t. nepolean dr. a. k. fazlullah khan - core
TRANSCRIPT
IDEN1'IFIC.ATION OF 0TL.s FOR YIE1.D .AND II'S C'OhlPONEYT TRAITS. .AND DOH'YY >III.DE\\' [Sclrrospora graminicolu (Sacc.) J . SchrBt.1
RESISTANCE IN PE.4RI. MIL.L.E'I' [P~nnisrrum gluurum (I..) R. Br.1
T. NEPOLEAN
Chairman Dr. A. K. FAZLULLAH KHAN Professor and tlead (Dcpannient ol.Millets) ('cntre lor Plant Hrecd~ng & ( i c n c ~ ~ c s 1 am11 Nadu Agricultural I J n ~ v r r s ~ t ) . Cormbatore- 64 1 003 Sam11 Nadu
C o - c h a ~ m a n Dr. C. TOM HASH l ' r~nc~pal Scienust. Pedrl M~llc t Hreedlng (~ loba l Theme-] Lntematlonal Crops Research lnsl~tute for Semi-And Trop~cs Patancheru. - 502 324 4ndhra Pradcsh
Thls study was formulated to ~mprove rhc yreld potential of hybnds of
1'1 '32.4.0, u hlch I S one ol' the ul~te arid imponant male-stenle llnes used In hyhnd
1)rcedlng programs In Tamll Nadu Ident~ficat~on ol' downy ~ n ~ l d e w resistance ycnomlc
reglons was also set as an add~ t~ona l object~be One hundred m d th~ny-srx F> denbed F4
self-bulks of a pearl m~l l e t mapplng populat~on (skeleton-mapped F! ~nd iv~dua l s ) dcnved
tiom PT 7328 x P 1449-2 were used as the has~c source populat~on for t h ~ s study.
P7' 1450, an elite poll~nator ~nbred was used for producrng tcstcross hyhr~ds for each o f
the 136 Fr self-bulks To rdentrfy the QTLs for y ~ e l d and 11s component traits, the
testcross hybnds were raised at two locat~ons In Tamil Nadu namely. a1 Tam11 Nadu
.Agricultural Un~ver s~ ty . Coimbalore and at Regional Research Sta t~on. Bhavanisagar
dunng October 2001. Disease resrstance screenrng was also conducted at these two
I o c ~ t ~ u ~ i s using sc1ir.d seeds ~)i [;; ,elf hulks Jurlny tkt0hr.r 0 0 1 t ' ~ ~ h r e ' ~ i 01 1.5 \ \C ' IC
~dcr i t~t ied iron1 he I\ \(> loc.it~orls lor nine .~g rono~ i i~c tralts uslng p l~ l i t I I C I ~ ~ ~ I . I I I I IC 10
5(1",) stlgnld cnicrgunce plant height togerhcr \rlth Itnit to 5 0 " a s[lgmd cniergmce as
predictors o f the renialnlng y~eld-related Iralts .\rnong these rilne traits. t1111c to 50""
stigma emergence. panlclc s~rcuml'erencc, plant hclght, panicle length ;und yr;lln ye ld per
wason reg~slered one Q7'1.. ~liousa~ld-gra111 riidss rcg~stcrcd 1 ~ 1 ~ O'l 1.s. graln yield per
day reg~stered three QTl-s ~ n d slngle-panicle grdlrl mass rey~stcrcd lour (Jl'l.s. Phe
.rcross-locations data set produced S I X QTLs for live traits ( ie~io~i i lc rcylons on L.(i 4 and
I.(; 7 controlled these tralts. For downy m~ldew resistance, tivc d~i ierent Q1.1-s were
~ i e ~ e c t e d on four linkage groups using d t s e s e lncldence percentage and arc-sln radians
\ d u e s . Of these two QTLs were ldentllied from the Coirnbatore data set on I-Ci 2. two
irorn the Rhavanlsagur data set on LC; 1 and 1-G 4 and one from the across-locations d a ~ a
set on L C 7 . Cira~n yield performance of hybrlds for 'Tam11 Nadu conditions can bc
inlprovcd by marker-ass~sted hack crosslny of these QT1.s rcglons tnto seed parent palr
I'T ' j 2 k B Marker-ass~sted transfer of resrstance QTLs arid py ra rn~d~ny of' reslslancc
Zenrs may Improve resistance to downy m ~ l d e w dlsease.
Dr. 4 . K . Fazlullah Khan. t'rofe\:;(>r And l iea~ ' , I ~ ~ T , I ~ I I I C ~ ! Ot \ IIIICI>~ < I'INi. I'NAI!
Dr. ('. 'Tom Hash. I ' r ~ n c ~ p ~ l I3rt.edt.r. Pearl \Illlet t3reedlng. I ( KIhA I
Dr. <'. Surendran. I)~rcctor. cI'H(i. I N h l l
Dr. N. Subbaraman. I'rofessor. RKC'. Kov~lankulan~
Dr. .A. Copalau. Proiesaor And I lead. Dep~.of Forages. ('PHt i. I N!\I '
Dr. i'. Bangarusam!. I'roiessor. I'NAI I
Dr. \I. (;ovindasarny. 1'roli.ssor. I'NAI i
Dr. \1. hlaheswaran. . \ss(x~ate t'rotssor. ('PMt3. rNAL1
Dr. .M. Kumar. Asslslant Professor. ('PRC;. 'fNAlI
Dr. C . Vanniarajan. r\ss~slant Professor. (.'PB(;. M A l l
Dr. Mcenatchi Geneson. Associale Professor. ('PHCi. INAI I
Dr. K. Kandasam?, I'rotkswr. 1'N,\c1
Dr. J . H. ('rouch, [lead. A(iL. ICRISAT
Dr. Maria Kolesnikova-Allen. Scientist, IClUSA'f
Mr. A.G. Rhaskar raj, Scientific Oficer. ICRISA'f
Mr. S. Sundar. Research scholar. INAlI
,Mr. V. Thiru\.cngadam. Senlor Research Fellow. ('I'MI3. I NAI'
Ms. H. Selvi. Research Scholar. INAII
Ms. K. Pushpam. Research Scholar. INAIJ
Mr. S. Senthilvel. Research scholar, INAIJ
Mr. P. Snthish Kumnr. Research scholar. mA1J
.Mr. Muthu, PG Scholar. M A l J
Mr. K. Kadirvel. Research scholar. 1YAIJ
Mr. S. Harikrishnan. Scient~fic officer, ICRISA'T
Ms. Aruna Rupakula. Research scholar. ICRISA'f
.Ms. Rupa. Research Associate. ICKISAT
Reglonal Research station. Bhavan~sagar
Zonal Research Centre, TNAU
ICRISAT. Patancheru. Andhra Pradcsh
John lnnes Centre. UK
(1.. Nepolean)
CONTENT
r i t l e
Inrroduct~on Re\ ~ c w ot I 11cralurc \ larrr~al \ .ind Z4rthoJs Kcsul~s L~lscussloll Summary Kcfcrences Ippen t i~k
- - -
Page No.
rille P a ~ e Y 0. -
:INOL'.A ior mdpplns p o p u l ~ t ~ o ~ i tcstcross Ii!hrlds lor ~l~ll'crctit tr;llts lront thc tr~al conducted at C'oirnbalore 1oc~11oli. 1001 2 0 0 2 50
.ISOL'.A ior Iiidpplny populdl~on testcross Ii>hrtd~ li)r ~ I~ l 'Sc t~~ i t tr311s liolll the 1rt3l conducted ;it Ijtiai anlsdgdr locattun. 2Uol -7002 5 I
ANOL'A ior mapping population tcstcross hybrids for dt(fi.rolr tr:llts lion1 moss- locat~ons pooled &la. 1001 2002 5 2
C'orrelallon matnx of rnapplng popular~on lestcross hybrlds dl ( olrnbatore. 55 Bhab anlsagar and across locat~ons. 2001 12002
:\NOVA for percentage and arcs~n-transformed \aiues for downy ~ n ~ l d e w lnc~dencr tiom tnals conducted at Cotmbatore, Hhavanlsagar and across-locat~ons. 2001 12002 5 7
()TL dssoclated w ~ t h gram yteld-detemi~n~ng trdlt5 of peal nllllet rndpplng progeny testcross hybrids uslng plant helght ds J pred~ctor of squde root-transforn~ed \ ~ I U C S of other tralts at Co~mbatorc 04
QI I. dssoc~ated w ~ t h gram y~eld-determ~n~ng trdlts ol pu r l ttllllel tndpplng progeny testcross hybnds uslny tlme lo 507% stlgrnd emergence ds a prcd~ctor of squdre root-translormed ralues of other tralts dt Cotrnbatore 0 5
QTL dssoclated H I I ~ yatn yteld-determ~n~ng crafts of pcdrl nllllet mapplng progeny testcross hybr~ds uslng plant hetghr together w ~ t h tlmc to 50'6 \tlgrna emergence as a pred~ctor of square root-translbrmed values of other trdtts at Co~rnbatore 65
QTL assoc~ated w ~ t h gram y~eld-determln~ng tralts of pearl m~llet mapplng progeny testcross hybnds using plant he~ght as d pred~ctor of log-transformed \slues of other tra~ts at Co~mbatore 68
QTL assoc~ated w ~ t h grdln yield-determ~n~ng tratls of pearl millet mdpplng progeny restcross hybrtds uslng tlme to 50% stlyma emergence as d prcdrctor of log-transformed values of other tratts at Co~mbatore 6 9
QTL assoc~ated w ~ t h gram y~eld-determ~n~ng tralts of pearl m~llet rnapplng progeny testcross hybnds uslng plant height together w ~ t h tlrne to 50% Sllgma emergence as a pred~ctor of log-transformed values of other tralts at Co~rnbatore Locat~on 69
. - -
Title Page Y 0.
QTL associated with $rain y~~ld-determin~ng traits ol' pearl n~illet nlepplny progenv testcross hybnds L I ~ I I I ~ plant height as a pred~ctor ul' square root-transformed \ alues of other traits at Rha\ anisagar 7 2
QTL assoelated uith goln yield-determin~ng traits of pearl in~llct mapping progeny testcross hybr~ds using time to 409:) stigma cmcrgence as a predictor of square root-transformed \dues ofothcr traits ;it Bhavanlsogar 7 3
QTL associated with gram y~eld-detcrmin~ng trdits of pearl ~riillet mdpping progeny testcross hybnds uslng plant he~ght together with tlme to 50% stlgma emergence as a predictor of square root-transformed values of other tralts dl Bhavanisagar 73
QTL associated with grain y~eld-determin~ng tralts of pearl m~llet napping
progeny testcross hybnds using plant height as a predictor ol log-transformed alues of other rralts dl Bhavan~sagar 7 0
QTL associated with grain-y~eld determin~ng traits of pearl millet mapping progeny testcross hybnds using tlme to 50% stigma emergence ds d predictor of log-transformed values of other tralts at Bhavanisagdr 7 7
QTL dssoc~ated w ~ t h grain weld-detenn~ning traits ol pearl m~llet mapping progeny testcross hybrlds uslng plant he~ght together w ~ t h tlnie to 50% stlgma emergence as a predictor of log-transformed values of other traits at Ehdvdnlsagar 77
QTL assoc~ated with gram yield-determining tralts of pearl m~llet mapping progeny testcross hybrids using plant height as a predictor of square-root transformed values of other tralts at across-locations 80
QTL associated with graln yield-determ~n~ng traits of pearl m~llet mapplng progeny testcross hybnds using time to 50% stigma emergence as a predictor of square root-transformed values of other traits at across-locations 8 1
QTL assoc~ated w ~ t h graln y~eld-determ~n~ng tralts of pearl millet mapplng progeny testcross hybnds using plant height as a predictor of log-transformed bdlues of other traits at across-locat~ons 84
QTL associated with grain yield-determining traits of pearl m~llet mapping
progeny testcross hybrids using time to 50% stigma emergence as a predictor of log-transformed values of other traits at across-locations 85
No. ... ~... - . .
Title . - ---- ~ ~.
QTL associated ~ l t h yralr~ yield-detern~ining lralts ol' pearl 1i111le1 ~iiapplng progeny testcross hybrtds using plant helgltt togetlter w~tli onie to SOU; sttgrna emergence as a predtctor of log-translomicd ialues of othcr traits at across-locations
QTL associated * ~ t h downy mtldex tnc~detice of pearl milist mapping progeny Fq self-bulls ustrig tl~sease reststance perccntagc and their arcs~~l- t ra~is lbrni~d balues at Co~mbatore
QTL ssociated W I I ~ downy mlldtlw inc~dence ot pearl ~ n ~ l l e t rliapplng progeny F5 self-bulks medn d~scase reslstance percentdge m d the~r arcs~n-transformed balues at Bhabanisdgar
QTL assoc~atedwlth downy m~ldew inc~dence of pearl rlilllet mapping progeny FS self-bulks using disease reslstance percentage and thelr arcsin-transformed ~ a l u e s at dcross-locauons
LOD scores and percentage of observed phenotypic vanance expla~ned by best QTL models for different types of data transformat~ons and traits used for regression for different locations data sets
P a ~ e No.
1.IS'I' OF FIGL'HES
Title - -- -- --
Page Yo.
RFLP-based generlc I~nkape niap of F, nlapptng populatton developed fiom thc cross PT 732B P 1.110-2 shoulng 1.6 1 and LC; 2 43
RF1.P-based Senetlc I~nkage map off:, mapping populatton de~elopcd i'roiii the cross PT 732B x I' IJJO-2 shouing 1.C; 3 and 1 Ci J 14
KFLP-based genetic ltnkage map o i Fa tnapping population ilcvclopcd front ihe cross PT 7328 x P 114')-2 shoutng LG 5. LC; 0, and l,(i 7 15
QTL LOD peaks for various tratts using plant heigbt and tinie to 50% stigma mergence together w ~ t h plant helght as predictors of log-tronsformcd values from Colmbatore yield tnal (16
QTL LOD peaks for varlous tratts uslng tlrne to 50" ; stlgma emergence as a predictor of square rool-transformed values from Coimbatore yteld trlal 06
Companson of QTL LOD peaks for vdnous tralts ustng dtfferent typr.i of predtctors of square root-transformed values from Colrnhatore yteld trtdl 67
QTL LOD peaks for various tralts using plant helght and ttnlc lo 50'6 stigma emergence together with plant height as predictors of log-transformed values from Coimbatore ylcld trial 70
QTL. LOD peaks for Lanous lrdits using ttme to 50% ctigrnd erncrycnce as o pred~ctor of logtransformed values from Colmbatore yleld trtdl 7 0
Comparison of QTL LOD peaks for various traits uslng different types of predictors of log-transformed values from Coimbatore yield tnal 7 1
QTL LOD peaks for vanous traits ustng plant helght and [!me to 50% sttgma emergence as a predictor of square root-transformed values from Bhavanlsagar yleld trial 74
QTL LOD peaks for varlous tratts using time to 50% stigma emergence as a predictor of square root-transformed values from Bhavanisagar yield trtal 74
Comparison of QTL LOD peaks for varlous tralts uslng dtfferent types of predictors of square root-transformed values from Bhavanisagar yteld tndl 7 5
- .~.. .. ~- . . . . -. . - - -
I'itle - . ~ . . . . - .. - . .
QTL LOD peaks b r idrlous rralts uslng plant hc~yht and time to 51)"" stlgnla emergence as a predictor of log-translbrnied \ alues from Blia\anlsdgar yeld rrlal
QTL LOD peaks for vanous traits using tlme to 50% stigma c?ieryence as a predlctor of log-transformed \alucs liom Bhavarllsagar yeld trlal
('ompar~son ot' QTL LOD peaks for vanous rra~ts using d ~ t ' k r e ~ ~ t types of pred~ctors of log-transl'im~ed ~ a l u e s fiorii Bhavalils~yar y~eld tr~al
QTL LOD peaks for tarlous tralts uslng plant he~yht as a predictor of square root-transformed values from across-locations
QTL LOD peaks for vanous tralts uslng trme to 50% stlgma emergence as a predictor of square root-transformed values from across-locat~ons
Companson of QTI. LO13 peaks for vanous tralts uslng different types of pred~ctors of square root-transformed values from across-locations
QTL LOD peaks for vanous tralts using plant helght and tlme to 50% stlgma emergence as a predlctor of log-transformed values from across-locat~ons
QTL LOD peaks lor varlous tralts uslng time to 5096 stigma emergence as a pred~ctor of log-transformed values from across-locat~ons
Compar~son of QTL LOD peaks for vanous tralts uslng dlffcrent types of pred~ctors of log-transformed values from across-locations
QTL LOD peaks for downy mlldew resistance from Colmbatore tnal
QTL LOD peaks for downy mildew resistance from Bhavan~sagar tnal
C o m p a n s ~ n QTL LOD peaks for downy mildew reslstance from Coimbatore, Bhavan~sagar and across-locations
Genetic linkage map of PT 732 x P 1449-2 showlng QTL pos~tions on LG 2 and LG 4 for agronomic traits
Genetic linkage map of PT 732 x P 1449-2 showlng QTL posltlons on LG 6 and LG 7 for agronomic traits
Page Yo.
- -
Title No.
10 I (3ene11c llnkaye map o f PT 732 x P 1339-2 shouing QT1.s positions on 1.C; I and LG 2 for downv n~~lde\c rcslstance I00
10 2 (ienetic linkage niap o f PT 7 3 2 K f ) 1340-2 stlowing QT1.s posl l io~~s on 1.G 4 and LC 7 for down) rnlldetv resls~ance 107
Pearl rn~llet (P r~ r~~rse r r r r~ l yllr~rc~rrrl ( L I R Br ) I S a princ~pal food cereal grown on dbout
27 r n ~ l l ~ o n ha ot drousht-prone s o ~ l s In the scrnl-and reglons of the Indlan ~uhcontlnent
~ n d Atncd ( F A 0 dnd lCRlS4T 1996) w ~ t h a grain yield a\eraglng 500-600 h g ~ h a I t Is
.ilso used as lorage In 4ustral1a. South Afr~ca and rhc b S A and ranks JS the liflh cereal In
urder ol slobal cconomlc lnipondnce Pearl rn~llet has the Lapdclty to rolerate drought and
low so11 fertlllty, but responds well to water and favourdble soil c o n d ~ t ~ o n s (Kumar and
Andrews. 1989) S o thls crop provldes scope for Increased product~on in reylons too drld
for sorghum (Burton. 1983)
Y ~ e l d IS the ult~rnate target of any heterosts-breed~ng program 4 major problem
ot economlc concel l w ~ t h the use o f Inbred CMS llnes In hybnd breedlng 1s thelr low
vleld In seed production plots Good y~eldlng ab~l i ty and seed set, particularly In A- l~ne ,
I S needed to prdct~cally dnd econom~cally mdlntaln and use such ~nbreds Increased and
stab~llzed pearl m~l l e t gram product~on I S essent~al for the well b e l n ~ of m ~ l l ~ o n s of
people who llve In these a r ~ d and sernl-and troplcdl reylons
Inhentance of the rnajonty of econorn~cally Important plant rralts such as y a l n
weld and ~ t s components can be class~fied as polygen~c or quanutatlve Even tralts
cons~dered to be stmply lnhented, such as d~sease resistance, may be o l ~ y o g e n ~ c or "semc-
qudntltat~ve" for whlch tralt expression IS governed by several genes ( e g , a major gene
plus several modlfrers) The challenge to strateg~cally use new tools (such as DNA-based
markers) to Increase the contrlbut~on of "science" to the "art plus sc~ence" equatlon for
plant Improvement therefore app l~es to most, ~f not all, tralts of Importance In plant
breeding programs
.Sclerosportr g~.trrnr~r~toltr (Sacc.) J . Schrot. 1s an obl~_u;~tc hlolrophlc pseudo
iunyus that causes donny mlidew disease on pearl ni~llet, ofien resulting In devastating
~ l e l d losses. The study o f host plant resistance to thls pathogen has been hlndered by the
fact that resistance In the host shows continuous varlatlon (Shlnde cr ( I / . , 1084) and
resistance IS regionally vanable (ICRISAT. 1989). S o breeding material has to he tested
In espenslve, tlme consuming and oflen unreliable mu l~~ loca t~ona l rralts. T h ~ s reglonal
banablllty has been found ro be principally due to generlc var~abllity of pathogen
populations rather than envlronmental difference between locations (Ball and Pike.
1984). Molecular markers llnked to host plant resistance genes would allow resistance to
different pathogen population to be selected for at a s ~ n g l e location In the absence o f the
pathogen vanants. Llnkage drag and the confounding effects of env~ronmenlal variation
associated wlth conventional breeding methods would also be reduced or elim~nated.
The establishment of saturated molecular maps using restriction fragment length
polymorphism (RFLP) and other DNA marker techniques make it poss~ble to dissect
Mendelian factors underlying coniplex tralts such as grain yield. Systematic studies on
mapping quantitative trait loci (QTL) have been conducted in a number of crop specles
(Paterson et al., 1991 ; Tanksley and Hewitt, 1988; Stuber el al., 1992) for various traits.
In this study, characterization was done for QTL for yield and its component traits
and resistance to downy mildew disease. The objectives of this study were:
f* Estimate the mean performance of mapping population testcross hybrids for yield
and its component traits
*:* Determine correlations between g a i n yield and its component traits
-3 Est~mate rhe number and locclt~on ot'QT1. s~g~uticantly atTccr~ng llie \marlon ol'
gram yield and irs cornponenr traits Across r\vo loca~lons In Tdrnll Nadu
Determ~ne the ~nagn~tudc o i the genetlc ~I'fects of QTL iur a11 c l~ le and
sconornically important rester and
f* ldent~fy QTLs for downy m~ldew d~sease resistance under field condition.
2. REVIEW OF LITERATURE
1.1. Pearl millel
f'earl mlllet [Petrnoer~rn~ ,qluucunt ( L . ) K. Br.] IS a cereal belonging to the genus
Pent~rserum, which contains about 140 grassy troplcal species. Pearl mlllet is prowh
~lmosr e.tclusibelv ds hunldn lood, dnd ~ndeed IS the stdple ceredl of 00 n~illlon people
\tho I I L ~ in dgrocl~ma~lc /ones uhere there dre severe dblotlc stress I~ni~ ta t~ons to crop
production malnly due ro heat, low and erratlc rainfall, md so11 rvpe (low ~nherent
lertlllty and moisture holdlng capacity, and In some cases low pH or high levels of
dlum~nlum saturation) Since fertlllzers are seldom used and cult~vdtion 1s ldrgely by hand
or anlmal tractlon actual grain ylelds are low In these reg~ons (500 to 600 kglha), yet In
the agroecologles where t h ~ s crop 1s grown, ~ t s leld is hlgher and more reliably obta~ned
than those from other posslble tropical dry land cereal crops such .is sorghum or maze
Gram IS dlways the pnnclpal object of cultlvatlon, but the stover 1s often secondarily
linportant as animal fodder, and stems can also be used as fuel, for fenclny, dnd roofing
2.2. Molecular marker importance
There is such an enormous amount of diversity in the DNA of higher plants that no two
organisms are likely to be identical in DNA base sequence. Thus, there is a tremendous
amount of DNA vanatlon present in natural populations of plants. These variations have
been detected in restricted (1.e.. enzymatically digested) genomic DNA of plants and have
paved way for the development of molecular markers (Winter and Kahl. 1995). Genetic
cnp~neering and biotechnology hold great potential for application in plant breeding as
they promise to reduce the time taken to produc,e crop varieties with desirable characters.
W~th the use of molecular techn~qurs. t t would now be possible to hasten the transfer ol'
desirable genes among iarletles ~ n d to introduce novel genes horn related species
(Mohan c.1 ul.. 1997). tvlolecular niarkers detect ununbiguous. single-site genetic
differences that can easily be scored and mapped in most segregating populations. It is
not Jit'ficult in populations of most crop species lo ~dentify and map 10-50 segregating
niolecular niarkers per chromosome palr (Kearsey. 1998). DNA markers can illcrease
ct'liclency In breeding programs in a number of ways.
t 'The abllity to screen In the seedling stage for traits that are expressed late in
the life of the plant.
~ i . The ability to screen for traits that are extremely difficult. expensive, or time
consuming to score phenotypically.
... 111. The ability to distinguish between the hnmozygous and heterozygous
conditions of many loci in a single generation without progeny testing.
I \ The ability to perform simultaneous, marker-aided selection to screen for a
character or complex of characters that could not prev~ously bc. included in the
program because of cost or difficulty of conventional methods based on
phenotypic screens.
Molecular markers can accelerate the generation of new varieties and allow
connection of phenotypic characters hith the genomic loci responsible for them. However.
the real advantage of using molecular markers is to permit efficient backcross transfer of
desirable alleles in a directed manner that would not be practical with conventional
phenorypic selection procedures.
plant breedmy methods can now be wadll) studled and 11 IS nc:u wlsti\ely c a y to cstnblish
genetic relationships k t u c e n e\en sewall\. ~ncompa~lhle crop apecies (Mohui el ul.. 1997).
I'he abll~l). IO map genes conmbut~tlg touards \anatlon In complex traits with enough
Accuracy to be usc.hl for plant brecdlng npplicatlons has bccn made posslhle through the
de\.clopment ot comprehensl\e mol~rular marher maps (Jones r l 111. 1997)
I he t o l lou~ng 1s a llsr ot DNA marher techn~ques that habe been developed over
the years (Mohan et ul . 1997, C~upta and Varshney, 2000)
I Acronvm I Technique 1 Reference 1
I AS-PCR Allele Soectfic PCR ! Sarkar el ul.. 1990 --.A
\ I I P Zmpl~fied Fragment Length 1 V O ~ e y l . I995 - -4 1 Pol) momhlsm
/ Cleaved Amplified Polymorphic I I CAPS 1.yamichev el 01.. 1993 i Seauence -- - - - -. -- _. . I I) A F , - 1 DNA Am~llfication Fintemrintinn Xaetano-Anol les el 01.. 19911
ALP - --.
I AP-PCR
Amollcon L.ennth Polvmoruh~sm Ghare a z ~ e el a / , 1995
Arb~trarlly Pr~med PCR Welsh and McClelland.
