givedevernalizin g temperatures longer theop ­ portunity to reduce thepotentia l number ofbolters .Therefore , especially photophase shouldb e atoo l toreduc e thetendenc y forboltin g to acceptable levels inplan tpopulations . Todetec t specific genotypic differences,breeder swil l have tod othei r selectionsunde r conditioned circumstances,whic hwil l make them lessdepen ­ dent on the weather conditions.Moreover ,th ebreede r will thenb e ablet o choose climatic conditionswhic h result intooptimu m adaptation to the cli­ mate ofa give nbeet-growin g area. To select forboltin g resistance, long inductionb y lowtemperatur e gives opportunity todiscar d easilyboltin g plants.Th ebreede r does notnecessa ­ rily need to select for strictly non-bolting plants.Whe nth e lowtempera ­ ture induction is so strong that all plants bolt indu e time,probabl ya very good selection will be made by selecting the late floweringplants . Selecting forth e latest 10% of floweringplant swoul d probably faster lead to a bolting-resistant population than discarding ofth eearlies t 10% of floweringplants . If one attempts to select for fast devernalization, periods ofhighe r temperature shouldb e applied to follow the lowtemperatur e period. Insuc h a way also, plants insensitive to devernalizing temperatures (ifany )ca n beremoved . Longden & Scott (1979)mentione d also that seed-testing institutes have reasont ob e interested incomparin g cultivars under conditioned circumstan­ ces. But also with this technique oftesting ,th eproble m ariseswha tkin d of 'climate' should be applied.Whe n specific differences inseparat e com­ ponents of the bolting resistance exist, one must expect thata specifi c testing climate overestimates the resistance ofon evariet y andunderesti ­ mates that of another. Temperature should then be similar to thato fth e beet-growing area. Moreover the assessment of bolting resistance isvali d only for thatarea . Although arelatio n certainly existsbetwee n growth andboltin g (Chapter 6), there isno tenoug hconvincin g evidence to statetha t selection towards bolting resistance will have anegativ e influence ongrowt h rate.Suc hdif ­ ferences in growth ratebetwee n selected andunselecte d genotypes mayonl y be detectable when the plants have been submitted to acol dperiod . This couldb e one ofth ereason s why no suchdifference s were found inth etria l described inSectio n 6.4. Sowingearly ,however , introduces theproble m that comparisons areobscure dbecaus e ofth edifference s inbolting . Measurements should be made then before the plants startbolting . If,however , further trials prove such arelation ,breeder s shouldb e reluctant todiscar d bol­ tingplant s thatotherwis e growvigorously . Boltingplant s that show asmal ­ ler than average plantweigh t should alwaysb e removed from theplan tpopula ­ tion,bu t laterboltin g plantshavin g adistinc t lead couldb e of advantage inmaintainin g productivity.

98 Besides theus eo fbolting-resistan t cultivars,boltin g inearl y sowings also could be reduced ifchemical swer e available to inhibitbolting .Cau ­ tion should be takentha tsuc hsubstance s dono treduc e growth.Becaus e of the relationbetwee n growth andbolting , ascreenin g ofchemical smigh tsho w up sometha t inhibitboltin gbu tb ymean s ofreduce d growth. Inth elitera ­ ture, bolting-inhibiting chemicals like maleic hydrazide also inhibited growth (Lasa& Silvan, 1976).

99 Summary

Bolting of sugar-beet (Beta vulgaris L.) was studied in a number of trials. In general, thebee tplan t is fairlywel lprotecte d fromprematur e bolting: at first,th eplan tha st ob evernalize db ymean s ofa perio d with relative low temperature and further theplant s have along-da y requirement (the longer the day after chilling, the soonerplant s startboltin g andth e higher theproportio n ofboltin g plants). Thepurpos e ofth e studywa s toclarif y the quantitative effects oftem ­ perature and daylength onboltin g ofsugar-bee t init s firstgrowin g season. A relative simple quantitative and dynamic model was developed based on trials ingrowt h rooms.Th emode l assumed ahypothetica l substanceV , which accumulated duringvernalizatio n and afina l substance F,whos e rate ofsyn ­ thesis was influenced by amounto fV , temperature anddaylength .Th e final amounto fF determine s when andho wman y ofth eplant swoul dbecom egenera ­ tive.

