ContentslistsavailableatScienceDirect
Flora
j o ur na l h o me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / f l o r a
Minor
differences
with
big
consequences:
Reproductive
patterns
in
the
genus
Gagea
(Liliaceae)
Martin
Schnittler
a,∗,
Angela
Peterson
b,
Jens
Peterson
c,
Saltanat
Beisenova
d,
Rakhmetkazhi
I.
Bersimbaev
d,
Tanja
Pfeiffer
aaErnst-Moritz-Arndt-UniversityGreifswald,InstituteofBotanyandLandscapeEcology,Soldmannstr.15,D-17487Greifswald,Germany bMartin-Luther-UniversityHalle-Wittenberg,InstituteofBiology,Weinbergweg10,D-06120Halle/Saale,Germany
cStateOfficeforEnvironmentalProtectionofSaxony-Anhalt,ReideburgerStr.47,D-06116Halle/Saale,Germany dL.N.GumilyovEurasianNationalUniversity,5MunaitpasovStr.,010008Astana,Kazakhstan
a
r
t
i
c
l
e
i
n
f
o
Articlehistory: Received17March2013 Accepted5September2013 Available online 5 October 2013 Keywords: Bulb Bulbils Morphology Seeds Vegetativereproduction Reproductiveswitch
a
b
s
t
r
a
c
t
ReproductivepatternsintenspeciesofGageaSalisb.werecomparedbycountsandmeasurementsof bulbs,bulbilsandflowersinlargecohortsincludingalllifestages.Twotypesofbulbilswerefound:taxa with“typeIbulbils”starttodevelopasingletoseveralbulbilsassoonasthereplacementbulbhasreached acertaindiameterandthencontinuetoformthemindefinitelythroughoutthelifeoftheplant.“TypeII bulbils”areonlytemporarilyproducedinimmature,non-floweringplantsofsomespecies,butnotinfully grown,floweringindividuals,aphenomenontermed“reproductiveswitch”.Patternsofbulbilformation arespecies-specific:G.davlianidzeae,G.nigra,G.peduncularis,G.pratensis,andG.spathaceaproduceonly typeIbulbils;G.angelae,G.fedschenkoanaandG.luteadeveloponlytypeIIbulbils.Bothbulbiltypesoccur simultaneouslyinG.fragiferaandG.villosa.Thequantitativeinvestigationsdemonstratetheexistence ofspecies-specificthresholdsforthedevelopmentofbulbilsaswellasflowers.Comparedtotheadult volumeofthereplacementbulb(where90%ofallplantsflower),bothtypesofbulbilshaveusually lowthresholds:0–5%(typeI,allbutonespecies)and3–13%(typeII).Inflorescencesdevelopifplants attainbetween38and60%oftheadultbulbvolume.Minorchangesinpatternsofbulbilformationand thresholdsfortheirdevelopmentmayensuresurvivalofhighlysteriletaxa(e.g.G.spathacea,G.fragifera). This,inturn,canfacilitatespeciationinthegenusdrivenbybothhybridizationandpolyploidization.
© 2013 Elsevier GmbH. All rights reserved.
Introduction
Like many monocotyledons, thespecies of the genusGagea Salisb. (Liliaceae) showrather strict growthpatterns withonly subtle differences. Thisalready attractedthe attention of early botanists,describingbranchingpatternsandindividual develop-ment(Buxbaum,1937;GraebnerandKirchner,1934;Horn,1874; Irmisch,1850).ThespeciesofGageaaresmallbulbousplantswith cryptodicyclicdevelopment(Levichev,2006)andoftenvegetative multiplicationbybulbils.Withrespecttotheirgeneral morphol-ogy,allspeciesofGageahavetwobasalleaves(Levichev,2006).The firstleafofthepresent-yearshootisafreeassimilatingbasalleaf, itsaxillarybuddevelopsthereplacementbulb,i.e.theparentbulb ofthenextseason.Thisistheminimumsetforveryyoungplants. Thesecondbasalleafcanbereducedtoascaleorisoften concaules-centwiththeinflorescencescape(manyspeciesinsectionsGagea andMinimae;Levichev,2006).Therefore,inmanyspeciesonlya
∗ Correspondingauthor.Tel.:+493834864123;fax:+493834864114. E-mailaddress:martin.schnittler@uni-greifswald.de(M.Schnittler).
singlebasalleafisvisible,butinothers,likeG.spathacea(Hayne) Salisb.,vigorouslygrowingvegetativeplantsdeveloptwofreebasal leaves.Asitiscommoninsect.Didymobulbos(Petersonetal.,2008; Tisonetal.,2013),oldervegetativeandfloweringplantsofG. vil-losa(Bieb.)Sweethavetwobasalleaves.Theaxilofthissecond leafcansupportasingletoseveralbasalbulbils.Ifmorethanone basalbulbilisformed,oneoftheseisoftenlargerthantheothers and/ormayhaveadifferentcolourorshape(Levichev,1999).Even moreadventivebulbils(onetomany)canbeformedintheaxils ofleavesontheinflorescenceshoot.Innon-floweringplants,the lattermaybesituatedabovethereplacementbulbonshorterect shoots(“Bulbillenköpfchen”;Levichev,1999:384;suprabasal bul-bils,compareTisonetal.,2013),whichistheusualcaseinGagea fragifera(Vill.)Ehr.Bayer&G.LópezGonzalez.Infloweringplantsof somespecies,bulbilsaredevelopedintheaxilsofthethirdand fol-lowingupperleavesbelowtheinflorescence(e.g.Gageagageoides (Zucc.)Vved.,AliandLevichev,2007).Detailedschematic draw-ingsofdifferentgrowthpatternsinrepresentativesofGageaare providedbyLevichev(2013:414,Fig.2).
Bulbilsarenotalwayssessile,butmayalsodeveloponstolon(s) ofvariouslengths(comparePlate6inLevichev,1999).Generally,
0367-2530/$–seefrontmatter © 2013 Elsevier GmbH. All rights reserved.
bulbilswillbereleasedwith thedecayof theirsupportingleaf at the latest, but often rupture of its base and bulbil release occursmuchearlier.Bulbilslay oftendormant,and inthiscase bulbilsfrompreviousyearsarerecognizablebyausuallydarker colour.
Inhisstudyonthemorphologyofthegenus,Levichev(1999) foundforallbutthreespeciesofGageavegetativereproduction viabulbils.Theirshape,quantityandlocationprovideimportant morphologicalcharacters for species differentiation,which was illustratedbyLevichev (1999,2013),Petersonet al.(2011) and Tisonetal.(2013).Differentnumbersofbulbilsformedperyearand plant,reachingfromone(G.nigra)toseveraldozen(G.spathacea, Schnittleretal., 2009), indicatethat therelative proportionsof vegetative vs. generativereproduction arelikely tobe species-specificandveryvariable.Extremesmaybeexclusivevegetative reproduction(G.spathacea;Pfeifferetal.,2011,2012a),or exclu-sivegenerativereproduction,whichcanbeexpectedforthefew specieslackingbulbils(compareLevichev,1999;G.robustaZarrei andWilkininZarreietal.,2010).