-- , 1990- -
IMP
I ISA=ISSR
1 MP-PCR
, MFLP
1 RAMS RAPD
REMAP
/ SNP 1 ~ i & l e Nucleotide Polvmoruhism I Nikiforov e/ ul.. 1994 i SSCP 1 Sinnle Strand Conformation I Orita el a/. . 1989
Inter-MITE (Miniature ~ n v ; n e d - r ~ ~ e a r Transwsable Elements) Polymomhism Inter-SSR Amplification = Inter Simple Seauence Reoeat Microsatellite-Primed PCR Microsatellite-anchored fragment length wlvmoruhism Kandomlv Amplified Microsatellite Random-Amolified Polvmomhic DNA Retrotransposon-M~crosatell~te I
, RFLP
i SAP
SCAR
hang el 01.. 2001 1 -- --i
Zietkiewiez ul,, 1994
Mever et 01.. 1993
Yany el 01.. 2002 1
Ender el ul.. 1996 Williams el 01.. 1990 ..,
Am~hf ied Polvmoruhtsm 1 Restriction Fragment Length Polvmorphism Specific Am~l i con Polvmorphism Sequence Characterized Amplified Region
Botstein el ul., 1980
Williams et ul., 1991
Williams ul,, 99,
)SLP fl~crosdtellite Simple Sequence I ength / I<nng\\el~ t 1 '11 1995 - ---_ ,P~!L~E&sE-- - _ _ _. 4 - - -
\SLP \tinisatellire Simple Sequence !.r.ngth I Jarrnan and Wellr. 1989 - . - -- - I Po~vrnorohism +_ -- _- - SSR . S I ~ I D I ~ ~euuence Re~ea t Ci!carnr.t.! G I . , I 992 . . . .
! STlVS , / Sequence Tagged Micro-satellite Sites _.ific,'I(m.cn-and Soller. 1994 S f S / Sequence I ' a n n e d a ' Fuhuoka r/.ul.. 1994 .' .
2.3. Importance of RFLP marker and its application
\mong the various DNA-based molecular markers. RFI-Ps were the lirst to he used in
human genome mapping (Botstein er ul.. 1980) and later they were adopted for plant
genome mapping (Helentjaris er a/., 1986a; Helentjaris. 1987: Paterson u / a/ . . 1988:
U'eber and Helentjaris. 1989). RFLP is the most reliable DNA polyn~orphism that can be
used for accurate scoring of genotypes. It has provided a relatively rapid means of
producing genetic maps of densely spaced marker loci in numerous crop species (Ellis.
1986; Helentjaris el 01.. 1986a; Landry er ul., 1987; Burr 01.. 1988; Mohan el 01..
1997) The four primary advantages of RFLP markers over morphological markers are
co-dominance. frequent polymorphism. absence or limited influence of Ihe environment.
and absence of pleiotropic effects (Botstein el a/.. 1980; Beckrnann and Soller. 1983).
Since RFLP markers have no known effect on the phenotype of the plant. they are ideal
for studying quan!itative traits (Stuber. 1992)
RFLP analysis employs cloned DNA sequences to probe specific regions of the
genome for variations that are seen as changes in the length of DNA fragments produced
by digestion with restriction endonucleases (Landrj er a/.. 1987). In plants, U L P s were
first been used in tomato, maize and rice to saturate their already extensive genetic maps
based on morphological markers ; l i d Isor! me markers ( HernarrA? and rmksley. 1986;
t le len~aris L*I '11 . 1086a. hlc('~)uch rr ~il . . 1988).
Prior to the ara~lability of SSR markers. two types of DNA markers have been
must commonlg used for most crop plant molecular marker-based linkage map
Je\elopment and subsequent quantrtati\e trait locus (Qf I . ) mapping. KF1.P markers
I Rotstein er ul.. 1980) and RAPD markers (Williams t.1 iil.. 1090). l3o1h detect IINA
polymorphism and monitor the segregation of a, DN.A sequence anlong progeny of a
senetic cross permitting construction of a genetic linkage map. I-lonever co-dominant
RFLP markers are more robust a id repeatable than RAPD markers. whlch are inherited
In a dominant manner.
RFLP and RAPD marker allelic differences between plants are inher~ted in the
hame fashion as conventional Mendelian genes. thus genetlc linkage maps of these
molecular markers can be constructed using conventional methods. Such RFLP linkage
maps indicate the locations of specific restriction site andor insertion/deletion
pol~morphisrns In chromosomal DNA relatlve to one another. Ellis ( 1086) reported that
simple consideration of RFLP mapping as a method of analyzing the inheritance of
quantitative characters suggests that there are several limitations to the utility of this
dpproac h.
RFLP and morphological markers have been used in pract~cal plant breeding
programs to map quantitative trait loci (QTLs) (Tanksley rr ul.. 1982; Edwards el 01..
1987: Stuber er ul.. 1987: Weller er ul., 1988: Mohan e l ul.. 1997) and to monitor
response to recurrent selection (Stuber er ul.. 1980, 1982). RFLP markers facilitate the
selection of progenies with desirable genotypes in a relatively short span of time.
aql 'suo!lelndod Su!ddeu lall!u p a d 30, uo!~ej!ld!~lnw ptm luaudolahap l o j 1 ~ ~ ~ 3 1
le pasn Bu~aq sainpajoid aql paqlissap ( t661) a q w o 3 1 ! ~ pue q m ~ .s!sAleue j!dA~ouaqd
pen? vNa J O ~ pasn ale suo!ieiaua% q s ~ q m pue 'pajueApe s~ s s o ~ s aql moq .uo!lelndcd
aqi JO a z ~ s aql .%u!sso~s io j uasoqs ale siuaied q s ~ q m uo luapuadap Alu~euc s! qs!qn j o
aJnl!ej JO ssasjns aql waiold 8 u 1 d d ~ w E JOJ s io i se~ lueuodw! aql pa.na!Aai (b66 1 ) Bun(?),
.lsarold Bulddew aqi JO 1eo8 aql s! q y m JO ~ueuodm! )sow aqi ~ ! U J u! ldaq aq lsnur
s i o i ~ e j [eia,+as .suo!s!sap asaql 8u!qew ul .uo!lelndod Bulddow aqi 8u!dolahap u! apeul
asoql are s ~ a ~ i e u VNU ~ I I M sdew asequ!l Su! l~rulsuo~ u! suo!s!sap IE?!lU3 lsolu aqj.
uo!~s[ndod 8u!ddem e Bu!dola~aa .t.z
(Lb()[ ' /fJ 1.7 lllOlJd :LO61 ' /" Is7 SaUOf) Sl!CJl ~ ~ l d l u 0 3 j l ) SWS!UEY5aUI 1~3!80101~iqd
aql J U Bu!punisiapun ino Ru~~o id ru ! i o j pup '(6661 ' 8 u n o ~ : ~ 6 6 1 " IP I,? ueqop,
'UEJl?yaq5I?JpUEq;) pUE W[Eq~EUTUv ! 1661 "Ifl 1" UOSJJlEd :()661 ' l a f iu !qs [a~ :L86[
/fl ra spmmp;j) SI!CJI a~! la i ! lu~nb puz ah!ia!lenh iueuodw! 8u!paa~q u! aBeiuz,%pr
ma15 c s~a.f.11) uayinw st? s , l l d y JO asn a" lnql 1ss88ns siaded 1umaJjo iaqwnu .1>[[1iu
[mad u! asmls!saJ mapl~lu Lu.nop 01 asuaJajaJ n l n s i w d ql1.n 8uIpaaiq ~ w l d u! d - l r i ~
j o asn le!iuaiod ig sa!~lunuoddo aql pa7!seqdrua (000;) xo.) - lamJt l pun q s q + p m (666 1
: ~ h h I ) ,117 1.7 q":1 1 '(;b(,l l a q l u ~ x ! , ~ ~ U C ale!) '( 166 1 ) q s c ~ .<IJIJF! \ I R I S J ; I L U ~ O J 01 U I ~ J I ,
asJnosaJ U I O J ~ uo~ssa i f fo i~u~ palslssa-~ayicul J I ~ ~ I pur 1301 ilrll > \ ~ ~ r i ~ l u m h .lo 3u1ddr~u
pur U O ! ~ E ~ ! ~ I ~ L I ~ ~ I . U O I ~ I : J ! I I I U ~ ~ I l ~ i ~ ~ i l - \ ffu!{)nlx~l SLI ( I I I I !> ! [~~P S ~ O I J I ! ~ JC) ,~ p > ~ o > s all, i r q ~
lr!np!A!pu! i.xl srus~q&oruil~)d i o laqwrlu pur slcnpt \!pu~ j o s u ~ ~ a ~ u! ( '861) ' a l p s pur
uueurys~f l Lq p?nw\-sl: >JJ\\ irl>ru>\oidtu~ >!iauaff 01 sdir l ) l SLII ilddr In slso;)
( ~ h 6 1 ' [ I ) I.' u r q q \ i ) ~ U I L U I I < U ~ J aiull pilr J \ I<U>IUI inoqcl SI slsilrur.
<{-I.~H 'sanh~utlwl J ~ ~ J ~ I L I Jrlrnaloul pAola\ap i [ lua>a~ aiotu oi pamduos , la \ar \u t l
parentage of mapplng pupulat~ons hen a\a~lable. and the tr;llrs I;>r \bIilch the) n11ght he
u\cd tor Q 1'1. mapping.
1.5. Mapping QTL using testers
\1t1lt ol' the earl! Q rl. htud~es publ~>Iied on grain > ~ c l d .ind > ~ e l d con~$>ricnrs ~n \o l \ ed
c~ther ~ n d ~ k ~ d u a l plant5 or rcpllcated progcnles t)f seprcgntlng pop~llat~on deriied tiom
h~parenral crosses by belting or hackcross~ng (1.ubbcrsredt cr t i / . . 1997) Ilowevcr In
11) brld pearl ln~llet bree~ling, the main selectlon criterion is testcross perthrmance and line
pcrtormance ptr ve is onl) of secondary importance because these measures are poorly
correlated. cspec~ally for yield characters. Use o f testcross progenies in (J'I'I. mapping
>tud~es prov~des ~nt'ormation about the intluence of the tester and hence. IS important for
both bas~c research and appl~cat~on of marker-aided selection (MAS).
Lubbcrstedt ul (1997) crossed 380 F, lines obtainrd bq selling I:: individuals
I'rom a malzr mapplng population. and the two parental lines. with two diverse dent
inbred resters to map (S'ILs afecting testcross perfbrmancc tbr Imponant lbraye maize
traits and to lnvesttgate their consistency across-environments and testers. I-hey detected
seven (starch yield) to 16 (plant height) QTLs in each testcross series. explain~ng between
i 2 to 7 I % of 0,' in a simulation test.
For forage maize. Lubberstedt er ul. ( 1997) found good agreement across testers
for drq matter concentration and plant height. but nor for other trails including dry matter
! ield and rn \,rrro digestibility of the %hole plant. Hence at least for most of the relevant
lbrage maize traits. 11 appears the separate QTL mapping is necessary for each tester.
Lubberstedt rr ul. (1998) evaluated four independently-derived mapping
populations crossed with same tester. in maize.'They observed that consistency for QTL
poiuon across a11 h u r populations. \\hich \\ere greater li)r d q matter concentration.
\r,lrch cuncentrarlon. .lnd plant he~ght than li>r d c mailer ! 1e1~1. r r r I , I I I . ~ J ,liyc.s~~hle organic
l11;ltler dnd proteln cc>ncenlrarion. Kc.\ults from rhe~r htud? ~nd~cated ( ) \ ' I s \\ere poorl?
consistent among crosses urthin the llint hetcrot~c pool. suggesting prlor to MAS, Qr I .
[napping must he perlimned separately I'or each population.
The consistency 01'QiL. mapping results across testers u ~ l l he larpel! retlected hy
the genot?pic correlaticln among testers and the prcdorn~nant type ol' gene action Ibr each
trait. I'hus. for a given sample selection response from MAS Ibr testcross performance of
traits with mainly additive gene action should he comparable for testcross progenies with
orher related testers. \lelchinger rr r r l . r 1998) lhund little evidence Ibr digenic epistasis
dmong the detected QTLs. particularly when re-examined in an independent sample. On
the contrary, differences in the testcross performance of FI lines with each tester were
due to the presence or absence of common QT1.s. I'his wgyests that nun-epistatic gene
clf'ects are malor determinants of general and specific combining ahility In hybrid
perli)rrnance. as was also concluded that numerous clasalcal quan t~ ta t i~e genetic
cupenments
Austin rr u l . 12000) reported that QTLs detected with only one tester were not
necessarily detected for the other testers especially for grain yield. Aust~n rr ul. (2001
used three different testers in maile. Results indicated that regions containing QTL
effects for a single tester appear to be less stable across test environme~its and less likely
to be detected for mean testcross performance across testers than those associated with
OTL effects for two to three testers. Mean testcross effects (MTC). however. appear 10 be
le>> ,enslli\ e I t ) en \ ~ r o n ~ i ~ c ~ i t a l ldct<~rs \\ ~ t h the Iiialorlt) 01' ( ) I I s \4 1111 1lie largest bl 1 C
etl'ects helng cuns~~ ten t l ! detccrcd across reat en\ ironnients
\'add\ t l r '11 (20U2a1 ubed pearl rn~llet testcross 1 . 1 Ii? hrlds tlrr plicnol)p~ny Q'f1.s
.irsoclated ui th traits de t e rn i~n~ng yrrun and 5tovr.r y ~ e l d under terni~nal drought stress
coridit~ons. rather than uslng 1nhrr.d proyerlles Ibr \c\eral reast)~is.
I to restore hetcrot~c \]your to the Inbred rnapplng pt)pulat~u~i that ~niglit uthewrse he
roo \beak tor et'fecti\e \creenlng under \tress condirions (pearl 1111llct IS highly
cross-poll~nated In nature and \ut'fers conslderahly tiom ~ n b r e c d ~ n g Jep res s~o~ i ) :
1 to use the domlnanrly inherited early tlowerrng of the rester lo reduce variat~on in
Ilowerlng rime m o n g the test unlts in order to locus the mapplng on specilic drought
tolerance traits rather than traits or responses associated wrth drought escape: and linally;
3 to have test units that approximate the genetic structure of' the I:I hybrids grown by
t i m e r s rather than F;) or l:1 inbred lines.
2.6. 1,inkage mapp ing
L~nkage mapping is putting marker loci (and O'1'I.s) in order. ~ n d ~ c n u n y thc relative
d~stances among them. and assrgning them to linkage groups on the bass of their
recombrnat~on values from all pair-wrse and three-pant combinat~ons. I'he lirst map of
the human genarne based on molecular markers (Botstein rr ul.. 1980) f'uelled the
derelopment o f molecular marker-based genome maps in other organisms. and has led to
the recent genomlc sequencing of humans. mice. tlruhidop.ri.~ and rice
The theory of linkage mapping is same for DNA markers as In classical genetic
mapping based on morphological markers, however. several new considerations must be
kept in m ~ n d . This is primarily a result of the fact that potentially unlimited numbers of
I)>:\ markers can be anal!red In a single mapping p>pulation. l)h;\-h;r.*d maps can IX
~ e l ~ t e d IU rtlstlng c!togenet~c maps through the use L I ~ aneuplold or \uhst~tut~un lines
I iielentl;lris rr Lri.. 19Xhh: Sharp er 01 . 1989. Young r.r ,ri.. 1987) or 11, ~ r r r r liyhr~d~/.atlon
I IS1 I ) ((.hang rr '11.. 7000)
Sincc 1)N:i ~iiarAer tech~lolog> \\as lirst appllcd to plants. there has been an
c \ p l ~ i \ ~ o n In [lie de\elopn~enr .~nd appIic.rt~on 01' yencuc Ilnhagc maps ~ h i u l ~ a n vl 111.
10071. I.~stng these new 1)N.A-hased markers. scient~sts have constructed maps in specles
\\here only poorly populated classical maps existed before (Bonierhalc 1.1 ul. . 1988;
(iehhardt ri ul.. I991 : Liu ec ~ 1 . . 1994). located genes governing quantitative characters
ottcn In great detall ~ n d taken the lirst steps towards yelie cloliing based on yeneuc map
position. Detailed genetic linkage maps are also fundamental tools Ibr studies on
>election. identification and organi7ation of plant genomes ('l'anksley. 1903: I3eckmann
and Suller. 1986. [.andry and Michelmore. 1987).
2.6.1. Achievements in different crops
I 'sing RFLPs as genetrc markers. Helentjar~s rr ul. (1986a) constructed linkage maps I'or
maize and tomato. The first true RFLP-based genetic linkage map in a crop plant
(tomato) was constructed in 1986 with only 44 F2 plants and 57 marker loci (Bernatzky
and Tanksley. 1986,. Since then. DNA marker-based genetic linkage maps tbr many
plant species hare been constructed (tielentjaris. 1987: McCouch rr (11.. 1988: Heun er
111. 1991 : Tanksley. 1993. Mohan rc (11 . 1997)
:\ detailed map of' lettuce was constructed by Landry e/ irl (1987) using 53
yenetlc markers. These included 41 RFLP loci. 5 downy mildew resistance genes, 4
ijozyme loci and 3 morphological markers covering 404 cM.
\lc('ouch (11 I l9XKI rcpmed the conbtructloti 0 1 ' 411 K1.I 1'-hased genetlc
linkage map oi rice I he map cornprlsed o i 135 loct correspo~id~ng to clone> selected
~ r o m a I'\/I ycnomlc Ilhran co\t.rlng I . ;XO cLl I I I ' the rice yenonlc ( ' ~ u . ( \ e r.1 (11. ( 10941
Je\eloped 3 rice yencuc map us~ng ca. X O O K1.l Ps ~lwt expanded the length o f the rice
I~nhage map to 1491 cbl. ('hao (11, c I V X V I attempted KI-I P mapplny in heyaploid wheat
( I I I L I L I ~ ~ I ur \rr i~~tm) ualng I8 cl)t4,1 clones: I 4 anonymous and 4 oikno\*n iutict~on. I'he
1oc1 Identi tied b) these probes were mapped on one or more of' \r hear homcologus group
7 chromosomes. (iraner ' ,1 111. (1991) analyzed two populations to consuuct an
KI-LP-based genetic linkage map of barley uslng 250 genomic and cI1NA markers. Maps
oichromosomes 3 h . 3 B and 31) of wheat and 3 K of rye were de\cloped hy I>evos el (11 .
( I ')92i us~ng 2 2 1)NA probes and 2 enzyme marker systems.
2.6.2. Computer software packages for constructing genetic linkage maps
\d\anccs In computer technology hate been essential to progress In D N A marker-based
genetre l~nkage maps. I'he theory hehind linkayc mapplny with L ) l i i \ markers 1s Identical
to mapping with classical genetlc markers. but the complex~ty of the problem has
dramatically increased because of the larger numbers of markers that must bc used. This
Increase in numbers of segregating l o c ~ (and the number of progcnles rn which they are
segregating) relative to \tudies of class~cal genetic markers has necessitated the
Je\elopment of complex computer algor~thms and software packages spcc~fically Sor t h ~ s
purpose.
Construction of a genetic linkage map from a DNA marker data set requires
computer software packages capable of running X 2 contingency table analysis. 7be
program, LINKAGE-] (Suiter et 01.. 1983) cam'es out this type of analysis automatically
~ n d .11w compares the obscried ~ l l c l i c disir~buuons to cspcctcd dis~r~hut ions . In :i
d~t'fereni strategy tbr upitmiring the use of DNA marker ~ntbrnmatlon. rlme computer
program "HyperGene" cumens genotypic data inlo a "graphical genutype" (Young and
I anksley. 1989a.h). in which a complete genome of an i n d i ~ i d ~ i a l l'ronl the mapping
populat~on I S disp1a)c.d.
\,lAPM.-IKER'ENP is a linkage analysis software packayc t i ~ r construcring
prlmap linkage maps of markers segregating in euperimcntal crosses. It performs full
rnultipo~nt l~nkage analysis for dominant. recessive and co-dominant (e.g. KFLP-like)
markers in B C I backcrosses. F2 and FI (self) intercrosses and recombinant inbred lines
(Lander et ul.. 1987. I-incoln e / 01.. 1992a. b).
rhe sotiware package Joinmap (Stam 1993; Stam and Van Ooijen. 1995) analyses
all types of mapping populations, and can combine maps of different mapping
populations prov~ded there are common markers. Another sofiware for linkage mapping
I S (;mendel from Oregon State IJniversity. lJSA (fiolloway and Knapp, 1994). The
package Mapmanager. with different versions such as QTX. QTXP and (J1.X-Classic for
Macintosh- and IRM compatible computers (Manly. 1993: Manly and Olsen. 1999). can
be used to analyse the results of genetic mapping experiments using backcrosses or
recombinant inbred lines.
In addition with these packages QTL cartographer and PLABQTL are seldom
used to carry out the genetic linkage analysis using molecular markers.
2.7. Pearl millet genetic map
fhe first detalled molecular marker-based genetic linkage map of pearl millet was
published in 1994. and was comprised primarily o f RFLP markers (Liu el ul., 1994).
'0261 a3u!s alqepene uaaq seq Bu!ddsu
laqlew %u!illapun hoaql ~ ! s e q a u .saddloua% JayJew aql ~ I I M sanleh adkouaqd
I!eJl aql J O UO!IFI~OSSE aql pue sJaTJeu J O Bu~ddew aql :saile~s le!luassa o w <e[l
11 ~uo!le[ndod du!ddetu aql u! 8u!leBa~8as salalle JayJew aql pue I!ell a,~ilr?l!luenb aql JOJ
sanlm s~dLlouaqd uaaMlaq suo!leposse uo palexpa~d s! s!s.(leue 710 '(6t.61 'laqlew)
sauafidlod st! 01 pallajal aJam s-110 'slaylelu lolnsalolu jo h a ~ o s s i p aql aJoja$j '(sib]
'ueuapla!)) slrell a,%!le)~luenh a ~ l ~ a p u n leql s2so1q aua8 JO saua% aql J O auo SI . ( r ~ o l
I!EJI a~!lm!luenb = s l j . 0 = le~nld) snsol ]!ell a~!lel!luenn JOJ mLuo~se aq] ',,l~().. y
(auepleH) wqo t .1 seal ie 01 pualxa dl~en~uaza
~ [ I M dew a8eyu!l 31iauaii lall!w llead aql leq] slsa88ns s!qj .(.mwos . s a d .so.haa .j.z;
uapley) q3ea jo spua aql des 01 palsadva a n l e q ~ sanuanbas s!laluolal p m sdnol;i
a8eyu1l uanas asaql u! isol JayJew aql u a a ~ i a q palsalap uaaq sr?q a8eyu!l ]uesy!ud~s
ou a ~ e p 01 lnq .fis auepleH OOL e s~!s 01 d e u aBeyu!l s11auaZ lal l~w Inad aql j o
ql8ual aql papualxa a.4eq salpnls iuanbasqn~ 'am!] ~ I I M a m p y 01 palsadxa s! asuaJajj!p
s!ql Inq .uoqs i ~ ~ e l n " ~ ~ c d SSC.M 11 1eq1 U I satuoua8 SSCJR 8uou1: lensnun sem dew aiieyu!~
LUOJJ pa1~1?% ( < ) - ] I sdnoJ8 a;iI?yuI~ [r?np!\!pu! aql ' W 2 ; lnoqr: s e n 1201 uaanlaq aJualslp
aBo~a,%r: aql puc ly? WAS ls~lrrd awosouOJqs llll!lu ]Jl!ad ua.%as aql 01 8u!puodsa~lo?
.sp![os alqnlos doj s l l o dnoj pue ssew ~ ! r q Bu!l lo~~uos s l l b
x!s lmal le %u!ddew ' o ~ e ~ u o l jo sso1sl3eq J ~ I ~ X ~ S - I ~ I U ! ue u! SIOIXJ ueqapuan a l a ~ ~ s ! p
01 s]!ed] a~!,el!]uenb pa.%losad d a q ~ '(8861) '10 la uosdaled dq olewol u! pauoda~ s2.n
S I J ~ I E L U d l d ~ uo paseq dew adeyu!l do13 alaldmos Alqeuosea~ e jo asn I S I ~ a q l
' ( ~ 6 6 1 'Aalpna) s ~ a y ~ e u Jnlnsalou ql1.n YJOM luasaj a ~ o m IOJ uo!iehdasqo pue ICloaqljo
puno~alseq e papl~odd sa!pnls da!llea asaqL '(1961 'MET) sawosoworqs [enp!h!pu! q11*
a p d l a~!la! iwnb JO uo!ie!sosse S!)uap! 01 pasn uaaq aheq s s ~ u o s o u o u ( . I wnnfrsan
wn3111d.~) l F a q pioldexaq U l '(1961 'UOSUaqOH PlIE UOSUJlOS-LRleLUJ!N) pauoda~
uaaq aAeq sd3y~eu d n o ~ 8 poolq 10s uo!~e%a~das q1l.n sl!erl a..\lie~!luenb JO suo!le!sosur
'sletu!ue p a i e ~ ~ ~ s a m o p ul .!sol laydew ~ I I M saua%i(lod jo aaeyu!~ f u ~ i s a ~ a p IOJ spoqlau
padola~ap (1961) depoqj. ~aLuosowodq3 ~elns!ued e uo daqunu alis!dq doj saua8blod
awsol 01 alqt! aJaM .~arsnXoun/aw u! s i u a u ~ ~ a d x a uo~lsalas Aq ( ~ 9 6 1 ) d e p o q ~ puc
uosq19 puv ( 0 ~ 6 1 ) JayleW pm? u o s u l e ~ .sauaddlod uaamlaq a8eyu!l j o asuap!Aa punq
pm! 'iaqwnu alls!dq lou~wopqn mol pug q81q doj palsalas ~a~mXounlaw a ~ o Saul[ passod:,
(1561) !IIeAej 'sl!edi a~!~e~!l t renb uo sawosowonp Ienp!i\!pui JO s ~ s a g a aql pale~lsuouap
(6P61 'uos1.11~11 pur ~ a q i a ~ ) ~arsnXoun/aw n/rydoso~u u! y i o ~ a ~ 1 s u a l x 3
. ( 1 .7."7X/n,1 una/1do/{)
Aalreq u! pla!i( ilu~isajjc uo!8a1 [eu~osouodq:, c payuell inql sJayJelu l e ~ ! 8 0 j o q l o u punoj
( ~ ~ 6 1 ) ~a l luqss pu1? U O S I ~ , ~ : ~ 'dnolos la,noll doj aus8 pa~l.isqu~ Xldmls E q11.n (.I urtl.t!111\
wtls!~) scad ul I I I ~ J I az!ic~!~ucnb e ) at1111 ;?ul~a.noll J O a3cyu1l palerlsuouap ( c c h l I
uossntuse)l , ' ( " I W'.II~SI'I,I .vfl/o.'v~qd) sue" U I daydam dnolos ico:, paas c JOJ uo!le3a~Ra\
q l ! ~ pal&?!3oSSr 2 I I S paas 191 ~011~iialfTa~ pJnl3sqo 211 u a q ~ slueld U I IIVJI a.z!lolllucnh
C q1IU S l ~ Y . ' L l l 3!1>~38 ~ S 1 1 J ~ l ~ 1 1 1 I(IJUIIS J 0 UUIICISOSSC pa~Od.71 )$][I (L;L6[ )
2.8.2. Mapping QTLs for grain jield and its related traits
(;rain yleld is generally controlled b! a numher ot'quantltat~\e tralt loc~ and is ~lf'ected by
env~ronmental factors. making i t d~fficult to manipulate and Improve In a breeding
program. Grain yield can be dissected into 3 numkr of component traits such as -
~ndividual grain mass. spikelet number. grain number per panicle. ett'ective filler number
per p!ant. and plant population dens~ty that depend upon the crop concerned. 'These
component traits are also under QI'L control and the effects of individual QTLs on
phenotyp~c varlatiun are relatively small. Some uf them. however. Lue less
envlronmentally sensitive and have higher heritabilities than grain yield itself (Bezant
(11.. 1997; Yano and Sasaki, 1997). I'heretbre. whlle looking for QrLs controlling grain
yield. QTLs for yield and yield components should also be determined to provide useful
~nformation.