Someaspect s ofth eboltin gproces s intrial swer e as follows. Ingrowt h rooms and inth e field,plant s inth ever y early stages couldb e vernalized somewhat less efficiently, butn o true juvenile stagewa s detected incon ­ trast to published observations. Therefore, the course of temperature in the Netherlands after sowing in most years suggests that flower induction by low temperature occursmainl y beforeemergence . Ina tria lwit h several temperatures duringchilling ,th e lowesttempera ­ ture in the trial (3 °C)wa s also most effective. The true vernalization process probably proceeds fast attha ttemperature , although theprocesse s followingvernalizatio nwil lperhap s startearlie rwit h ahighe r temperature. Vernalization proceeded also attemperature s ashig h as 15 °C;whe n daylength isthe n extremely long,al lplant swil l finallybolt .Ther ewer e no reasons to assume an interaction between degree ofboltin g resistance and optimum vernalization temperature. The duration of cold and daylength after chilling strongly influenced finalproportio n ofbolters .A certain exchangebetwee n cold period andday - lengthbecam e apparent: longer coldperio d and shorter daylength could give the same proportion of bolting plants as shorter chilling and longerday - lengthafterwards . High temperatures (25 °C) after chilling reduced bolting considerably (devernalization), especially immediately after chilling.Ye t an important interaction with daylength became visible because high temperature could

100 still devernalizelate r afterchilling ,provide d that inth eperio d between chilling and high temperature a shorter daylength were applied. Sother e seemst ob en o fixation ofth evernalize d conditionunde rmoderat etempera ­ tures incontras tt oreport s forothe rplan tspecies . A reasonable interpretation of the model istha tth ehypothetica l sub­ stance V, accumulating during chilling, represents acertai nconditio no f the plant: the intensity of deblocking offlower-hormone-formin ggene si n leaves, originating from avernalize d growingpoint .Thi s intensity ofde ­ blocking together with thedaylengt h determinewhethe r andho wmuc h ofth e final floweringhormon e istranslocate d toth egrowin gpoint ,wher ediffer ­ entiation inste m and flowerbud s canfollow . Another aspect investigated waspossibl ephysiologica l and geneticcou ­ pling between growth and bolting. For otherplan tspecie s (e.g. spinach), fast-growing cultivars tendt obol tearlie r thanslow-growin g cultivars. If sucha relatio n applies also forsugar-beet ,breeder shav e aseriou sdiffi ­ culty in creating higher-yielding cultivars: selecting forboltin gresis ­ tance would mean loss ofproductivit y and selecting forproductivit y would mean anincreas e insusceptibilit y tobolting . Incipientlyboltin gplant s indeed had alarge rplan tweigh ttha n (still) vegetative ones. Further,growth-stimulatin g factors (likenitroge nfertil ­ izer and irrigation) increased bolting. So the larger weight of bolting plants could be caused by locally better conditions of growth e.g. more space andN o fsom eplant s inth epopulatio n ofa field.Wit h sucha nexpla ­ nation, it is notnecessar y toassum e genetic couplingbetwee n fastgrowt h andbolting .Althoug h chilling and longday s assuc hals o stimulated growth in a trial in a conditioned room, theconclusio nwa s drawntha tthes eef ­ fectswer e almostindependen to fth e flower(-inducing)proces s andwer epho - tomorphogenetic effects.Als o adecrease d productivity ofa numbe r ofgeno ­ types after selection forboltin g resistance didno tsho wu p inthre e field trials.

From thepoin to fvie wo fbreeders ,th e suggestionwa sgive nt o letday - length play a role in the selectionprocedure s forboltin g resistance. If cultivars arepossibl e thatreac t slowlyt odaylength , atwofol d aim canb e reached.A slowerdaylengt h reactionwil l directlypreven t someplant s from reaching thethreshol d forboltin gbu twil l alsoexten d theperio d inwhic h devernalizing temperatures can be active,s ostrongl y reducing thepropor ­ tiono fboltin gplants . A lastaspec to fth e studywa s toaccoun t forth evariatio n inproportio n ofboltin gplant swithi n andbetwee nyear s (1966-1979),base d onth ecours e of temperature after sowing.A regression approachshowe d that temperature immediately after sowing already influencesbolting , inagreemen twit hth e trials inwhic hn otru e juvenilephas ewa s found.Maximu m daily temperature was aparticularl y goodpredictiv evariable .Fo rth ecultivar s used (Monohil

101 indPolykuhn) , inductionb y cold couldb e fairlywel l estimatedb y counting thenumbe r ofday swit h amaximu m temperature up to1 2 °C.Wit h thisnumber , 60% (Monohil)t o 80% (Polykuhn)o fth evariatio n inboltin g couldb eac ­ counted for.Whe ntemperatur e above2 0 °Cwa s alsotake nint o consideration from sowingt o1 5Jul y (Monohil)an d from sowingt o 15Ma y (Polykuhn), this percentage was raised to 80% (Monohil) and 95% (Polykuhn). Ina secon d method, based on anoptimizatio nprocedure , about 76% ofth evariatio n in bolting for Monohil couldb e accounted for.Wit h thismethod , the complete course of temperature from sowing tomi d Junewa s included inth ecalcula ­ tions.