LevichevandJezniakowsky (2008)stated thatthespecies of Gageafollowratherrigidspecies-specificgrowthpatternsduring ontogenesis.Althoughthesegrowthpatternsarewelldocumented foranumberofspecies(e.g.Levichev,2013),quantitativeanalyses ofvegetativeandgenerativereproductivetraitsarelargelymissing. InvestigationsinG.lutea(L.)KerGawl.andG.spathacea(Schnittler etal.,2009)revealedthatbulbilformationintheformerspecies showsareproductiveswitch:non-floweringplantsformbulbils, floweringonesdonot.Incontrast,G.spathaceadevelopsbulbils inallontogeneticstages.Thisdifferenceiscrucialforthebalance betweensexualandvegetativereproduction(Pfeifferetal.,2011; Schnittleretal.,2009)andgrantssurvivalofthelatter,sexually sterilespecies(Pfeifferetal.,2012a).
Inthepresentstudyweprovideaquantitativeassessmentof thereproductivestrategiesineightGageaspecies:wemeasured diameterandcountednumbersofbulbsandbulbilsofflowering andnon-floweringindividualsinnaturalpopulationsofG.angelae LevichevetSchnittler,G.davlianidzeaeLevichev,G.fedtschenkoana Pascher,G.fragifera,G.nigraL.Z.Shue,G.peduncularis(J.Presl&C. Presl)Pascher,G.pratensis(Pers.)Dumort,andG.villosa.In addi-tion,theresultsofanearlierstudy(Schnittleretal.,2009)onG. luteaandG.spathaceaareconsidered.Ouraimsareto(i)illustrate andquantifyspecies-specificpatternsofbulbilandflower devel-opment,(ii)deriveestimatesfor resource-dependentthresholds forthedevelopmentoftheseorgans,(iii)togeneralizethefindings tocharacterizetherealizedreproductivestrategiesofthestudied taxa.
Materialsandmethods Studyspecies
ForeachoftheeightinvestigatedspeciesofGageaonetoten populationstotallingbetween166and572individualswere mea-sured(seeAppendix1).Todeterminetherespectivethresholdsfor bulbilandflowerformation,wecollectedalllifestages,although extremelylargeplantsarenaturallyrareandseedlingsaredifficult tofindinalltaxa.
Morphologicalanalysis
Forthisstudy,wedistinguishthreetypesofbulbs.Parentbulbs (from the previous year) nourish the leafy present-year shoot, whichgrowsintotheinflorescence,or,injuvenileplants,is rep-resentedsolelybyone,morerarelytwobasalleaves.Replacement bulbsdevelopintheaxilofthefirstbasalleafandwillcarrytheplant
intothenextyear(becomingthentheparentbulb).Inall investi-gatedspeciesexceptforG.pratensisthereplacementbulbdevelops withintheparentbulb,sometimesaccumulatingremnantsof pre-viousyearstorageleaves(tunics).Smallerbulbs,calledbulbils,can beformedintheaxilsofthefollowingleaves,andactasvegetative diasporesaccountingforclonalmultiplication.
Allstudieswerecarriedoutonfreshlyexcavatedplants.Using arulerand/oradigitalcalliper(precision±0.02mm)thefollowing organswerecountedandmeasuredinfreshcondition:(1) replace-ment bulbs:diameter(or lengthand widthfor ellipsoid bulbs; usuallywiththethintunicoftheformerparentbulb);(2)bulbils: diameter,number,position(onlybulbilsofthecurrentyearwere considered;thenumeroussmallbulbilsofG.fragiferaandG.villosa wereonlycounted);(3)flowers:number.
Todeterminethenumberofbulbilsperlifestageandthe thresh-olds for the formation of bulbilsand flowers, respectively, we constructedplots based onclass widthsof 0.5mm diameterof thereplacementbulbandcalculatedtheproportionsof individ-uals developing therespectivestructures. Using a routinebuilt intoSigmaPlot3.0(SyStatGmbH);thesegraphswerefittedagainst equationsintheformy=1/[1+e−(x−x0)/b]whichdescribeasigmoid
functionwithx0asthethresholdforbulbdiameter,i.e.theturning pointofthegraph’sslope(compareSchnittleretal.,2009).For bul-bilsformedexclusivelybyjuvenileplants(seebelow)onlythepart ofthecurveuptothemaximumproportionofplantswithsuch bul-bilswasconsideredforregression.Forthegraphicalcomparisonof thresholdsinFig.2,weestimatedthediameterofthereplacement bulbs,assumingaround(ellipsoid)shape.Wealsoincludedresults ofsimilarinvestigationsforG.luteaandG.spathaceapublishedinan earlierstudy(Schnittleretal.,2009).Asaproxyforstoredresources, we calculated thevolume correspondingtothemeasured bulb diametertoderivetherelationshipspicturedinFig.2.Round bul-bilsweremeasuredonce,withvolumescalculatedaccordingtothat ofasphere.Forovatebulbils,likethoseofG.pratensis,lengthand widthweremeasuredandtherespectivevolumewascalculated accordingtoanellipsoid.Averagemeasurementsarealwaysgiven withthestandarderrorofmeans(SEM).
Results Typesofbulbils
Followingbulbildevelopmentthroughoutalllifestages, two typesofbulbilswererecognizedintheinvestigatedGageaspecies (seeTable1).
TypeIbulbils
These bulbils develop as soon as the replacement bulb has reachedacertaindiameterandarepersistentlyformedthroughout thelifeoftheplantinnon-floweringaswellasflowering individ-uals.Theyaretypicallydevelopedasbasalbulbilsintheaxilofthe secondbasalleaf(Levichev,2013).Foraveragenumberandaverage diameterofbulbilsseeTable1.WedetectedthesetypeIbulbilsin sevenofthestudiedspecies.Gageadavlianidzeaedevelopsrather smallcollateralbulbils,arrangedaroundthereplacementbulblike garliccloves(max.2.39mm,464bulbilsfrom119plantsmeasured). InfloweringplantsofG.fragifera,especiallyinplantswithmultiple flowers,thesinglebasalbulbil(max.2.70mmdiameter,309plants arecounted)isoftensmallerthaninlargenon-flowering individ-ualsandhardtodetect.Gageanigradevelopsonebulbiloverits entirelifeandistheonlyspecieswherethethresholdsforbulbil andflowerproductionareveryclosetoeachother(Table2). Vir-tuallyallfloweringplantsofthisspecieshavethisbulbil,whichis largeinrelationtothereplacementbulbdiameter(max.3.99mm, 252plantsmeasured).Onaverage2.6flowersareformedperplant,
Table1
Systematicaffiliation,reproductivepatterns,meannumbersandsizesforbulbilsand/orflowers(±SEM)andchromosomedataoftheanalyzedGageaspecies. Species SectionofGageaa Reproductive
pattern
Meannumberofbulbils/flowersb Meansizeofbulbils(mm) Chromosome
numbers Flowers BulbilstypeI BulbilstypeII BulbilstypeI BulbilstypeII (2n)c
G.angelae Gagea B 3.70±0.23 – 2.94±0.15 – 1.44±0.03 – G.davlianidzeae Minimae A 3.25±0.42 3.90±0.29 – 1.22±0.02 – 24,28 G.fedtschenkoana Gagea B 2.20±0.09 – 1.00±0.00 – 1.34±0.07 18e G.fragifera Didymobulbos A+B 2.67±0.15 1.00±0.00 10.39±0.57 0.90±0.03 0.3–1.0d 60,84 G.luteaf Gagea B 3.95±0.12 – 8.65±0.35 – 1.49±0.01 72 G.nigra Minimae A 2.58±0.11 1.00±0.00 – 2.12±0.03 – – G.peduncularis Didymobulbos A 1.67±0.12 1.00±0.00 – 2.04±0.04 – 36 G.pratensis Gagea A 1.88±0.07 1.32±0.03 – 4.56±0.07 – 60 G.spathaceaf Spathaceae A 1.51±0.06 9.29±0.47 – 1.27±0.02 – ∼102,106 G.villosa Didymobulbos A+B 5.00±0.42 1.00±0.00 11.46±0.70 2.14±0.05 0.4–1.3d (36),48
aForinfragenericclassification,seePetersonetal.(2008,2011),Peruzzietal.(2008)andPeruzzi(2012b). bAveragenumbersforallindividualsdevelopingtherespectivestructure(flowersand/orbulbils). c ChromosomedataaccordingtoPeruzzi(2003,2008b),Peruzzietal.(2011),andliteraturecitedtherein. d Bulbilsizeestimatedfromoccasionalmeasurements.
