全 文 ：广 西 植 物 Guihaia Aug．２０１５,３５(４):５５６－５６３ http://journal．gxzw．gxib．cn
DOI:１０．１１９３１/guihaia．gxzw２０１４０３０１７
杨翠凤,张春雅,杨丽涛,等．不同基因型甘蔗种质资源的表型遗传多样性[J]．广西植物,２０１５,３５(４):５５６－５６３
YangCF,ZhangCY,YangLT,etal．Phenotypicandgeneticdiversityofsugarcane(SaccharumL．)protospeciesandlandraces[J]．Guihaia,２０１５,３５(４):
５５６－５６３
Phenotypicandgeneticdiversityofsugarcane
(SaccharumL．)protospeciesandlandraces
YANGCuiＧFeng１,ZHANGChunＧYa１,YANGLiＧTao１,２,LIYangＧRui２∗
(１．AgriculturalCollege/StateKeyLaboratoryofConservationandUtilizationofSubtropicalAgroＧbioresources,GuangxiUniversity,
Nanning５３０００５,China;２．SugarcaneResearchCenter,ChineseAcademyofAgriculturalSciences/KeyLaboratoryofSugarcane
BiotechnologyandGeneticImprovement(Guangxi),MinistryofAgriculture/GuangxiAcademyofAgricultural
Sciences/GuangxiKeyLaboratoryofSugarcaneGeneticImprovement,Nanning５３０００７,China)
Abstract:Inordertoexploregeneticdiversityandrelationshipamongdiferentsugarcanespecies,selectexcelentsugＧ
arcanegermplasmandparentalcombinationforhybridization,thepresentstudywasconductedforgeneticdiversityaＧ
nalysisusing１８genotypesofprotospeciesandlandracesofsugarcane(Saccharumspp．)with１４quantitativetraits．
Theresultsshowedthatthephenotypicandgeneticdiversityofsugarcanewashighlysignificantinsucrose％cane,
stalkweight,leafwidth,stalkdiameterandfiber％canebycoeficientofvariationandcorrelationanalysisof１４
quantitativetraits．Fourprincipalcomponentsofquantitativetraitsnamedqualityfactor,growthfactor,maturityfacＧ
torandphotosyntheticfactor,respectively,wereextractedfromprincipalcomponentanalysis,andthecumulativeconＧ
tributionrateoffourprincipalcomponentswas８３．４８％．Tendesirablegenotypesofsugarcanehadhighervaluesof
quantitativetraitsthantheaveragesinfactoranalysisinturnasfolows:Sampana→Tiangeda→HeqingcaoGanzhe→
GuilinZhuzhe→Tanzania→Mango→Guzhizhe→Dadaozailai→Tuojianghong→Chunnee．The１８accessionscouldbe
dividedintofiveclustersbasedonthegeneticdistance,andpotentialhybridmaybedevelopedbycrossingSampana
withTiangedaand/orSampanawithHeqingcaoGanzhe．ItsuggestedthatweshouldnotonlyconsiderthecompleＧ
mentaryofmaintraits,butalsogeneticdistancewhenweselectparentsforcrossinginsugarcanebreeding．ThepresＧ
entstudyprovedthatfactoranalysiswasgoodtohelptheparentalselectionanddeterminecrossingcombinationsin
sugarcanebreedingprogram．
Keywords:sugarcane;quantitativetraits;geneticdiversity;factoranalysis
CLCnumber:Q９４４,Q９４５．４ Documentcode:A ArticleID:１０００Ｇ３１４２(２０１５)０４Ｇ０５５６Ｇ０８
不同基因型甘蔗种质资源的表型遗传多样性
杨翠凤１,张春雅１,杨丽涛１,２,李杨瑞２∗
(１．广西大学农学院/亚热带农业生物资源保护与利用国家重点实验室,南宁５３０００５;２．中国农业科学院甘蔗研究中心/
农业部广西甘蔗生物技术与遗传改良重点实验室/广西农业科学院/广西甘蔗遗传改良重点实验室,南宁５３０００７)
摘 要:为探明甘蔗原种和地方种的遗传多样性和亲缘关系,以期筛选出优良甘蔗种质和优良杂交亲本.该研
究对１８份甘蔗原种和地方种的１４个数量性状进行了表型遗传多样性分析.结果表明:通过１４个数量性状的
收稿日期:２０１４Ｇ０８Ｇ２０ 修回日期:２０１４Ｇ１１Ｇ１０
基金项目:国家“８６３”计划课题(２０１３AA１０２６０４);国家国际合作项目(２００９DFA３０８２０,０S２０１３ZR０１３０);广西自然科学基金(２０１１GXNSFF０１８００２);广
西科学研究与技术开发计划项目(桂科产１１２３００８Ｇ１);广西农科院团队项目(桂农科２０１１YT０１).
作者简介:杨翠凤,女,博士研究生,广西桂林人,研究方向为甘蔗生理生化与分子生物学,(EＧmail)yangcuifeng１９８８＠１６３．com.
∗通讯作者:李杨瑞,教授,主要从事甘蔗研究,(EＧmail)liyr＠gxaas．net.
变异系数(coeficientofvariance,CV)和性状之间的相关分析,１８份甘蔗原种和地方种的数量性状遗传变异主要
来自甘蔗蔗糖分、单茎重、叶宽、茎径和纤维分;对１４个数量性状进行主成分分析提取获得了４个主成分因子,
分别命名为“品质因子”、“生长因子”、“成熟度因子”和“光合因子”,主成分因子累积贡献率达８３．４８２％;进一步
通过对主成分因子开展综合评价分析,获得数量性状综合表型高于平均水平的１０份材料,依次为Sampana→甜
圪塔→合庆草甘蔗→桂林竹蔗→坦桑尼亚→芒戈→古芝蔗→大岛再来→托江红→春尼;聚类分析基于不同的遗
传距离可将１８份种质聚为５个类别,潜在的优良杂交组合是Sampana和甜圪塔或Sampana和合庆草甘蔗,表明
在甘蔗遗传育种亲本选择上既要考虑各性状主要因子的互补,又要保持一定的遗传距离.该研究认为,在甘蔗
育种工作中,利用因子分析法进行表型遗传多样性分析,将更加有助于亲本和杂交组合的选择.