The advent of molecular markers, and in particular RF1.P has greatly facilitated
the detection of QTLs controlling yield components and the relationship between grain
! irld and its components. (!sing molecular linkage genetic maps. i t is possible to estimate
the number of loc~ controlling statistically significant portions of genetic variation in a
segregating population and to characterize these loci with regard to map position, gene
action. phenotypic effects. pleiotropic effects and epistatic interaction with other QTLs
(Xiao rr ul.. 1996). It has been demonstrated that correlated components of yield or other
traits often have QTLs mapping at similar locations. This has been observed in maize
(Abler et ul.. 1991; Veldboom er ul.. 1994: Austin and Lee. 1996). tomato (Paterson. et
al.. 1991). barley (Tinker el ul.. 1996: Bezant el ul.. 1997). rice (Xiao el ul.. 1996). and
pearl millet (Yadav el ul.. 2002a).
In potato. tuber 5tarch content ~ n d tuber yeld are quantltdtl\e {rafts that are easy
ro determlne under field cond~trons Schafer-Pregl trl (1998) ~vapped Q'1'l.s for tuber
\trlrch content and tuber yield In two F I populat~ons d e r ~ ~ c d from crosslng non-~nbred
J ~ h a p l o ~ d potato breed~ng llnes :4 total ot' I8 putatlte (2 l'Ls tor luher starch content were
idrnt~licd on J I I 12 potato linkage groups and 8 putatlw Q1l.s for tuhcr yield were
idcntlfjed on t.lght linkage eroups *-\lso. twenty-slx putatlve Q1 I.> Lrcrc reproduc~bly
detected In two envwonments and/or mapplng populat~ons
Orf r t al. ( i 999) measured and compared QTLs for agronomic tralts of soybean In
.I large RI population derlved from crosses between three d~fferent sets of population.
QTLs were ident~fied for all the primary and der~ved tralts w~th a s~yn~ficance level
LOD 3. on 17 of the 20 soybean linkage groups and these QTLs tended to be clustered
on only three of the linkage groups. QTLs wlth major effect (R' > 10%) were ~dentified
lor all the observed characters and for many of these characters expla~ned more than half
u t the observed hentable varlatlon.
Campell ri 01. ( 1999) conducted a study to determlne assoclatlons between kernel
tralts and molecular markers and to ~dent~fy QTLs affecting kernel traits In a soft x hard
wheat cross. They identified QTLs for kernel traits located on chromosomes IA. 2B, ZD,
3B. 7A and 7B. Particularly the prnB marker on chromosome arm 5Ds explained over
60% of the phenotypic variat~on for kernel texture
Shah er L I I . (1999) were able to locate QTLs for a number of agr~nomicaily
Imponant traits such as grain yield, kernel number per spike. 1000-grain we~ght. spike
number. grain volume weight. plant height and anthesis date to the long arm of 3A
chromosome using a subst~tution line.
t\\o-ro\v barley cross Harrington~CrR 306 on the hasis ol'e\aluation ol' 1-45 1111 line in 30
tield eiperiments (Spaner ~ r l . 1999) ['hey compared among groups $)I' lines with
iontratlng markers gcnor!ptts un chromcrsome 7 ( 5 1 1 ) and confirmed that a QI'I. on thc
"plus" arm of that chromosome affects graln y~eld and plant height
2.9. QTL * environment interactions
One 01' the major goals fbr plant breeders 1s to develop genotypes w ~ t h a high yield
potcnt~al and the ability to Inamtam yield across-environments. fhe clTect of
(>TI. x environment interaction has been addressed in several studies in which QTL have
been mapped in the same population In different cnvironmrnts (l'aterson ~ ' r trl. 1991:
Stuber rr ul 1991; Hayes rr ul. 1993: Yan rr (11.. 1999; Yadav rr ul.. 2OO2b).
Paterson er 01. (1991) investigated the prediction value of Q'r1,s across-environments
in tomato by comparing QTL maps of an Fz population and its denved FI families. They
>bowed that only 1 out of 29 QT1.s were detected in all testing environments. Stuber er (11.
I 1992) studied genotype x environment ~ntenction for QTLs of mam by field ctaluation of
backcross families in six diverse environments. but limited evidence was found.
Zhuang er a/. (1997) repeated studies of an F2 and two equivalent FI populalions
of an rndicu-rndico cross of rice grown in three different environments. In all three trials
QTLs for yield components were frequently detected in the same intervals. They
identified 23 of the 29 QTLs for yield and its component traits and 9 of the 15 QTLs for
plant stature in more than one trial. They indicated that detection of chromosomal
segments harboring QTL was hardly affected by environmental factors, perhaps because
the environmental difference themselves were small.
\ doubled I1;lploid rlce Inapplrlp populatio~l ,I!' I?: li~ies troll1 IK o.(~:I/~icena \ V ~ S
ilsed to anal! se the genet) p ~ c . enrironrnental intrracuc~n t i~ r tight dil'ferent plant-type
t ra~ts ~n rice I )'an r,/ trl.. 19991 l:our to nlne O I L S ctffccting dil'fcrcnt plant-t!pe traits
ncrc detected They wggesred rhat (21'1.5 \\ith suhstant~al :n:tin eflkcts could he used in
\I.1S cross-en\~runr~icnts. O'II. i environment 111rcraction etfccts \\ere detected more
thdn ()'I1. mall1 effects tbr plant height. uhich might indicate thar gene eupresslon Ibr this
rralt could he preatl? afkcted b\ e11\1r011met1ts
In order to identtfy ()ll.s controlling Jgronomlc tratt Larlatlon and thelr
consistency under Mediterranean conditions in barley. J progeny of 167 K1L.S and their
parents I'admore and EriAPbl were grown under SIX environments I I'eulat c.1 ul.. 7001 ).
\ total of ZJ 01'1. consistent across all the testtng envlronments were detected using
muluple environment analysis. Out of these Ql'1.s. 1 I presented main efTects. seven
presented QTI- x cnkironrnenr Interaction. and six presented both effects
LLlao ? r rrl ( 2 0 0 1 ) used ,I ricc doubled haplo~d populat~on and a ricc recomb~nant
trihred l ~ n e population der~ked from crosses between a tropical 11111un1cu ra le ty .
\/ucena. and 1-0 indrcu rarietles. IR b4 and IR 1552. ~n both field and pot cuperlmenls.
tbr detect~ng QTLs and eptstasis for rice panicle number in different genetic backgrounds
and different lowland irrigated rice production environments. rheir results indicated that
the effect of genetic background on 07'Ls was greater than that ot'environments, and
cpistasls between QT1.s I S more sensltl\e to genetlc backgrounds and en\lronments than
rnaln effect 0 I I . s M a ~ n effect QTl s and epistatlc Q1l.s could be interchangeable
depending on the genetic backgrounds and probably on the envlronments where they are
~denri tied.
lsom s! )] .[e!lalelu )sal Bu~fios alojaq syaam aaq) s r e ~ ~ l l n s alq!ldaxnsjo aInlx!w e
~ I I M (MOI qlu!u lo q y g h a ~ a ) SMOI ~olsaju! jo Bu!fios aql sa.qo.\u! anb!uqjal s ~ q j '( 1861
"1" / a sme!ll!y,) mnlnsou! le!8uelods ap!ho~d leql SMOI ~olsaju! u . ~ o s - a ~ d uo paseq
'padolahap uaaq aju!s a,\eq s a n b ~ q s a l Bu!uaa~x plag aless-a81el ,110s aql u! salodsoo
aql bq pale!l!ut seM uo~lsaju! pue slold asaq~ U I u.nos a l a ~ sle!lalew lsal a q l '(9161
.qBu!~ pue auaN, sna.i l e m a s ~ o j paqBnold uaaq psq slueld la l l~w p a d %u!~eaq-a~odsoo
.paljaju! q s ~ q ~ OIU! slold '.a,! ,.slold ys!~, . uo papuadap map[lm bumop lalpw iread
01 asuels!sa jo ssslnos loj uaa13s ol s ~ d w a ~ l c i(11eg ' v a ~ q b e p laye Buol L a z lq alqnt\
u ! ~ w s ~ IOU op .i[lc~aunX put? sqnt u s 8 s el4 aieuruuaR e l d u e ~ o d ~ .o/,,oL M O ~ Y S [ ~ \ I I
bl!p!wnq a~11e1s.1 IC psp lom s! uo!lelnlods o~ ..q~wo~Fi lall!w [lead IOJ lunlut~do aql
molaq llam 1 s ~ ~ ut s! l e q ~ a~nleladmal e J 0;. Ie s1nJso uo~lsnpo~d e!%ue~ods wnwlxelq
.&p!wnq a ~ ~ l e l a ~ q8!q p m s a ~ n ~ e ~ a d m a ~ alelapow jo suo!l!puos lapun 1q8!u IE pasnpo~d
an! BIUWI(K~S Ivnwasc aq.1 '(6661 'UOSIIM) s u o ~ ~ ~ p u o s Oolc~oqel lapun sreai ~1 ol
sqluow 8 wo!l s.zlNns uss pm anss!l lsoy pazluolos U I uo!leu!qwosa~ Innxas Xut%\ollo,j
WOJ salodsoo ,Islaum!p u! wrl ol ;.;. pur ,mollad q s ~ u w o ~ q 'lcs~laqdz 'palleh\-ysrqi
are sa~odsot, lenxag (1x61 'I01aqg) uosws qsca wnlnsou!jo axnos i r e u ~ ~ d aql a p l t o ~ d
PUF a w q paas J O 110s 21s y s ~ q ~ .sa~odsoo sasnpo~d a8els [onwas aq.1 'saqqd lcnvasr pur
lenwas yloq J C I ~ ~ U I J L L I O , S I . l n ~ y ~ ? ; . r ( . > > P S I I ~ I . ~ I I ( I I U I ~ . I ~ t l l O i ~ . ~ 0 . l . ~ / . 7 ~ ~ o a1~i . l I L I I
sanb!uq~a~ % u ! u a a ~ ~ ~ - mapl~w ,Cu.nou . lsO1.;
' ( s s ~ ~ d u! 0002 'aqWojl!M p m qEH :L661 'qseH :L661 " I D IV~U!S)
~ a p l ! t u A u ~ o p iall!ur [lead 01 a ~ ~ s ! s a i iueld isoq u! sasuaiaufp yauaX aql isaiap
01 a l q e p e spoqlalu Wu!uaaiss aql u! sieaX O[ wed aqi 1x0 iuamahoidw! snopuauail
uaaq seq aiaql snq.L .slnoq Xuew JaAo ainssald aseasrp q%!q dlaienbape pue uuoj!un
~ A E % asealai aiodsoo7 aJojaq palpqs uaaq peq leqi e!%ueiods jo uo!suadsns e qlrm
sfluilpaas t l u ~ k e ~ d s .e18ue1ods j o uo!suadsns pall!qs e Bu!ieids Aq umlnsou! j o le!iua~od
uo!ljaju! u!eiu!elu oi sdem ahllsajja passnx!p ( [ O O i ) '1" f a sauor SMOJ ~ o l s a ~ u ! uo pascq
sanb!uq~a1 ~ u ! u ~ > J % p[a!l w "a,% SF uo!ielnmu! lam01 8ur11ias pue .uo!ielnsou! uo~isaru~
'uo!lslnsou! d o ~ p .uo!lclnsou! i e ~ d s 'uo!lelnsout d ~ p au!pnlsu! aseas!p s1q1 ioj alqelletr
sanb!uq~ai flu!uauss asnoquaai8/L1ois1oqel [[e pau!eldxa ( ~ 6 6 1 ) IP I? q % u ! ~
. ( I . + ~ t u ~ ~ X u e ~ o d s 0 1 InOqe) ei8ue~ods pareda~d ilqsaij jo uo!suadsns
snoanbe UF q i ! ~ pa1e1nnou!-,4e~ds aiam atleis pal-auo 01 al!idoa[os aqi ie sau
2.10.2. QTL for downy mildew resistance in pearl millet
I'he first hirly detailed molecular marker map for pearl ln~llet \\as cu~rstructed by I-iu
L I (11 (1943) so thal Q'fI. anal\sis 1s no\\ poss~ble 111 t h ~ s crop 1l.s l i ~ r host-plan1
reslstance to downy mildetr. caused hy S ,grtr~i~rtlic~olcr pathogen populat~ons from India.
k ~ g e r ~ a . N~ger. and Senegal \rere mapped using the cross [.(ill-I-R-10 (susceptible)
ICklP 85410 (resistant Jones 6.1 111 . . 1005). I lost-plan( resistance 0 I'1.s were detected
that were effective against each of the four pathogen populations. To locale genes in
[napping populations other than those for which RFLP maps exist, a skclecon map needs
to be transferred to the new mapping population. In pearl millet less than 40 single-copy
probe-enzyme combinations will produce such a map. ~ i l h an avcrage map distance of
less than 15 cM between marker loci (Liu el 01.. 1994).
Jones el 111. (2002) demonstrated that field screening and greenhouse pot
screening of secdl~ngs detect the same QTLs for host-plant resistant to pearl millet downy
mildew using F2 derived F 4 self bulks of a mapping population derived from a cross of
resistant line P 7-3 and susceptible 7042 (S).
Howarth rr 01. (unpublished) ident~fied QTLs for downy mildew resistance and
seedling heat tolerance from pearl millet mapping populations produced from crosses
lCMP 451 x H 771833-2 and H 771833-2 x PRLT 2/89-33. Hash rr d l . (unpublished)
uorked with mapping populations from crosses PT 732B x P 1449-2. 81B x ICMP 451
dnd 841 B x 8638 to locate Qr1.s for reslstance to pearl millet downy mildew. Q'fLs for
host-plant resistance effective against downy mildew African and Indian pathogen
populations were identified in new mapping population based on cross W 504 x P 310
tKolesnikova 2001). and Tift 238D1 x IP 18293,(Azhaguvel. 2001). To date over 65
( ) f l . s for pathogen-populatm-spzc~tic host plan1 rcststance I C ~ pearl nilllel downy
nl~lde\r, have been detectrd (c' K . tiash. pers comm.)
2.10.3. QTL mapping for disease resistance in other crops
h'ith DNA markers and QTI. mapptng. coniplev tbrnls of dtsease resictance m d their
~~nderlping genes are now tir more access~ble to applicd plant hrerders and patholog~sts.
Ouantltative genetics IS unsuited for dissecting polygen~c reststance characters into
dixrete genetic loci or defining the roles of individual genes in disease recistance. With
OF1. mapping. the role of spec~fic re5istance loci can be described, race-specificity of
partial resistance genes can he assessed. and interactions between resistance genes.
growth stage of plant development m d the environment can be analyzed (Melchinger,
1990; Young, 1996).
'The quantitative host-plant resistance system for rice blast caused by Pyriculuria
o):~.zut has been espectally well characterized (Wany t.r (11.. 1994). 'Two dominant
qualitat~ve resistance loci were ~denttfied on chromosomes 4 and I I o f rice (Yu er ul. .
1991 ). Another disease system that has been studied with QTL mapping I S late blight ot'
potato caused by Phyrophrhoru infesrans, an oomycetic pseudo-fungus distantly related to
Sclero.sporu c~rcrminicolu. !.eonards-Schippers C I ul. (1994) identified eleven genomic
segments on nine chromosomes that were associated with host plant resistance to potato
late blight.
Inheritance of disease reaction to leaf spot caused by Cercu.~poru :rue-rnuydis in
three maize F? populations was examined to study quantitative resistance using RFLP
marken (Bubeck er ul.. 1993). One Q'TL on maize chromosome 2 was found to be
significantly associated with resistance in all three populations.
:I stud) o i resistance to bacterla1 \4111 caused Pst~rr~fot~ro~~rrs iolo~tlrcenrum in
tonlalo was reported by Danesh rr ul. ( 1904) using L)Nr\ marker genot!pes ,lnd d~sease
reststance reactions tbr 71 k: indiv~duals. 1\+0 genornlc reglolls irere s~gn~licantl!
~ssociated with resistance. one on chromosome h jnd anorher on chronioson~e 10. LOCI
contributing towards quantitat~\e \artatton In litseast. reststonce have been mapped in
ionlato tbr reststance ,~gainbt Insects (Nlenhu~s '11 01.. 1087). 111 p ~ ~ r o for rcslstance
Jyalnst cyst-nematode ( l i re~he e1 01.. 1903). In peas I'or resistance ayatnst ~scochyta
bl~ght (Dirlewanger rr ul.. 1994). and in maize for northern corn Ieat' blight (Freymark
rr 01.. 1993) and stalk and ear rot (Pe cr 01. . 1993).
Manzanares-1)auleux rr (11. (2000) identified Q'1'I.s iigainst clubroot disease of
Brusslca napus caused by Plosmodiophora hrussicue. IrIh~rilaYIct. o f ('tlrco.~poru leaf
spot resistance in sugar beat was studied by Nilsson er 01. (1999) and they identified
QI'Ls tbr this trait. In sugar beet. four Qr1.s associated with ('rrcos/>uru resistance on
chromosomes 111. IV. VII and IX Ncrc rekealed using composite ~nlerval mapping
~Setiawan rt ul.. 2000). Four ()'TLs were locallzed for the leaf rust tl'rcccrntu horde,)
reststance in barley. which explained 96.196 o f . the segregating yenetlc variation
(Kicherer er 01.. 2000). Brown stem rot (Phiulophoru greguru) resistance QTLs were
Identified by Lewers rr ul. (1999) in a RII, mapping population of soybean using I46
KFLPs. 760 AFLPs and 4 probes for reslslance gene analogs (RGAs)
2.1 1. QTL analysis: Statistical methods
Ja>akar (1970) suggested mathematical-statistical methods for the detection and
estimation of linkage between a qualitative marker gene and a locus influencing a
quantitative character. Since then, experimental designs for determination of linkage
ialndruo~ aql u! paluawaldlu! are l q l pue SILO %u!ddem JOJ hoaql ~ E J ! S ~ ~ J aql pualxa
pue k~!pow l q i spoqlaw les!li(~eue jo ias e paquxap (686 1 ) u !~moa pw J ~ P ~ I
'(C66I
'llouy pue Xa le~ : L R ~ I '~allafl 11961 'depoql) alq!ssod osle ale saqseoidde ialdwis
qanoqlle '(066 1 "1" 1.' ddeuy :686 1 'U!alSIO~ pUE JapUe-1 :686 1 'UaSUJl T:L86 1 'JallaM )
'8u!ddew [ehlalul. paua l a inpa~o~d E ' 1 ~ 0 E ~ U ! I ~ ~ J E I ~ slayJew JO 11ed e jo asn
ayl aAlOAU! leql sisal o!iei pooq~lay!~ jo sueaw 6q io ( 9 ~ 6 1 l a 1alloS) s laynu al8u1s
uo paseq slsal-I qlnolql palJnpuos aq u e ~ uo~lsalap uo!~e!sosse ~ l t ) - i a y r e h
.suo!lelnw!s alox-a;ilel dq pauleiqo a l a ~ qs!q*n 's!silnur
110 10j sploqsalql asuusy!u8!s (10-1 01 suo~lew~xoidde aie~nmr: klqeuosea~ a p ~ t o ~ d
leqi spoylslu paluasa~d (6661) ua[!oo tm,j .slJaga 110 ~ u e x l ~ u d ~ s Bu!nlsap ioj san1e.z
p(oqsa~ql Bu!lew!lsa ioj 'sisal uo!lelnuuad jo ldax~os aql uo paseq .poqiaro Ie>!lrdrrls
m! p?q!J3Sap (0661) filao(] pU€' i[!q3lnqn '(L661 "1" 12 1nOIld) SnJOI lJYJelu te[nslued
E le sassel:, addioua% a q ~ JO smaw s!dLiouaqd aqi uaadlaq sasualajj!p lue~!j!u;i~s
am uaql iagaqm lsa, 01 s! VAONV aqi JO ald!~uud aqj xpoqlalu l a y n u ald!llnm pm
VAONV ia1~0w-dq-~aynw am uo!lsaiap 71 0 ioj pasn saqsco~dde ~ s ~ ~ s s e ~ s OMJ
.passnss!p osle SIIEJI uaawaq suolle!sossr: is31 01 sls180101s4qd iq
asn u! dlluauns spoqiaiu laqw lano s!s.(leur: 71.0 1 0 .suoliellw!l awos pue . s ~ ~ E I u B \ ~ \ .
~s!s.i(eur I.[,() 8i11sn lauuow a t~~e i l~uenh e ul p111~aqur aJr Irt[i z11s11 uaaml.y suollet>o<<t,
~ ~ ~ I s w " ~ T ~ ~ I : u v (11 p3~11iha~ spoq ix~~ 1rslisua;i aqi paq~~ssap ( ~ h h l I 11' I,? 1n011d
'(0661 ' 1" 1" d d h ~ :686 1 'U!2lSlOQ PUT
IapuF.1 :6Xbl 'UXklaf :Oh(>[ 'EX61 'UUULUyx>cI PllC 19llOS :<Lhl 'II!k{ 'EL61 'UwiapIa!)
: I L ~ [ ' w b \ > I ~ ptfr Lloly~l p ~ l i 3 ~ ? ~ [ a p l + \ Llaaq ?\el{ -110 puc ~xol laymu uast\pq
>ali\hare packagc Xl:\P?rl:\KI:KQTI.. In thls. ~titer\al nupplnp I S .~ppl~cd in 4 ..straight
ior\+ard" fashion to sc.\cral populntion t!pes I:ach ~nter\r+l het\\t.cn ;~dj,icent pairs of
markers along a chronlosome 1s scanned and the likelihood protile o i ; ~ 01.1. helng at any
particular point in each intenal is detemlined.
Michelmore r t (11 ( IqOl ) used a nioditicatton of con\entional ( J I l . mapping to
Jetcct VI'Ls tbr do\\ny nilldeu resistance In lettuce in a procedure they called "hulk
>cgregant analysis". tbhich is remarkably sltnilar to that previously descr~hed by l3unon
and Wells (198 1 ) for assessing the value of a trait in near-isogenic F1 popu:ations.
Particularly in the case of cross-pollinating crop populations. interval mapping
has been enhanced to "all marker mapping". To calculate the likelihood of a segregating
QTL. the segregation information of all linked markers is employed. Each segregating
marker may follow a different segregation type, with two to four alleles (Maliepaard and
Van Ooijen, 1994).
An alternate approach was developed by Knapp el 01 ( 19901 rind Haley and Knott
(1992) for QTL analysis uslng regresston. It produces results very similar to interval
mapping both in terms of accuracy and precision, but has the advantage of speed and
51mplicity of programming. This method uses the coefficient of regression of the
phenotype on the genotype of the different markers (Martinez and Curnow. 1992; WU
and Li. 1994). A significant regression coefficient is indicative of an association between
the marker locus and gene(s) contributing to phenotypic differences. The significance of
the association is affected by the degree of linkage between the marker and the QTL and
the type and magnitude of genetic effects of the QTL.