eThiswouldbeananeuploidstate(1.5×),butflowcytometricmeasurementsofrelativeDNAcontentcomplybycomparisontoCentralEuropeantaxawiththetetraploid
level,i.e.2n=48.
f SeeSchnittleretal.,2009fordetaileddata.
butwithlargedifferencesthatseemtobecorrelatedwith habi-tatmoisture(smallandoftensingle-floweredplantsinthesteppe zonebutlargemulti-floweredplantsinthemountains).Plantsof G.peduncularisstartsoontodevelopasingle,comparativelylarge bulbil(max.4.53mm,325plantsmeasured).ForG.pratensis, vir-tuallynothresholdforbulbilproductionexists,eventhesmallest plantswerealreadyfoundwithbulbils.Notableistheirlargesize (max.12.9mm×7.9mm,595bulbilsfrom452plants analyzed). Themajorityofnon-floweringplants(76%)hadonlyonebulbil; for144floweringplantswecountedonaverage1.3bulbilsand1.9 flowers.ThenumeroussmallbulbilsofG.spathaceabelongaswell totypeI(Schnittleretal.,2009).SimilartoG.fragifera,G.villosa developsasinglebutlargerbasalbulbil(max.4.05mm,205plants measured).
TypeIIbulbils
The second type of bulbils develops temporarily in non-floweringplantsoffiveoftheinvestigatedspecies.Thesebulbilscan beformedonstolons(G.angelae),appearasbasalbulbilssolitary(G. fedtschenkoana)oringroups(G.lutea),orareclustersofverysmall suprabasalbulbilsinG.fragiferaandG.villosa.Non-floweringplants ofGageaangelaeformbulbilsonastolonofupto7.9cmlength (mean2.3cm).Oneofthesebulbilsislargerthantheothers(max. 5.85mm,average1.94±0.07mmdiam.,239plantswith702 bul-bilsanalyzed).However,virtuallyallplantsstoptodevelopastolon whenflowering.SimilarisG.lutea(Schnittleretal.,2009),butin thisspeciesthegroupofbulbilsissessile(max.2.90mm,largest
bulbilonaverage1.86±0.03mmdiam.,226plantswith1955 bul-bilsanalyzed).Largenon-floweringplantsofG.fedtschenkoanaform asinglebulbil(max.2.88mm,53plantsmeasured),whichceases tobedevelopedinfloweringindividuals.
Thresholdsforbulbilandflowerformation
Thegraphicsontheleft-handsideofFig.1(AandB)showthe proportionofplantswithtypeIand/ortypeIIbulbilsandflowersin dependencefromthediameterofthereplacementbulb;thefigures ontherightsideofbothfigurepanelsdisplaynumbersofbulbils andflowers,respectively.Fromtheplots,athresholdforthe diam-eterofthereplacementbulbcanbedetermined,where50%ofall plantscanbeexpectedtodeveloptherespectivestructure.Table2 liststhesethresholdsforbulbilandflowerformation,respectively, andthecorrelationcoefficientsoftheregressionwithasigmoid function.InallspeciescharacterizedbytypeIIbulbils,largerplants hadeitherthesebulbilsorflowers,withveryfewexceptions(G. angelae:sixlargeplantswithinflorescencesplusinadditionone stolonwithbulbils;G.lutea:oneplantwithagroupofbulbilsin anotherwise normalinflorescence; G.villosa: threeplants with mixedinflorescences;G.fedtschenkoana:one plantwithflowers andabulbil).
Ifadultplantsaredefinedasthoseindividualsattainingabulb diameterwhere90%ofallplantsareexpectedtoflower,wecan takethisvaluefromtheregressionfunction(Fig.1,leftcolumnsof partialfigures)andcalculatethecorrespondingbulbvolume.This
Table2
ThresholdsfordevelopmentofflowersandbulbilsintenspeciesofGagea.Givenisthemeanaveragediameter[mm]ofthereplacementbulbwhere50%oftheplantsdevelop therespectivestructure(equaltotheparameterx0ofthesigmoidalfunctionsshowninFig.1).ThegoodnessoffitRaccordingtoasigmoidalfunctionisgiven(seeSection
“Materialsandmethods”forexplanation).
Species Thresholdforbulbil/flowerformation(mmreplacementbulbdiam.)
BulbilstypeI R BulbilstypeII R Flowers R
G.angelae – 3.23±0.13 0.987 7.43±0.42 0.974 G.davlianidzeae 1.90±0.01 0.999 – 3.88±0.10 0.990 G.fedtschenkoana – 1.82±0.93 0.945 3.27±0.02 0.999 G.fragifera 2.55±0.07 0.985 3.47±0.09 0.996 5.74±0.07 0.995 G.luteaa – 2.55±0.09 0.980 6.94±0.04 0.998 G.nigra 3.56±0.06 0.995 – 3.87±0.02 0.999 G.peduncularis 1.03±0.04 0.995 – 3.63±0.04 0.998 G.pratensis 0.00 – 8.74±0.35 0.979 G.spathaceaa 2.44±0.10 1.000 – 5.34±0.13 1.000 G.villosa 2.77±0.16 0.935 3.01±0.25 0.984 7.05±0.20 0.964
Fig.1.(AandB)ReproductivepatternsintheinvestigatedspeciesofGagea.Left:proportionsofplantsformingtypeIbulbils(whitecircles),typeIIbulbils(grey)and/or flowers(black)inclassesof0.5mmdiameterwidthofthereplacementbulb.Theincreasingpartsofthegraphsarefittedagainsttheequationy=1/[1+e−(x−x0)/b]withx0as
thethresholdfortheformationoftherespectivestructure(seeTable2).Verticaldottedlinesindicatethepositionofthisthreshold.Right:numbersofbulbilsand/orflowers perplantintherespectivediameterclasses;barsdenotethestandarderrorofmeans.ForgraphsofG.luteaandG.spathaceaseeSchnittleretal.(2009).
allowstoexpresstherespectivethresholds(fortheformationof thetwobulbiltypesandflowers,respectively)asafractionofadult bulbvolume(Fig.2).ExceptforG.nigrawhichstartstogrowits singletypeIbulbilonlyifthereplacementbulbhasreachednearly thesizewhenfloweringstarts,thethresholdsfortheproduction ofalltypesofbulbilsarelow(typeI:0–5%,typeII:3–13%ofadult bulbvolume)andwellcomparablebetweentheanalyzedspecies. Thedifferenceiswithrespecttoflowering:inflorescencesstartto
Fig.2. Proportionofvolume(definedasthereplacementbulbvolumewhere90%of allplantsofaspeciesflower)whichmustbereachedthat50%ofallplantsdevelop typeIbulbils(opencircles),typeIIbulbils(greycircles)orflowers(blackcircles).
developwhenbetween38and65%oftheadultbulbvolumeare reached.