关键词:甘蔗;数量性状;遗传多样性;因子分析
Sugarcaneisanimportantcashcropofthetropics
andsubtropics,andisgrowninmorethan９０countries
between４０°Nand３２°５′S,encompassinghalfofthe
globe(Waclawovskyetal．,２０１０;Silvaetal．,２０１２)．
Itprovides６０％－７０％oftheworld’ssugar(Prabuet
al．,２０１１;Silvaetal．,２０１２)andproducesnumerous
valuablebyＧproductslikeethanolusedasafuel,alcohol
usedinpharmaceuticalindustry,bagasseusedinpaper
industryandaquaculture,andpressmudusedasarich
sourceoforganicnutrientsforcropproduction (Li
２０１０;Yangetal．,２０１２;AlonsoＧPippoetal．,２０１３)．It
hasbeendemonstratedthatmodernsugarcanevarieties
arehighlycomplexpolyploidsderivedfromafewcrosＧ
sescarriedoutmorethanacenturyagobetweenthe
highsugarcontentspeciesSaccharumofficinarum (x
＝１０,２n＝８x＝８０)andthewildspeciesS．spontaneＧ
um (x＝８,２n＝５x－１６x＝４０－１２８)andbackcrossed
withS．officinarum (Bhatetal．,１９８５;D’Hontet
al．,１９９６,２００２;D’Hont２００５)．However,previous
studieshavealsoshowedthatsugarcanebelongstothe
genusSaccharumL．composedofhybridsderivedfrom
S．officinarum (Nobleclones),S．sinense (Chinese
clones),S．barberi (NorthIndianclones)andS．
spontaneum (Haetal．,１９９９;Irvine１９９９;Brownet
al．,２００２;Selvietal．,２００３;Cuadradoetal．,２００４)．At
presentnotmorethan１５varietiesofS．officinarum
specieshavebeenusedasparents,andonly２genotypes
ofS．spontaneum andS．barberihavebeenutilized
(Li,２０１０)．Consequently,mostofthesugarcanevarieＧ
tiesaroundtheworldhavethesameancestors．The
factthatonlyafewinnovativeparentshavebeenused
insugarcanebreedingleadstohighgeneticsimilarityin
commercialsugarcanecultivarsandthegeneticbasis
tendedtonarrow(Selvietal．,２００３;Cordeiroetal．,
２００３;Hemaprabhaetal．,２００６;Lavanyaetal．,２０１０)．
ParentselectionisthefirststepinplantbreedingproＧ
gramthroughhybridization,soitisurgenttointroduce
protospeciesorlandracestobroadenthegeneticbase
andthenimprovethepotentialtraitsofmodernsugarＧ
canevarieties(Salemetal．,２００８)．
Morphologicaltraitsarevaluabletoolstoidentify
genotypesastheyreflectexternalresemblancemore
thangeneticrelatedness(Demeyetal．,２００３;Pereraet
al．,２０１２;Ahmedetal．,２０１２)．Morphologicaltraits
arephenotypiccharacteristicsofgeneticvariations,and
thereareaseriesofcomplexaspectslikegeneexpresＧ
sionandregulation,andontogenyintheprocessfrom
genotypetophenotype(Lietal．,２０１０;Liuetal．,
２０１０)．TheregulationofgeneticvariationcanbereＧ
vealedbyreasonablesamplingandadoptingefective
mathematicalstatisticalmethods(Liuetal．,２０１０;
Hussainetal．,２０１０;Pereraetal．,２０１２)．ManyreＧ
searchershavestudiedthemorphologicalgeneticdiverＧ
sityincropssuchassorghum,wheat,rice(DurrishahＧ
waretal．,２０１２;Khodadadietal．,２０１１;Rampantet
al．,２０１１)．Principalcomponentanalysis(PCA)and
clusteranalysisasthemainmeanshavebeenwidely
usedindiversityanalysisandsatisfactorytraitsselecＧ
tion(Ahmedetal．,２０１０;Iiyasetal．,２０１０;Smiulahet
al．,２０１３)．
Mostofthereportsonmorphologicalandgenetic
diversityofsugarcanehaveaimedatthecommercial
varietiesinoneparticularregionorcountry(Wagihet
al．,２００４;Sajadetal．,２００９;Pereraetal．,２０１２),while
referencesonstudyingprotospeciesorlandracesare
verylimited．Inthepresentstudy,１８sugarcaneprotoＧ
speciesorlandraceswereemployedtoinvestigatethe
geneticdiversitybasedon１４quantitativetraitsbyuＧ
７５５４期 杨翠凤等:不同基因型甘蔗种质资源的表型遗传多样性
singPCA,factoranalysisandclusteranalysismethods
toprovideareferenceforappropriateparentselection
insugarcanebreeding．
１ MaterialsandMethods
１．１Plantmaterials
Eighteensugarcaneaccessionsofprotospeciesand
landracesconservedinSugarcaneResearchCenter,ChiＧ
neseAcademyofAgriculturalSciences/SugarcaneReＧ
searchInstitute,GuangxiAcademyofAgriculturalSciＧ
ences,including７sugarcaneprotospeciesnamelyManＧ
go(S．barberi),Chunnee(S．barberi),Loethers(S．
officinarum),Robustum (S．robustum),Heqingcao