102 Samenvatting

Hoewel de suikerbiet (Beta vulgaris L.)to td etweejarig eplante ngere ­ kend wordt, kunnen inbepaald e jarenbi jd eteel tvee l schietersoptreden , planten die reeds inhe teerst e groeiseizoen generatiefworden .D eplan ti s tamelijk goedbescherm d tegenvoortijdi g schieten.Voorwaard evoo r schieten is een gevernalizeerde toestand, dieto tstan dkom ti nee nperiod eme tbe ­ trekkelijk lage temperatuur, vervolgens moet nog aand ezogenaamd e lange- dagbehoefte worden voldaan. Hoegrote rd edaglengt e isn ad ekoudeperiode , hoe sneller en hoe meer planten er schieten.Normaa l gesprokenza l alleen na overwintering van deplante n aanbeid evoorwaarde nvoldaa nworden .Toc h kan ook na vroege zaai een aanzienlijk aantalplante n inbloe i komen.Di t kanto tmoeilijkhede nbi jd eoogs te nvolgteelte n leiden. Deze studie beoogt vooral de kwantitatieve effectenva ntemperatuu r en daglengte ophe t schietenva nd eplante n (speciaal inhe teerst egroei-jaar ) duidelijker temaken . Nadathe tmerendee l vand eexperimente nwa suitgevoerd ,wer d eenkwanti ­ tatiefrelatiemode l ontwikkeld watd egevonde n resultatenko nverklaren . In het model werd voorlopig aangenomen,da te ree nhypothetisch e substantieV onder invloed van lagetemperatuu r gevormdword ttijden she tvernalizatie - proces. Inhe tmode lwer dverde r aangenomenda td esynthesesnelhei d vanhe t uiteindelijkebloeihormoo n Fpositie fbeïnvloe d wordtdoo rd eaanwezig e hoe­ veelheid V, de temperatuur en de daglengte.D e hoeveelheid Fdi e gevormd wordt,bepaal tof ,e nhoeveel ,plante nuiteindelij k generatiefworden . Inklimaatkamer s eni nhe tvel dbleek ,da tplante n ietsminde r effectief in jonge stadia gevernalizeerd konden worden. Uit deze proeven bleekdu s nietda td ebietenplante n ind e jonge stadiaongevoeli g zijnvoo r deverna - lizerendewerkin gva n lageretemperature n (juveniele fase), zoalsda ti nd e literatuurword tvermeld .Verwach tma gworde n datonde rNederlands eomstan ­ digheden, gezien ook het temperatuurverloop in demeest e jaren,d ebloei - inductie door deze lagetemperature nvoora l ind etij dvoo ropkoms tplaat s zalvinden . Inee nproe fme tverschillend e temperaturen tijdens dekoudebehandeling , bleek de laagste temperatuur (3°C )he tmees teffectief .D eindru kwer dver ­ kregenda tweliswaa r hetvernalizatieproce sbi jdez e lagetemperatuu r opti­ maal plaats vindt, maar datbi jvernalizati ebi j ietshoger e temperaturen, processendi egewoonlij k opd evernalizati evolge n (daglengtereactie)reed s op gang kunnen komen,wa tuiteindelij k positiefka nwerken .He tvernaliza ­ tieprocesbleek ,indie ntegelijkertij d eenextree m langedaglengt eheerste ,

103 zich zelfs af te kunnen spelen bij temperaturen van 15 °C.E rwerde n geen aanwijzingen verkregen, datrasse n dieverschille n in schietergevoeligheid, interacties vertonenvoo rwa tbetref t deoptimal evernalizatietemperatuur . De lengteva n dekoudebehandelin g end edaglengt e na de koudebehandeling, blekenbeid ehe tpercentag e schieters sterk tebeïnvloeden . Eenzeker e uit­ wisselbaarheid tussen koudeperiodee ndaglengt e kwamnaa rvoren :ee n langere koudeperiode enkorter e daglengte konee n zelfdepercentag e schieters geven als eenkorter e koudeperiode enee n langeredaglengte . Hogere temperaturen (25 °C)n a dekoudeperiod e beperktenhe tpercentag e schieters aanzienlijk (de zgn.devernalizatie) , vooral vlakn a dekoudepe ­ riode. Toch bleek een belangrijke interactie met de daglengte, omdatoo k geruime tijd na dekoud eperiod ed ehog etemperature n nogeffectie f konden zijn,mit s ind eperiod e tussen lage enhog e temperatuur eenkort e daglengte werd toegepast.Bi jbiete n lijkte rdu sgee n sprake tezij nva nee n fixatie van degevernalizeerd e toestand onder gematigde temperaturen, zoalsva nan ­ dere gewassen gemeldis . Bijherinterpretati e vanhe tmode lwer dhe t aannemelijkergeacht ,da td e hypothetische stofV , die zicho p zouhope ntijden s hetvernalizatieproces , meer eenbepaald e toestand representeertva nd eplant ;ee n toestand diedoo r middel van celdeling kanblijve nbestaan .Omda t de lage temperatuurspeci ­ fiek op het groeipuntwerk te nhe tverde r aannemelijk is,da tbladere n die uitee ngevernalizeer d groeipunt komenverantwoordelij k zijnvoo r desynthe ­ se van het bloeihormoon, werd verondersteld datdez e toestand demat eva n deblokkering van de genen die het bloeihormoon vormen, representeert.Of , en inwelk ehoeveelheid , bladeren uitee n gevernalizeerd groeipunt inderdaad bloeihormonen exporteren, hangt dan afva n dedaglengte . Indienhe tbloei ­ hormoon inhe tgroeipun t aankomt,volg tdifferentiati e instenge l enbloem ­ knoppen.