Discussion
Reproductivepatterns
Ourstudyisinagreementwithearlierstudies(e.g.Levichev, 1999,2013;Schnittleretal.,2009)showingthatrelative propor-tionsof vegetative vs. generative reproduction are likely tobe species-specificandveryvariablewithinthegenusGagea. Com-paringthe species,we candifferentiate two mainreproductive patterns (see Table 1): (A) species with type I bulbils only (G. davlianidzeae,G.nigra,G.peduncularis,G.pratensis,G.spathacea), whichlatterareindefinitelyformedinimmatureaswellas sex-uallyreproducingindividuals;andothers(B)formingonlytypeII bulbils,whicharenolongerdevelopedbyfloweringadultplants(G. angelae,G.fedtschenkoana,G.lutea).Inaddition,bothstrategiescan occursimultaneously(G.fragifera,G.villosa).Thelattertwotaxa, combiningbothbulbiltypes,possessasingle,basalbulbilwhich developsalreadyinratherweakplants.Withmoreresources,plants formadditionaltypeIIbulbils.Inthesetwospecies,thetwotypes ofbulbilscanbedistinguishedmorphologically:typeIbulbilsare basal,mostlikelydevelopingintheaxilofthesecondleaf (com-pareLevichev,2013:414),whereastypeIIbulbilsaresuprabasal andarisemostlikelyfromtheaxilsofthethirdandupperreduced leavesonashort(G.fragifera)toextremelyshort(G.villosa)scape.
Duetothis reproductiveandmorphogenetic variability,it is currentlytoospeculativetosynthesizeourfindingsinto a gen-eralgrowthand reproductionschemeapplicableatsection(see Table1)orevengenuslevel(seeSection“Introduction”).Recently, somespecies-specific growthandontogenetic patterns,derived fromobservationsofwildandcultivatedplants,werewell illus-tratedbyLevichev(2013:414),includingG.lutea,G.fragifera(=G. liotardii(Sternberg) Schult.and Schult.), G. pratensisand G. vil-losafromourstudyspecies.Theseschematicdrawingsillustrate thedifferentrealizedreproductivestrategies,i.e.patterns(A)in G.pratensis,(B)inG.luteaandthemixtureofbothinG.fragifera, respectively.
Thresholdsforreproductivestructuresandsizeofbulbils Throughoutthisstudywefoundcleardependenciesforthe for-mationof all reproductivestructures from thediameter ofthe replacementbulb.Thelatteristheorganstoringresourcestocarry theplantintothenextyear.Thisfactprovidesstrongevidencethat itistheresourcestatusoftheplantthatdeterminestheformation oftherespectivereproductivestructures.Thedetectedthresholds arespecies-specific(compareBlodgett,1910,forlikewise species-specificthresholdsforfloweringinErythronium).Bulbilformation isgenerallylesscostlythanflowering(Schnittleretal.,2009), irre-spectiveofthebulbiltype.
ForGageaspeciesofdryhabitats,weexpectthatalargersizeof bulbilswillbeadvantageoustosurvivelongperiodsofdroughtin summer,thuslimitingthenumberofbulbilsperplant.Fromour setofstudyspecies,thisespeciallyappliestoG.nigra,G. peduncu-larisandG.pratensis,allthesespeciesinhabitingratherdryhabitats comparedtootherspeciesofGageafoundintherespectiveregions. ThisstrategyisespeciallyprominentinG.nigra,aspeciesfromthe ChineseTian-Shanmountains,inhabitingaswellthestepperegion ofthefoothills.Herethesinglebasalbulbiliscomparativelylarge (>2mm),andthethresholdforitsdevelopmentstayshigh(Fig.2). ExtremelylargebulbilsarealsoformedinG.pratensiswithone basalbulbil(rarelymore).Inthisspecies,bulbilformationinupper leaves(atleastthethird)wasreportedtooccurregularlybysome earlyauthors(e.g.,Horn,1874;Irmisch,1850),butthiswasdoubted byotherauthors(Martinovsk ´y,1930).Wehaverarelyseensuch aerialbulbils,andtheywerenotconsideredherefortheanalysis. Forthisspecies,wewereunabletodetectaneffectivethreshold forbasalbulbilformation.Hence,weassumethatplantsgrowing frombulbilscandeveloptheirownbasalbulbilalreadyintheirfirst year.AccordingtoIrmisch(1850)(seealsoGraebnerandKirchner, 1934),thebasalbulbilisevenabletoproduceaninflorescencein thenextyear,i.e.itcanfloweralongwithitsparentplant.Asseed setisvirtuallyabsentinthepentaploidpopulationsofthisspecies investigatedbyPfeifferetal.(2012b),verysmallplantsgrownfrom seedswerenotfound.However,populationsinsouthernCentral EuropeandtheMediterraneanregionarereportedtobetetraploid andcanthusbeexpectedtosetseed(GutiérrezandVázquez,2010; Peruzzi,2003).
Resource-dependentbulbilformation
WhereastypeIbulbilsareformedconsecutivelyassoonasa cer-tainsizeofthereplacementbulbisreached,typeIIbulbilsnormally ceasetobedevelopedifaplantislargeenoughtoproduceflowers. Withgrowingbulbsize,thedecreaseintheproportionofplants withtypeIIbulbilsmirrorstheincreaseinproportionofflowering plants(Fig.1).Wehavetermedthisresource-dependentchange betweenbulbilandinflorescenceformationtobea“reproductive switch”(compareSchnittleretal.,2009),a termalready coined byWangandCronk(2003)fora studyonmeristemfateinthe
GesneriaceaeTitanotrichumoldhamii,whereyoungplantsdevelop bulbilsintheinflorescenceinsteadofflowers.