Ganzhe (S．sinense),Luzhe (S．sinense),Guilin
Zhuzhe(S．sinense)and１１landracesnamelySampaＧ
na,Jiagudao,Tanzania,Tuojianghong,HekouQingpi,
Luoguzhe,Guzhizhe,Senbaina,Dadaozailai,Tiangeda,
HongpiS１７,wereselectedforexperiment．
TheexperimentwascarriedoutattheSugarcane
ResearchInstitute,GuangxiUniversity,Nanning,China
(２２°１３′－２３°３２′Nand１０７°４５′－１０８°５１′E)during
January２０１１andDecember２０１２．Theexperimentwas
laidoutinarandomizedblockdesignwith３replicaＧ
tions．Eachvarietywasaccommodatedinaplothaving
３rowsof５minlengthwithspacing１．２mbetween
rows．ThefieldmanagementwasdoneasforcommerＧ
cialsugarcaneproduction．Observationswererecorded
for１４agronomictraitsviz．leaflength,leafwidth,leaf
length/width,plant height,stalk diameter,stalk
weight,fieldbrix,juicebrix(aftermiling),gravitypuＧ
rity,juiceyield,reducingsugar％juice,sucrose ％
cane,sucrose％juiceandfiber％cane．ThestandardiＧ
zationofdatawasaccordingtoCaietal．(２００６)．
１．２Statisticalanalysis
Coeficientofvariationwascalculatedaccordingto
folowingformula:CV＝
SD
X ×１００
WhereCV＝Coeficientofvariation;SD＝StandＧ
arddeviation;X＝Mean．
TheoriginaldatawereorganizedbyusingExcel
program,andthenanalyzedbyusingSPSS(１８．０)for
ANOVA (AnalysisofVariance),correlationanalysis,
principalcomponentanalysis(PCA),factoranalysis
andclusteranalysis(Steeletal．,１９８０;Xue,２００６)．
２ ResultsandAnalysis
２．１Analysisoncoefficientofvariation
CoeficientofvariationforquantitativetraitisaＧ
doptedtoreflectthediscretedegreeandphenotypicdiＧ
versity．Thecoeficientsofvariationforthe１４quantiＧ
tativetraitsinthepresentstudywereshowedinTable
１．Thehighestvaluewasrecordedforfiber％cane
(３４．０７％),folowedbystalkweight,stalkdiameter,
sucrose％juice,sucrose％cane．Plantheightshowed
relativelylowcoeficientofvariation,andjuiceyield
thelowest．Thisindicatedthattheheritablevariation
ofquantitativetraitsmainlycamefromfiber％cane,
stalkweight,stalkdiameterandsucrosecontent．This
viewwasconsistentwithChaudhary(２００２),Sajadet
al．(２００９),andAhmedetal．(２０１２)．Sajadetal．
(２００９)foundthatfiber％caneisthehighestcontribＧ
utortogeneticdiversity,however,whichwasopposite
totheresultsreportedbyAhmedetal．(２０１２)that
canefiber(％)showedlowgeneticadvanceandheritaＧ
bilityvalues,indicatingdificultyofimprovementof
thistraitbythemeansofdirectselection．ThediferＧ
encesmaybeduetodiferentgenotypesusedandexＧ
perimentalconditions．Inthepresentstudy,plant
heightshowedalowcoeficientofvariation．Sajadet
al．(２００９)andAhmedetal．(２０１２)alsoreportedsimiＧ
larresultsintheirstudies．
２．２Correlationanalysis
Correlationanalysisamongphenotypictraitsmay
reflectbiologicalprocessesthatareofconsiderableevoＧ
lutionaryinterest,correlationcanbetheresultofgeＧ
netic,functionalandphysiologicalorevendevelopmenＧ
talcharacters(Wagneretal．,２０００;Soomroetal．,
２００６)．
ThedatainTable２revealedthatthecorrelations
ofleafwidthwithfieldbrix,juicebrix,sucrose％juice
andsucrose％canewerepositivelysignificantat０．０５
level,andthatwithstalkweightwaspositivelyand
highlysignificantat０．０１level．Stalk weightwas
positivelyandhighlysignificantlycorrelatedwithleaf
８５５ 广 西 植 物 ３５卷
Table１ Geneticdiversityevaluationonquantitativetraits
Trait Minimum Maximum Variationamplitude Mean Standarddeviation Coeficientofvariation(％)
Leaflength(cm) １０６．９ １８３．９３ ７７．０３ １５３．６３ １９．６２ １２．７７
Leafwidth(cm) ３．８１ ６．６９ ２．８８ ５．１３ ０．７７ １５．０１
Leaflength/width ２１．７３ ４０．６９ １８．９６ ３０．４０ ５．０４ １６．５７
Plantheight(cm) ２３４．１ ３４４ １０９．９ ２９９．８８ ２８．１５ ９．３９
Stalkdiameter(mm) ９．６６ ２９．３７ １９．７１ １８．０２ ５．１７ ２８．７１
Stalkweight(kg) ０．４９ ２．３ １．８１ １．４４ ０．４５ ３１．０９
Fieldbrix(％) １２．０７ ２０．０３ ７．９６ １６．０６ ２．８８ １７．９１
Canebrix(％) ９．７２ １７．１４ ７．４２ １４．４９ １．９０ １３．１３
Juiceyield(％) ５３．３８ ７７．１４ ２３．７６ ７０．６０ ４．７３ ６．７
Juicegravitypurity(％) ５３．０４ ８９．１２ ３６．０８ ７９．０３ ９．７９ １２．３９