.8u!ddew IeAJalul J ~ J slapow alnlx!w lsuuou paiig pue a e p aPev!l uaamlaq 1uawaa~8r
aql ssasse 01 pasn a n a s a u ,slapow 1 ~ 0 8u!ug JOJ s~ua!sgjaos uo!ssa~8a~ pur
sa lgo~d pooq!laq!l 's[r"p!SaJ UO paSEq SlOOl 3!lsouse!p pJq!JJSJp (1661) llJS3VH
. l . l O
a ~ ! l ~ ] n d aql puc s n q aql usam1aq isusnbaq uolleulqwosai aql jo uo!iJunj aql 1su1e5r
snJol E 1e S U E ~ U Xl/iloua8 J ~ X J ~ U a q ~ uaamlaq asuaiajjlp aA!l!pps aqi 8u!ssar%a~ saAloAul
poqlaw a q j 'uo!lclnLu!s i q JO salenbs )seal palq%!am i q sisal asuesg1u81s su!elqo
pua 'dlsnoauvi[nw!s awosowoIq3 uah18 e uo sueaul Jaqrem aqi [[r 01 laporu 01 sidlualir
11 .qmo~dde uo1sss~8a~ JalJEUI aql pado[a,iap JaqunJ ( ~ 6 6 1 ) sub^ pue desnay
'7.1.0
a~!lslnd aql jo uo!l!std q 3 s ~ IE lapow ieau!l E 01 %u!peal qJomameq s~loidwXse lesol
aqi U! padola,\ap 'ia~aiuend uo!ieJol ,110 aql .(o [c.uaiu! asuapguos aql 8u1lsruisuos
~ o j poqlaur r! paq!Jssap ( ~ 6 6 1 ) '1" I,? u 1 8 u s ~ .1cua1u! asuap!luos aql 40 uo!lsrulsuo?
aldw!s D a \ 1: 01 spoal sap! slq.1 'suol~lsod o w uaawaq lsal O I ~ C J pooq!layl aql , O
uo!inquls!p . X aql jo S I I U I ! ~ uu paseq Ir u a ~ u l asuap!lucn r j c r asn aq1 pasodo~d .s!s.ilrur
affcv[u!l JO p ~ q ! ~ J ~ I ul ' ( ~ ~ 6 1 ) I.? i11aauuo3 ,1001 ~rsr~.ileut. aql se lapom U O I S S ~ ~ ~ ~ J
t: 8u!sn s s o l ~ ~ ~ q t? JO qloeaumj a q ~ 111 ( ~ 6 6 1 ) \\own;) pur z a u ~ u r ~ y i q passnas!p
s ~ l y n u l 8ulyut:~l 8ulsn s . [ . [ o i o s l s ~ . ~ 2 aqllo ar!s aql pur uo!~cjol aql Rullaw!ls4
~ m m y o s L O l d n N D aqi B u y STLO jo suo!~!sod dew sdelds!p 11 .suo!~elndod
s s o ~ x p e q io Z.J jo s!sdlem q u a d )I ,Su!ddew [ehlalu! al!sodwos pue '8u!ddew
~ w a l u ! 'uo!ssai8a1 ia!frew-qSu!s srwojiad )I .swaisds %u!leiado iaindwo:,
s m o p u ! ~ JO s0a 'qsoiu!~e~y ' X I N ~ 1Oj ua)l!Jm alemijos S u ! d d ~ w - - ~ b
e s! 6 . j , . iaqde18ou1?3 IIQ
.8u!ddew Iehlaiur
al~sodwo:, ioj asn oi aJemyos lsa~sea aql s! waiSo~d s ~ q i .Xliuaun3 .sp!oldaq
alqnop j o u0113np01d 10 %UIJI~S )Cq S S O J ~ lauarod!q e woij pahpap suo!ielndod
Bu!ddaw u! s-1j .0 szii3i~emqs pur: az!lv301 01 si a s d n d u ~ e w sl[ '(000;
'./n ra z io :gbbl 'iaau!q:,la# p u ~ ziil) s-1j.0 JO 8u1ddew leualu! a ldw~s p u ~
Bu!ddrw IWNJ)UI a~!sodwos IOJ w r ~ S o ~ d ~aindwos alq1iedwo3-lyal palnq!~ls!p
d l a a ~ e s! a 7 1 b 8 ~ 7 d
'paIpmq IOU aie (:..I 9.a) sassel:, iayrew OMI mql a ~ o w q i i ~ sadXl Xua80ld
.(sau!l palqu! lueu!qwo~a1 lo 'sploldeq palqnop) iua8oid sn08.l;zowoq UOIJ elep
JO s!s61rm a q ~ oi pal:,!~isa~ s! - 1 ~ O f i " X i u a ~ ~ n j ,BuldJew ~ J J ~ U ! a l d y s uuoj~ad
o s p UB3 11 '(9661 ' p m d a ! [ e ~ pen! uar!og WA) s)uauruoJl,%ua ald!llnw u! 8u!ddew
I W ~ I U ! ~ I ! S O ~ L U O ~ 104 ~ 1 ~ 0 1 d 1a1nd~u05 a l q ! l ~ d ~ 0 ~ - ~ ~ 1 ue S I I . l h ~ . . 'slJnp350id 31rlddriu II?\JJlU! pmpurls 8utsn vivp SSOIJ)(:,E~ JO ;.J az,(leur! III,+\
11 ~ s w q i e l d 11" I ~ ~ L U I E uo sun] p m .palnq!lis!p ilaaq 'ilpua!q-iasn s! 11 '(9&6[
"1" l a U ~ D J U I - I ) s~alndwo., bv.VHI Pin qsolu13eI ~q 3Lt2MljOS %u!ddew 71.0
1.1- Canographrr \bas developed by the group of Lrng at Nonh Carolina Slate
University (Zeng. 1993. 1994: Basten ci ' 11 . 1994. 1097). I t dllous rniirkers lo be
chosen 3s cofactors to reduce the background genetic noise and lncreilse the
resolution of OTL detection, This provides an effective strategy fix improving the - ability to detect QTLs of small effect provided that the numbcr of progenies in the
mapplng population is reasonably large.
MapQ'fL. (- , , . . ). .4 conlposite interval mapping
method similar to that :mplemented in QTL Cartographer has been developed by
Jansen and co-workers at U'ageningen Ilniversity (Jansen. 1993: Jansen and
Stam. 1994) called multiple QTL. modeling (MQM).
Multimapper (Sillanpaa and Arjas, IY98), based on Bayesian modeling
and inference. treats the number of quantitative trait loci as an unobserved random
variable using ideas similar to composite interval mapping. This method i s
introduced for inbred lines and it can he applied also in situations involving
frequent missing genotypes.
Qgene is a QTL mapping and marker-aided breeding package written for
Macintosh computer operating systems. It has a user-friendly graphical interface
and produces graphical outputs. QTL mapping is conducted hy either singlc-
marker regression or interval regression.
QTLSTAT is based on interval mapping using nonlinear regression for F2,
backcross. RIL and DH populations and outputs results in graphical form (Knapp
et ul.. 1992; Liu and Knapp. 1992).
PCiRl calculates hased on the iunct~ons o i I-test. cond~tlonal !-test. linear
regression. multiple QTL. modeling and permutation tests (1.u and L iu 1995). It is
for Fz. backcross. RIL. heterot\.gous F I and open-poll~nated populations.
SAS (SAS. Ic)99) is a general statistical analysis .;oit\rare package. It can
detect QTL by identihing assoclatlons beltbeen marker g e n o t > p and qunntitauve
trait phenotype by single-marker analks~s approaches such as ANOVA. 1-test. and
regression (e.g. PROC ANOVA. PROC (i1.M or I'ROC RE(;)
2.12. Reliability of QTL mapping
Kearsey and Farquhar (1998) reported that the available analytical methods locate QTL.
w t h poor precision unless the her~tability of phenotypic data used Ibr mapping a
particular trait is high. Also the estimates of the QTL effects. part~cularly dominance
effects. tend to be inflated because only large estimates are detec~ed as being stntistically
significant. This IS espec~ally problematic where mapping population sire is less than optimal
(as 11 usually is).
Darvasi et (11. ( 1993) showed that the power of detecting a (JPL. was virtually the
same for a marker spacing of 10 cM as for an infinite number of' markers and was only
,lightly decreased for marker spacings of 20 cM or 50 cM. However. a very imponant
consideration is the confidence intenal for the Q'TL position on the linkage group.
Effective utilization of molecular marker technology to manipulate loci controlling
quantitative traits is considered to be dependent on tight linkage between the marker (s)
and the QTL (Dudley, 1993). but in fact. even loose linkages can be exploited in an
applied breeding program (Sharrna, 2001 ).
3. 3IATERIAl.S AND SIETHODS
F: denved Fa self-bulks of a pearl m~llet nrapprng population (skclrton-mapped F!
~ndrviduals) obtalned from a cross of t ~ o pearl rnlllet inbreds. PT 7328 and P 1449-2,
\\ere used as the bas~c rnater~al. PT 7328 (.Appatlura~ 1.1 I , / . . I')H!). In e l~ te ti: dwarf
liybrld seed parent rnalntalner line drbelopcd at Tdmil Nadu Agrrcultural Ulllversity
( T N A U ) and P 1440-2 (ICRISAT, 1997; Singh. 1900) is a tall, downy mrldew resistant
parent, which is a selection developed at ICRISAT from a germplilsrn accession
onginatlng from Mali. PT 4450. an elite poll~nator inbred was used as a male parent to
produce testcross hybrids on each of the F4 self-bulk mapplng populal~on progenies.
PT 4450 I S an elite restorer line being used to produce the commercial hybrid CcHCU-8
(PT 732A x PT 4450) in Tan111 Nadu.
3.1. Test units
One hundred and th~rty-SIX F2 plants were derived from a slngle FI plant from the cross
PT 732 x P 1449-2 were previously selfed at ICRISAT and skeleton mapped at John Innes
Centre, LX. The FI plants were advanced to the FA seed generation at ICRISAT without
selection. For this study the F4 self-bulks of this mapping population were crossed with
pollen from elite restorer line PT 4450, and the resulting 136 tcstcross hybrids, along with
control hybrid CoHCU-8, and testcross of the two mapping population parental lines,
were evaluated in replicated field trials.
3.1.1. Seed multiplication of testcross h!brids
One hundred and th~ny-SIX F, self-hulks along n ~ t h lie t\ro parental Inbred lines
IPT 732B and P 1449-2) and the pollinator (P.S 4450) \\ere sown In Apr~l 2001 (summer
season at TNAU. Colmbatorc).
Seeds were sown In a well-prepared nursery. Emerged seedlings were
transplanted to the maln field. The rnapplng populat~on uas ra~sed in plots
~ccommodating three rows each hac~ng I m length. She adopted spac~ny was 30 cm
between plants and 60 cm between rows. The pollinator was raised along w ~ t h the
mapping population. Two sowings were taken for the F4 self-bulks so as to makc
effective crosslng of all the lines. To get the synchron~sation of flowering multiple
sowlng were taken of the pollinator line: one week before the F1 lines. two weeks
accompanying the F4 lines, and one week after the second sowing of the F4 lines. This
plan provided suffic~ent time to make crosses as well as allow synchronisat~on of
flowering. During flowering, pollen fiom the PT 4450 was collected and used lo pollinate
protected stlgrnas of multiple panlcles of all the 136 F 4 sclf-bulks. Standard package of
agronomic practices were carried out during the entlre crop growth penod.
3.1.2. Selfing
In addition to crossing, selfing was also carr~ed out in all the F, self-bulks so as to get F5
self-bulk seeds for field screening against pearl millet downy mildew.
3.1.3. Evaluation of test cross hybrids
Testcross hybr~ds were evaluated for phenotyping gram yield performance and its
component traits during October 2001 (Rainy season at Tamil Nadu, 2001). Field trials
were conducted in two environments. one at TNAU. Co~mbatore itself and another at the
TNAU Regional Research Station (RRS), Bhavan~sagar. Testcross seeds from all 136
Ilnes, their parents and the commercial hybrid control COHCU-8 were evaluated In an
dpha design with 18 x 18 plots. The testcross hybrids were sown in plots of 2 rows x 4 m
w ~ t h three replications. Inter-row spacing was maintained at 0.6 m and plots initially over
sawn. were thinned w~thin two weeks of seedling emergence to a unlform plant stand of
dpproxlmately 12 plants per row (30 cm spacing between plants within the row) in both
environments, for an average final plant population density of 50,000 per ha.
Recormended cultural practices were followed during the entire crop growth period.
3.1.4. Screening for downy mildew resistance
Selfed seeds from the F4 self-bulks mapping population progenies were used for
screening against downy mildew in both locations (1.e. TNAU, Coimbatore and RRS,
Bhavanisagar) during October 2001. screening was done in sick plot conditions i.e.
fields, having sufficient oospore inoculum. The infector-row technique was followed
(Williams et al . , 1981, as modified by Singh et al.. 1993) to screen against downy
mildew.
The disease screening was done in the following way:
The line 7042 (S) was sown as an infector in every 5Ih row, 3 weeks prior to
sowlng of the test materials to develop a viable sporangial load for the test matenals. At
two-leaf stage the infector rows were spray inoculated with a viable sporangial
suspension (10' sporangia m ~ . ' ) during the late evening hours, afler irrigation. Frequent
furrow irrigation was given during the first 15 days after inoculation to promote high
humidity favoring a higher frequency of infected plants at an early growth stage.
The Fc selfed seed bulks produced by selfing of F, self-bulks mapping population
progenies were sown three weeks after the infector rows sown in the intercrossing rows
after the infection rows have developed 50-60% disease incidence. A well-known - susceptible control (HB 3 ) genotype was also sown along w ~ t h the test material after
every 20 entries to monitor variation in the level of disease incidence across the field.
Test materials and controls were sprayed with viable sporangial ~noculum ( lob
sporangia m ~ " ) when they reached two-leaf stage to increase the likelihood of disease
development in genetically susceptible individuals.
All the test lines and controls were sown In rows of 4m length with two
replications. Standard package of practices were followed.
3.2. Observations recorded in mapping population testcross hybrid yield trials
The following observations were noted in the F, testcross hybrids from both locations.
Time to 50% stigma emergence in days (ET)
Flowering time was recorded as the number of days fTom sowing until 50% of the plants
in each plot produced stigmas on their main stem panicles.
Plant height (PH)
Plant height was measured from the base of the stem to the tip of the panicle at maturity.
Data was recorded on five random plants from the middle of each row, and was recorded
In cm.
Productive tiller number (PT)
Number of productive tillers per m2 was taken by counting the panicles from individual
plants occupied per m' area from the mlddle portion of the rows.
Panicle length (PL)
Length of panicle on the main stem was measured for the same plants considered for
plant height in each plot and recorded In cm.
Panicle circumference (PCR)
Girth of the panicle was measured in cm using vernier caliper on all those plants for
which panicle length was measured, and this was converted to circumference by
multiplying girth by x.
Grain yield per season (GY)
Panlcles were threshed and their grains cleaned. Weight of the grains in grams was
recorded from each plot.
Thousand-grain mass (TGM)
One thousand grains were counted and their weight (in grams) was recorded for all the
entries.
Grain yield per day (GYD)
This is calculated by dividing plot grain yield per season with total number of days taken
to attain physiological maturity (approximated as time to 50% stigma emergence + 25)
and expressed in grams per plot per day.
Single-panicle grain mass (SPGR.1)
rh l s IS the ratio between plot grain yield and the number of productive tillers per plot and
was expressed in grams.
Single-panicle grain number (SPGN)
This is derived from the ratio of slngle panicle gram mass and thousand grain mass and
expressed in numbers.
Grain number per unit area (GNPS)
Grain number per panicle surface unit area is obtained by the following formula:
Panicle gram number
GNPS = .............................................
Panicle c~rcumference x panicle length
3.3. Scoring of disease incidence for downy mildew screening trials
Dlseased plants were ~dentified by the scoring method developed at ICRISAT (Singh
er al., 1997).
3.4. Statistical analysis
The statistical analyses were done using the program, GENSTAT s ' ~ edition (1993).
Analysis of variance, F-ratio and heritability (mean and plot basls) were calculated for
each observed or calculated trait for single-stte data sets from Coimbatore and
Bhavanisagar, and across-locations, for both yield trials and downy mildew screening
trials.
3.4.1. Linkage map construction
.A previously constructed RFLP marker-based genetic linkage map for the cross
PT 732B P 1449-1. developed at John lnnes Centre by Dr. Katries Devos and
co-workers, using the 136 progenies in the current study has used to locate the QTLs.
Thls map consists of seven linkage groups wlth different lengths, which b%ry from 27.6 to
177.6 cM (Haldane), and accommodates a total of 60 RFLP markers (Figures 1.1-1.3).
The linkage map was constructed using the program MAPMAKERJEXP 3.0 (Lander
er ul., 1987).
3.4.2. QTL analysis
3.4.2.1. Data processing for yield trials
Plot values for grain yield and yield components data from Coimbatore, Bhwanisagar
and across-locations were subjected into square root and log-transformations before
regression analysis. Time to 50% stigma emergence, plant height and time to 50% stigma
emergence together with plant height were used as predictors of plot yield performance.
All the traits were regressed with these predictors individually and the residuals from this
analysis were then used to map QTLs for grain yield and its component traits. This
procedure was adopted after initial QTL analyses suggested very strong effects of
flowering..time and plant height QTLs (perhaps linked) on nearly all other agronomic
tralts studied.
3.4.2.2. Data processing for downy mildew screening trials
Data recorded from Coimbatore, Bhavanisagar and across-locations were converted into
percentage disease incidence values and these were subjected to QTL analysis. These
Liukage group 5 Linkage group 6 Linkage group 7
3 3 9 chl
22.3 ih l
Figure 1.3: RFLP-based genetic l~nkage map o f F, mapping population developed from the cross PT 732B x P 1449-L show~ng LG 5. LG 6 . and LC 7
data were also transfomied Into arc-srn \slues (rad~ans) and used for detecting downy
mrldew resistance QTLs.
3.4.2.3. Mapping QTLs for.yield trials
Res~dual data from Cormbatore. Bhavanrsagar and across-locations from 136 mapping
population testcross hybrids were sorted into progeny order corresponding to the marker
genotype data set. QTL mapplng was then carried out using MAPMAKERIQTL version
I . l b (Lander and Botstein, 1989; L~ncoln el ul., 1992a). An additive genetic model from
the program was used because testcross progenies derived from a heterozygous FZ plant
are a sample of the two parental alleles in combination with the tester allele, and the
average of the heterozygote is the average of the two homozygotes (Cowen, 1988; Beavis
et al., 1994; Yadav et a/., 2002), so only additive effects are detected in such testcrosses
and dominance effects can not be detected.
3.4.2.4. Mapping QTl,s for downy mildew screening trials
Percentage of mean disease rncidence and radians from arc-sin transtbrmat~on were used
for detecting downy mildew resistance QTLs from screens using the Coimbatore,
Bhavanisagar and across-locations data sets. MAPMAKERIQTL version I . l b was used
to identify these QTLs. A free genetics model was considered as suitable because
phenotyping was done in the F2-derived Fr self-bulk population.
Eleven important agrononilc tralts, ~ncluding grain y~eld dnd 11s components. were
phenotyped and the~r mean perfbrniance were recorded, .4nalysis of variance was
calculated for all the tralts at Coimbatore. Bhavanrsayar and across-locations. lndiv~dual
locat~on data and pooled data showed sign~ficant difference Ibr all the characters under
study and interactions between genotypes and the locations were not significant for any
of the characters, permitting Interpretation of these traits using only the across-location
means (Tables 1-3). Heritability calculation showed significant higher values for most of
the traits (>50%), which IS a prerequisite for effectwe QTL mapplng.
4.1. Mean performance for different traits
Time to 50% stigma emergence
Testcross hybrids at both locations took a minimum of 40 days for completing 50%
stlgma emergence. Sim~larly 47 days was the maxlmum for completing 50% stigma
emergence at both trial sites. Heritability for this trait was only 51% at Coimbatore and
reached its maximum value (79%) when the statistical analysis was performed using
pooled data from across the two test srtes.
Plant height
The tr~al at Coimbatore had shorter statured plants than that conducted at Bhavanisagar,
but maximum height was almost the same for both locations (177 cm). Heritability for
this trait reached maximum at Coimbatore. while Bhavanisagar had lower heritability
values.
Panicle length
Highest mean values for panicle length uas obtained in Bha\an~sagar. This location had also
highest maximum values fbr panicle length (32.4 cm). Bhavanisagar and cross-locat~ons
data showed maximum heritability values.
Panicle circumferences
The two locat~ons had slrnllar mlnlmum mean values for panlclc circumference, but
Coimbatore had highest maximum panicle circumference (10.7 cm), where as Bhavanisagar
registered the highest mean values for this trait (8.6 cm). Heritabll~ty for thls trait was more
lhan 90% at both locations and across-locations.
Productive tiller number
Data from both locations revealed that mean performance for n i~nimum number
productive tiller number were same. The maxlmum number o fp roduc t~ve l~l lers was also
same for both locations. Across-locations data had the highest broad sense hentability
(73%) for this trait.
Thousand-grain mass
It was observed that thousand-grain mass reached minimum value (6 . lg) at Coimbatore and
had a maximum value of 12.6 g at Bhavanisagar. Individual locations and across-locations
had high broad-sense heritability values for the trait (97 to 9896).
Single-panicle grain mass
Bhavanlsagar had htghcst Inem \ d u e (8.9 g) but Coimbatore rtg~stcrcd the maximum
observed value (14.4 g) for t h ~ s trait. Broad-sense heritabil~ty calculated acrcss-localions was
the hi$er (839'0) than that !?om indib~dual locations data.
Single-panicle grain number
Values in Bhavan~saga ranged from 685 to 1302 g and the mean values attained the
maximum of 953 g. Heritability (plot basis) was very low in both Locations (3 1 and 21%, for
Coimbatore and Bhavanisagar respectively) but broad-sense hentab~lity was more than 50%
ibr [he across-locations analys~s.
Grain yield per day
Bha~anlsagar had maximum values for ga in yield per day and ~t also had the highest mean
values. Where as Coimbatore had the mintmum value for thls trait. Hentab~lity (broad-sense)
for p n yield per day was 8796 when pooled data were taken for consideration.
Grain number per unit panicle surface area
Both locations registered similar minimum and maximum values for g a i n number per unit
panicle surface area. Also both locat~ons had low plot-basis heritability but broad sense
heritability at across-locations had higher values (55%).
4.2. Correlation studies
Grain yield
G r a ~ r ~ yleld per season 1s the ultlnlate trait that was taken first as an explanatory vanable
and correlated with other traits to lind the relat~ve contribution of cach constant tralt to
the observed yield variat~on. The results are shown In the Table 4.
Plant he~ght, psn~cle c~rcunifercnce. thousand-yra~n mass. s~nyle-pan~cle grain
mass, single-panicle grain number. grain yield per day and g a i n nuinber per unit pan~cle
surface area had positive correlations with graln yield per season. Tra~ts like time to SO%
stlgrna emergence, panicle length and productive t~ller number were correlated negatively
w ~ t h grain yield at both locations. Coimbatore showed the highest positive correlations
for plant height (0.620), pan~cle circumference (0.642), gram y~eld per day (0.983) and
s~ngle-panicle grain number (0.283). For grain number per unit panicle surface area.
Bhavanisagar registered the higher correlation. In both locations the trait grain yield per
Jay was closely correlated with grain yield per season.
Coimbatore had higher values for traits negatively correlated w ~ t h grain yield per
season 1.e. time to 50% stigma emergence (-0.520) and productive tiller number (-0.421)
than did Bhavanisagar where panicle length was highly negat~vely correlated (-0.527)
w ~ t h grain y~e ld per season.
Time to 50% stigma emergence
Productive tiller number and panicle length were the two traits showing positive
relationships with time to 50% stigma emergence at both locations. Other characters
rncluding g a i n yield per season showed a negative correlation with time to 50% stigma
emergence.
aql aio,eqrn!oD ul ' (%~ 'zs ) mappu Aumop 01 Bu!qmn~>ns uo!lelndod aql joj leq ~ I ! M
alnssa~d aszas!p ~aqB!q e peq le8es1ue~eqa n q l palea,\aJ v/zONV mOJj Sllnsall
' S a1qeL u! umoqs ale uoi~eue,t
jo slua!3y~ao3 .11aq1 pue suo11e3ol-sso~3e pue n%es!ue.zeqa 'alo]eqm!o3 1e a3ueuuojlad
uo!lzal aseas!p JOJ V/ZONV moq sllnsa~ aqL .suo!leln31~s le311s11els Jaqunj loj
p s n aJam elep asaq l ,sue!pej u! laqmnu s!qljo u o ~ ~ e u u o j s u e i ~ rl!s-sle pue (%) a3uapl3ul
aseasip oiu! p a u a ~ u o 3 aJam ]old lad lunos lueld paseas!p pue [el01 ~ o j san!ea [eu!Suo a q ~
a>uap!au! assas!p
Joj s(sp1 % o ! u a a ~ ~ s mapljcu 6umop o! so!talndod jo ~ > U E W J O J J ~ ~ o a a w \ ; ' ~ ' p
1q81aq lueld qllm SllCJl p5lF15U03 ~ [ ~ A I I P ~ J U Jql J l J M ql8ual a l31u~d
pue Jaqmnu J ~ [ { I I a ~ ~ l m p o ~ d ears llun lad laqlunu ~ 1 ~ 1 % pue iaqmnu ule18 al31wd-a[lu1s
JOJ sanleA ~ A I I I S O ~ 1aq81q peq n 8 e s 1 u e ~ e q a puoq laqlo a q ~ uo Aep lad pla~d U I ~ J R
pue sselu u1e.18 al31ued-al4u1s 'ssem UI~J%-puesnoq~ ' a ~ u a l a j m n ~ ~ ~ : , alslued JOJ sanlez
~a@rq peq a.101eqm103 S I I ~ J I a saq~ Buome pup 1q81aq lueld ~ I I M s11e.11 palelal Ala~rllsod
a q ~ aUm vam llun lad Jaqlunu u1e.18 pue dep lad p l a ~ i u 1 e ~ 8 'daqmnu urn14 alsrwd-al8u1s
'sselu u1e.18 als1wd-a~8u1s 'ssem u ~ e ~ l - p u e s n o q ~ ' J ~ U J J J J L U ~ ~ J ~ ~ Jl3lUEd SIIEJI J J ~ I O
~ I I M peq uosess lad plalX ulelS lo 1rq1 sr sdlqsuol~ela~ Jr!lluls p a l ~ q ~ q x s lq81aq ]UPld
l@!aq Il'eld
~1~4es1ueneqa ie san1e.t Ja.no1 pamoqs ' ~ a q u n u
u!ell-puesnoq~ ~dasxa eaJr llun lad Jaqmnu ule~S pue Xep lad plaid u ~ e B .Jaqmnu U ! C J ~
al"ued-al4uls 'ssrul tr1e~4 s[slued-alRuls 'asuaJajurnsjls alslued .]qfilaq lueld a111 slleJl
~ a q r o 11e JOJ '~111x1 ps~r l suos Allnllefisu JOJ saI1le.i ~ a q R ~ y pa.noqs a~o leqm103
Allllqeqo~d j o la\al I aql re IU~JIJIUYIS .. '(1111qeqwd j o 1.~31 j IE III~JIJIU~IS.
n (YXJJnS 331ued 10-UJ lad) I2qlunU UIElZ) :SdNZ) :(.<ep!.lu@) (ep ld pl>l,< UlRl9 a d A 9
.,-. l a q ~ ~ i n u u1e~3 1131ued-a(3u1q NWS .(a) smru U I ~ a ~ ~ r d - a l l u ~ q CYjdS . ( Z ) ssplu u ! e f i - m ~ . ,~91 .(.lu;d) ~aqwnu ql!~ a~!t,npmd :u (UIJ) ~JU~~~IIII~JII~ ~IJIUE~ 8~ ~ ~ 1 1 ) 414~3 4 ~ 1 u e d . i d ~ ~ 1 2 ) 1q81aq )weld ~d .(s(ep) a~ua&um elu81rs -60Z'o~ a r u l ~ :lj .(;tui4)pla!.( 111133 :AD
I m l w ' W F I I .L91 O ..LLBO ..!f$O .Wl V LC1 0- ..Iff V m - - 561 0 01100- .iL10
a~laqmlo.) I S M O ..6r80 ..zno .SLI v rol n ..lot v .&so- rsoo w n s m n
c ~ ~ ' ' f ~ r ' l ~ ' ' V I f f I 0 .+OW0 ..68PO e691 0- Oi l f t ..91ftt ..009V 8PI 0 6ROV I f1 0 WNL) 1~2mtrn(t-q(1 I ..19iO ..VSLO .tfSLO ..SWO ..f950- ..L8f V ..8IVO ..8<4V ..iRhIl
>~o~rqto!n ) I .+8biO ..f08O ..f6i 0 ..$:911 5 t S t ..6$911 ..SWll- ..fRhll I , I r.hLif l ..nun t ..lwn ..ltso- ..litct ..IIW~ .,8<VlV ..lnho ad.\!)