For the genus Gagea,two explanationsfor the only tempo-raryformationoftypeIIbulbilsappearpossible,bothrelatedto resources stored in replacement bulbs. The first of these pos-sibilities would be a resource-dependent conversion of flower meristems into bulbilslike in Titanotrichum (Wang and Cronk, 2003).If we assumethat floral meristemsdo not develop into (sexual and resource-consuming)flowersbut into (asexualand cheaper;Schnittleretal.,2009)bulbils,weakplantswillproduce bulbilliferousinsteadoftrueinflorescences.Thisappearslikelyto bethecaseforG.fragiferaandG.villosa.Inbothspecies,weakplants developasinglebasalbulbiland,inaddition,groupsofsuprabasal bulbilsintheaxilsofupperreducedleaves,i.e.atthesameposition whichistypicallyassumedbyinflorescence(s)inflowering indi-viduals.ThisismostobviousinG.fragifera,sincethebulbilgroup resemblesatrueinflorescence:ashortscapeelevatesthebulbils tosurfacelevel,thelowermostbractisaratherlargegreen assimi-latingleaf,followedbyreducedfurtherbracts(seeLevichev,2013: 414).Wedemonstratedthatsuchbulbilheads(whichcanbe inter-pretedasbulbilliferousinflorescences,followingthismechanism) occurinweaker,immatureplants,tobereplacedbytrue inflore-scencesinwell-provisioned,largeplants.Thisphenomenonseems tobecommoninseveralspeciesofGageasect.Didymobulbos.Here groupsofsuprabasal(typeII)bulbils(compareTisonetal.,2013)are likelytorepresentconvertedinflorescences.InG.villosa,aspecies closelyrelatedtoG.fragifera(Peruzzietal.,2011),theadventive suprabasalbulbilsarelocatedonextremelyshortscapes,but occa-sionallyplantsdevelopmixedinflorescenceswithbulbilsin the axilsoflowerleavesandflowersontop.AlreadyIrmisch(1850) statedthatsometimesflowerscanbereplacedbybulbils(seealso Fig.180inGraebnerandKirchner,1934;Horn,1874).Inagreement withtheirtransientnature,suchmixedinflorescencestendtohave shorterscapesthantrueones(compareFig.179inGraebnerand Kirchner,1934).
TheGageaspecieswithtypeIIbulbilswhichdo notdevelop insuprabasalposition,butbasal(G.angelae:bulbilsatthetipof astolon, G.lutea:agroupofbulbils,andG.fedtschenkoana: sin-glebulbil,allfromsect.Gagea)showasimilar,usuallycomplete, reproductiveswitch.Inthesecasesasecondexplanationforthe resource-dependentformation ofbulbilsisconceivable: flower-ingplantsmaylackresourcestoproducebulbilsinadditiontoan inflorescence.
Areproductiveswitchbetweenvegetativereproductionby bul-bilsinjuvenilestagesandsexualreproductionthroughseedswith theonsetoffloweringalsooccursinotherLiliaceaes.str.:in imma-tureTulipasylvestris,anadventivebulbisformedonanextending (rarelyveryshorttoabsent)stolon,itsapexistheterminalbudof thenon-floweringmotherbulbandconsequentlysubstitutedby aninflorescenceinadultindividuals(Irmisch,1850,1863a;Jäger, 1973).InErythronium,thebudproducingthebulbilsinimmature stagesisreplacedbythebudformingtherenewalbulbinthethen sympodiallygrowing,floweringplantsinmostspecies(Blodgett, 1910).However,there mightbeanadditionalresource-induced component,as“undercultivationaxillarybudsinadultbulbsin mostspeciesdevelopintobulbs”(Blodgett,1910:footnoteonpage 350);forculturedE.dens-canisthedevelopmentof ratherlarge adventivebulbs(oftenabletoflowerinthenextseason;compareG. pratensis)wasalreadyreportedbyIrmisch(1863b).Anexceptionis E.propullans,wheretheswitchismissingandfloweringplants con-tinuetoreproducevegetatively.Buttheadditionalproductionof adventivebulbsonelongatedlateralstolon’siscostly,andprobably explainsthereducedflowersizeinthisspecies(Blodgett,1910).
ManyspeciesofAllium(Alliaceae),anotherspecies-richgenus withadiversitycentreinCentralAsia(Gregoryetal.,1998), pro-ducevariableproportionsofbulbilsandflowers.WhereasinGagea
atleastsomespecieshaveanearlycompleteswitch,i.e.produce eitherallbulbilsorallflowers,inAlliumbothbulbilsandflowers usuallycoexistinasingleinflorescence.Therespectivediaspores mayhave different dispersal options and reproductivesuccess, dependingonenvironmentalconditionsinagivenyear(Ceplitis, 2001).However,astudybyRonsheimandBever(2000)proofed theoccurrence ofheritable allocationpatternstoseeds, subter-raneanandaerialbulbilsinA.vineale,thuslimitingthepotential ofthesegenotypestoadapttheirreproductivestrategyinresponse toresourceavailability.
Bulbilsasanescapestrategyforsexuallysteriletaxa
Inmanyplanttaxa,variousmeansofvegetativereproduction supplementsexualreproductionbyseeds.InallinvestigatedGagea species,theoftenprolificproductionofbulbilsenablingextensive clonalmultiplicationoccursatleastinimmaturelifestages. Espe-ciallyin taxawithefficientsexualreproductionof adultplants, thevegetativereproductionlosesitsformerimportanceoriseven completelyabandoned(reproductiveswitchintaxawithtypeII bulbils).PrimeexamplesareG.lutea(hexaploid,groupsoftypeII bulbilsbutcompletereproductiveswitch),whichsetsseed regu-larly(Pfeifferetal.,2012b;TakahashiandTani,1997)andG.nigra (withasinglebasaltypeIbulbil),whichwasalsoregularlyobserved withseeds(99%ofallplants,n=95,mean19.5seedspercapsule; Petersonetal.,2011).However,notallspeciesofGageacanuse bothbulbilsandseedsforreproduction;manyspecieswithhigh and/oroddploidylevels(comparePeruzzi,2003,2008a,2012afor chromosomedata)seemtohavelimited(e.g.G.pratensis;Pfeiffer etal.,2012b)tonoseedset(Bianchi,1946;Garganoetal.,2007; Tisonetal.,2013).Insuchcasesofsexualsterility,theproduction ofmanybulbils(especiallyoftypeI,produced indefinitelyover theentirelifetimeofaplant)couldbeaneffectiveescapestrategy. InnonaploidG.spathaceawithvirtuallyabsentseedset,survival isensuredonlythroughnumeroustypeIbulbils(without repro-ductiveswitch).Theentirespeciesismadeupbyvirtuallyasingle clone(Pfeifferetal.,2012a).Likewisein G.davlianidzeae, where seedsappeartobeaborted(Petersonetal.,2011),severaltypeI bul-bilsareformedandgrantsurvival.However,thefewinvestigated speciesdonotallowforageneralconclusion.
TheoftensterileG.fragifera(Bianchi,1946)doeslikelyswitch frombulbilliferoustotrueinflorescences(compareGraebnerand Kirchner,1934:342).Inaddition,theformationofonebasaltypeI bulbilaccountsforminimalvegetativemultiplicationevenin flow-eringindividuals.Ifweassumeseedsterility,thefitnessofaplant (measuredasnumberofbulbilsformedperplant)shouldreachits optimumaround5mmdiameterofthereplacementbulb(Fig.1), butshouldstronglydecreaseforlargerplantseffectively reprodu-cingonlybythesinglebasalbulbildespiteofflowering.Thesame mechanismisprincipallyvalidforG.villosa;in thistaxon occa-sionallymixedinflorescencesoccur(ourobservations;Horn,1874; Martinovsk ´y,1930).