Fiber％cane １．２４ ３．５８ ２．３４ ２．１０ ０．７１ ３４．０７
Reducingsugar％juice ０．１４ ２．６７ ２．５３ ０．８７ ０．１５ １７．２４
Sucrose％juice ７．４ １７．４２ １０．０２ １３．０５ ２．７４ ２１．０２
Sucrose％cane ５．８３ １４．６９ ８．８６ １１．１５ ２．３０ ２０．６５
Table２ Correlationanalysisamongquantitativetraits
A B C D E F G H I J K L M N
A １
B ０．３３２ １
C ０．４９３∗Ｇ０．６３７∗∗ １
D Ｇ０．０４４ ０．０１６ Ｇ０．０４ １
E ０．０４８ ０．２３ Ｇ０．０８７ ０．４ １
F ０．３０６ ０．６７９∗∗ Ｇ０．３６２ ０．５５５∗ ０．５８９∗ １
G ０．５８６∗ ０．５５７∗ Ｇ０．０２５ Ｇ０．１８７ ０．１３２ ０．３０１ １
H ０．０７９ ０．４７７∗ Ｇ０．４２３ ０．１９３ ０．１０１ ０．５６３∗ ０．４４６ １
I Ｇ０．２８９ ０．２５４ Ｇ０．４７９∗ ０．３７４ ０．３５１ ．６４２∗∗ Ｇ０．２６８ ０．５３５∗ １
J Ｇ０．１１３ Ｇ０．０７５ Ｇ０．０３６ Ｇ０．３６４ Ｇ０．３９５ Ｇ０．４１１ Ｇ０．２９２ Ｇ０．４６７ Ｇ０．３ １
K ０．２３３ ０．４６８ Ｇ０．２４９ ０．０９５ ０．３７９ ０．６０７∗∗ ０．３１１ ０．５６２∗ ０．６５０∗∗Ｇ０．５１３∗ １
L ０．３６６ ０．５８０∗ Ｇ０．２７４ ０．１６４ ０．２８ ０．６７９∗∗ ０．５７７∗ ０．８４４∗∗ ０．５３１∗ Ｇ０．５５７∗０．８６３∗∗ １
M ０．２９２ ０．５５０∗ Ｇ０．３０９ ０．２ ０．２８ ０．６９１∗∗ ０．５３４∗ ０．８９２∗∗ ０．５８４∗ Ｇ０．５７６∗０．８３７∗∗０．９９２∗∗ １
N Ｇ０．３２１ Ｇ０．３１３ ０．１１５ ０．００８ ０．５１７∗ Ｇ０．２９２ Ｇ０．１８８Ｇ０．５９８∗∗ Ｇ０．３５３ ０．０５８ Ｇ０．３６３ Ｇ０．５４０∗ Ｇ０．５５３∗ １
Note:LettersAＧNrepresentleaflength,leafwidth,leaflength/width,plantheight,stalkdiameter,stalkweight,fieldbrix,canebrix,juiceyield,reducingsugar％juice,juice
gravitypurity,sucrose％juice,sucrose％cane,fiber％cane,respectively．∗significantat０．０５level; ∗∗highlysignificantat０．０１level．
width,juiceyield,gravitypurity,sucrose％juiceand
sucrose％cane,whilepositivelyandsignificantlycorＧ
relatedwithstalkdiameter,plantheightandjuicebrix．
Juicebrixwassignificantlyandpositivelycorrelated
withleafwidth,stalkweight,juiceyield,gravitypuriＧ
ty,sucrose％juiceandsucrose％canebutwassignifＧ
icantlyandnegativelycorrelatedwithfiber％ cane．
Juiceyieldwassignificantlyandpositivelycorrelated
withstalkweight,gravitypurity,juicebrix,sucrose％
juiceandsucrose％canewhilesignificantlyandnegaＧ
tivelywithleaflengthandleafwidth．Gravitypurity
waspositivelyandsignificantlycorrelatedwithstalk
weight,juiceyield,jucebrix,sucrose％juiceandsuＧ
crose％canewhilenegativelyandsignificantlywith
reducingsugar％juice．Sucrose％canewaspositively
andsignificantlycorrelated withleaf width,stalk
weight,fieldbrix,juicebrix,juiceyieldandgravitypuＧ
ritybutwasnegativelyandsignificantlycorrelatedwith
reducingsugar％juiceandfiber％cane．Thevariation
ofsucrose％juicewasinaccordancewiththatofsuＧ
crose％cane．Fiber％canewassignificantlyandnegＧ
ativelycorrelatedwithcanebrix,sucrose％juiceand
sucrose％cane．Reducingsugar％juicewassignifiＧ
cantlyandnegativelycorrelatedwithgravitypurity,suＧ
crose％juiceandsucrose％cane．
Basedonthepresentstudy,sucrosecontent,stalk
９５５４期 杨翠凤等:不同基因型甘蔗种质资源的表型遗传多样性
weight,leafwidthandbrixaredesiredselectionparamＧ
eterstoevaluatetheagronomicperformanceofthe
sugarcanevarieties,folowedbyjuiceyield,gravitypuＧ
rity,fiberandreducingsugar．Khanetal．,(２００７,
２０１２),Mahmoodetal．,(１９９０),Ramdoyal(１９９１)and
Soomroetal．,(２００６)alsoreportedsimilarresultsin
sugarcane．
２．３Principalcomponentanalysis
Theeigenvalueofcorrelation matrix,variance
contributionrateandprincipalfactorswereobtained
viaprincipalcomponentanalysis(Table３)．Basedon
theprincipleofeigenvaluesgreaterthan１,fourindeＧ
pendentfactorswereextractedfromthe１４quantitative
traits,whichexplained８３．４８％ofthetotalvariationin
thedependencestructure．Factorloadingsthatgreater
than０．５wereconsideredimportant．Asummaryofthe
compositionofvariablesofthefourextractedfactors
withloadingaregiveninTable４．
FactorIincludedsixvariableswhichaccounted
for４４．１６％ ofthetotalvariation．Thesixvariables
weresucrose％cane,sucrose％juice,juicebrix,fiber
％cane,gravitypurityandjuiceyield．Thesignofthe
loadingvaluesindicatesthedirectionoftherelationship
betweenthefactoranditsrelatedcharacters．ThenegＧ
ativesignoffiber％caneindicatedthenegativecorreＧ
lationcoeficientswitheachoftheothervariablesin
factorI(Table４)．Eachvariablehadahighloadingin