Een ander onderzocht aspect was de mogelijke genetische fysiologische koppeling van groei en schieten.Bi j andere gewassen (bv.spinazie )hebbe n snelgroeiende rassen een sterkere neiging totschiete n dantraa g groeiende rassen. Indien ditverban d ookbi jbiete nbestaat ,zo udi tvoo r kwekers een ernstige handicap betekenen omto tproduktiever e rassen te komen: selectie opprodukti e zou leiden totschietergevoelighei d en selectie opschieterre ­ sistentie totverminderd e produktiviteit. Uitveldwaarneminge n bleek inderdaad dat juistschietend e planten veelal aanzienlijk zwaarder zijn dan (nog)vegetatiev e planten.Verde r bleek echter datoo kgroeibevorderend e maatregelen zoals stikstof (over)bemesting enbe ­ regening tot meer schieters leidde.He tovergewich tva n schietende planten zou dus veroorzaakt kunnen zijndoo r toevallig betere groei-omstandigheden van deze planten (bv.mee r ruimte,stikstof , enz.). Zowel degroe i alshe t schietenword tda nbevorderd .

104 In deze verklaring is het niet nodig een genetische koppelingva n snelle groei enschietneigin g aant enemen .We lbleek ,da tee nkoudeperiod e enlan ­ gedage no pzic hgroeibevorderen d kunnenwerken ,maa r opgron dva nd eproe ­ venwer d geconcludeerd, datdi teffec tmi no fmee r losstaa tva nhe tproce s van (bloei) inductie enmee r temake nheef tme t fotomorfogenetischeeffec ­ ten. Ook een verminderde produktiviteitva n eenaanta l genotypengeselec ­ teerd op schieterresistentie, kon in een drietalproeve nnie t significant aangetoondworden . Vanuithe tgezichtspun tva nd eplantenveredelaar ,dien toverwoge nt ewor ­ deno fd edaglengt e nietee nro l kanspele ni nd e selectieo pschieterresis ­ tentie. Indien een genotypemogelij k isda ttraa g op langedage nreageert , kanee ntweeledi g doelbereik tworden .Doo r dezetrag e daglengtereactie zul­ lenminde rplante nd edrempelwaard evoo r schietenoverschrijden , anderzijds zald eperiod evoo rhe tbereike nva ndez edrempelwaard e langerworden ,zoda t devernalizerende temperaturen een grotere kans krijgen om het potentiële aantal schietersteru gt edringen . Het laatste aspect indez e studiewa s om devariati e inschieterpercenta ­ ges tussen en binnend e jaren 1966to te nme t 1979t everklare nme tbehul p van het temperatuurverloop na zaai. Met behulp van regressieberekeningen bleek, dat de temperatuur onmiddelijk na zaai al invloed heeft,hetgee n in overeenstemming isme t de inproeve ngevonde n afwezigheid vand e juveniele fase. Speciaal de dagelijkse maximumtemperatuur kon goed als verklarende variabele gebruikt worden.Voo r detwe egebruikt e rassen (Monohil enPoly - kuhn)ko n het effect van koude-inductie goedworde nbenader d doorn azaa i het aantal dagen te tellen met een maximumtemperatuur kleiner dan 12°C . Hiermee kon6 0% (Monohil)to t8 0% (Polykuhn)va nd evariati e inschieter ­ percentages verklaardworden . Alshe taanta l dagenme tee nmaximumtemperatuu r vanmee r dan2 0 °Ci nd e periode van zaai tot bv. 15jul i (Monohil)o fto t1 5me i (Polykuhn)i nd e berekeningen werd opgenomen, kon het percentage verhoogd worden tot 80 % (Monohil)e n9 5% (Polykuhn). Eenander emethode ,m.b.v .ee noptimaliseringsprocedur e leverdevoo rMo ­ nohil eenpercentag e vanca .7 6% op .Hierbi jwerde n dagelijkse temperatuur- waarnemingengebruik tove rd egehel eperiod eva nzaa i tothal f juli.