Themanifold variation in patterns of bulbil production and hencereproductivestrategiesgeneratedbymerechangesinthe respective thresholds for the formation of reproductive struc-turesopensanewwindowofopportunityforevolutionthrough hybridization (see Peruzzi, 2008b; Peterson et al., 2009), since divergentparentalreproductivestrategiesmightgeneratenovel reproductivecombinationsinhybridoffspring.Bulbilproduction maythusbethekeyforthesurvivalofsterilehybridsbetween G.luteaandG.pratensis(Pfeifferetal.,2012b)andseveral hybri-dogenicspeciesofsect.Didymobulbos(Tisonetal.,2013).Witha strongselectionforincreasingfertilityofnewlycreatedhybrids (RamseyandSchemske,2002),vegetativereproductionviabulbils allowssurvivalofinitiallysterilehybrids,whichmayextenteven totheleveltoberecognizedasspecies.Such“transientspecies”
mayregain fertility bychanges in ploidy level, thus effectively preservingtheir“geneticconfiguration”forfurtherevolution.This phenomenonseemstobeespeciallycommoninspecies-richplant generawhere hybridizationand sometimesreticulateevolution includessexuallyhighlysteriletaxa(seeexamplesinWangand Cronk,2003).SpeciesofAllium(Alliaceae)withoddploidylevelsare otherexampleshowbulbilproductionensuressurvivalof popula-tionsincaseofloworabsentgenerativereproduction(e.g.Mathew, 1996;A.monathum:KawanoandNagai,1975;A.oleraceum:Fialov andDuchoslav,2013).
Conclusion
Wedemonstrated forthetenanalyzed speciesofGageathat thebalancebetweenreproductionthrough generative(flowers) and/orvegetativemeans(bulbils)canbedramaticallyalteredby slightchangesinafewtraits,namely(i)numberofbulbilsand(ii) thewayandtimingoftheirformationduringtheplant’sindividual life.Whereasinflorescencesstarttodevelopatca.40%oftheadult bulbvolume,bulbilsdevelopatlessthan13%(exceptforG.nigra) ofvolumeoftheadultreplacementbulb.Dataforresource allo-cationtobulbilsandseedsinG.lutea(Schnittleretal.,2009)may explainthemuchlowerthresholdsfortheformationofbulbilsin comparisontoinflorescences.Thethresholdsfortheformationof therespectivereproductivestructuresseemtobehighly species-specificanddeterminethesurvivalpotentialofsteriletaxa.Hence, quantitativeanalyses ofreproductive traits,as provided in this studyforasmallselectionoftenGageaspecies,areapromising approachtoanalyzereproductivepatternsandunderstandtheir importanceforspeciation,especiallyviahybridizationand poly-ploidization,inGagea.Moreinvestigationscombinedwithanalyses ongeneticdiversityarenecessarytoelucidatewhichfactorsshape thesepatterns,especiallytheinteractionbetweenenvironmental factorsandphylogeneticconstraints.
Acknowledgements
Wethankanumberofcolleaguesandstudents,especiallyK. Münse, A. Petzold, S. Starke and D. Zajac, for help with mor-phologicalmeasurementsduringfieldtripsandvariouspractical courses,Dr.I.LevichevforthedeterminationofG.fedschenkoana andseveralfundingagenciesenablingcollectioninwesternChina as a side activity within the respective projects (Volkswagen Foundation,I/78636;DAADpartnershipprogramme,D/05/06946; Deutsche Forschungsgemeinschaft, SCHN 1080/1-1). T. Pfeiffer wassupportedbyaKäthe-Kluthscholarshipofthe Ernst-Moritz-Arndt-University Greifswald, which is gratefully acknowledged. Furthermore,wewouldliketothankananonymousreviewerfor helpfulcomments.
Appendix1. Localitiesandnumbersofinvestigatedplants Gageaangelae
China:1–76 XinjiangProv.,EasternTian-Shan, HeavenLake, Jurte settlement, stony, open slopes, 88◦0602E, 43◦5331N ±250m,2173±50ma.s.l.,30.4.2007;77–140EasternTian-Shan, HeavenLake,roadthroughJurtesettlement,stony,openslopes, 88◦0602E, 43◦5331N ±250m, 2173±50m a.s.l., 26.4.2007; 141–162EasternTian-Shan,HeavenLake,sidevalleyaboveJurte settlement,poorsoilonalpinemeadows,88◦0542E,43◦5325N ±250m,2224±50ma.s.l.,28.4.2009;163–279EasternTian-Shan, Heaven Lake, side valley above Jurte settlement, richer soil, alpine meadows, 88◦0542E, 43◦5325N ±250m, 2224±50m a.s.l.,28.4.2009;280–352EasternTian-Shan,HeavenLake, Jurte
settlement,gravelly,openslopes,88◦0602E,43◦5331N±250m, 2173±50ma.s.l.,29.4.2009;353–379EasternTian-Shan,Heaven Lake, S-exp. slope in a valley 5.5 km SE of the lake,near log-house camp, alpine meadow,88◦1014E, 43◦5018N ±100m, 2170±50ma.s.l.,26.4.2009.
Gageadavlianidzeae
China:1–166XinjiangProv.,EasternTian-Shan,HeavenLake, fringesofspruceforestbelowthelake,88◦0650E,43◦5435N ±250m,1920±75ma.s.l.,27.4.2009.
Gageafedtschenkoana
Kazakhstan,Astanaregion:1–180Korgalzhynnationalreserve ca.115kmWSWAstana,slightdepressionsinsteppearoundthe stationofthereservenearalake(oldfishingfactory),69◦3244E, 50◦2838N±100m,305±25ma.s.l.,18.4.2012;181–201Astana, city limits, mown lawns with some steppe remnants in front of the administrative building of the L.N. Gumilyov Eurasian National University, Munaitpasov Str., 71◦2523E, 51◦1052N ±50m,355±25ma.s.l., 20.4.2012;202–245Astana,citylimits, mown lawnswith some steppe remnants,scattered tree plan-tationsalong MunaitpasovStr.,71◦2524E,51◦1052N±50m, 355±25ma.s.l.,22.4.2012.
Gageafragifera
China:1–19 XinjiangProv.,EasternTian-Shan, HeavenLake, alpine meadows near melting snow, 88◦0516E, 43◦5314N ±250m,2452±50ma.s.l.,20.4.2007;20–173EasternTian-Shan, HeavenLake,88◦0529E,43◦5310N±250m,2348±50ma.s.l., 20.4.2007;Italy: 174–250Piemont, Cuneo, settlement St.Anna 4km SW Sampeyre, alpine meadows, 44◦3352E, 7◦1015N ±100m, 1458±50m a.s.l., 6.5.2010; 251–351 settlement Pon-techianale 5km W Casteldelfino, alpine pastures, 44◦3710E, 7◦0137N ±250m, 1668±50m a.s.l., 7.5.2010; 352–412 set-tlement SantAnna, near Meleza, ca. 8km WSW Casteldelfino, alpine meadows, 44◦3427E, 6◦5831N ±150m, 1852±50m a.s.l.,8.5.2010.