thefactorandalofthemwereassociatedwithcane
quality,therefore,thisprincipalcomponentcouldbe
caledqualityfactor．
FactorIIwasresponsiblefor１６．４１％ofthetotal
variationinthedependencestructure．Itincludedfour
characters,namelystalkdiameter,plantheight,stalk
weightandreducingsugar％juice．ThehighestloadＧ
ingvalueinthefactorisstalkdiameter,folowedby
plantheight,stalkweightandreducingsugar％juice
inturn．Thevariablesofthisprincipalcomponentwere
regardedasgrowthfactor．TheseresultswereinagreeＧ
mentwiththefindingofElＧGeddawietal．,(１９９２)and
Husseinetal．(２０１２)．
FactorIIIwasmadeupoffieldbrixandleaf
length．Themaximumloadingvaluewasfieldbrix,
whichwasslightlyhigherthanthesecond,leaflength．
AsthevariablesoffieldbrixconcernedwithcanemaＧ
turity,itwasreferredtoasmaturityfactor．ItaccounＧ
tedfor１３．１４％ofthetotalvariationinthedependence
structure．
FactorIVcontainedtwovariables,namelyleaf
lengthandleafwidth．BothvariableshadahighloadＧ
inginthefactor．FortheleafisthekeyorganforphoＧ
tosynthesisanditplaysanimportantroleonplant
growthandmetabolism．Thisprincipalcomponentwas
feasiblycaledphotosyntheticfactor．
Principalcomponentanalysisbasedondiferent
agronomictraitswouldgetdiferentresults．Many
componentanalyseshavebeenperformedforcomplex
traitsbasedonmorphologicalandphysiologicalcharacＧ
terizations(Petrasovitsetal．,２００７;Khanetal．,２０１２)．
Forsugarcane,previousstudiestendedtofocusoncane
yieldcomponents (Husseinetal．,２０１２;Smiulah,
２０１３),whilethisstudyliedemphasisoncaneitsown
traitsinordertoexplorethegoodtraitsofcaneitself
andrevealedthathighersucrosecontentandstalk
weight,endowedwithhigherfiber％cane,gravitypuＧ
rity,leaflengthandleafwidthweretheimportant
characterswhichshouldbeconsideredforsugarcane
genotypeselection．
Takingadvantageofprincipalcomponentanalysis
byplantbreedershasthepotentialofincreasingthe
comprehensionofthecasualrelationshipofvariables
andcanhelptodeterminethedesirabletraitstobeseＧ
lectedinabreedingprogram (Husseinetal．,２０１２;
Smiulah,２０１３)．
２．４Factoranalysis
Thefactoranalysisof１８accessionswasdoneby
utilizingtheComprehensiveEvaluationModelF＝a１F１
＋a２F２＋a３F３＋．．．＋amFm(Xue２００６),wherea１＝
４４．１５６％,a２＝１６．４１３％,a３＝１３．１４１％,a４＝９．７７２％,
obtainedfromTable３,andFmmeansfactorscore(“０”
isthemeanoffactorscoreandstandarddeviationis
“１”．PositivevaluesindicateaboveＧaverage,andnegaＧ
tivevaluesindicatebelowＧaverage)．Thefactorscoreof
eachaccessionwasusedinaboveequationforcalculatＧ
ingthecompositescoreFvalue(Table５)．TheperＧ
formancesofthe１８accessionsinphenotypictraits
wereevaluated,andrankedindescendingfromsuperior
０６５ 广 西 植 物 ３５卷
Table３ Eigenvaluesinprincipal
componentcorrelationmatrix
Principal
component
Characteristic
root
Variance
contribution
(％)
Cumulative
contribution
(％)
１ ６．１８２ ４４．１５６ ４４．１５６
２ ２．２９８ １６．４１３ ６０．５６９
３ １．８４０ １３．１４１ ７３．７１０
４ １．３６８ ９．７７２ ８３．４８２
５ ０．９５９ ６．８５２ ９０．３３４
６ ０．７１６ ５．１１２ ９５．４４６
７ ０．３１５ ２．２５３ ９７．６９９
８ ０．１９０ １．３５７ ９９．０５６
９ ０．０８７ ０．６２０ ９９．６７５
１０ ０．０２１ ０．１４９ ９９．８２４
１１ ０．０１４ ０．１００ ９９．９２４
１２ ０．００６ ０．０４５ ９９．９６９
１３ ０．００４ ０．０２９ ９９．９９８
１４ ０．０００ ０．００２ １００
Table４ Principalcomponentloadmatrix
Trait
Thefirst
principal
compoＧ
nent
ThesecＧ
ondprinＧ
cipal
compoＧ
nent
The
third
principal
compoＧ
nent
The
fourth
principal
compoＧ
nent
Sucrose％cane ０．８９３ ０．２４１ ０．２５０ ０．２０７
Sucrose％juice ０．８６７ ０．２３４ ０．３２９ ０．２０６
Canebrix ０．８５３ ０．０６１ ０．０７７ ０．２５７
Fiber％cane Ｇ０．８０８ ０．４８６ Ｇ０．０５３ ０．０４６
Juicegravitypurity ０．７３７ ０．３３４ ０．１５２ ０．１８７
Juiceyield ０．６６２ ０．３５０ Ｇ０．５１７ ０．２５１
Stalkdiameter Ｇ０．０４６ ０．９２７ ０．１１２ ０．２１６
Plantheight ０．２２６ ０．６４５ Ｇ０．２７８ Ｇ０．１５７
Stalkweight ０．５６６ ０．５８７ ０．１４６ ０．３６０
Reducingsugar％juce Ｇ０．４９９ Ｇ０．５５９ Ｇ０．１３０ ０．２３５
Fieldbrix ０．２６３ ０．０１７ ０．８５７ ０．２３６
Leaflength ０．２０５ ０．０１８ ０．８４６ Ｇ０．２４４