IP~~PIIIIPIPII~I 1 .+I990 W1l1)- 9f111 .IMII)- ..tsi(l- r.lSi0 0;Ilb- ..X9cll vwrq<toln I ..ts911 t t ~ w . + t ~ 1) .i91 1 1 ..nt~ t t ..;f;ll ..it: 0- ..i~<ll
, t~v t~ -dq $ % ) t y I ..RS91) GLOlt .tXI 11 .VLI lt i it^ w ,+(t i lb .!XI 11- ..VLill N!*IS
rram~ar rrq$l I ..lWO ..9<S 11 ..RtS 0- ..oL¶ 0- ..SLS I1 .rLXPll- ..')tL 11
. v , v ~ ~ t , t t u ) I ..81~n ..LWO . . t c s ~ ..~980- ..xsvn ..ftso- ..kin \ttot~cwl ,VU?\J I ..ih90 ..Z190 ..bfSIt ..WRO- .rhlV1l . . l l S I l - r 4 9 L O W%Is
~rd=~ur\vq i l I ..fi90 ..:t<ll- . . 8 l t ~ . . I~SO . . ~ x f l - . * i 5 ~ 0 . v ~ % z c q ! u ~ ~ ~ ) I ..IX911 L S I 8 I- ..fLV 11 ..ills 0- ..IKL 0
,,,,,,lc7*,l q,u,7,, I ..fwo s s t ..MWO I IS^- ..ZVLII W ~ L IP?IIIUP\PI((I I .sibs (I- e.861 V ..t8S 1) ..llht(l- .rlLSII
.3x,lrqtt>w ) 1 ..~090- ..9fs11- .. f f~O ..Liiw ..~WII \I!OIIP%,~ ~~~~~~~ I ..RbS 0- .+L9t0- ..IW 0 .riiSO- .rMW11 X.Xl
IP~PCIIIP\P~$I I ..iOtO ..MS 0- ..t8b 0 ..LiSll- ZIOIFL(IIIIII ) I ..O5PO +.8590- ..9Ttll ..6Vtll-
~ t t , ~ t l ~ ~ n l \s~~,,,, I ..$it n ..LZV 11- ..OL t II ..mm ~d
;I.~P\IIIP 1~1111 I ..9Ti'O- ..OTf O ..CVCO- >I,IIPqllllo ) I r+tcs 0- ..hit 11 .. l i tW
~ t r , ~ o l r 7 ~ ~ l %c,J7,. I ..n~to- ..LL\ 11 ..;t4 0- .).a
J ~ . ~ ~ ~ L I ~ , , ~ , ~ ~ ~ I ..s~so- ..tusn .Irc?s2qntn,,) I ..!LC 11- rrOZ9 0
S L , , ~ ~ I P X ~ < y ~ ~ , " I ..ors o- ..ctw n H,I ,,.d,.\,"~,P,l" I ..LOSO-
;I>nlPqtclty) I ..nzs o- \IIIIIIFXI( C C I > J Z ~ I ..hO5 0- L3
!"3mr,l"\!.,~~~ I 3,,,l,~q,,,,,>,) I
,,,<>, lP~'~~ %<<>,y I A!)
- d W!)rlS 1 - 834 Xi Ld I id I I -. -~
. . - ;()0;, I ()(I; "uc'111!3o[ ~ 0 1 ~ ~ pur I~RFFIIIR.\V~~ 'aJoll?qullo,) IP %p!~q ((1 ?SOJ~IS~I 111111rlndod 3 i11dd l?~1 J O ~11)1?111 ~IOIIP~~UOJ :t, alqej,
disease incidence was also severe. reach~ng nearly lit)',. Hcrltability lor the disease
~ n c ~ d e n c e values (for both percentage diseased plants and the arcsin-transfbrmed data)
were high enough for use to niap QTLs. D~sease ~ n c ~ d e n c e ( O O ) at Bhavanisagar had the
highest heritability (plot basis). T h ~ s locat~on also showed highest herita6ility for arc-sin
transformed disease incidence
4.4. Information on linkagemap
A Previously constructed RFLP linkage map was used for this study. This linkage map
was developed From the cross PT 732 R x P 1449- 2 using 58 RFLP probes detecting 60 loci.
Seven linkage groups (LGs; singular = LC) were constructed using MAPMAKEWEXP
multipoint analysis with the LOD threshold value of 2 0 and a recornbinallon fraction of 0.5.
The mlnirnum and maximum length of linkage groups varied from 27.6 cM (LG 3) to
177 6 cM (LG 1 ). These linkage groups were used Ibr mapping QTL for both in test cross
hybrids for yield and its components tralts and In F5 population lvr mapplng downy
rn~ldew resistance QTLs.
Linkage group 1
LG I has a length of 177.6 cM (Haldane units) and t h ~ s is thc lengthiest LG in pearl
rnlllet. It accommodates 15 markers with d~fferent marker intervals. LOD score of this
group was very high (-365.0), much higher than the other linkage groups.
Table 5 : ANOVA for percentage and arcs~n-lransiormed \slues for do\*ny ni~ldew tncldcnce from tnals conducted at Colmbatore. Bha\onlsagsr drld across-loca~~o~is. 2001,2002
- -- - --- - - Paramerers Co~mbarore Bhd~dn~sdgdr Across 1 acaMons
- - .. -- --. . -- . -.. -- ~
percentage radlans percentage radlarls percentage radians . . ~. .- - SEF) 12.70 0 20 1 1 57 0.17 857 0.11
Mean 49.2 0 57 51.6 0.01 5; 9 0.55 CV (9'0) 36.52 49 45 31.10 38.01 23 81 27.94 F,rat~o 3 83** 3 03** 4 73** 4 $2'. 401** 3 86.' 11: (plot bass) 59 5 0 (15 Oh 00 5 9 i i (mean bass 74 07 ' 9 7) 7 5 7 4
n, ' SE oe ' 3 1** J 2** 3 0 3.3'' 3 X b * l . l O *
S~gn~f icant at the 0 1 level of probab~l~ty
Linkage group 2
f h ~ s group has S I X markers w~th the total d~stance of 87 ') cM She order ot' [Ire markers
on t h ~ s l~nkage group 1s .Ypsn1708ri. .Yp~n1706. .Ypsm?s. .Ypsnr.SY?, .Qsnt32l and
Ypsnr70Rb The LOD score of t h ~ s l~nkape group \\As -170 39
Linkage group 3
Tlic total length of t h ~ s l~nkage group I S 27 6 cM Th~s 1s the shortest pearl rn~llet l~nkape
group although 11 accommodates I0 markers. The LOD score of this group IS -162.28.
Linkage group 4
T h ~ s group has the length of 100 0 cM and has I I markers w~th optlmum ~nter-marker
~ntervals to detect QTLs.
Linkage group 5
SIX markers occup~ed LG 5 The maxlmum l~kcl~hood positlon of the marker ~ntervals I S
Ypsm815, .Ypsm328, .Ypsm73.4, ,Yrmll-I. .Ypsm719 and .Ypsm735o The total length of
thls linkage group IS 30 2 cM and its LOD score I S -137 84.
Linkage group 6
LG 6 accommodates seven markers In a length of 83 1 cM This group has the LOD score
of -205 52
Linkage group 7
Thls IS the smallest linkage group In terns of number of markers. I r has only f i ~ e markers
~ n d their correct order 1s .\psttr!69, .'t'rrtr9-!h. .\psnlblS. .Yps1117l; and .Yps1tr834. The
length of this group 1s 37.6 cM tc~th LOD score of -1 43 03
4.5. Mapping QTLs
I'he constructed linkage map from the cross P'T 7 3 1 0 x P 1449.2 uslng 136 individuals
was used for mapplng QTLS lor y~eld and 11s related tralts. Software package
MAPMAKERIEXP version 3.0b was used for constructing linkage groups and
MAPMAKERIQTL version I . I b was used for detecting QTLs.
4.5.1. MAPMAKERIQTL
The interval mapping method as implemented In MAPMAKEWQTL was used with a
LOD of 2.0 as threshold value for detecting significant QTLs. The additive genetic model
tiom this software package was used as the phenotyp~ng was done In testcross hybrids.
For thls, the command "sequence [all: additive]" was used to restnct the genetic model
only to additive effects.
For mapping downy mildew resistance QTLs, phenotyping was done in the Fp self
bulks. So, all possible genetlc models (additive, dominant and recessive) were
considered. This is carried out by uslng the command "sequence [all]". Combined effects
of multiple QTLs were calculated by multiple QTL models for two QTLs, three QTLs
etc. The qualifying criteria for accepting a multiple QTL model was a LOD score of two
units more than the highest LOD score of the best model having one less QTL.
1-OD = Min~nium qual~fylng [.OD score for acceptance of a multiple
QTL model with (11) QT1.s.
= ? + L O D ( n - I )
LOD ,,.I, = Max~mum LOD score for observed model witk (n- I ) QTLs.
4.6. QTLs for agronomic traits
.4 total o f 18 QTLs were '~dent~iied across seven linkage groups for nine traits, but
yenomlc regions flanked by only seven markers loci controlled all these QTLs. The
details of the QTLs detected on different linkage groups are shown in Tables 6-22.
Graphical representation of LOD values obtained from different types of transformation
for diffcrent traits are shown in Figures 2.1-7 3.
Mean values at Coimbatore, Bhavanisayar and across-localions from the mapping
population testcross [(PT 7328 x P 1449-2) x PT 4450) consisting of 136 hybrids were
used for mapping QTLs for the different tralts. Square root and log-transformed values
from Coimbatore. Bhavanisagar and across. locations were used to map these QTLs in an
attempt to reduce distribution abnormalities in the trait data set. Plant height, time to 50%
stlgma emergence, and plant height together with time to 5096 stigma emergence were
used as predictors of other traits using liner regression, and the res~duals from these
regressions were used to locate QTL positions. Most of the detected QTLs are situated on
LG 4. LG 2, LG 6 and LG 7 are the other groups having QTLs. No QTLs were detected
on LG I , LC 3 and LC 5.
Time to 50% stigma emergence
.A slngle QTL was ldentlfied for time lo jOoo stlgia enicrgericc 21 C.u~mbatore, uslng
log-transfom~ed data. T h ~ s QTL for tlme ro 5090 s t l q a micryencc IS sltuatcd on LG 4 and
explained 7.8% of observed phenotypic vanation with a LOD value of 2 . I . The additive
effect of the P 1449-2 parent allele at this QTL decreased flowering by 0.5 day
Plant height
.A slngle QTL was mapped on LG 81 for plant height. This QTL liad its minimum LOD
score of 2.83 at Bhavanisagar when l o ~ transformed values were used. But the maximum
LOD value of 6.95 was obtalned at Colmbatore when square root transformed values
were used. At this maximum LOD a maximum explanation of observed phenotypic
variance was (23.9%) also obtained. Additive genetic model gave the maximum value of
0.7984 for thls maximum LOD score, which corresponds to an increase of plant height by
one cm when the P 1349-2 parent allele is present.
Panicle circumference
For panlcle clrcurnference one QTL was ident~fied on the bottom of LG 4. The panicle
clrcumference QTL was observed between the marker loci XpsmSl2 and Xpsm344 when
regressed against plant hcight and time to 50% stigma emergence at both locations. The
significant LOD score for this QTL ranged from 2.46 to 7.46. The phenotypic variance
ranged from 10.1 to 26.6%, depending upon the data manipulations used prior to QTL
mapping. At the maximum LOD value (7.46) the additive effect of the allele from P
1349-2 increased the panicle circumference by 6.5 cm.
Panicle length
ihls tralt had d s~ngle QTL. tvhlch IS locdted betireen the marker IOCI .Ypsmj68 and
Ypvmjl? on LG 3 Nearly all types of transformat~on of data from horh locat~ons and all
[he res~duals from different types of funct~ons detected t h ~ s QTL T h ~ s @I'L dt the LOD
score of 6 5 2 evpla~ned 22 750 of the ohscned phenotyp~c Landnce dt Colmbatore
rhousand-grain mass
Two QTLs were ~ d e n ~ ~ f i e d for thousand-gram mass These QTLs are both located on
LG 4 but at d~fferent Intervals (,Ypsm306- .Ypsm4?lc and .Ypsm568- .Ypsm512). A
rnaxlmurn LOD score of 7 4 was obtalned for thls tralt by using square root
transformat~on of data from Bhavan~sagar when plant helght used bs a pred~ctor But the
maximum portlon of observed phenotyp~c vanance ( 1 1 6%) was expla~ned when tlrne to
50% stlgrna emergence was used as a pred~ctor of t h ~ s tralt
Grain yield per season
One QTL for gram y~eld per season was mapped at the bottom of LG 3 T h ~ s QTL was
detected when gram y~eld per season was regressed on tlme to 50% stlyma emergence
from both types of tra~sfotmat~on. Square root transformat~ons and log transfonnat~ons
gave more or less slrnllar LOD scores ( 2 6) and R' values (10 0) They also exhlblted
slmllar addltlve effects (0 SI), wh~ch correspond to an Increase of grain y~eld per season
by 0 3 g/m2 when a P 1449-2 allele replaced that of PT 7328
Grain yield per day
rhree QTLs \\ere ldentlfied for prun );~eld per dab JI \a~.lous ~ n t c r \ ~ l s on l.(j 1. These
~ntervals are .'fpsf?rYJ to .Y)s~~rtiI. .\pstnj68 to .Yps1115l? ~ n d .\j1~1rr306 to .\psn142lc. The
middle QTL between marker loc~ .\psrrr568 and .Ypst~rSI?, recorded thc maxllnurn LOD
(7.71) and explained the largest portlon of the obsened phenotypic ~anGice (10.7). This
\vis obta~ned by regressing grarn yleld per day J_qalnst llnlc to 5O"i, stlgrna emergence
uslng log-transformed data from Bhavanisagar.
Productive tiller number
:I maxlmum of four QTLs were obtalned for this tralt on four d~fferent linkage groups
(LG 2, LG 4. LG 6 and LG7). The maximum LOD peak of 2.02 was found at Coimbatore
uslng log transformation together with time to 50% stigma emergence as a function. A
rnaxlmum of 15.44'0 for R' was explained by a slngle QTL, which was located on LC 2
(between .Ypsr1132/ and .Ypsrn70Rh) with the additive effect of 0.6900 corresponding to a
decrease of t~ller number by 0.3 m" when the PT 1449-2 allele was replaced for that of
PT 732B at this locus.
Single-panicle grain mass
.Ypsrn84- .Ypsm6/2, ,Ypsm579- .Ypsrn613b and ,Yrrn9_2h- Xpsw618 are the three marker
loci intervals accommodating QTLs for this trait on LC 4. LG 6 and LG 7 respectively.
The QTL on LG 6 explained more of the observed phenotypic variance (15.4%) than
other QTLs, and had a LOD value of 2.19. However the QTL on LG 4 had the highest
LOD score (3.58) and explained 12.4 % o f observed phenotypic variance.
2 5
2
-PT SEPH 1
0 5
0 2 3 4 5 6 7
LOD scores across seven pearl millet linkage groups
[rgend P('R-parr~cle circumference. PI.-pan~cle.iength. SP(iM-s~ngle-paoiclr gram nlasc. 1'1'-producl~ve 11lier ~ n ~ m h e r . PH- plant he~ghl as predictor. _SEPH- time lo 50°c atig~nn emergence and plant he~ght as predictors
Figure 2.1 : QTL LOD peaks for various traits using plant height and ti~rle to SO%, stigma cnlergcnce together with plant height as predictors of log-kansformed values from Coirnbatort: yield trial.
2 3 4 5 6 7
LOD scores across seven pearl miltal linkage groups
Legend PCR-panicle circumference: pH-plant he~ght; PI.- panicle length: PI - product~ve t~ller number; SPGM- Flngle-panicle grain mass. -SE- time to 50% stigma emergence as predictor
Figure 2.2: QTL LQD peaks for various lraits using time to 50% sligma emergence as a predictor of square root-transformed values from Coirnbatore yield trial.
1 2 3 4 5 6 7
LOD scores across seven pearl millet linkage groups
I (;end 1 7 1 ticink! to 50% stigma emergence: PCK- p m ~ c l e c~rcumkrencr , PL-pan~cle length. S I J ( i M - ~ ~ n g l c - p a n ~ c l e gram .,,I,.. PT- productlvr tlller number. P H - plan1 he~ght as pre+ctor. _SF.PII- tllllr to 50% s!xgInn rlllergrl~ce and plant h e ~ g !.predictors
I ilgure 3 1 : QTL LOD peaks for various traits using plant height and time to 50% stigma enlergence ~wgethcr with plant height as predictors of log-lransfornled values li.orn Coimbatore yield trial.
I
1 2 3 4 5 6 7
LOD scores across seven pearl millet linkage groups
]Legend. GYD- grain yield per day. PCR- panicle circumference; PH- plant height, PI.- panicle length; PT- productive tiller number: SPFM- single-panicle grain mass: -SE- tlme lo 50% stigma emergence as predictur
Figure 3.2: QTL LOD peaks for various traits using time to 50% stigma emergence as a predictor of logtransformed values from Coimbatore yield trial.
2 3 4 5 6 7
LOD scores across reven pearl mlllst linkage groups
Lagend: PC'K- p a ~ ~ ~ c l r c~rcumference: PL.- ~)aoiclr length. 1'OM- thousand-g~a~n mass. _SF.PH- ti111e to 50% stigma emergence and plant height as pred~ctors
Figure 4.1 : QTL LO11 peaks for various traits using plant height and time to 50% stignla emergence as a predictor of square root-lransfonned values from Bhavanisagar yield trial.
2 3 4 5 6
LOD scores across seven pearl millet linkago groups
]&lend: GY D- grain yield per day; GY- grain yield; PCK- panicle circumference, PH- plant height: PI,- panicle length. TGM- thousand gram mass: _SE- time to 50% stigma emergence as predictor
Figure 4.2: QTL LOD peaks for various traits using days to 50% stigma emergence as a predictor of square root-transformed values from Bhavanisagar yield trial.
l'able 16: QTL assoc~ated w1tl1 grab11 y~c ld-dr teml~n~~,g !I-alts of pearl millet Inapplng progeny rcstc~oss 11yh11ds us~ng clnic tv SO?,, stlgnla emergence as a predictor of log-translbrmed values of other tralts at Bhavan~sagar.
. .- 11nk) ,live ac -transToiiGd Trait Marker interval moue " *,&IS "ad:~tive effects - . --
Grain yield per day Xpsn1568-Xpsm5 12 4 4.0 2.71 10.7 0.5208 3.3 g
Grain yield Xpsn1568-Xpsm5 12 4 4.0 2.68 10.6 0.51 77 3.3 g
Panicle circumference Xpstn5 12-.Ypsn1344 4 2.0 5.89 21 4 0.7378 5.5 cnl
Plant height Xpsn15 12-Xpsnr334 4 4.0 2.83 9.9 0.5 148 3 3 c m
Panicle lennth Xpsn15 12-Xpsn1344 4 0.0 5.9 1 20.1 -0.6947 0.2 cm
Table 17: QTL associated with grain yield-determining traits of pearl millet mapping progeny testcross hybrids using plant heigl~k togethrr with time to 50% stigma emergence as a predictor of log-transformed values of other traits at Bhavanisagar.
-- L~nkage Position LOD i t~ve ac -trans o r n i i
Trait Marker interval WOUP " Z e c t s "ad~itive etftkts - -
Panicle circumference Xpsr115 12-Xpsr11344 4 2.0 2.90 11.3 0.5391 3 5 cnl
Panicle length Xpstr15 12-.+st11344 4 0.0 ~- - 3.36 12.0 -- - 0.5381 0.3 cm - -
1 2 3 4 5 6 7
LOD scores across seven pearl m~llet linkage groups
lrgend. GYD- gram yleld per day, PTR- pan~cle c~rcumference: PI: panicle length. 1'C;M- thousand-grain mass. .PI{- plant helght as predictor. SEPI I - tlme lo 50% stigma emergence and plant height as predictors
Figure 5.1 : QTL LO11 peaks for various traits using plant height and time to SO% stigma emergence as a predictor of log-transformed values from Rhavanisagar yield trial.
8 '- G Y D SE
' GY-SE
1 PCR SE I PH SE
2 3 4 5 6 7
LOD scores acroas seven pearl millet Ilnluge groups
Legend: GYD- gram y~eld per day, GY- grain yield, PCR- pan~cle circumference. PH- plant height: PI.. panicle length: -SE- time to 501 stigma emergence as predictor
Figure 5.2: QTL LOD peaks for various traits using time to 5090 stigma emergence as a predictor of log-transformed values from Bhavanisagar yield trial.
2 3 4 5 6 7
LOD scores across seven pearl milkt linkage groups
Legnd: GYD- grain yleld per day:PCR- pmcle cucumference: PL- panicle length: TGM- ihousand grain mass: GY- grain yield; PH- plant height; -pH- plant height as predictor: -SE- time to 50% s t i p a emergence as predictor: -SEPH- time to 50% stigma emergence and plant height as predictors
Figure 5.3: Comparison of QTL LOD peaks for various traits using different types of predictors of log-transformed values fmm Bhavanisagar yield trial
Table 19: QTL. associated witl&gra~n y~eld-detemi~nlng l ~ d ~ t s ofpearl m~llel nlapplng progeny testcross 11yb11ds uslng Ilnle to SO':.. st~gnia eniergence as a predictor of square I-ool-lransfonlred values of other trails at across-localions.
Marker interval L'nkage Position LOD ltlve axitrans o r n l T Trait
group R' ",":ccts 'addit~ve eLects
Panicle circumference Xpssz5 12-Xpsnz344 4 0.0 5.97 21 -0 0.7059 0.5 cm
Plant height Xpsn15 12-Xpsnz344 4 4.0 6.48 22.5 0.7752 0.6 cnl
Panicle length Xpsnz568-Xpsn15 12 4 6.0 6.22 22.1 -0.7406 0.5 can
Single-panicle grain mass Xpsnz84--l(psn1612 4 4.0 2.78 -. . - -
9 0 0.4812 0.2 g -
1 2 3 4 5 6 7
LOD scores across seven pearl millet linkage groups
legend PCR- pan~cle c~rcumferance: PL.- panlclz length: PT- product~ve tlllrr numher. SPGM- s~nglz-patiiclr grain ~nass. _PI+- plant he~ght as predictor
Figurc 6.1 : QTL LOD peaks for various traits using plant heighl as a predictor of square root-transformed values from across-locations
2 3 4 5 6
LOD scores across seven pearl millet linkage groups
Legend: PCR-panicle circu~nference; PH- plant height; PANICLE LENGTII- pan~cle length: SPCM- single panicle grain mass: _SE- time to 50% stigma emergence as predictor
hgure 6.2: QTL LOD peaks for various traits using time lo 50% stigma emergence as a rredictor of square root-transformed values from across-locations
1 2 3 4 5 6
LOD scores across seven pearl millet linkage groups
Izgend. PCR- pamcle clrcumferencr: PI.- pan~cle length. PI'- product~ve t~ller numbel. SPOM- s~ngle-panlclr gram mass, _pH- planl he~ght as predictor: _SEPII- tlme to 50% sngma emergence and plant height as predictors
Figure 7.1 : QTL LOD peaks for various traits using plant height and time to 50% stigma emergence as a predictor of log-transfornmed values from across-locations
2 3 4 5 6 7
LOD scores across seven pearl millet linkas4 groups
k g e n d : PCR- panicle c~rcumference. PH-plant height. PL- panlcle length, SPGM- single-panicle gram mass: _SE- h e to SOW stigma emergence as predictor
Figure 7.2: QTL LOD peaks for various traits using time to 50% stigma emergence as a predictor of log-transformed values from across-locations
2 3 4 5 6 7
LOD scores anou seven pew1 millet linkage groups
Legend. PCR- pan~cle cucumference: PL- pan~cle length: PT- productive tiller number: SPGM- s~ngle-panicle gain mass; PH- plant height; -pH- plant height as predictor: _SE- tirnr to SO% stlgma emergence as predictor: -SEPII- time to 50% stigma emergence and plant height as pd~Ct06
Figure 7.3: comparison of QTL LOD peaks for various traits using different types of predictors of log-transformed values from across-locations
1.7. QTLs for downy mildew resistance
Data on total and d~seased plant counts per plot were converted to disease incidence (%)
and arc-sin of this number in radians. These values were used for mapping QTLs for
downy mildew resistance. A total of five QTLs were obtained from Coimbatore,
Bhavanisagar and across-locations. LOD peaks are shown in Figures 8.1-8.3.
Coimbatore
Two QTLs were mapped uslng downy mildew screening results from Coimbatore. Both
mapped to LG 2 at different intervals. The QTL located between the marker loci
.Ypsm708a and Xpsm706 had the maximum LOD score (4.77) and explained as much as
18.9% of the observed phenotypic vanance. This was obtained from arc-sin transformed
values. The P 1449-2 allele at this locus mean was associated with lowerir~g of disease
incidence by 2%. This QTL for disease resistance behaved largely as if it was dominantly
inherited (Table 23).
Bhavanisagar
This location also had two QTLs but situated on different linkage groups, i .e, on LG 1
and on LG 4. The QTL at LG 4 had a higher LOD score (3.69) and explained a greater
portion of the observed phenotypic variance (41.5). This QTL is mapped between marker
loci Xpsm464 and Xpsm716, and was inherited recessively. The P 1449-2 parental allele
at this QTL had an additive effect of 2% mean disease incidence (Table 24).
Across-locations
Totally two downy mildew resistance QTLs were identified for across-locations data.