Kazakhstan:413–480IliAlataurange,Almaty,roadfromthe citytoMedeustadium,ca. 1 kmabove mudslideshelterdams, richdeciduousforest,77◦0225E,43◦1009N±25m,1498±40m a.s.l.,22.04.2013;481–535Almaty:Medeuvalleyca.1.2kmWNW Shymbulakskiresort,alpinemeadowonashallowNNE-exp.slope, 77◦0351E,43◦0830N±25m,1941±50ma.s.l.,21.04.2013. Gageanigra
China:1–110XinjiangProv.,Urumqi,SouthernMountains, Sar-qokka,WestBiayangouvalley,openalpinemeadow,87◦0993E, 43◦2546N ±250m, 2034±20m a.s.l., 16.4.2007; 111–270 Urumqi,Southern Mountains,Sarqokka,WestBiayangouvalley, openalpinepasture,87◦1017E,43◦2546N±250m,2096±20m a.s.l., 16.4.2007;271–369EasternTian-Shan, HeavenLake, pas-tures near Jurte settlement, 88◦0602E, 43◦5331N ±250m, 2173±50ma.s.l.,20.4.2007;370–393EasternTian-Shan,Heaven Lake,meadowsaboveJurtesettlement,88◦0542E,43◦5325N ±250m, 2224±50m a.s.l., 20.4.2007; 394–447 Urumqi, city limits,Jarmalakhill,N-exp.slope withtreeplantationson rem-nants of natural steppe vegetation, 87◦3451E, 43◦4801N ±100m,942±20ma.s.l.,23.4.2009;449–457EasternTian-Shan, HeavenLake,driermeadowsaboveJurtesettlement,88◦0542E, 43◦5325N±250m,2224±50ma.s.l.,20.4.2007.
Gageapeduncularis
Greece,Crete,southernslopesofthePsilioritismassive:1–31 Mires, rocky outcrops within a Kermes oak forest ca. 600m SSW Miamou,24◦5617E,34◦5757N±50m, 490±25ma.s.l., 18.03.2013;32–244Mires, phyrganaonloamysoil below lime-stone,W-exp.shallowslopeabovetheroadtoLentas,24◦5646E, 35◦0016N ±50m, 173±25ma.s.l., 18.03.2013; 246–342 Psil-ioritismassive,Anogia:shallowloamysoilaccumulatedbetween dolomiticlimestonerockswithAcersempervirens,heavilygrazed, ca.3kmSAnogia,neartheroadtotheNidaplateau,24◦5309E, 35◦1515N±50m,1186±50ma.s.l.,20.03.2013.
Gageapratensis
Germany:1–45Mecklenburg-WesternPomerania,Behrenhoff, park of the former estate, N and W of the lake, 54◦0013E, 13◦2311N±100m,20±5ma.s.l.,1.5.2006;46–213Greifswald, Botanical Garden W of the old city, 54◦0537E, 13◦2203N ±30m, 10±4m a.s.l., 9.5.2008; 214–236 Behrenhoff, access road from B110, between Behrenhoff and Müssow, field mar-gins, 54◦0008E, 13◦2442N ±100m, 20±5m a.s.l., 8.5.2009; 237–345 Wrangelsburg, access road to castle, field margins, 54◦0040E,13◦3558N±100m,20±5ma.s.l.,8.5.2009;346–356 RügenIsland,Trent,churchyard,54◦3114E,13◦1528N±50m, 10±5ma.s.l., 20.4.2009;357–368RügenIsland, Zirkow,church yard, 54◦2317E,13◦3223N ±50m, 10±5m a.s.l., 20.4.2009; 369–419Groß Mohrdorf, churchyard,54◦2318E, 12◦5649N ±50m,20±10ma.s.l., 25.4.2009;420–472RügenIsland, Trent, nearachurchyard,54◦3114E,13◦1528N±50m,10±5ma.s.l., 20.4.2009;473–547 RügenIsland, Gingst,churchyard, rightof thechurch underlinden trees, 54◦2723E,13◦1539N ±50m, 12±5ma.s.l.,21.4.2010;548–572Greifswald,Krullstrasse,park, 54◦0508E,13◦2402N±50m,10±4ma.s.l.,11.04.2010. Gageavillosa
Germany: 1–35 Mecklenburg-Western Pomerania, Rügen Island,Trent,churchyard,alongthewaytocemetery,54◦3114E, 13◦1528N ±50m, 15±5m a.s.l., 21.04.2010; 36–63 Rügen Island, Samtens, cemetery, near the parking lot, 54◦2125E, 13◦1743N ±50m, 15±5m a.s.l., 21.04.2010; 64–77 Rügen Island, Gingst, church yard, 54◦2723E, 13◦1539N ±50m, 12±5m a.s.l., 21.4.2010; 78–128 Greifswald, Krullstrasse, road marginsaround trees,54◦0508E, 13◦2402N±50m, 10±4m a.s.l., 11.04.2010; 129–180 Rügen Island, Patzig, church yard, 54◦0500E, 13◦2400N ±800m, 10±4m a.s.l., 21.04.2010; 181–199Thuringia,Ingersleben,woodlandfringenearriverGera, 50◦5516E, 10◦5748N ±500m, 222±25m a.s.l., 10.04.2010; 200–218Molsdorfcastle,park,50◦5358E,10◦5738N±150m, 240±20m a.s.l., 10.04.2010; 219–251 Molsdorf Castle, slope to the castle, 50◦5402E, 10◦5737N ±50m, 240±20m a.s.l., 10.04.2010.
References
Ali,S.I.,Levichev,I.G.,2007.Gagea.In:Ali,S.I.,Qaiser,M.(Eds.),FloraofPakistanNo. 215Liliaceae.Dept.ofBotany,UniversityofKarachi&MissouriBotanicalPress, Karachi&St.Louis,pp.17–82.
Bianchi,R.,1946.UntersuchungenüberdieFortpflanzungsverhältnissevonGagea fistulosa(Ramond)Ker-GawlundLloydiaserotina(L.)Rchb.Ber.Schweiz.Bot. Ges.56,573–582.
Blodgett,F.H.,1910.TheoriginanddevelopmentofbulbsinthegenusErythronium. Bot.Gaz.50,340–373.
Buxbaum,F.,1937.DieEntwicklungslinienderLilioideaeII.Diesystematische Stel-lungderGattungGagea.Bot.Arch.38,305–398.
Ceplitis,A.,2001.Geneticandenvironmentalfactorsaffectingreproductivevariation inAlliumvineale.J.Evol.Biol.14,721–730.
Gargano,D.,Peruzzi,L.,Caparelli,K.F.,Cesca,G.,2007.Preliminaryobservations onthereproductivestrategiesinfiveearly-floweringspeciesofGageaSalisb. (Liliaceae).Bocconea21,349–358.
Fialov, M., Duchoslav, M., 2013. Response to competition of bulbous geophyte Allium oleraceum differing in ploidy level. Plant Biol.,
http://dx.doi.org/10.1111/plb.12042.
Graebner,P.,Kirchner,O.,1934.GattungGageaSalisb.Goldstern.In:Kirchner,O.,von Loew,E.,Wangerin,W.,Schröter,C.(Eds.),LebensgeschichtederBlütenpflanzen Mitteleuropas.Band1.,Abt.3.UlmerStuttgart.,pp.326–355.
Gregory,M.,Fritsch,R.M.,Friesen,N.,Khassanov,F.O.,McNeal,D.W.,1998. Nomen-clatorAlliorum:AlliumNamesandSynonyms–AWorldGuide.RoyalBotanic Gardens,Kew.
Gutiérrez,M.,Vázquez,F.M.,2010.Liliaceae.In:Marhold,K.(Ed.),IAPT/IOPB Chro-mosomeData.Taxon,vol.59,pp.1298–1302(inpress).