Leaflength/width Ｇ０．２３２ ０．０１０ ０．３３０ Ｇ０．８８６
Leafwidth ０．３６０ ０．０８７ ０．４０８ ０．７７４
toinferiorbasedontheFvaluesinTable５．
TengenotypesshowedpositiveFvalues,which
couldberegardedasexcelentgermplasminthisreＧ
search,andthelistswereSampana→Tiangeda→
HeqingcaoGanzhe→GuilinZhuzhe→Tanzania→ManＧ
go→Guzhizhe→Dadaozailai→Tuojianghong→Chunnee．
ELＧGeddwayetal．,(２００２)andArainetal．,(２０１１)
statedthatsugarcanevarietiesaregreatlyafectedby
geneticmakeＧup．AccordingtoKeerioetal．,(２００３),
unlessthegeneticpotentialitiesofavarietyarehigh,
mereprovisionsofgrowingconditionssuchasfertiliza
Table５ Factoranalysis
Species
Factorscore
F１ F２ F３ F４ F Ranking
Sampana １．２８１ ０．６０６ ０．８１６ Ｇ０．３２８ ０．７４ １
Tiangeda ０．８９４ Ｇ０．０３７ １．０９８ ０．８４２ ０．６１５ ２
HeqingcaoGanzhe １．１９ Ｇ０．７１ ０．５８ Ｇ０．１３３ ０．４７２ ３
GuilinZhuzhe ０．４７５ ０．６６３ ０．７７１ ０．１５９ ０．４３６ ４
Tanzania ０．４２４ １．２０３ ０．２７４ ０．０４２ ０．４２５ ５
Mungo ０．８８６ Ｇ０．６３１ ０．９５６ Ｇ０．４９４ ０．３６５ ６
Guzhizhe ０．９９７ Ｇ０．１０５Ｇ０．５６４ Ｇ０．５ ０．３ ７
Dadaozailai Ｇ０．８７１ ２．１３４ ０．１５８ ２．２０２ ０．２０２ ８
Tuojianghong ０．１３１ Ｇ１．５６９ ０．４０６ １．８０９ ０．０３ ９
Chunnee ０．２３２ ０．６２４ Ｇ０．３９４Ｇ１．４０７ ０．０１５ １０
Loethers Ｇ０．１１２ ０．２２ Ｇ０．２７６ ０．１９３ Ｇ０．０３１ １１
HongpiS１７ ０．３１４ Ｇ０．８５７Ｇ１．１７９Ｇ０．２７５Ｇ０．１８４ １２
Luoguzhe Ｇ０．３３８Ｇ０．１９７Ｇ０．６７３Ｇ０．１５３Ｇ０．２８５ １３
Robustum Ｇ０．５９７ １．２７ Ｇ０．３９９Ｇ１．９９９Ｇ０．３０３ １４
HekouQingpi Ｇ１．１１８ ０．６３１ Ｇ１．６７７ ０．６７６ Ｇ０．５４５ １５
Senbaina Ｇ０．６０６Ｇ１．２１９Ｇ０．８９１ ０．３８５ Ｇ０．５４７ １６
Luzhe Ｇ０．３９４Ｇ１．１８３Ｇ１．２８８Ｇ０．１３３ Ｇ０．５５ １７
Jiagudao Ｇ２．７８８Ｇ０．８４２ ２．２８１ Ｇ０．８８８Ｇ１．１５６ １８
tion,irrigationetc．wil notleadtoappreciableimＧ
provementincaneorsugaryield．Hence,thetopaccesＧ
sionsshouldbepaidmoreattentiontoinpractical．
２．５Clusteranalysis
In Q clusteranalysisthe１８genotypes were
groupedintofiveclustersA,B,C,DandEbasedon
theEuclideansquareddistanceof５Ｇ１０(Fig．１),while
clustersA,B,CandDgottogetherintoonegroupat
Euclideansquareddistanceof１５,andclusterEasan
independentgroup．Cluster A included Heqingcao
Ganzhe,Guzhizhe,Senbaina,Loethers,HongpiS１７and
Chunnee．ClusterBconsistsofTiangedaandMango．
Sixclonesviz．Tanzania,Robustum,Luoguzhe,Guilin
Zhuzhe,SampanaandDadaozailaifelintoClusterC．
TwogenotypesLuzheandHekouQingpicameinto
ClusterD．ClusterEconsistedoftwogenotypesTuoＧ
jianghongandJiagudao．
Clusteringanalysiscanclearlyrevealthegenetic
relationshipbetweendiferentaccessionsandprovide
thebasisforevaluationofscreening(Kashifetal．,
２００７;Sajadetal．,２００９)．Sajadetal．(２００９)suggesＧ
tedthatoneparentmayselectfromeachclusterto
produceahybridhavingdesirableresults．Itisnotwise
tocrossinintraＧclusterfortheaccessionsarenotgeＧ
１６５４期 杨翠凤等:不同基因型甘蔗种质资源的表型遗传多样性
neticalydiverseenoughtoproduceagoodhybrid．AcＧ
cordingtotheresultoffactoranalysis,theknowndeＧ
sirablegenotypesviz．Sampana,TiangedaandHeqingＧ
caoGanzhebelongtodiferentclusterssotheseaccesＧ
sionscanbecoupledtodevelopapotentialhybrid．
WhileotherwelＧbehavedgenotypeslikeTanzania,
GuilinZhuzheandSampanagatheredintoonegroup,
whichdenotedthatthesegenotypesaregeneticaly
moresimilarsotheyarenotsuggestedtobeusedtoＧ
gethertoproduceahybridcombination．ThemostdisＧ
tantclusterisE．Asthegenotypesinthisclusterare
notverygood,therefore,theyarealsonotsuggestedto
becrossedwithotherclusterstobreedasugarcanehyＧ
bridcombination．Kashifetal．(２００７)statedthegenoＧ
typeswithhighindexscoresandfalingintodiferent
clusterscanbecrossedtohavemaximumvariabilityof
goodcombinationsofcharacters．Similarly,ifoneisinＧ
terestedinimprovingaspecificcharacterwhichisunＧ
desirableorotherwiseweakongenotype．Thegenetic
divergenceinformationwilbehelpfultoselectparent
combinationsinsugarcanebreeding (Douleetal．,
１９９７;Kadianetal．,１９９７;Vermaetal．,２０００)．Inthis
study,apotentialhybridcombinationmaybedeveloped
bycrossingSampanawithTiangedaand/orSampana
withHeqingcaoGanzhe．
Fig．１DendrogrambasedonEuclideansquareddistance