105 References

Bachmann, L., P. Curth & H.J. Rö'stel, 1963.Untersuchunge n über Gibberellin-Verabreichung in verschiedenen Wachstumsphasen der Beta-Rüben. Züchter 33 (1):50-58 . Barendse, G.W.M., 1964.Vernalizatio n in Cheiranthus allionii Hort. Meded. Landbouwhogesch., Wageningen, 64-14. Behaeghe, T.J., 1975.Eco-fysiologisch e grondslagen van de grasgroei. Agricultura 23(4) : 19-46. Behaeghe, T.J., 1978.Th e seasonal discrepancies between potential and actual grass growth. An essay of explanation by eco-physiological constraints. Repr. 7th Gen. Meeting Eur. Grassl. Fed., Gent. Bernier, G., 1976. La nature complexe du stimulus floral et des facteurs de floraison. In: R. Jacques (Ed.): Etudes de Biol, végétale. Hommage au prof. Chouard. Paris. Besnard-Wibaut, C, 1977.Histoautoradiographi c analysis of the thermoinductive processes in the shoot apex of (swedish variety)o f Arabidopsis thaliana L. Heynh, vernalized at different stages of development. Plant Cell Physiol. 18 (5):949-962 . Bhargava, S.C., 1964. Photoperiodism, floral induction and floral inhibition in Salvia occidentalis. Meded. Landbouwhogesch., Wageningen, 64-12: 1-70. Bliss, C.I., 1973.Plan t Protection no. 12,Leningrad . Bornscheuer, E., 1972. Influence of different production regions in sugar-beet seed. 35th I.I.R.B. Winter Congress 10/11 Febr. 1972 session II.Repor t no. 2.1. Bosemark, N.O., 1970. Influence of seed crop environment on root crop characteristics. J.I.I.R.B. (4):193-206 . Chouard, P., 1960. Vernalization and its relation to dormancy. Annu. Rev. Plant Physiol. 11: 191-238. Chroboczek, E., 1934.A study of some ecological factors influencing seed stalk development inbeet s (Beta vulgaris, L.) Mem. Cornell Univ. Agric. Exp. Stat. 154,8 4p . Curth, P., 1955.Temperatu r und Licht als blühinduzierende Faktoren bei der Zuckerrübe I. Zuchter 25: 14-17. Curth, P., I960. Der Übergang in die reproduktieve Phase bei der Zuckerrübe in Abhängigkeit von verschiedenen Umweltfaktoren. Wiss. Abh. Dtsch. Akad. Landwirtschaftswissensch. Berlin, nr. 46. (Beitr. Z. Rübenforsch, nr.4) . Curth, P., 1962. Spezielle Untersuchungen zu Vernalization und Photoperiodismus der Zucker­ rübe. Z. Pflanzenzucht. 47:254-276 . Curth, P., 1963. Über die photoperiodische SummationsWirkung intermittierend gebotener nächtlicher Zusatzlichtes unterschiedlicher Periodenlänge bei Zuckerrüben. Biol. Zen- tralbl. 82:485-487 . Curth, P. & K. Fürste, I960. Eine neue Prüfungssmethode für Schoszresistentz bei Zucker­ rüben. Wiss.Abh . Dtsch. Akad. Landwirtschaftswiss., 48: 29-37. Curtis, G.J. & K.G. Hornsey, 1964. Graft-transmissible induction of elongation and flowering in scions of sugar-beet bred for resistance to bolting. Nature 202 (4938): 1238. Curtis, G.J., 1964.Graf t transmission of the flowering stimulus from awil d Beta species to a line of beet selected for resistance to bolting. Nature 203:201-202 . Curtis, O.F. & H.T. Chang, 1930.Th e relative effectiveness of the temperature of the crown as contrasted with that of the rest of the plant upon flowering of celery plants. Am. J. Bot. 17:1047-1048 . Davies, A., 1971. Growth rates and crop morphology invernalize d and non-vernalized swards of perennial ryegrass in spring. J. Agric. Sei. (London) 77(2): 273-282. Devay, M., E . Paldi & I. Kovacs, 1976. Thermolabile and thermostable interphases in the vernalization of winter wheat. (Triticum aestivum var. bankuti 1201). Acta Bot. Acad. Sei. Hung. Tomus 22(1-2): 9-16. Dowker, B.D. & J.C. Jackson, 1975. Bolting in some carrot populations. Ann. Appl. Biol. 79: 361-365. Fife, J.M. & C. Price, 1953. Bolting and flowering of sugar-beet in continuous darkness. Plant. Physiol. 28:475-480 . Forrester, J.W., 1961. Industrial dynamics. MIT-press, Boston.