One was similar to that of the QTL found both at Coimbatore and Bhavanisagar, which
was located in LC 2. Another QTL. a new one was mapped from across-locations data.
which was not found In ~ndividual locations. This QTL has Identified using multiple
QTL model by fixing the previously mapped QTL, wh~ch located between the marker
loci Xpsm 7080 to Xpsm 706. The command "sequence [.Ypsm 7080- .Yp'psm 706:additivel
[all]" was used to get this new QTL. The program fixed first QTL at ;his location and
identified the'second QTL wlth the LOD value of 4.67, which was more than 2 to that of
the fixed QTL (Table 25).
1 2 3 4 5 6 7
LOD scores across seven pearl millet linkage groups
1,egznd. C'O- C'oimbatore locat~on. _PER- percentage. _ARC'- r td~ans ol arc-s~n
Figure 8.1: QTL LOD peaks for downy mildew resistance from Coimbatore ttial
1 2 3 4 5 6 7
LOD scores across seven pearl millet linkage groups
Legend BHA- Bhavan~sagar locatton _PER- percentage. .AR('- radlans ot arc-sln
- BH4PER
BHA ARC
Figure 8.2: QTL U)D peaks for downy mildew resistance from Rhavanisagar trial
LOD scores across seven pearl millet linkage groups
U g m d 0. Cauobdmre location: B H A Bhannirapr locanon; A C - across locat~ons. P E R - pmml.; - a C - rad~ua of arc-sin: MULTI- multiple QTL model for radians of arc-sin values
F i s h 8.3: Comparison QTL UID peaks for downy mildew resistance from Coimbato~. Bhavuusaga and acms-lmtioar
5. DISCUSSION
Mean performance of the mapping population testcross hybrids
Mapping population testcross hybrids were raised at two locations in Tamil Nadu,
namely Coimbatore and Bhavanisagar. The mean performance for all eleven observed
tram was ri~ore or less s~milar at both locat~ons. Analys~s of variance study indicated high
sl~nlficant vanatlon for the mean performances of individual entrles for all the traits
under study at both test locations, b i t that the interaction between these two locations and
the individual entries (genotypes) was not significant for any of the traits. This may be
due to the physical closeness of the places where trials were conducted and the similarity
In sowing dates and agronomic practices used for the two tnals. The first trial location
Coimbatore, located at 1 I' latitude and 77' longitude and the second trial location
Bhavanisagar. located at 11' 08' latitude and 77' 29' longitude. The soil types and
packages of agronomic practices employed were simllar. so that the environment may not
have had much opportunity to differentially influence the mean performance of the
mapplng population testcrosses.
Trials were conducted at different locations to identify or elucidate the effect of
rhe environment and assess the relative importance of genotype x environment
~nteraction effects and genotype effects, because differential expression of a phenotypic
trait by genotypes across environments, or genotypic x environment interaction is an old
problem of primary importance for quantitative genetics and plant breeding trials
(Eberhard and Russel, 1966; Falconer, 1981; Via and Lande, 1987; Tiret et ul., 1993).
Though the mean performances were similar for the two test locations. data from
Bhavanisagar showed a slight increase in mean performance for the following traits: plant
height, panicle circumference, productive tiller number, grain yield per season, grain
yield per day, single-panicle grain mass and single-panicle gram number. Other traits like
time to 50% stigma emergenc,e. panicle length and grain number per unit area registered
very similar mean values at the two test locations.
Though the data from two location trials were not significantiy different, the
existing variation between two location trials may give different results on QTL mapping.
With this precaution, QTL analysis was done for individual locations entry means as well
as pooled means across-locations.
It is important to realize that heritability is a property not only of a character but
also of the population in which this character was measured and of the environmental
circumstances to which individuals are subjected prior to this measurement (Falconer,
1960). Also estimating heritability for a particular trait is the prime-most concern for
even a simple selection scheme. This is applicable for QTL mapping also. The reliability
of the QTL mapping depends very highly on the heritability of the individual traits
(Kearsy and Farquhar. 1998).
Heritability (plot basis) studies from the individual location data sets revealed that
all the traits registered heritability (plot basis) values greater than 50% excluding for
single-panicle grain number and grain number per unit panicle surface area. At
Coimbatore thousand-grain mass registered the highest heritability (plot basis) value of
93% followed by panicle circumference (75%). Other traits namely time to 50% stigma
emergence, plant height, panicle length, ~roductive tiller number, grain yield per season,
Qraln yield per day and single-panicle grain mass showed moderate heritability values
ranging from 40 to 70%. Single-panicle grain number and p i n number per unit panicle
surface area were the two tralts having poor heritability values. .4t Bhavanisagar plant
height recorded the highest heritability (plot basis) value of 95% followed by
thousand-grain mass, which ,had a heritability value of 93%. When compared to
Coimbatore location, the her~tability values for all the traits were higher in Bhavanisagar.
- Broad sense heritability values for pooled locations were higher for all the traits than plot
bass heritability values obtained from individual locations data sets. Single-panicle grain
number and gram number per un~t panicle surface area too had higher broad-sense
hentability values (>SO%) from the pooled data sets.
Correlation studies
Correlation studies provide indications of the extent of linkage and pleiotropism of genes
controlling the different traits. Grain yield per season was taken as the dependant variable
and the other traits were correlated with this. Plant height, panicle circumference,
thousand-grain mass, grain yield per day, single-panicle grain mass, single-panicle grain
number and grain number per unit panicle surface area were the traits positively
correlated with grain yield per season. Among these traits, ga in yield per day registered
the highest significantly positive correlation with grain yield per season at individual
locations as well as with entry means from pooled analysis of data. This was followed by
single-panicle grain mass, thousand-grain mass, panicle circumference and plant height
which had similar values towards the contribution to grain yield per season. Improvement
of grain yield per season may be achieved by increasing values of these positively
correlated traits.
Time to 50% stigma emergence, productive tiller number and panicle length were
the three traits associated negatively with grain yield per season. Among these traits time
to 50% stigma emergence had the strongest negative correlat~on w~th grain yield per
season. Selection of early flowering lines may enrich the hybrid yield potential of pearl
m~llet genotypes in this mapping population,
When time to 50% stigma emergence was taken as a dependent variable,
productive tiller numbers and panicle length were associated positively wTth this trait. All
other traits exhibited negative relationships with time to 50% stigma emergence. Plant
he~ght was also considered as a dependent variable and its relationship with other traits
was assessed. Grain yield per day and panicle circumference had higher positive
associations with plant height. Time to 50% stigma emergence, productive tiller numbers
and panicle length were the three traits associated negatively with plant height.
From these correlation studies it can be concluded that time to 50% stigma
emergence, productive tiller numbers and panicle length had strong associations with one
another. This may be due to the linkage or pleiotropism among these traits. As they have
negative associations with grain yield per season, it can be concluded that selection of
early flowering types with shorter panicles and less number of panicles per plant will
improve the total grain yield of hybrids produced by crossing PT 4450 with progeny 6om the
cross of PT 732B and P 1449-2. Correspondingly, selection of tall genotypes having good
panicle circumference, single-panicle grain mass and thousand-grain mass would likely
directly improve the grain yield potential ofhybrids on this mapping population.
Mean performance of downy mildew screening trials
Mean disease incidence values were converted into disease incidence percentages and
radians of arc-sin transformed percentage values. The disease incidence percentage was
more severe at Bhavanisagar than at Coimbatore. Both locations registered significant
differences for disease incidence percentages between mapping population progenies, as
well as for radian values. Heritability (plot basis) values for these two measures of
disease reaction were also high, giving confidence of success in mapping QTLs
conferring resistance against downy mildew disease.
Mapping QTLs
Knowledge gamed from QTL mapplng expenments is of greatest interest to plant
breeders if the results are directly applicable to practical breeding programs. Therefore,
when such experiments are initiated, one of the most important questions is the choice of
population for phenotyping experimental materials. For field trials. we used testcross
progenies related to a commercially important hybrid, looking for opportun~ty to improve
upon this combination, in a manner that comes closest to the applied plant breeder's
situation. In applied breeding programs, the tester is often an elite inbred line chosen
because of its use as a commercial hybrid parent. Therefore mapping of QTL for such
testcrosses promises (i) an insight into the relative importance of additive effects with
regard to testcross performance and their underlying genetic factors and (ii) the design of
a more efficient breeding strategy (Schan er al., 1994). So, this study formulated to
~dentify opportunities for favourable contributions in terms of additive effects of the
inbred P 1449-2 towards improvement the hybrid performance of seed parent PT 732 in
combination with PT 4450.
Grain yield in cereals is generally controlled by a number of quantitative trait loci
(QTLs) and is affected by environmental factors, making it difficult to manipulate and
improve in plant breeding programs. Grain yield can be dissected into a number of
component traits such as thousand grain number, productive tiller number, panicle length.
etc depending upon the crop concerned. These component traits are also under the control
of QTL and the effects of individual QTLs on phenotypic variation are relatively small.
Some QTLs however are less environmentally sensitive and have high heritabilities than
grain yield itself (Bezant et al.. 1997). Further, the standard relationship between various
yield component traits are not found for all QTLs, so it should be possible to identify
specific QTLs that can be manipulated wlthout adversely affecting otherwise correlated
traits (Hash. 2000; Yadav er al. , 2002b). Therefore, while looking for QTLs controlling
grain yield, QTLs for yield components should also be determined to provide more useful
~nformation.
QTL for time to 50% stigma emergence
From the MAPMAKER program, a single QTL was identified on linkage group (LG) 4
for time to 50% stigma emergence (Figure 9.1). This QTL had an additive effect of 0.5
days with the earlier flowering allele inherited from parent P 1449-2. So this parent
contributed early flowering to the testcross hybrids with elite. pollinator PT 4450. Using
this genomic region it may be possible to transfer the early flowering allele from
P1449-2 to PT 732A/B. Yadav et al., (2002a) mapped two QTLS for time to 50% stigma
emergence in pearl millet with one situated near the bottom of LG 4 and the other
mapped to the bottom of LG 6.
QTL for plant height
A single QTL was mapped for mapping population testcross hybrids plant height. This
QTL mapped to the bottom of LG 4 (Figure 9.1). A significant LOD score of 6.95 was
recorded from the Colmbatore data set. The allele from P 1449-2 had s positive additive
effect of 6 cm for this trait, and increased the plant height of hybrids. This particular QTL
explained 24% of the observed p;~enotypic variance for testcross hybrids plant height at
Coimbatore.
This QTL is likely to be considered as d2 dwarfing gene (~zhaguvel . 2001), for
which mapping population parents PT 7328 and P 1449-2 have the dwarf and tall alleles,
respectively. Dwarf phenotypes can be considered as a consequence of mutations that
occur in genes involved in plant height expression (Lin et al., 1995). Dwarf mutants of
pearl millet have been studied (Kadam er al., 1940; Burton and Forston. 1966; Appa Rao
er ul., 1986) and at least four single recessive genes have been reported i .e., dl and d~
(Burton and Forston, 1966) and dl and d , (Appa Rao el al., 1986). with possible presence
of additional modifying factors. The d2 dwarfing allele may have a pleiotropic effect
since d2 near-isogenic lines have longer and narrower panicles, wider leaves and smaller
seeds then their tall counterparts (Rai and Hanna, 1990).
The d2 dwarfing gene mapped to the bottom of the LG 4 in the mapping
population IP 18293 x Tifl 238D1 (Azhaguvel, 2001). The d2 dwarfing was inherited
from the parent 1P 18293. The parent Tiff 238Dl had one more dwarfing gene, d,, which
mapped to LG I. From the current study, it was clear that the d2 dwarfing gene from PT
732B had a significant contribution to height reduction in the testcross hybrids with elite
pollinator PT 4450, suggesting that PT 4450 and its hybrid with PT 7328, i.e., CoHCU-8
are infact also genetically dwarf at this locus.
QTL for panicle circumference
For panicle circumference. one QTL was identified in the middle of LG 4. This QTL was
detected using different rransfoimations and using Coimbatore. Bhavanisagar and
across-locations data sets, after regressing out the effects of time to 50% stigma
emergence, plant height and time to 50% stigma emergence together with plant height as
predictors of panicle circumferences. The highest LOD peak was oblained from the
Co~mbatore data set using time to 50% stigma emergence as a predictor. At this LOD
score, the QTL had an additive effect of 0.9, with increased panicle circumference
inherited from parent P 1449-2. This QTL has a favourable effect on hybrid performance
so it will be useful to introgress this genomic segment from P 1449-2 into PT 7328.
QTL for panicle length
Panicle length was also observed to be largely under the control of a single QTL and this
QTL also mapped to LG 4. Nearly all the predictors allowed detection of this QTL in the
interval between markers Xpsm.568 and XpsmSI2. The P 1449-2 allele for this QTL
decreased the panicle length of hybrids, in agreement with the observations of Rai and
Hama (1990) on the effect of the tall allele at the nearby d2 dwarf gene locus on this
character.
QTLs for productive tiller number
For productive tiller number, up to four QTLs were obtained, which were mapped to four
different linkage groups (Figures 9.1 and 9.2). All four QTLs explained similar portions
of phenotypic variation, but the QTL located on LG 2 explained a comparatively higha
Ponlon ( 15 4%) For this QTL the allele from P 1449-2 reduced the number of productive
t~llers
QTLs for thousand-grain mass
Two l~nked QTLs on LG 4 were mapped for thousand-gram mass Thesetwo QTLs were
detected using Bhavanlsagar data and both types of transformat~on The QTL s~tuated In
the marker interval between Ypsm568 and Xpsm512 expla~ned the h~gher proportion
( 11 6%) of observed phenotyp~c vanatlon The a d d ~ t ~ v e effect of thls QTL IS 0 3 g, wh~ch
IS tnhented from the tall parent P 1449-2 Usually dwarf plants, reduced the gram mass In
the hybnds lead to reduct~on In y~eld Desp~te lower gram mass and gram y~eld In the
dwarf plants, ~t IS posslble to produce dwarf hybnds w ~ t h y~elds equal to the tall hybnds
by selection of su~table poll~nator Breed~ng programs on dwarf pexl mlllet should be
successful ~f they are des~gned to take advantage of poslt~ve ~nteract~ons between the
dwarf hab~t and speclfic genet~c background (B~d~nger and Raju, 1990)
QTL for grain yield per season
For gram y~eld per season a s~ngle QTL was mapped near the bottom of LG 4 It
explained 10% of observed phenotyp~c vanance The favourable allele for thls QTL was
~nhented from parent P 1449-2, whlch had the a d d ~ t ~ v e effect of 0 Sl that 1s equal to a
gram y~eld Increment of 3 3 g m 2 At the plant populat~on denslty used In thrs study
(50.000 plantsiha) t h ~ s corresponds to a y~eld advantage of 3,300 giha = 33 kgha
Transfer of t h ~ s genomlc segment may be useful to Improve the gram y~eld of the hybnd
of PT 732A x PT 4450 = CoHCU-8, but would clearly be associated w ~ t h an Increase In
plant height due to the strong l~nkage of this QTL to the tall allele at the t i : dwarfing gene
locus.
QTLs for grain yield per day
Grain yield per day can be considered to be an important trait, where i t explains the
source and ,sink relationship aAer flowering. Up to three QTLs were found to be
associated with g a i n yield per day in this study. These QTLs were distributed on LC 4 at
different positions. Of these QTLs one was detected from the Coimbatore data set and the
other two were detected from the Bhavanisagar data set. QTLs from Bhavanisagar using
log transformation and time to 50% stigma emergence as a predictor, explained a higher
portion of the observed phenotypic variances (10.7%). The corresponding additive effect
for this locus was 0.52, i .e., 0.3 g per day of grain yield with the favorable allele inherited
from the parent P 1449-2.
QTLs for single-panicle grain mass
Four QTLs for single-panicle grain mass were obtained on LG 4 (2 QTLs), LG 6 (1 QTL)
and L C 7 (1 QTL) (Figures 9.1 and 9.2). Data from Coimbatore detected all the three
QTLs, where as across-locations data produced only two QTLs, r .e . , those mapping to
LG 4 and L C 7. In all the cases, the QTLs for single-panicle grain mass co-mapped with
QTLs for productive tiller number, with the parental alleles associated with increased
single-panicle grain mass appearing to have negative pleiotropic effects on productive
tiller number. This negative relationship between the two traits is commonly observed in
pearl millet.
* M 0
L - 0 a
C - 0 L u0 0 n - -
5 3 L a 0 Q I-, '0 . cc "1 3
3 - 0 0 - , a * > 7 2 I\
1 :-;? 2 2 2 : 2 % .? . - - : : a r -T
T
a a \ \ / /
g w Y w
5 .- - 5 9 5 3 5 3 3 5 5 5 m - S. m c m m m o U - - f " m r m - u , -,
Linkage group 6
Linkage group 7
0 - Coinlhatorc Figure 9.2: Genetic linkage map of PT 732 x P 1449-2 showing e - productive tiller number QTL positions on LG 6 and LG 7 for agronomic traits i - single-panicle grain mass @ - Across locations
QTLs for downy mildew resistanre
Pearl mlliet downy m~ldew has hrstor~cally been considered to be a quantltatlve trart,
slgnlficantly affected by the 2nvlronment (whlch 1s often confounded W I I ~ pathogen~c
vanablllty differences) Host plant reslstance agalnst downy mlldew was cont~nuously
dlstnbuted In the F? Fq progenies used In t h ~ s study as has been found in most prevlous
studles on the genetrcs of pearl m~llet downy mildew reslstance (S~ngh et a l . 1980,
Basavara~u el al , 1981, Dass er a1 . 1984, Shlnde et a!. 1984, Jones er 01, 1995)
At least five different QTLs were mapped for downy mlldew resrstance on four
linkage goups uslng dlsease rncrdence percentage and radlan values (Frgures 10 1 and
10 2 ) Of these, two QTLs were ldentrfied from Colmbatore data mapped to LC 2, two
QTLs from Bhavanlsagar data mapped to LC 1, and LC 4 and one QTL From across-
locations data mapped to LC 7
QTLs from the Colmbatore data and across-locat~ons means were lnhented In an
additlve fashlon Alleles from P 1449-2 contnbuted t h ~ s reslstance However, the two
QTLs detected from Bhavanlsagar were lnhented recess~vely and parent PT 732 was the
contnbutor of thls reslstance The majonty of prevlous research on the genetlcs of downy
m~ldew res~stance In pearl mrllet has found dominance to be an ~mportant component of
reslstance (Appadura~ el (11, 1975, G111 et 01, 1978, Pethanl er u l , 1980, Basavaraju
el a1 , 198 1, Shlnde et a1 , 1984, Mehta and Dang, 1987) and over dom~nance has also
been detected (Slngh er a1 , 1978, Basavaraju et a1 , 1981, Dass er a1 , 1984) However,
the inheritance of downy rnlldew reslstance In pearl millet 1s at least occaslonall~ found
to be recessive (Slngh er 01, 1978) and recesswe reslstance genes, although unco-on,
have been found In other plant-pathogen systems (Day. 1974, De Wltq 1992)
Unfortunately. whlle useful In hybr~d seed product~on plots, such recessively lnhented
resistance 1s unl~kely to contr~bute pos~t~vely to hybr~d performance In tamers field
General discussion
total of 18 QTLs were obtained from t h ~ s study, uslng square root transformat~on and
log transfomatlon of agronomlc data sets from Colmbatore, Bhavamsagar and
across-locations Between these two transfonnat~ons, the square root transfonnat~on gave
74 QTLs and log transformation gave 23 QTLs (Table 26) Out of three pred~ctors ( I e ,
time to 50% stigma emergence, plant helght and tlme to 50% stlgma emergence together
wlth plant helght), time to 50% stlgma emergence produced 34 QTLs followed by plant
helght whlch revealed 21 QTLs and tlme to 50% stlgrna emergence together w ~ t h plant
helght mapped 6 QTLs Out of these 18 QTLs, only seven genomlc Intervals were
respons~ble for all the QTLs controlllng agronomlc traits ( I e , some of the genomlc
reglons were respons~ble for controlllng more than one tra~t) Out of these seven genomlc
reglons, LG 4 had four and LG 2, LG 6 and LG 7 each had one genomlc reglon
contnbut~ng to the detected QTLs
In LG 4, the Interval flanked by marker loc~ Xpsm568 and Xpsm512 had the
control over five tralts, lncludlng graln y~eld per season The other tralts controlled by
this genomlc reglon were panlcle circumference, panlcle length, thousand-gram mass and
gram y~eld per day Marker Interval Xpsm84 to Xpsm612, wh~ch 1s also on LG 4
controlled three tralts, 1 e , productive tlller number, graln yleld per day and
s~ngle-panlcle gram mass It seems hlghly l~kely that genes or gene blocks in these two
regions may have ple~otrop~c effects on these traits SO transfemng these particular
reglOns, which are controll~ng major tralts, to the parent PT 732 may be advantageous
~ n d reasonable since In most cases P 1449-2 was found to contribute the favorable allele
But from the correlat~on stud~es 11 was found In the mater~dl studled that pan~cle
length and productlve t~ller numbers were assoc~ated w ~ t h each other and both are havlng
negative relatlonshlp with y~eld So refinement of these genomlc reglo-ns may prov~de
more ~nfomlatlon about ~ n d ~ v ~ d u a l tra~ts. whlch may be controlled by d~fferent QTLs Of
course. refinement of the map posltlons of QTLs controlling these tram w ~ l l requlre
genotyp~ng and phenotyplng a substant~ally larger mapplng populat~on So further
analysis of the exlstlng data sets may be requlred to just~fy the substantla1 costs that thls
refinement would requlre
Although d~fferent QTLs were obtaned from the two d~fferent test locat~ons, it IS
better to restnct d~scuss~on of ap?l~cat~on to the QTLs from the across-locat~ons due to
stat~st~cal constraints Across-locat~ons data set produced only SIX QTLs, which mapped on
LG 4 and LG 7 and only four genomlc reglons [three on LG 4 (Xpsm568-Xpsm512,
Ypsm84-Xpsm612 and 4psmjl.?-Xpsm344) and one on LG 7 (Xrm9-2b-Xpsrn618)l were
respons~bie W ~ t h respect to traits, plant he~ght, pmcle crrcumference, pmcle length,
productlve tlller number and angle-pan~cle gram mass were the tralts for whlch QTLs
could be mapped from analys~s of the across-locat~ons entry means
Tralts such as plant he~ght, panlcle circumference and single-~anlcle grain mass
were posltlvely correlated w ~ t h gram y~eld and gram yleld per season and these regions
were controlled by three different reglons (two controlling slngle-panlcle grain mass and
the other controlling plant helght, pan~cle length, and panicle circumference) So
transferring these genomic regions may offer the chance to Improve grain yield
performance of hybrlds of PT 7 3 2 8 .
Opportunities for Marker-Assisted Selection (MAS) to improve CoHCU-8
The advent of molecular-marker based techniques has had a large impact on quantitative
genetics. Marker-based methods applied to segregating populations have provided us
with a means to locate quant~tative trait loci (QTLs) to chromosomal regions and to
estimate the effects of QTL allele substitution (Lander and Botsteln, 1989). The ability to
estimate gene effects for a quantitative trait can be very useful for the design and
dpplication of new, more efficient, breeding strategies. .A new selectlon strategy,
marker-asslsted selectlon (MAS), has been proposed by many authors as a way to
increase gains from selection for quantitative traits (Tanksley, 1993; Lee, 1995; Kearsey - and Pooni, 1996). In backcross breeding programs, it has been shown that MAS can be
effective in reducing linkage drag and optimtzing population sizes, by permitting
effective selection against the donor genome except for allele(s) in the genomic region to -.-. -
be Introduced from the donor MAS can also Improve selectlon for qudntltat~ve tralts by
selecting for the presence of spec~fic marker alleles that are llnked to favorable QTL
alleles that would be othenv~se difficult to select for phenotypically (Berloo and Stam,
1998) Published reports of successful appllcatlon ofthls strategy to Improve hybnd yleld
performance are just beginning to appear.
From this study, it is clear that different regions of pearl millet genome are
specifically associated with grain yield component traits such as plant height, panicle
circumference and single panicle-grain mass when across-locations data was considered.
These traits were positively correlated with grain yield and grain yield per season. These
three genomic regions also explained significant portions of observed phenotypic
~ariance for their respective traits and these traits all had high heritab~lity values. Soller
and Beckmann (1990) stated !hat MAS for quantitative traits w ~ t h high h2 would not
necessarily be as efficient as conventional breeding. However, for a quantitative trait with y--
hlgh-e, MAS could still be effective after major QTL controlling the tr'lit are fixed and
the h2 of remaining genetlc variation is reduced (Paterson et al., 1991).
The parent PT 732A serves as the female parent for producing conunercial hybrids
(hybrids X6 and CoHCU-8) that have been released h m Tamil Nadu Agricultural
University. So improving the yield potential of PT 732B (maintainer of PT 732A) may
usher in new ways for increasing :he grain yield of hybrids that can be produced for
Tamil Nadu using this seed parent. The QTLs identified from this study can be used in a
maker-assisted backcrossing program for hybrid parental line improvement because their
effects have already assessed in testcrosses to the best available hybrid produced from
crosses onto PT 732A. So marker-assisted backcrossing of P 1449-2 alleles for putative
QTLs on LG 4 and LG 7 into the PT 732B background may be effective to improve the
y~eld potential of hybrids produced on PT 732A. Yadav et al. (2002a) suggested a similar
strategy in pearl millet to transfer the drought tolerance QTLs into elite pollinator inbred
H 771833-2 (male parent of early-flowering released hybrid HHB 67) using marker-
assisted backcrossing to introgress genomic segments from donor PRLT 2/89-33.