(Qclusteranalysis) １８genotypesweregroupedintofiveclusＧ
tersA,B,C,DandEatEuclideansquareddistanceof５Ｇ１０,while
clustersA,B,CandDgottogetherintoonegroupatEuclidean
squareddistanceof１５,andClusterEbecameanindependentgroup．
３ Conclusions
Basedonthepresentstudy,coefficientofvariＧ
ation,correlationanalysis,principalcomponentaＧ
nalysis(PCA),factoranalysisandclusteranalysis
wereeffectiveforinvestigatingthe phenotypic
traitsindifferentgenotypesofsugarcane．EspecialＧ
lyfactoranalysiswasprovedtobehelpfulforpaＧ
rentalselectioninsugarcanebreeding．
Bycomprehensiveconsiderationoftheresults
ofcoefficientofvariation,correlationanalysisand
principalcomponentanalysis,thephenotypicand
geneticdiversityofsugarcanemainlycamefrom
sucrose％cane,stalkweight,leafwidth,stalkdiＧ
ameterandfiber ％ cane．Itinferredthatthese
traitscouldbeusedasdesiredselectionparameters
toevaluatetheagronomicperformanceofsugarＧ
canegermplasms．Fourprincipalcomponentsof１４
quantitativetraitswereobtained,viz．qualityfacＧ
tor,growthfactor,maturityfactorandphotosynＧ
theticfactor．
Combiningconsiderationoffactoranalysisand
clusteranalysismightbeeffectivetoselecttheadＧ
mirableparentsforhybridization．Severaldesirable
genotypeswereobtainedsuchasSampana,TiangeＧ
da,HeqingcaoGanzhe,GuilinZhuzhe,andpotential
hybridcombinationsmightbedevelopedbycrossＧ
ingSampanawithTiangedaand/orSampanawith
HeqingcaoGanzhe．Itisconcludedthatscreening
parentsnotonlyneedstoconsiderthecomplemenＧ
taryofmaintraits,butalsoneedstokeepcertain
geneticdistance．
References:
AlonsoＧPippoW,LuengoCA,AlberterisLAM,etal．２０１３．PractiＧ
calimplementationofliquidbiofuels:thetransferabilityofthe
Brazilianexperiences[J]．EnergyPolicy,６０:７０－８０
AhmedAO,ObeidA．２０１２．Investigationonvariability,broad
sensedheritabilityandgeneticadvanceinsugarcane(Saccharum
spp．)[J]．IntJAgricSci,２(９):８３９－８４４
ArainMY,PanhwarRN,GujarN,etal．２０１１．Evaluationofnew
candidatesugarcanevarietiesforsomequalitativeandquantitative
traitsunderThattaagroＧclimaticconditions[J]．Anim &Plant
Sci,２１(２):２２６－２３０
２６５ 广 西 植 物 ３５卷
AhmedAO,ObeidA,DafalahB．２０１０．TheinfluenceofcharＧ
actersassociationonbehaviorofsugarcanegenotypes(SacＧ
charumSpp)forcaneyieldandjuicequality[J]．WorldJ
AgricSci,６(２):２０７－２１１
BhatSR,GilBS．１９８５．Theimplicationof２negggametesinnobilＧ
isationandbreedingofsugarcane[J]．Euphytica,３４:３７７－３８４
BrownJS,SchnelRJ,TaiPYP,etal．２００２．Phenotypicevaluation
ofSaccharumbarberi,S．robustum,andS．sinensegermplasm
fromtheMiami,FL,USAworldcolection[J]．SugarCaneInＧ
tern,９:３－１６
CuadradoA,AcevedoR,MorenoDES,etal．２００４．GenomereＧ
modelinginthreemodernS．offcinarum,S．spontaneumsugarＧ
canecultivars[J]．JExpBot５５(３９８):８４７－８５４
CordieroGM,PanYB,HenryRJ．２００３．Sugarcanemicrosatelite
fortheassessmentofgeneticdiversityinsugarcaneGermplasm
[J]．PlantSci,１６５:１８１－１８９
CaiQ,FangYH．２００６．SugarcaneGermplasmResourceDescription
StandardandDataStandard[M]．Beijing:ChinaAgriculture
Press,７:７－４０
ChaudharyRR．２００１．GeneticvariabilityandheritabilityinsugarＧ
cane[J]．NepalAgricResJ,４&５:５６－５９
D’HontA,GrivetL,FeldmanP,etal．１９９６．Characterisationof
thedoublegenomestructureof modernsugarcanecultivars
(Saccharumspp．)bymolecularcytogenetics[J]．MolGenGenＧ
et,２５０:４０５－４１３
D’HontA,PauletF,GlaszmannJC．２００２．OligoclonalinterＧspecifＧ
icoriginof‘NorthIndian’and‘Chinese’sugarcanes[J]．Chrom
Res,１０:２５３－２６２
D’HontA．２００５．UnravelingthegenomestructureofpolyploidsuＧ
singFISHandGISH;examplesofsugarcaneandbanana[J]．
CytogeneGenRes,１０９:２７－３３
DemeyJR,ZambranoAY,FuenmayorF,etal．２００３．RelaciónenＧ
trecaracterizacionesmolecularymorfologicaenunacolecciónde
yuca[J]．Interciencia,２８(１２):６８４－６８９
DouleRB,BalasundaramN．１９９７．Variability,heritabilityandgeＧ
neticadvanceforyieldandqualityattributesinsugarcane[J]．In