106 Gaastra,P. , 1959.Photosynthesi s of cropplant sa sinfluence db ylight ,carbo ndioxide , temperature,an dstomata ldiffusio nresistance .Meded .Landbouwhogesch. ,Wageningen ,5 9 (13):1-68 . Gaskill,J.0. ,1952 .Inductio no freproductiv edevelopmen ti nsugar-bee tseedling sb ypho ­ tothermaltreatmen to fyoun gseedlings .Proc .Am .Soc .Sug .Bee tTechn .7 :112-120 . Gaskill,J.0. ,1957 .A preliminar y reporto nth eus eo fgibberellin eaci dt ohaste nrepro ­ ductivedevelopmen ti nsugar-bee tseedlings .J .Am .Soc .Sug .Bee tTechn .IX(6) :521-528 . Gaskill,J.0. ,1963 .Influenc eo fag ean dsupplementa lligh to nflowerin go fphotothermally - induced sugar-beet seedlings.J.Am . Soc.Sug .Bee tTechn .1 2(6) : 530-537. Gorodnii,N.G . 5.E.P . Sereda, 1975.Boltin gi ndifferen tform so fsugar-bee ti nrelatio n toclimati ccondition san dnutritio nlevel .Nauchn .Tr .Ukr .Skh .Akad .146 :65-6 9 [Rum]. Gregory.F.G .& O.N .Purvis ,1938a .Studie s invernalizatio no fcereals .II .Th evernaliza ­ tiono fexcise dmatur eembryos ,an do fdevelopin gears .Ann .Bot .N.S .2 :237-251 . Gregory,F.G . &O.N .Purvis ,1938b .Studie si nvernalizatio no fcereals .III .Th eus eo f anearobiccondition si nth eanalysi so fth evernalizin geffec to flo wtemperatur edurin g germination.Ann .Bot .N.S .2 :753-764 . Hanus,H .& 0 .Aimiller ,1978 .Ertragsvorhersag e ausWitterungsdaten .Fortschr .Acker -un d Pflanzenbau5 . Heide,O.M. ,1973 .Environmenta lcontro lo fboltin gan dflowerin gi nre dgarde nbeets .Meld . Nor.Landbrukshoegsk .52 :(15 )17p . Highkin,H.R. , 1956.Vernalizatio ni npeas .Plan tPhysiol .31 :399-403 . Hoekstra,0. , 1960.D ewerkin gva nnitraat -e nammoniumstiksto fo paardappele ne nsuiker ­ bieten.Landbouwk .Tijdschr .72(10) :403-410 . Hunt,R . & I.T.Parsons ,1974 .A compute rprogra m forderivin ggrowth-function s inplan t growth-analyses.J .Appl .Ecol .11 :297-307 . Jones,H.G. ,M.A .For d &R .Plumley ,1975 .Th eeffec to fvernalizatio no nphotosynthesi s inwheat .Photosynthetic a 9(1) :24-30 . Jorritsma,J. , 1978.Kwalitei tva nsuikerbieten .Uitgav e Instituutvoo rRationel eSuiker - produktie,Berge no pZoom . Junges,W. , 1959.Abhängigkei t des Schossens bienner Gemüsepflanzenvo nihre mAlte run d vonde rDaue rde rEinwirkun gniedrige rTemperaturen .Arch .Gartenb .7 :485-504 . Kilsdonk,A.C.M. , 1977.Non-linea r optimisation package0PTPAC3 .Progra mdescriptio nan d referencemanual .ISA-DS AScientifi cCompute rApplications . Kinet,J.M. ,M .Bodson ,A .Jacqmar d& G .Bernier ,1975 .Th einhibitio no fflowerin gb yab - scisicaci di nSinapi salb aL .Z .Pflanzenphysiol .77 :70-74 . King,R.W .& J.A.D .Zeevaart ,1973 .Flora lstimulu smovemen ti nPerill aan dflowe rinhibi ­ tioncause db ynoninduce d leaves.Plan tPhysiol .51 :727-738 . Kloen,D. , 1952.Jarowisati eva nbietenzaad .Zaadbelange n6 (15) : 177-178. Lane,H.C. ,H.M . Cathey & L.T.Evans ,1965 .Th edependanc eo fflowerin gi nsevera llon g day plants onth espectra lcompositio no fligh textendin gth ephotoperiod .Am .J .Bot . 52 (10):1006-1014 . Lang,A- ,1952 .Physiolog yo fflowering .Annu .Rev .Plan tPhysiol .3 :265-306 . Lang,A. ,1965 .Physiolog yo fflowe rinitiation :In :W .Ruhlan d (Ed.):Handb .de rPflanzen ­ physiol.Bd .XV/1 .p .1300-1536 .Berlin -Heidelberg -Ne wYork . Lang,A . & G.Melchers ,1947 .Vernalizatio n und Devernalization bei einerzweijährige n Pflanze.Z .Naturforsch .B2 :444-449 . Lasa,J.M. , 1977.Sugar-bee tboltin gi nth eroo tcrop .J .Agric .Se i8 9 (1): 223-228. Lasa,J.M .& A .Silvan ,1976 .Factore squ einfluye ne ne lespigad od el aremolach aazucare - ra,Bet avulgari sL .Fundacio nJua nMarch .Convocatori sEspan a -1.974.Ciencia sagrarias . Lenton,J.R. , T.0.Podoc k &M.E .Radley ,1975 .Roth .Exp .Station .Repor tfo r1975 .Par t I,p .45 . Lexander,K. ,1969 .Increas ei nboltin ga sa neffec to flo wtemperatur eo nunrip esugar-bee t seed.32n dWinte rCongres so fth eI.I.R.B. ,Pape rno .1.8 . Longden,P.C .& R.K .Scott ,1979 .Effect so fsowin gdat eo nsugar-bee tseedlin gemergence , sugarproduction ,boltin g and subsequent seedproduction .Int .not eBroom' sBar nExp . Stat. Longden,P.C. ,R.K .Scot t& J.B .Tylderley ,1975 .Boltin go fsugar-bee tgrow ni nEngland . OutlookAgric .8(4) :188-193 . Lüdecke,H. , 1938.Einflus s derErnährun g auf dasSchosse nde rZuckerrübe .Dtsch .Land - wirtsch.Press e65 :115-116 . Lysgaard,C.P .& S.V .N0rgaar dH0lm ,1962 .Th eeffec to fboltin go nth equantit yan dqualit y ofswed ean d foddersugar-beet .Arsskr .K .Vet.-Landbohoejsk. ,Copenhagen ,p .94-123 . Lysgaard,C.P. ,1978 .Selectio nfo rreduce dboltin gsusceptibilit yi nbee tan dswedes ,an d the influence of environmental factors onbolting .Arsskr .K .Vet.-Landbohoejsk. ,p . 138-158.Par to fa collectiv edocument .