To obtain durable resistance for downy mildew there are two ways. One is
pyramiding genes from all known sources (Jones et al., 1995) and the second possibility
is the production of hybrids that are genetically heterogeneous for disease resistance, thus
mimicking the durable resistance of open-pollinated cultivars (Witcombe and Hash,
2000) This could be accompl~shed by produc~ng a set 01 backcross I~nes, each d~ffenng
for d ~ l n g l e resistance gene. and allow~ng these l~nes to recomblne durlng mult~pllcation
of the male sterile Ilne breeder seed In the hybr~d seed product~on cham Incorporating
more than one dom~nant gene effective agalnst pathogen populdt~on Into each component
line may be expected to Increase reslstance durab~llty
Hash el (2000) d~scussed an alternative procedure of marker-ass~sted transfer
of QTLs In pearl mlllet I he first successful appllcat~on of marker-asslsted select~on for
pearl m~llet has been enhancement of downy m~ldew reslstance of Inbred poll~nator
H 771833-2 (male parent of popular early-matunng pearl m~llet hybnd HHB 67) Several
Improved verslons of thls poll~nator have been developed at ICRISAT using thls "fast
track" marker-assisted backcross procedure (Sharma, 2001) T h ~ s has been demonstrated
to be a tlme and cost eftic~ent route for the appllcatlon of marker-based downy mlldew
res~stance breed~ng In t h ~ s crop Such approaches may be warranted to Improve the
d~sease reslstance of e l~ te hybr~d parental l~ne PT 732B and 11s male-stenle counterpart
PT 732A
We have ~dentltied the preclse locat~on of QTLs by ordinary l~nkage mapp~ng,
wh~ch has become a standard startlng polnt for map-based clon~ng (Tanksley er a l ,
1995) In plants, several econom~cally ~mportant genes have been Isolated by map-based
cloning, lncludlng a photopenod-sens~t~ve gene (flowering gene) In Arahldo~sls
(Puttem111 et a / , 1995) However, ~t would be reasonable now to confirm these QTL
locations using CIM (Composlte-~nterva1 mapp~ng) methods as lm~lemented In the
QTL Cartographer and PLABQTL software packages
Most genetlclsts and breeders consider QTLs to be chromoson~al locations of
indicldual genes or goups of genes that influence complex traits (Stuber er ul., 1999) -- Although it is often tac~tly assumed that a QTL represents a single genetic determinant
(Or factor), there are examples of Individual QTLs that have been resolved into multiple
genetlc factors by recombination (Graham et 01.. 1997; Yamamoto et ul: 1998). For the
rnanlpulation of the vast majority of QTLs in plant breeding programs. it may not be
Important to determine whether the QTL represents a single genetic factor or a cluster of
tlghtly linked genes. However, ~f cloning of specific QTLs is paramount to their
ut~lization, then the chromosomal location must be limited to a manageable piece of DNA
(Paterson, 1998)
Recent advances in molecular genetics have promised to revolutionize
agricultural practices. As stated by Lande and Thompson (1990) "There are, however,
several reasons why molecular genetics can never replace traditional methods of
dgncultural improvement. but instead should be integrated with conventional methods to
obtaln the maximum improvement in the economic value of domesticated populations."
Their analytical results, as well as the more recent computer simulations and the limited
empirical results, however, are encouraging and support the use of DNA-based markers
to achieve substantial increases in the efficiency of artificial selection.
6. SUMMARY
Pearl millet [fet~t~rserlrt?r glrrarto?~ ( L . ) R. Br.] is an Important staple food crop for
rnlllions of rural people living in semi-arid regions of tropical and sub-tropical Asia and
Africa. In parts of USA. South Amenca and Southern Africa it is cultiva_ted for feed and
forage purposes. Pearl m~llet is a crop that can be grown in adverse agro-climatic
conditions like drought. heat and infertile so~l . It is the only crop that gives assured
harvest to the farmers whose subsistence is totally dependant on farming in hot, dry
marginal environments. Among the diseases affecting pearl millet, downy mildew is the
most devastating. This is caused by the pseudo fungal pathogen Sclerospora graminicola
(Sacc.) J . Schrot..
Improvement of yield and breeding for resistance to pests and diseases are the
prime concerns of the breeders. This study was designed to identify genomic regions
from donor P 1449-2 with the potential contribute to yield increments in the genetic
background of released hybrid CoHCU-8 and also for downy mildew disease resistance.
One hundred and thirty-six F2-derived Fd mapping population progenies of a pearl
millet mapping population (skeleton-mapped F2 individuals) obtained from a cross of
PT 732B and P 1449-2 were used as a source population for this study. PT 4450 is an
elite pollinator line that produces an agronomically superior released hybrid (CoHCU-8)
when crossed to PT 732A. It was used as a pollen source for crossing with these FJ
self-bulks. Testcross hybrids produced from these crosses were raised for the purpose of
phenotyping during the rainy season (October, 2001) at two locations in Tamil Nadu:
Tamil Nadu Agricultural University, Coimbatore and Regional Research Station,
Bhavanisagar. For downy m~ldew screening, selfed seeds from the FA self-bulks were
raised at the two above-ment~oned locations during the rainy season of 2001.
Results from the yield.trials showed that there was significant variation for all
observed traits within the set of mapplng population testcrosses at each location, and
there was no sign~ficant genotype x environment interaction for any of tfie 11 agronomic
tralts considered in this study. Hentab~lity estimates for individual traits from the y~eld
trials at two different locations and pooled data across these two locations had reasonably
h ~ g h values (>SO%), which were sufficient to permit QTL mapping procedures to identify
genomic regions contributing to the observed variability. Grain yield per season was
positively correlated with most of the observed tram including plant height. But time to
50% stigma emergence, productive tiller number and panicle length were associated
negatively with grain yield per season.
From the downy mildew screening trials, the data set 6om the two locations each
rvh~bited s~gnificant genetic differences, but there was also significant genotype
x environment interaction indicating that the vimlences constitutions of the pathogen
populations at these two locations were different. The heritability (plot basis) values were
also high enough to do the QTL analysis.
Yield trial data from the two locations were subjected to two types of
transformations namely, square root and log, so as to minimize the heterogeneity in the
data sets. Improvement of yield is a complex process. To minimize this complexity and
facilitate identification of QTLs that d ~ d not directly correspond to major genes affecting
plant height and flowenng time (which are relatively simply inherited traits known to
grain yield and its components), plant height, time to 50% stigma emergence and plant
height together with t ~ m e to 50% stlgma emergence were used as pred~ctors I e . all the
dgronomlc traits were regressed w ~ t h these pred~ctors. and the res~duals trom these
regressions for each abronomlc tralt were subjected to QTL mapplng
A previously constructed l~nkage map uslng 60 RFLP markers for the [(PT 732B
P 1449-2)- based] mapplng populat~on were used for locatlng QTLs QTL analys~s w ~ t h
the MMMAKERJQTL program showed d~fferent QTL posltlon for d~fferent tralts In
total, 18 QTLs were obta~ned for nlne d~fferent tralts from the Colmbatore. Bhavan~sagar
and across-locat~ons data sets Among these nlne tralts, tlme to 50% stlgma emergence,
pan~cle c~rcurnference, plant helght, pan~cle length and gram y~eld per season each
reglstered one QTL, thousand-gram mass reglstered two QTLs, gram y~eld per day
reglstered three QTLs and s~ngle-pan~cle gram mass reglstered four QTLs However,
these 18 QTLs were under the control of only seven genomlc reglons, suggesting roles of
t~ght llnkage andlor ple~otroplsm In the lnhentance of these oRen correlated tralts Of
these seven genomlc reglons, LC 4 had four reglons, LC 2, LC 6 and LC 7 each had one
genornlc reglon contnbutlng QTLs In LC 4 the reglon flanked by marker loc~ XpsmS68
and .YpsrnSIZ contnbuted to control over five tram ~ncludlng gram y~eld per season
Across-locations data produced SIX QTLs for agronomic tralts studled They were
on LC 4 and LC 7 Totally four genomlc reglons vl: , three on LC 4 and one on LC 7
shared these SIX QTLs The tralts controlled by these QTLs Included plant he~ght, panicle
~~rcumference, pan~cle length, product~ve t~ller number and s~ngle-~anlcle graln mass
For downy mildew. five d~fferent QTLs were mapped on four linkage groups by
using disease lncldence percentages and then arc-s~n transferred radians values Of these,
two QTLs were detected from the Colmbatore data set on LC 2. two QTLs from the
Bhavanlsagar data set on LC I and LG 4, and one QTL from across-locat~ons on LG 7
Marker-ass~sted select~on prov~des an opponunlty to Improve the effectiveness of
quantltatlve t ram by selecting for the presence of spec~fic marker alleles that are llnked to
favorable QTL alleles From thls study. ~f we consjdered only the acros's-locations data
set, different reglons of pearl m~llet genome were detected as spec~lically associated with
gram y~e ld per season. plant he~yht. pan~cle c~rcumfcrence and s~ngle-pan~cle gram mass
QTLs for these tralts also explained slgnlficant portlons of observed phenotyp~c vanatlon
So marker-ass~sted backcross~ng from P 1449-2 to move putatlve QTLs on LG 4 and
LG 7 Into the PT 732NB background may be effective to Improve the yleld potentla1 of
hybnds of e l ~ t e seed parent PT 732, at least those hybnds produced wlth ellte pollinator
PT 4450
For lmprovlng the reslstance agalnst downy m~ldew marker-ass~sted transfer
~ n d \ o r pyramldlng of the reslstance genes (or QTLs) may glve good results
REFERENCES
'Abler. B.S.B.. M.D Edwards and C.S. Stuber. 1991. Isozymatic identification of
quantitative trait loci in crosses of elite maize ~nbreds. crop Scr.. 31: 267-274.
A P P ~ Rae, S., M.H. Mengesha and C.R. Reddy. 1986. New sources of dwarfing genes in
pearl millet (Penn~sertim unrerrcrrntrttr). Theor. Appl. Genet.. 73: 170-1 74.
Appadurai, R.. C. Parambaramani and V.S. Natarajan. 1975. Note on the ~nheritance of
susceptibility to pearl millet to downy mildew. Indian J Agrrc. Scr.. 45: 179-180.
Appadurai. R., T.S. Raveendran and C . Nagarajan. 1982. A new nrale-sterility system in
pearl millet. Indrun J. . Igr~ . Scr., 52: 832-834.
.Arunachalam. V. and S. Chandrashekaran. 1993. RFLP approach to breeding for
quantitative traits in plants. A critique. 1 Genet.. 72: 73-83.
.&ustin D.F., M. Lee.. L.R.Veldboom and R. Hallauer. 2000. Genetic mapping in maize
with hybrid progeny across testers and generations. Crop Sci., 40: 30-39.
Austin, D.F, M. Lee and L.R. Veldboom. 2001. Genetic mapping in maize with hybrid
progeny across testers and generations: plant height and flowering. Theor. Appl.
Genet.. 102: 163- 176.
Austin, D.F. and M. Lee. 1996. Comparative mapping in F 2 , and F67 generations of
quantitative tralt loci for grain yield and yield components in maize. Theor. Appl.
Genet.. 92: 817-826.
Azhaguvel, P. 2001. Linkage map construction and identification of QTLs for downy
mildew (Sclerospora grami~ticolu) resistance in pearl millet (Pertnrsetum glaucum
(L.) R. Br.). Thesis submitted to Tamil Nadu Agricultural University, Coimbatore.
Ball. S.L. and D.J. Pike. 1984. intercontinental variation of Sclerospora graminicola.
Ann. Appl. Biol.. 104: 41-51.
Basavaraju, R., K.M. Safeeulla and B.R. Murty. 1981. Genetic variance and heritability
for resistance to downy mildew in pearl millet. IndianJ. Genet.. 41: 137-143.
Basten. C.J.. S.B. Zeng and B.S. Welr. 1994. ZMAP-A QTL Cartographer. 111:
Proceedings of the 5"' World Congress on Genetlcs Applied to Livestock
Production: Computing strategies and Sof,ware. ( S n ~ ~ t h . C . J.S. Gavora, B.
Benket. J. Chesna~s. W. Fairfull. J.P. Gibson. B.W. Kennedy and E. B. Burns~de.
Eds.). Guelph, Ontario. Canada, 33: 65-66.
Basten. C.J., S.B. Zeng and B.S. Weir. 1997. QTL Cartographer: A reference manual and
tutorial for QTL mapping. Department of Statist~cs. Nonh Carolina State
University, Raleigh. North Carol~na: USA.
Bcav~s. W.D.. O.S. Smith, D. Grant .and R. Fincher. 1994. Ident~fication of quantitative
trait loci using a small sample of topcrossed and FA progeny from maize. Crop
SCI . , 34: 882-896.
Beckmann. J. and M. Soller. 1983. Restriction fragment length polymorphisms in genetic
improvement: methodologies, mapping and costs. Theor. Appl Genet., 67: 35-43.
Beckmann, J . and M. Soller. 1986. Restriction fragment length polymorphisms and
genetic improvement of agricultural species. Euph.vrrca. 35: 11 1-124.
B e c h a m , J. and M. Soller. 1990. Towards a unified approach to genetic mapping of
eukaryotes based on sequence tagged microsatellites. Bio Technology. 8: 930-932.
Berloo, V. and P. Stam. 1998. Marker-assisted selection in autogamous RIL populations:
a simulation study. Theor. .4ppl. Genet.. 96: 147-154.
Bematzky, R. and S.D. Tanksley. 1986. Towards a saturated linkage map in tomato based
on isozymes and random cDNA sequences. Generrcs. 112: 887-898.
Bezant, J., D. Laurie, N. Pratchett, J. Chojecki and M. Kearsey. 1997. Mapp~ng QTLs
controlling yield and yield components in a spring barley (Horderrm vulgare L.)
cross using marker regression. Mol Breed.. 3: 29-38.
Bhat, S.S. 1973. Investigations on the biology and control of Sclerosporu grut?tinicola on
bajra. Ph.D. thesis. University of Mysore, Mysore. Kamataka. India.
Bidinger, F.R. and D.S. Raju. 1990. Effects of dz dwarfing gene in pearl millet. Theor.
Appl. Genet. . 79: 521-524.
Bonlerbale. M.W.. R.L. Plalsted and S.D. Tnnskley. 1988. RFLP maps based on common
sets of clones reveal modes of chromosome evolutron In potato and tomato.
Get1etic.s. 120: 1095- 1 103
Botstein. D.. R.L. White. M. Skolnick and R.W. Davis. 1980. Constructron of a genetrc
map in man using restnctlon fragment length polymorphisms. ..lmerrcun J Hum.
Genet.. 32: 3 14-33 1
Bubeck. D.M.. M.M. Goodman, W.D Beavls and D. Grant. 1993. Quantitat~ve trait l o c ~
controlling resistance to gray leaf spot in maize. Crop Set.. 33: 838-847.
Burr. D., F.A. Burr. K.H. ~ h o r n ~ s o n , M.C. Albertson and C.W. Stuber. 1988. Gene
mapping with recombinant ~nbreds in maize. Generics. 118: 5 19-526.
Burton. G.W. 1983. Breeding pearl m~llet. Plant Breed. Rev.. 1 : 162-182
Burton. G.W, and H.D. Wells. 1081. Use of near-lsogenlc host populatrons to estlmate
the effect of three foliage diseases on pearl millet forage yield. Ph~~oputhol . . 71:
311-333.
Burton, G.W. and J.C. Fortson. 1966. Inheritance and utilization of five dwarfs in pearl
millet (Pennisetum rvphordes) breeding. Crop Sci.. 6. 69-72.
Butler. E. J. 1907. Some diseases of cereals caused by Sclerosporri grtrn~ittrrolu. Mem.
Dept. Agri. Indla Bor. S e n . -7: 1-24
Caetano-Anolles. G., B.J. Bassam and P.M. Gresshoff. 1991. DNA amplification finger
printing using very short arbitrary oligonucleotide primers. BioTechnology, 9:
553-557.
Campell, K.G., C.J. Bergman, D.G. Gualberto, J.A.Anderson, M.J. Giroux. G. Hareland,
R.G. Fulcher, M.E. Sorrells and P.L. Finney. 1999. Quantitative trait loci associated
with kernel traits in a soft x hard wheat cross. Crop Sci.. 39: 1 184-1 195.
Cao, G., J . Zhu, C. He, Y. Gao, J . Yan and P. Wu. 2001. Impact of epistasis and
QTL x environment interaction on the development behaviour of plant height in
nce ( O t y a sutivu L.). Theor. Appl. Genet., 103: 153-160.
CauSse. M.A.. T ?A. Fulton. Y.G. Cho. S.N. Ahn. J . Chunwongse. K Wu. J . Xiao. Z. Yu.
P.C. Ronald. S.E. Harr~ngton. G. Second. S.R. McCouch and S D. ranksley.
1994. Saturated nlolccular map of the rice genome based on an ~nterspec~fic
backcross popular~on. Get~cr~cs. 138: 125 1 - 1274.
Cavalli, L.L. 1952. An analysis of linkage in quantitative inheritance. In: Quantitative
inheritance (Reeve. E.C.R and C.H. Waddington. Eds ). His Maresty's Stationary
Office, London. Pp. 135- 144.
Chang, R.Y.. L.S. O'Donoughue and T. E. Bureau. 7001. Inter-MITE polymorphisms
(IMP): a high throughput trdnsposon-based genome mapping and fingerprinting
approach. Theor. Appl. Genet.. 102: 773-781.
Chao. S., P.J. Sharp, A.J. Worland, E.J. Warham, R.M.D. Koebner and M.D. Gale. 1989.
RFLP based genetlc maps of wheat homeologus yroup 7 chromosomes. Theor.
.Appl. Gener.. 78: 495-504.
Churchill, G.A. and R.W. Doerge. 1994. Empirical threshold values for quantitative trait
mapping. Genetrcs. 138: 963-971.
Conneally, P.M., J.H. Edwards, K.K. Kidd, J.M. Lalouel and N.E. Morton. 1985. Reports
of the committee on methods of linkage analysis and reporting. Cibtogenet. Cell
Genet., 40: 356-359.
Cowen. N.M. 1988. The use of replicated progenies in marker-based mapptng of QTLs.
Theor. Appl. Gener., 75: 857-862.
Danesh, D., S. Aarons, G.E. McGill and N.D. Young. 1994. Genetic dissection of
oligogenic resistance to bacterial wilt in tomato. Mol. Plunt-Microbe Interact.. 7 :
464-47 1
Darvasi. A,. A. Weinreb, V. Minke, J.I. Weller and M. Soller. 1993. Detecting
marker-QTL linkage and estimating QTL gene effect and map location using a
saturated genetic map. Genetrcs, 134: 943-951.
Dass, S., R.L. Kapoor. R.S. Paroda and D.S. Jatasra. 1984. Gene effects for downy
mildew (Sclerosporn grarnirzicola) resistance in peari millet. Indran J. Genet.. 44:
280-285.
Dave, H.R. 1987. Pearl millet hybnds. I,,: Proceedings International Pearl millet
Workshop. (Witcornbe. S R. and S.R. Beckerman. Eds.). ICRISAT. Patanchem.
India. Pp. 12 1 - 1 26
Day. P.R. 1974. Genet~cs of host paraslte interactions. W.H. Freeman and Company, San
Franstsco. USA.
De Wit, P.J.G.M. 1992. Molecular characterization of gene-for-gene systems in
plant-fungus lnteractlons and the application of avirulence yenes in control of
plant pathogens. Ann. Rev. Phvtop~rthol., 30: 391-418.
Devos, K.M., M.D. Atkinson, C.N. chinoy, C.S. Liu and M.D. Gale. 1992. RFLP-based
genetic map of the homeologus group-3 chromosomes of wheat and rye. Theor.
Appl. Genet.. 83: 93 1-939.
Devos. K.M.. T.S. Pittaway. C.S. Busso and M.D. Gale. 1995. Molecular tools for the
pearl millet nuclear yenome. Int. Sorghum and M~llets Newsl.. 36: 64-66.
Diers, B.W., P. Keim, W.R. Fehr and R.C. Shoemaker. 1992. RFLP analysis of soybean
seed protein and oil content. Theor. Appl. Genet., 83: 608-612.
Dirlewanger, E., P.G. Isacc, S. Ranade, M. Belajouza, R. Cousin and D. de Vienne. 1994.
Restriction fragment length polymorphism analysis of loci associated w ~ t h disease
resistance genes and developmental tra~ts in Pisunl sativum L. Theor. Appl.
Genet., 88: 17-27.
Dudley, J.W. 1993. Molecular markers in plant improvement: Manipulation of genes
affecting quantitative traits. Crop Sci.. 33: 660-668.
Eberhard, S.A. and Russel. 1966. Stab~lity parameters for comparing varieties. Crop Scr..
6: 36-40
Edwards, M.D., C.W. Stuber and J.F. Wendel. 1987. Molecular marker facilitated
investigations of quantitative trait loci in maize. I . Numbers. genomic
distributions and type of gene action. Genetrcs, 115: 113-125.
Ellis, T.H.N. 1986. Restriction fragment length polymorphism markers in relation to
quantitative characters. Theor. Appl. Genet.. 72: 1-2. '
Elston. R.D. and J . Stewart. 1971. A general method for the genetic analysis of pedigree
data. Hum. Hered.. 1 : 523-542.
Ender. A.. K. Schwenk. T. Stadler, 8. Streit and B. Scl~iewater. 1990. RAPD
ldentificat~on of mlcrosatellites In Daphi~lu. Mol Ecoloy.. j: 437-447.
Everson, E.H. and C.W. Schaller. 1955. The genetlcs of yield difference associated with
awn barbing in barley hybrid (Lion x Altas") x Altas. .,lgrotr. J . 17: 276-280.
Falconer, D.S. 1960. Introduction to Quantitative Genetics. The Ronald Press Company,
N Y . Pp.165-183.
Falconer. D.S. 1981. Introduct~on to Quantitative Genet~cs. Second Edition. Longman
Group Limited, London, LJK.
F A 0 and ICRISAT. 1996. The world sorghum and millet econornlcs: Facts, trends and
outlook. FAO. Rome. Italy and ICRISAT, Patanchem, AP, India.
Fatokun. C.A., D.I. Menancio-Hautea. D Danesh and N.D. Young. 1992. Evidence for
orthologous seed weight genes in cowpea ar~d mung bean based on RFLP
mapping. Genetics. 132: 841-846.
Freyrnark, P.J., M. Lee, W.L. Woodman and C.A. Martinson. 1993. Quantitative and
qualitative trait loci affecting host-response to Exserohilum rurcicum in malze
(Zea mavs L.). Theor. Appl. Genet.. 87: 537-544.
Fukuoka, S., T. Inoue, A. Miyao, L. Monna, H.S. Zhong, T. Sasaki and Y. Minibe. 1994.
Mapping of sequence tagged sites in rice by single strand conformation
polymorphism. DNA Res.. 1: 271-277.
Gale, 1M.D. and J.R. Witcombe. 1992. DNA markers and marker-mediated applications in
pearl millet breeding. In: Biotechnology and crop improvement in Asia. (Moss,
J.P. Ed.). ICRlSAT (International Crops Research Institute for the Semi-Arid
Tropics), Patanchem, Andhra Pradesh 502 324. India. Pp.323-332.
Gebhardt, C., E. Ritter, T. Barone, T. Debener, B. Walkemeier, U. Schachtschabel, H.
Kaufman, R.D. Thompson, M.W. Bonierbale, M.W. Ganal, S.D. Tanskley and F.
Salamini. 1991. RFLP maps of potato and their alignment with the homeologous
tomato genome. Tlieor. .lpp/. <;er~et.. 83: 19-57.
(;eldeman, H. 1975. Investigation on ~nher~tance of quantitative characters in animals by
gene markers. I . Methods. Tireor .Appl. Genet. -16: 300-319.
GENSTAT 5 Committee. 1993. GENSTAT 5 reference manual. Clarendon Press,
Oxford. UK.
Ghareyazie, B., N. Huang, G. Second. J . Bennett and G.S. Khush. 1995. Classification of
rice gemplasm 1 . Analysis using AFLP and PCR-based RFLP Theor. Appl.
Gener.. 91: 218-227.
Gibson, T.B. and J.M. Thoday. 1962. Effect of disruptive selection VI. A second
chromosome polymorphism. Heredity. 17: 1-26.
Gill. B.S.. P.S. Phul. S.S. Chahal and N.B. Singh. 1978. Inheritance of resistance to
downy mildew disease In pearl millet. Cereal Res. Comm., 6: 71-74.
Graham. (3.1.. D.W. Wolff and C.W. Stuber. 1997. Characterization of a yield
quantitative trait locus on chromosome five of maize by fine mapping. Crop Sci..
37: 1601-1610.
Graner. A.. A. Jahoor. J. Schondelmaier, H. Siedler. K. Pillen. G. Fischbeck. G. Wenzel.
and R.G. Hemnann. 1991. Construction of an RFLP map of barley. Theor. Appl.
Genet.. 83: 250-256.
Gupta, P.K. and R.K. Varshney. 2000. The development and use of microsatellite
markers for genetic analysis and plant breeding w ~ t h emphasis on bread wheat.
Euph.vtrcn. 113: 163-185.
Hackett, C.A. 1997. Model diagnostics for fitting QTL models to trait and marker data by
interval mapping. Heredity, 79: 3 19-328.
Haley, C.S. and S.A. Knott. 1992. A simple regression method for mapping quantitative
trait loci in line crosses using flanking markers. Heredrtv, 69: 3 15-324.
Harrison, B.J. and K. Mather. 1950. Polygenic variability in chromosomes of Drosophillu
melunoguster obtained from the wild. Heredrty, 4: 295-312.
Hash. C.T. 1991. ICRISAT pearl m~llet breeding and the potentla1 use of RFLPs. In.
Rockefeller Foundat~on Conference: The establishment of a sorghum and millet
RFLP network to support breeding In ~levelop~ng countries. 6-10 May 1991.
Report of meetlng. Pp.28-29.
Hash. C.T. 1997. Research on downy mildew of pearl m~llet. Irr: Integrating research
evaluation efforts: Proceedings of an International Workshop, r4-16 Dec 1994,
(Ba i~ t~ lan . M.C.S. and P.K. Josht. Eds.). ICRISAT. Patanchem. India. Pp.21-128.
Hash. C.T. 2000. Concepts for application of marker techn~ques In Africa. In:
Application of molecular markers in plant breeding: Train~ng manual for a
seminar held at IITA, Ibadan, Nigeria, 16-1 7 Aug 1999. (Haussmann, B.I.G.,
H.H. Geiger. D.E. Hess, C.T. Hash and P. Brarnel-Cox. Eds.). International Crops
Research Institute for the Semi-Arid Tropics (ICRISAT), Patanchem, A.P. 502
324, India. Pp. 155-1 69.
Hash, C.T. and J.R. Witcombe. 1994. Pearl millet mapping populations at ICRISAT.
In: Use of Moiecular Markers in Sorghum and Pearl Millet Breeding for
Develop~ng Countries. Proceedings of an ODA Plant Sciences Research
Programme Conference, (Witcombe, J.R, and R.R. Duncan. Eds.). 29Ih March -
I'' April 1993, Norwich, UK. ODA: London, UK. Pp.69-75.
Hash, C.T. and J.R. Witcombe. 2000. Gene management and breeding for downy mildew
resistance. Invited paper presented in Guanajuato, Mexico In Sep 2000 at the
"Global 2000: Sorghum and Pearl Millet Diseases 111" conference. (In press).