Sugar,４７(７):４９９－５０２
Durrishahwar,Noor M,HidayatＧurＧRahman,etal．２０１２．
CharacterizationofsorghumgermplasmforvariousmorphoＧ
logicalandfodderyieldparameters[J]．AfrJBiotechnol,１１
(５６):１１９５２－１１９５９
ELＧGeddawayIH,DarweshDG,ElＧSherbinyAA,etal．２００２．
Effectofrowspacingandnumberofbuds/seedsettsongrowth
charactersofratooncropsforsomesugarcanevarieties[J]．Pak
SugarJ,１７:７－１４
ElＧGeddawiIH,NasrSM,AboＧDouhAM．１９９２．Factoranalysisof
yieldcomponentsofsugarcane[J]．PakSugarJ,３:１１－１４
FaivreＧRampantO,BruschiG,AbbruscatoP,etal．２０１１．AssessＧ
mentofgeneticdiversityinItalianricegermplasmrelatedtoagＧ
ronomictraitsandblastresistance(Magnaportheoryzae)[J]．
MolBreed,２７:２３３－２４６
HaS,MooreP,HeinzD,etal．１９９９．Quantitativechromosome
mapofthepolyploidSaccharumspontaneumbymulticolorfluoＧ
rescenceinsituhybridizationandimagingmethods[J]．Plant
MolBiol,３９:１１６５－１１７３
HemaprabhaG,NatarajanUS,BalasundaramN,etal．２００６．STMS
basedgeneticdivergenceamongcommonparentsanditsuseiniＧ
dentifyingproductivecrosscombinationsforvarietalevolutionin
sugarcane(Saccharumspp．)[J]．SugarcIntern,２４(６):２２－２７
HussainK,NisarMF,NawazK,etal．２０１０．Morphologicaltraits
vs．geneticdiversity:reliablebasisforsugarcanevarietiesidentifiＧ
cation[J]．TheBIOL (EＧJ．LifeSci),１(２):４１－４３
IrvineJE．１９９９．Saccharumspeciesashorticulturalclasses[J]．
Theor&ApplGenet,９８:１８６－１９４．
IiyasMK,KhanFA．２０１０．DeterminingsucroserecoveryinSaccaＧ
rumofficinarumL．usingregressionandcorrelationanalyses[J]．
AfrJBiotechnol,９(２０):２９８５－２９８８
KhanIA,BibiS,YasminS,etal．２０１２．CorrelationstudiesofagroＧ
nonmictraitsforhighersugaryieldinsugarcane[J]．PakJBot,
４４(３):９６９－９７１
KhanFA,IqbalMY,SultanM．２００７．Morphogeneticbehaviourof
someagronomictraitsofsugarcane(SaccharumofficinarumL．)
[J]．PakJAgricSci,４４(４):６００－６０３
KadianSP,ChanderK,SabharwalPS．１９９７．Geneticvariabilityand
heritabilityinsugarcane[J]．IndSugar,４６(１２):９７３－９７５
KashifM,KhanFA．２００７．Divergenceinsugarcane(Saccharum
officinarumL．)basedonyieldandqualitytraits[J]．PakJBot,
３９(５):１５５９－１５６３
KeerioHK,PanhwarRN,MemonYM,etal．２００３．Qualitativeand
quantitativeperformanceofsomepromisingandcommercialsugＧ
arcanevarietiesunderagroＧclimaticconditionsofThatta[J]．PaＧ
kistanJApplSci,３(１０－１２):６７０－６７３
LavanyaDL,HemaprabhaG．２０１０．AnalysisofgeneticdiversityaＧ
monghighsucrosegenotypesofsugarcane(Saccharum spp．)
derivedfrom CoC６７１usingsugarcanespecificmicrosatelite
markers[J]．ElectrJPlantBreed,１(４):３９９－４０６
LiL,LiuT,LiuB,etal．２０１０．NorthernTianshanArabidopsis
thaliananaturalpopulationphenotypicchangeenvironmentdeＧ
pendonthecharacteristics[J]．PlantActaEcolSin,３４(１１):
１３１０－１３３１
LiuXL,MaL,CaiQ,etal．２０１０．Yunnansugarcanevarieties
phenotypictraitsanalysisofgeneticdiversity[J]．JPlantGene
Resour,１１(６):７０３－７０８
LiuXL,MaL,CaiQ,etal．２０１０．Sugarcanevarietiesresources
phenotypicgeneticdiversity[J]．Biodiversity,１８(１):３７－４３
LiYR．２０１０．ModernSugarcaneScience[M]．Beijing:ChinaAgriＧ
culturePress,６:１－１２５
MahmoodT,NazirMS,AshfaqM,etal．１９９０．CorrelationinsugＧ
arcane[J]．JAgricRes,２８:３５９－３６３
MiransariM,HabibiM,NajiAM,etal．２０１１．Geneticdiversityof
wheat(Triticumaestivum L．)genotypesbasedonclusterand
principalcomponentanalysesforbreedingstrategies[J]．AustrJ
CropSci,５(１):１７－２４
PrabuG,KawarPG,PagariyaMC,etal．２０１１．Identificationof
waterdeficitstressupregulatedgenesinsugarcane[J]．Plant
MolBiolRep,２９:２９１－３０４
PereraMF,AriasME,CostilaD,etal．２０１２．Geneticdiversity
assessmentandgenotypeidentificationinsugarcanebasedon
DNAmarkersandmorphologicaltraits[J]．Euphytica,１８５
(１０):４９１－５１０
PetrasovitsLA,PurnelMP,NielsenLK,etal．２００７．Productionof
polyhydroxybutyrateinsugarcane[J]．PlantBiotechnolJ,５:
１６２－１７２
RamdoyalK．１９９１．GeneticcorrelationandrepeatabilityforagronoＧ
(下转第６０２页Continueonpage６０２)
３６５４期 杨翠凤等:不同基因型甘蔗种质资源的表型遗传多样性