107 Mann, H.H., 1951.Th e effect of manures on the bolting of the beet plant. Ann. Appl. Biol. 38: 435-443. Margara, J., 1960. Recherche sur le determinisme de l'élongation et de la floraison dans le genre Beta. Ann. Inst. Natl.Rech . Agron., Serie B. Ann. Amélior. Plantes 10:361-471 . Margara, J., 1967.Recherche s sur les gibberellins et le developement floral chez labet ­ terave. J.I.I.R.B. 2 (4):242-254 . Margara, J., 1968. Physiologie de l'induction florale chez la betterave. J.I.I.R.B. 3(1) : 17-28. Melchers, G., 1939.Di e Blühhormone. Ber. Dtsch. Bot. Ges. 37: 29-48. Milford, G.F.J. & J.R. Lenton, 1976. Effect of photoperiod on growth of sugar-beet. Ann. Bot. 40: 1309-1315. Mugnier, C, 1977.A n attempt to replace vernalization by application of gibberellic acid inbienna l Hyoscyamus niger L. Biol. Plant. 19 (1):40-47 . Napp Zinn,K. , 1957.Untersuchunge n über das Vernalizationsverhalten einer winterannuellen Rasse von Arabidopsis thaliana. Planta 50:177-210 . Napp Zinn. K., 1961. In: W. Ruhland (Ed):Handb . Der Pflanzenphysiologie Bd. XVI (1961). Springer Verlag, Berlin, p.24-75 . Napp Zinn, K., 1973. Low temperature effect on flower formation: Vernalization. In:H . Precht (Ed):Temperatur e and life. Springer Verlag, Berlin, p. 175-194. Naylor, A.W., 1961. The photoperiodic control of plant behaviour. In: W. Ruhland (Ed): Handb. der Pflanzenphys. B. 16.p . 331-389. Nicholls, A.0.& D.M. Calder, 1973. Comments on the use of regression analysis for the study of plant growth. New Phytol. 72:571-581 . O'Connor, L.J., 1970. Environmental influence during beet seed production in bolting and quality characteristics of the subsequent root crop. J. Int. Inst. Sugar Beet Res. 4 (4): 207-216. Oorschot, J.L.P., 1960.Effect s of daylength upon growth and development of spinach (Spi- nacea oleracea L.) Meded. Landbouwhogesch., Wageningen, 60 (18):1-10 . Parlevliet, J.E., 1966. The influence of external factors on the growth and development of spinach cultivars (Spinacia oleracea L.).Meded . Landbouwhogesch., Wageningen,67-2 . Purvis, O.N., 1944. Studies invernalizatio n of cereals XIII. The role of carbohydrate and nitrogen supply in the vernalization of excised embryo's of Petkus winter rye. Ann. Bot. 8: 285-314. Purvis, O.N. & F.G. Gregory, 1952. Studies in vernalization XII. The reversibility by high temperature of the vernalized condition in Petkus winter rye. Ann. Bot. 16: 1-21. Purvis, O.N., 1961. In:W . Ruhland (Ed):Handb . der Pflanzenphysiol, 16. Springer, Berlin- Gottingen- Heidelberg, p. 76-122. Röstel, H.J. ,1968. Probleme und Ergebnisse der Züchtung auf Schossresistenz bei monokarpen Zuckerrüben in Kleinwanzleben. Bull. Res. Inst.Agric . Plant Breed. Plant Prod., Sopron- horpacs, nr. 4,p . 3-17. Salisbury, F.B., 1963. The flowering process. Pergamon Press,Oxford , London, New York, Paris. Schäfer,E. , 1976.Th e high-irradiance reaction. In: H. Smith (Ed):Ligh t and plant develop­ ment, p.45-59 . Schneider, G., 1960. Biochemische und entwicklungsphysiologische Untersuchung zur Frage des Schossens und der Blütenbildung der Zuckerrübe. I. Planta 55:669-685 . Scott, R.K. & K.W. Jaggard, 1978. Theoretical for maximum yield. Proc. 41st Winter Congress of the I.I.R.B. Scott, R.K. & P.C. Longden, 1970.Polle n release by diploid and tetraploid sugar-beet plants. Ann. appl.Biol . 66: 129. Shiomi, N. & S. Hori, 1973. Changes in some enzyme activities and the nicotinamide nucleotide coenzymes in barley seedlings during vernalization. Soil Sei. Plant Nutr. (Tokyo) 19 (2): 132-138. Smith, H., 1975. In: Phytochrome and photomorphogenesis. McGraw Hill,Ne w York. SPSS-manual, 1975. Statistical package for the social sciences. 2nd Edition by N.H. Nie et al., McGraw-Hill, New York. Steyvoort, L. van & R. van Stallen, 1973. Het probleem van de wilde bieten. Publikatie HI/1973. Belg. Inst, tot verbetering van de biet. Tienen. Stoddart, J.L., S.M. Tapster & T.W.A. Jones, 1978.Temperatur e dependance of the gibberellin response in lettuce hypocotyls. Planta 141 (3):283-288 . Stout, M., 1946. Relation of temperature to reproduction in sugar-beet. J. Agric. Res.72 : 49. Suge, H., 1970. Changes of endogenous gibberelling in vernalized radish plants. Plant Cell Physiol. 11 (5):729-735 .