Horn,P.,1874.BeiträgezurKenntnissderZwiebelbildungenderGattungGagea Salisb.Arch.Ver.FreundeNaturgesch.Mecklenburg28,30–47.
Irmisch,T.,1850.DieGattungGageaSalisb.mitihrenverbreitetsteneinheimischen Arten.In:Irmisch,T.(Ed.),ZurMorphologiedermonokotylischenKnollen-und Zwiebelgewächse.Reimer,Berlin,pp.20–54.
Irmisch,T.,1863a.BeiträgezurvergleichendenMorphologiederPflanzen.Tulipa. Bot.Z23,177–181.
Irmisch,T.,1863b.BeiträgezurvergleichendenMorphologiederPflanzen.Abh. Naturforsch.Ges.Halle7,175–227.
Jäger,E.J., 1973. ZurVerbreitungund Lebensgeschichte derWildtulpe (Tulipa sylvestrisL)undBemerkungenzurChorologiederGattungTulipaL.Hercynia N.F10,429–448.
Kawano,S.,Nagai,Y.,1975.Theproductiveandreproductivebiologyofflowering plants.I.LifehistorystrategiesofthreeAlliumspeciesinJapan.Bot.Mag.(Tokyo) 88,281–318.
Levichev,I.G.,1999.ZurMorphologieinderGattungGageaSalisb.(Liliaceae).I.Die unterirdischenOrgane.Flora194,379–392.
Levichev,I.G.,2006.FournewspeciesofthegenusGageaSalisb(Liliaceae)from WesternHimalayasandtheadjoiningregions.Pak.J.Bot.38,47–54.
Levichev,I.G.,2013.StructuralfeaturesofshootsinLloydia,Gagea,Kharkevichia (Liliaceae)asevolutionaryvariabilityofthemodulesofmesomenaturein mono-cotyledons.Bot.Zh.(St.Petersburg)98,409–452(inRussian).
Levichev,I.G.,Jezniakowsky,S.A.,2008.HistoriaGagearum,http://www.binran.ru/ infsys/gagea/index-eng.html(accessed20.01.12).
Martinovsk ´y,J.O.,1930.BeitragzurMorphologie,Phylogenesisund Entwicklungs-geschichtederGattungGageaSalisb.Beih.Bot.Cbl.,1.Abt.Anat.,Morph.Physiol. Pfl.46,435–459.
Mathew,B.,1996. AReviewofAlliumSectionAllium. Royal BotanicGarden, Kew.
Peruzzi,L.,2003.ContributiontothecytotaxonomicalknowledgeofGageaSalisb. (Liliaceae)sect.FoliataeA.Terracc.andsynthesisofkaryologicaldata.Caryologia 56,115–128.
Peruzzi,L.,2008a.Contributiontothecytotaxonomicalknowledgeofthegenus GageaSalisb.(Liliaceae).III.Newkaryologicaldatafromthecentral Mediter-raneanarea.Caryologia61,92–106.
Peruzzi,L.,2008b.HybridityasamainevolutionaryforceinthegenusGageaSalisb (Liliaceae).PlantBiosyst.142,179–184.
Peruzzi,L.,2012a.ChromosomediversityandevolutioninGagea(Liliaceae). Boc-conea24,147–158.
Peruzzi,L.,2012b.NomenclaturalnoveltiesatsectionallevelinGagea(Liliaceae). AttiSoc.Tosc.Sci.Nat,Mem.ser.B118,23–24.
Peruzzi,L.,Peterson,A.,Tison,J.-M.,Peterson,J.,2008.Phylogeneticrelationships ofGageaSalisb.(Liliaceae)inItalyinferredfrommolecularandmorphological. PlantSyst.Evol.276,219–234.
Peruzzi,L.,Peterson,A.,Tison,J.-M.,Harpke,D.,2011.Newlightonphylogenyand taxonomyoftheEurasianGageavillosa–Gfragiferacomplex(Liliaceae).Nordic J.Bot.29,722–733.
Peterson,A.,Levichev,I.G.,Peterson,J.,2008.SystematicsofGageaandLloydia (Liliaceae)andinfragenericclassificationofGageabasedonmolecularand mor-phologicaldata.Mol.Phyl.Evol.46,446–465.
Peterson,A.,Harpke,D.,Peruzzi,L.,Levichev,I.G.,Tison,J.-M.,Peterson,J.,2009.
Hybridizationdrives speciationin Gagea (Liliaceae).Plant Syst. Evol. 278, 133–148.
Peterson,A.,Levichev,I.G.,Peterson,J.,Harpke,D.,Schnittler,M.,2011.Newinsights intophylogenyandtaxonomyofGagea(Liliaceae)inWesternChinaRecent speciationthroughhybridization.Organ.Div.Evol.11,387–407.
Pfeiffer,T.,Klahr,A.,Heinrich,A.,Schnittler,M.,2011.Doessexmakesa differ-ence?Genotypicdiversityandspatialgeneticstructureintwoco-occurring Gageaspecieswithcontrastingreproductivestrategies(G.lutea,G.spathacea, Liliaceae).PlantSyst.Evol.292,189–201.
Pfeiffer,T.,Klahr,A.,Peterson,A.,Levichev,I.,Schnittler,M.,2012a.Nosexatall? ExtremelylowgeneticdiversityinGageaspathacea(Liliaceae)acrossEurope. Flora207,372–378.
Pfeiffer,T.,Harter,D.E.V.,Formella,N.,Schnittler,M.,2012b.Reproductive iso-lationvsinterbreedingbetweenGagealutea(L.)KerGawl.andG.pratensis (Pers.)Dumort.(Liliaceae)andtheirputativehybridsinMecklenburg-Western Pomerania(Germany).PlantSpeciesBiol.28,193–203.
Ramsey,J.,Schemske,D.W.,2002.Neopolyploidyinfloweringplants.Annu.Rev. Ecol.Syst.33,589–639.
Ronsheim,M.L.,Bever,J.D.,2000.Geneticvariationandevolutionarytrade-offsfor sexualandasexualreproductivemodesinAlliumvineale(Liliaceae).Am.J.Bot. 87,1769–1777.
Schnittler,M.,Pfeiffer,T.,Harter,D.E.V.,Hamann,A.,2009.Bulbilscontraseeds. ReproductiveinvestmentintwospeciesofGagea(Liliaceae).PlantSyst.Evol. 279,29–40.
Takahashi,H.,Tani,T.,1997.LifehistoryofthespringephemeralGagealutea (Lili-aceae)inSapporo,Hokkaido.Miyabea3,17–26.
Tison,J.-M.,Peterson,A.,Harpke,D.,Peruzzi,L.,2013.Reticulateevolutionofthe criticalMediterraneanGageasectDidymobulbos(Liliaceae)anditstaxonomic implications.PlantSyst.Evol.299,413–438.
Wang,C-N.,Cronk,Q.C.B.,2003.MeristemfateandbulbilformationinTitanotrichum (Gesneriaceae).Am.J.Bot.90,1696–1707.
Zarrei,M.,Wilkin,P.,Ingrouille,M.J.,Chase,M.W.,2010.Gagearobusta(Liliaceae)a newspeciesfromFloraIranicaarea.KewBull.65,327–336.