10? Tateyama,M. ,H.A . Ishikawa &K . Ishikawa, 1978.Nuclei caci dcontent san dpolyribosom e formationi nwhea tembryo sdurin ggerminatio nan dvernalization .Plan tCel lPhysiol .1 9 (3):411-418 . Voss,J. , 1936.Experimentell e Auslösung desSchossen sun dPrüfun gde rSchossneigun gde r Rübensorten (Betavulg .L. )Angew .Bot .XVIII :370-408 . Voss,J. , 1940.Zu r Schossauslösung und Prüfung derSchossneigun gvo nRübensorte n (Beta vulg. L.) und Brassica napus L.var .Napobrassicc a [L.]Reichenb .Zuchte r12 :34-44 , 73-77. Warne,L.G.G. , 1949.Spacin gan dboltin go fvegetables .Natur e 164:969-70 . Wellensiek,S.J. ; 1964a.Dividin gcell sa sth eprerequisit efo rvernalization .Plan tPhysiol . 39:832-835 . Wellensiek,S.J. , 1964b.Roo t vernalization of Cichorium intybusL .Repr .K .Ned .Akad . Wetensch.C6 7(5) : 307-312. Wellensiek,S.J. , 1977.Principle so fflowe rformation .Act aHortic .68 :17-27 . Wellensiek,S.J. , 1979.Th emechanis m ofvernalizatio ni nSilen earmeri aL .Z .Pflanzen - physiol.9 4(3) : 263-265. Wellensiek,S.J .& K .Verkerk ,1950 .Vernalizati ee nphotoperiodicitei tva nbieten .Meded . Dir.Tuinbou w1 3(6) : 341-357. Wiebosch, W.A., 1965.Jarowisati e bij enige groente- en aanverwante gewassen.Thesis , Wageningen. Withrow,A.P. ,R.B .Withro w& J.P .Biebel ,1943 .Inhibitin g influenceo fth eleave so nth e photoperiodical responseo fNobe lspinach .Plan tPhysiol .18 :294-299 . Wood,D.W . &R.K .Scott ,1975 .Sowin gsugar-bee ti nautum ni nEngland .J .Agric .Sei .84 : 97-108. Yusubov,A.M. , 1977.Boltin gan dth eimprovemen to ftetraploi d sugar-beeti nthi srespect . SB.Nauch .Fr .Vsero sNil .Sakha rSvekl yI Sakhara :7 (2) : 124-134. Zeevaart,J.A.D. , 1971.Effect so fphotoperio d ongrowt hrat ean dendogenou sgibberellin s inth elong-da y rosetteplan tspinach .Plan tPhysiol .47 :821-827 . Zeevaart,J.A.D. , 1976.Physiolog y of flower formation.Annu .Rev . Plant Physiol.27 : 321-348.

109