Hash, C.T. and P.J. Bramel-Cox. 2000. Marker applications in pearl millet.
In: Application of molecular markers in plant breeding: Training manual for a
semlnar held at IITA, Ibadan, Nigeria, 16-17 Aug. 1999. (Haussmann, B.I.G.,
H.H. Geiger, D.E. Hess, C.T. Hash and P. Bramel-Cox. Eds.). International Crops
Research Institute for the Semi-Arid Tropics (ICRISAT): Patanchem. A.P. 502
324, India. Pp. 112-127.
Hash, C.T., J.R. Witcombe, R.P. Thakur, S.K. Bhatnagar, S.D. Singh and J.P. Wilson.
1997. Breeding for pearl millet disease resistance. In: Proceedings of the
International Conference on Genetic Improvement of Sorghum and Pearl Millet,
September 21-27. 1996. INTSORMlL Publication No. 97-5. INTSORMIL and
ICRISAT. Pp.337-372.
Hash. C.T.. R.S. Yadav, G.P. Cavan, C.J. Howarth. H. Liu. X. Qi. A. Sharma, M.A.
Kolesnikova-Allen, F.R. Bidinger and J.R. W~tcombe. 2000. Marker-assisted
backcrossing to improve terminal drought tolerance in pe'arl millet. In:
Proceedings of a Strategic Planning Workshop on "Molecular Approaches for the
Genetic Improvement of Cereals for Stable Production in Water-limited
Environments". June 21-25. 1999. CIMMYT, El Batan. Mex~co (Rlbaut. J-M and
Poland, D. Eds.). CIMMYT, Mexico, D.F., Mexico. Pp.114-119.
Hash, C.T., S.D. Singh. R.P. Thakur and B.S. Talukdar. 1999. Breeding for disease
resistance. In: Pearl Millet Breeding (Khairwal, 1,s.. K.N. Rai, D.J. Andrews, and
G. Hannarayana. Eds.). Oxford & IBH: New Delhi, India. Pp.337-379.
Hayes, P.M., B.H. Liu, S.J. Knapp, F. Chen. B. Jones, T. Blake, J. Franckowiak, D.
Ramusson, M. Sorrels. S.E. Ullrich. D. Wesenberg and A. Kleinhops. 1993.
Quantitative locus effects and environmental interaction in a sample of North
American barley germplasm. Theor. Appl. Genet.. 87: 392-402.
Hayes, P.M., B.H. Liu. S.J. Knapp, F.Q. Chen, B. Jones, T.K. Blake, J.D. Franckowiak,
D.C. Ramusson. M. Sorrels. S.E. Ullrich, D. Wesenberg and A. Kleihofs. 1993.
Quantitative trait locus effects and environmental interactions in a sample of
North American barley germplasm. Theor. Appl. Genet., 87: 329-401.
Heame, C.M, S. Ghosh and J.A. Todd. 1992. Microsatellite for linkage analysis of
genetic tralts. Trends Genet., 8: 288-294.
Helentjaris, T. 1987. A genetic linkage map for maize based on RFLPs. Trends Genet.. 3:
217-221.
Helentjaris, T., D.F. Weber and S. Wright. 1986b. Use of monosomics to map cloned
DNA fragments in maize. Proc. Natl. Acad. Sci., USA. 83: 6035-6039.
Helentjaris. T. , M. Slocum. S. Wnght, A. Schaefer and J. Nienhuis. 1986a. Construction
of genetic linkage maps in maize and tomato using restrlctlon fragment length
polymorphisms. Tlzeor .-Ippl Genet.. 73: 761-769.
Heun. M.. A.E. Kennedy. J.A Anderson. N.L.V. Lapt~ton. M.E. Soller and S.D.
Tanksley. 1991. Construct~on of a restriction fragment length polymorphism map
for barley (Hordeum vulgure). Genome, 34: 437-447.
Hill, A.P. 1975. Quantitative linkage: A statistical procedure for its detection and
estlmatton. .4nn. Hum. Genet. , 38: 439-449.
Holloway, J.L. and S.J. Knapp. 1994. Gmendel 3.0 User Guide. Department of Crop and
Soil Science, Oregon State University: Corvallis, OR 97331, USA.
ICRISAT (International Crops Research Institute for the Semi-Arid Tropics). 1989.
Annual report 1988, PatanChe~. AP 502 324, India: ICRISAT. p. 30.
ICRlSAT (Intemational Crops Research Institute for the Semi-Arid Tropics), 1997.
Screening techniques and sources of resistance to downy mildew and rust in pearl
millet. lnformatlon Bulletrn no. 48: 1-30.
Jansen R.C. 1993. Interval mapplng of multiple quantitative trait loci. Genetics, 135: 205-21 1
Jansen, R.C. and P. Stam. 1994. High resolution of quantitative traits into multiple loci
via interval mapping. Genetics, 136: 1447-1445.
Jarman, A.P. and R.A. Wells. 1989. Hypervariable minisatellites: recombinators or
innocent bystanders? Trends Genet.. 5: 367-371.
Jayakar, S.D. 1970. On the detection and estimation of linkage between a locus
influencing a quantitative character and a marker locus. Bron~etrics. 26: 441-464.
Sensen, J . 1989. Estimation of recombination parameters between a quantitative trait
locus (QTL) and two marker gene loci. Theor. Appl. Genet.. 78: 613-618.
Jones, C.J., K.J. Edwards. S. Castaglione, M.O. Winfield, F. Sala. C. Van de Wiel, G.
Bredemeijer, B. Vosman, M. Matthes, A. Dalx, R. Brettschneider, P. Bettini, M.
Buiatti. E. Maestri, A. Malcevschi, N. Marmiroli, R. Aert, G. Volckaert, J. Rueda,
R. Linacero, A. Vazquez and A. Karp. 1997. ~ e ~ r o d u c i b i l i t ~ testing of RAPD,
AFLP and SSR markers in plants by a network of Furopean laboratories. Mol.
Breed.. 3: 381-390.
Jones. E.S.. C.J. Liu. M.D. Gale. C.T Hdoh and J.R. W~tcombe. 1995. Mapp~ng
quantitative trait loci ibr downy mildew resistance In pearl millet. Theor. Appl.
Gener., 91: 448-456.
- Jones. E.S., W.A. Breese and D.S. Shaw. 2001. Spray inoculation of pearl millet with the
downy mildew fungus. Sclerosporc! grumrnicolu: Chilling of sporangial inoculum
prevents premature release of zoospores and their subsequent spray damage. PI.
puthol.. (In press).
Jones, E.S., W.A. Breese, C.J. Liu, S.D. Singh, D.S. Shaw and J.R. Witcombe. 2002.
Mapping quantitative trait loci for resistance to downy mildew in pearl millet:
Field and glasshouse screens detect the same QTL. Crop Sci., 42: 13 16-1 323.
Kadam, B.S, S.M. Pate1 and R.K. Kulkami. 1940. Consequences of inbreeding in bajra.
J. Heredity, 31: 201 -207.
Kalendar, R. , T. Grob. M. Reg~na, A. Suoniemi and A. Schulman. 1999. IRAP and
REMAP: two new retrotransposon-based DNA fingerprinting techniques. Theor.
Appl. Genei., 98: 704-71 1 .
Kato, K., H. Miura and S. Sawada. 2000. Mapping QTLs controlling grain yield and its
components on chromosome 5A of wheat. Theor. Appl. Gener.. 101: 1 1 14-1 121.
Kearsey, M.J. 1998. The principles of QTL analys~s (a minimal mathematical approach).
J. Exp. Bot.. 49: 619-1623.
Kearsey, M.J. and A.G.L. Farquhar. 1998. QTL analysis in plants; where are we now?
Heredity. 80: 137- 142.
Kearsey, M.J. and H.S. Pooni. 1996. The Genetical analysis of quantitative traits.
Chapmanand Hall, London.
Kearsey, M.J. and V. Hyne. 1994. QTL analysis: A simple 'marker-regression' approach.
Theor. Appl. Genet.. 89: 698-702.
K e n . P . B.W. Diers. T.C. Olson and R.C. Shoemaker. 1990. RFLP mapplng in soybean:
dssociation between marker lac! and variation in quantitative traits. Gener~cs. 126:
7.'S-711.
Kerns. M.R., J .W. Dudley and'G.K. Rufener 11. 1999. Tester and type of progeny affect
QTL detect~on in maize. A ~ u ~ ~ I ~ c I . 44: 69-83.
Klcherer. S.. G. Backes. U. Walther and A. Jahoor. 2000. Localising Q T L ~ for leaf rust
resistance and agronomic traits in barley (Horcie~ertm \.rtlgure L.). Theor. .-lppl.
Genet.. 100: 881-888.
Knapp, S.J., W.C. Bridges and B.H. Liu. 1992. Mapping quantitative trait loci using non-
simultaneous and simultaneous estimators and hypothesis tests. In: Plant
Genomes: Methods for Genetic and Physical Mapping. (Beckmann, J.S., and T.S.
Osbom. Eds.). Kluwer Academic Publishers, Dordrecht. The Netherlands.
Pp.209-237.
Knapp, S.J., W.C. Bndges Jr and D. Birkes. 1990. Mapping quantitative trait loci using
molecular marker linkage maps. Theor. Appl. Genet., 79: 583-592.
Kolesnikova Maria Alekssandrovna. 2001. Mapping new quantitative trait loci (QTL) for
downy mildew resistance in pearl m~llet. PhD thesis submitted to Russian
National Academy of Sciences. Moscow.
Krelke, C.M.. J.R.A. de Koning, J.H. Vinke, J.W. Van Oijen, C. Gebhardt and W.J.
Stiekema. 1993. Mapping of loci involved in quantitatively inherited resistance to
the potato cyst-nematode Globodera ros~ochiensis pathotype Rol. Theor. Appl.
Genet.. 87: 464-470.
Kumar, K.A. and D.J. Andrews. 1989. Pearl millet: current status and future potential.
Outlook Agr~c. , 18: 46-53,
Lande, R. and R. Thompson. 1990. Efficiency of marker-assisted selection in the
improvement of quantitative traits. Genetics, 124: 743- 756.
Lander, B.S. and D. Botstein. 1989. Mapping Mendelian factors underlying quantitative
traits using RFLP linkage maps. Genetics, 121: 185-199.
Lander. E.S., P. Green. J . Abrahamson. A. Barlow. M.J. Daly. S.E. Lincoln and L.
Newburg. 1987. MAPMAKER: An interactive computer program for constructing
genetic linkage maps of c.xper~r.lental and natural populattons. Generics, I : 174.
Lnndry. B.S. and R.W M~chelrnore. 1987. Methods and applications of restriction
fragment length polymorphism analysis to plants. In: Ta~loring genes for crop
improvement: 4 n agricultural perspective. (Bruening, G.. J. Harada and A.
Hollaender. Eds. ). Plenum Press, New York.
Landry, B.S., R.V. Kesselt. Barry Farrara and R.W. Michelmore. 1987. A genetic map of
lettuce (Luctucu surivu L.) with restriction fragment length polymorphism,
isozyme, disease resistance and morphological markers. Genetrcs, 116: 33 1-337.
Law, C.N. 1967. The location of factors controlling a number of quantitative characters
In wheat. Genet~cs. 56. 145-461.
Lee. M. 1995. DNA markers in plant breeding programs. Adv. .4gron.. 55: 265-344
Leonards-Schippers, C., W. Gieffers, R. Schaufer-Pregl, E. Ritter, S.J. Knapp, F.
Salamini and C. Gebhardt. 1994. Quantitative resistance to Phvlophthora
infestans in potato: A case study of QTL mapping in an allogamous plant species.
Genetics, 13 7: 67-77.
Lewers, K.S., E.H. Crane. C.R. Bronson, J.M. Schupp, P. Keim and R.C. Shoemaker.
1999. Detection of linked QTL for soybean brown stem rot resistance in 'BSR 101'
as expressed in a growth chamber environment. Mol. Breeding, 5: 3332.
Liao. C.Y., P. Wu, B. Hu and K.K. Yi. 2001. Effects of genettc background and
environment on QTLs and epistasis for rice (Otyza sot~va L.) panicle number.
Theor. Appl. Genet.. 103: 104- 1 1 1.
Lln, Y.R., K.F. Schertz and A.H. Paterson. 1995. Comparative analysis of QTLs affecting
plant height and maturity across the Pouceae, in reference to an intraspecific
sorghum population. Genetics. 140: 391-41 1 .
Lincoln, S., M. Daly and E. Lander. 1992a. Mapping genes controlling quantitative traits
with MAPMAKERIQTL I . I. Whitehead Institute Technical Report. 2nd edition.
Lincoln. S . M. Dalv and E Lander 1992b. Constmctlng genetlc maps w ~ t h
MAPMAKEWEXP 3 O Whitehead lnstltute Techn~cal Report 3Id cdlt~on.
LIU. B H and S J. Knapp 1992. QTLSTAT I 0 a software for m,ipplng complex tralt
uslng nonl~near models. Oregon State Un~vers~ty
Llu. C.J.. J.R. Wltcombe. T S Plttaway, M. Nash, 13 T. Hash. C S. Busso and M.D. Gale. - 1994 An RFLP-based genetlc linkage map of pearl m~llet (Pe~lnrserum glaucum)
Theor .4ppl Genet 89 481-487
Lu. Y Y and B H. Llu 1995 PGRI, a software for plant genome research Plant Genome
111 conference, San D~ego, CA Pp 105
Lubberstedt, T , A.L. Melchlnger, D. Kle~n, H. Degenhardt and C. Paul. 1997. QTL
mapping In testcrosses of European fllnt l~nes of malze: 11. Companson of
d~fferent testers for forage qual~ty tram Crop Scr . 37 1913-1922
Lubberstedt, T.. A L Melchlnger, S. Fahr, D. Kleln, A. Dally and P. Weathoff. 1998.
QTL mapplng In testcrosses of flint lines of maze: 111. Companson across
populations for forage traits. Crop Scr.. 38: 1278-1289.
Lubberstedt, T.. A.L Melch~nger, Chris C.Schon, H.Fnednch Utz and D Kle~n. 1997.
QTL mapplng In testcross of European Fl~nt lines of malze 1 Companson of
d~fferent testers for forage y~eld tralts. Crop SCI . 37 921-931
Lyamlchev, V , M.A.D. Brow and J.E. Dahlberg. 1993. Structure-speclfic
endonucleolyt~c cleavage of nucle~c acld by eubactenal DNA polymerases.
Scrence. 260 778-783
Lynch, M. and B. Walsh. 1998 Genet~cs and Analys~s of Quantitative Tra~ts. Sinauer
Assoc~ates, Sunderland, MA. USA
Mal~epaard, C and J.W. Van Oo~jen. 1994. QTL mapplng In a full-s~b fam~ly of an
outcrossing species. In: B~ometncs in Plant Breeding: Applicat~ons of molecular
markers. (Van Ooijen, J.W and J . Jansen, Eds.). Proceedings of the nlnth meeting
of the EUCARPIA, Sect~on B~ometncs In Plant Breeding, 6-8 July, 1994.
Wagen~ngen: The Netherlands. Pp. 140- 146.
Mang~n. B., B. Goffinet and ~ 4 . Rebai. 1994. Construct~ng confidence intervals for QTL
location. (~ene1rc.s. 138 1301 - 1308.
Manly. 1C.F 1993. A ~Mac~ntosh program for storage and analys~s ofexpcrimentcl genetic
mapping data. Murnnl. &eno,ne. 1 : 123-1 26.
Mdnly, K.F. and J.M. Olsen. 1999. Overview of QTL mapping software and introduction
to Mapmanager QT. itlrr~n~n. Getlon~e. 10: 327-334.
Manzanares-Dauleus. M.J.. R . Delourme and F. Baron. 2000. Mapp~ng of one major
gene and QTLs Involved in resistance to club root in Br~~ssrcu nuptls. Theor. Appl.
Genet., 101: 885-889.
Martin, B., J . Nienhuis, G. King and A. Schaefer. 1989. Restriction fragment length
polymorphisms associated with water use efficiency in tomato. Science. 243:
1725-1728.
Martinez, 0. and R.N. Cumow. 1992. Est~mation the locations and sizes of the effects of
quantitative trait loci using flanking markers. Theor. Appl. Genet.. 35:480-488.
Mather, D.E., N.A. Tinker, D.E. LaBerge. M. Edney, B.L. jones. B.G. Rossnagel, W.G.
Lcggc, K.G. Briggs, R.G. Irvinc, D.E. Falk and K.J. Kasha. 1997. Regions of the
genome that affects grain and malt quality in North American two-row barley
cross. Crop Scl.. 37: 544-554.
.Mather, K. 1949. Biometrical genetics, I" edn. Methuen: London
Mather, K. and R.J. Hanison. 1949. The manifold effect of selection. Herediy. 3 : 1-52.
McCouch. S.R., G. Kochert, Z.H. Yu. Z.Y. Wang, G.S. Khush. W.R. Coffman and S.D.
Tanksley. 1988. Molecular mapping of rice chromosomes. Theor Appl. Genet.
76: 815-824.
Mehta, N. and J.K. Dang. 1987. Studies on the inheritance of downy mildew. ergot and
smut of pearl millet. Indian J. Mycol. PI. Pathol., 17: 200-203.
Melchinger, A.E. 1990. Use of molecular markers in breeding for oligogenic disease
resistance. PI. Breed.. 104: 1-19.
klelchinger. A E. H Frlednch Utr dnd Chr~s C Schon 1998 Qudntitatlve trait locus
(QTL) mdpplny using different testers and Independent populat~on samples In
rnalre reveals low power of QTL detect~on and large bids In estimates of QTL
effects Gettetrcs I49 383-103
Meyer, W., T.G. Michell, E.Z. Freedman and R. Vilgalys. 1993. Hybridization probes for
conventional DNA fingerprinting used as single pnmers in the p6lymerase chain
reaction to distinguish strains of Cnptococtts tleoforn~ntts. J. Clrt~icul Biol.. 31:
7274-2280.
Michelmore. R.W.. I. Paran and R.V: Kesseli. 1991. Identification of markers linked to
disease resistance genes by bulked-segregant analysis; a rapid method to detect
markers in specific genomic regions by using segregating populations. Proc. Null.
Acad. Scr. USA. 88: 9828-9832.
Mohan, M., S. Nair, A. Bhagwat, T.G. Krishna, M. Yano, C.R. Bhatia and T. Sasaki.
1997. Genome mapping, molecular markers and marker-assisted selection in the
improvement of quantitative traits. Mol. Breeding, 3 : 87-103.
Nene, Y.L. and S.D. Singh. 1976. Downy mildew and ergot of pearl millet. PANS, 22:
366-385.
Niemann-Sorenson. A, and A. Robertson. 1961. The association between blood groups
and several production characters in three Danlsh cattle breeds. Acta Agric.
Scand.. 11: 163-196.
Nienhuis, I., T. Helentjaris, M. Slocum, B. Ruggero and A. Schaefer. 1987. Restriction
fragment length polymorphism analysis of loci associated with insect resistance in
tomato. Crop Sci.. 27: 797-803.
N~kiforov, T.T., R.B. Rendle. P. Goelet. Y.H. Rogers, M.L. Kotewicz, S . Anderson, G.L.
Trainor and M.R. Knapp. 1994. Genetic bit analysis: a solid phase method for
typing single nucleotide polymorphisms. Nucl. Acids Res.. 22: 4167-41 75.
Nilsson, N.O., M. Hansen. A.H. Panagopoulous, S. Turesson, M. Ehlde, M.
Christiansson, T.M. Rading, M. Rissler and T. Kraft. 1999. QTL analysis of
Cercospora leaf spot resistance in sugar beet. PI. Breed., 118: 327-334.
Orf. J.H.. K.Chase, F R. Alder. L.M. Mansur and K.G. Lark. 1999. Genetics of soybean
agronomic tralts: 11. Interactions between yield quantitative tralt loc~ in soybean.
Crop Scr.. 39: 1652-1657
Orita, M., H. Iwahana, H. Kanazawa. K. Hayashi and T. Sekiya. 1989. Detection of
polymorphisms of human DNA by gel electrophoresis as single-strand
conformation polymorphisms. Proc. Natl. Acad. Scr. USA. 86: 2766-2770.
Paterson, A.H. 1998. Prospects of cloning the genetic determinants of QTLs.
In: A.H. Paterson (ed.) Molecular detection of complex traits. CRC press. Boca
Raton, FL. Pp.289-293.
Paterson, A.H., E. Lander, J.D. Hewin, S. Peterson, S.E. Lincoln and S.D. Tanksley.
1988. Resolution of quantitative traits into Mendelian factors by using a complete
linkage map of restriction fragment length polymorphisms. Nature. 335: 7 - "1-726.
Paterson, A.H., S. Damon. J.D. Hewitt, D. Zarnir. H.D. Rabinowitch. S.E. Lincoln, E.S.
Lander and S.D. Tanksley. 1991. Mendelian factors underlying quantitative traits
in tomato: comparison across species, generations, and environments. Genetics.
127: 181-197.
PC, M.E.. L. Gianfranceschi, G. Taramino. R. Tarchini, P, Angelini, M. Dani and G.
Binelli. 1993. Mapplng quantitative trait loci (QTLs) for resistance lo G~berella
zeae infection In maize. Mol. Gen. Gene[.. 241: 11-16.
Pethani, K.V., R.L. Kapoor and S. Chandra. 1980. Gene action and phenotypic stability
for incidence of downy mildew disease in pearl millet. Indian J .4grl. Res.. 14:
217-223.
Phillips, R.L. 1999. Unconventional sources of genetic diversity: de trovo variation and
elevated epistasis. In: Plant Breeding in the Tum of the Millennium. (Borem, A,,
M.P. Giudice and N.S. Sakiyama. Eds.). University of Vicosa, Brazil. Pp.103-131.
Pnoul, J.L., S. Quanie, M. Causse and D. de Vieme. 1997. Dissecting complex
physiological functions through the use of molecular quantitative genetics. J. Exp.
Bot.. 18: 1151-1163.
Puttemill. J.. F. Rob~nson. K. Lee. R. Simon and G. Coupland. 1995 .The CONSTANS
gene of :lrtthrllopsrs pronlotes flowering and encodes a protein showing
similar~ties to zinc finger transcription factors. Cell, YO: 847.857.
Ral. K.N, and W.W. Hanna. 1990. Morpholog~cal characteristics of tall and dwarf pearl
millet isolines. Crop Scr.. 30: 23-25.
Rasmusson. J.M. 1935. Studies on the inheritance of quantitative characters in Pisrrm: 1.
Preliminary note on the genetics of flowering. Hereditas. 20: 161-180.
Rongwen. J. M.S. Alikaya. A.A. Bhagwat. U. Lavi and P.B. Cregan. 1995. The use of
microsatellite DNA markers for soybean genotype ~denttficatlon. Theor. .4ppl.
Genet.. 90: 43-48.
Sarkar, G., J . Cassady, C.D.K. Bottema and S.S. Sommer. 1990. Characterization of
polymerase chain reactlon amplification of specific alleles. Anal. Biochem.. 186:
64-84.
SAS version 8.0. 1999. SAS institute Inc., Cary, NC, USA.
Sax, K. 1923. The association of size differences with seed coat pattern and pigmentation
in Phaseolus vulgaris Generics, 8: 552-560.
Schafer- Pregl, R., E. Ritter, L. Concilio, J . Hesselbach, L. Lovatti. B. Walkemeier, H.
Thelen. F. Salamini and C. Gebhardt. 1998. Analysis of quantitative trait loci
(QTLs) and quantitative trait alleles (QTAs) for potato tuber yield and starch
content. Theor. Appl. Gener., 97: 834-846.
Schon, C.C., A.E. Melchinger, J . Boppenmaier, E. Bmnklaus-Jung, R.G. Hermann and
J.F. Seitzer. 1994. RFLP mapping in maize: Quantitat~ve trait loci affecting
testcross performance of elite European flint lines. Crop Sci.. 34: 378-389.
Setiawan, A,, G. Koch, S.R. Barens and C. lung. 2000. Mapping quantitative trait loci
(QTLs) for resistance to Cercospora leaf spot disease (Cercospora heiicola Sacc.)
in sugar beet (Beta vulgaris L.). Theor. Appl. Genet.. 100: 1 1 76-1 182.
Shah. M.M.. K.S. Gill. P.S. Baenz~gcr. Y.Yen. S.M. Kaeppler and H.M. Ariyathne. 1999.
Molecular mapplng 01' l o c ~ for agronomtc tralts on chromosome 3A of bread
wheat. Crop Scr.. 39: 1728-1 732.
Shams. A. 2001. Marker-assisted Improvement of downy mildew resistance In elite pearl
millet (Pennlsetum gluucunr) parental line H 771833-2. Ph.D. thesis, Chaudhary
Charan Singh Haryana Ayrtcultural University, Hisar, India. - Sharp, P.J.. S. Chao. S. Desal and M.D. Gale. 1989. The isolation, characterization and
application in the Trit~ceae of a set of wheat RFLP probes identifying each
homoeologous chromosome aim. Theor. .+@pi. Genet., 78: 342-348.
Shetty, H.S. 1987. Biology and epidemiology of downy mildew in pearl ntillet. In:
Proceeding of the Internattonal Pearl Millet Workshop. (Witcornbe, J.R. and S.R.
Beckeman. Eds.). ICRISAT, Patanchem, AP, India. Pp.147-160.
Shinde. R.B., F.B. Patil and R.A. Sangave. 1984. Resistance to downy mildew in pearl
millet. J Maharashlra Agric. Univ.. 9: 337-338.
Sillanpaa, M.J. and E. Ajas . 1998. Bayesian mapping of multiple quantitative trait loci
from incomplete inbred line cross data. Genetics, 148: 1373-1 388.
Singh, F., R.K. Singh, R.M. Singh and R.B. Singh. 1978. Genetic analysis of downy
m ~ l d e w (Sclerospora graminicola) resistance in pearl millet (Pennisetum
t.vpho~des ( B u m . ) S. & H.). Z. Pflanzen~iicht. 81: 54-59.
Singh, F., R.M. Singh, R.B. Singh and R.K. Singh. 1980. Genetic studies of downy
mildew resistance in pearl millet. In: Trends in Genetical Research on
Penniserurns. (Gupta. V.P. and J.L. Minocha. Eds.). Punjab Agricultural
University, Ludhiana.
Singh, S.D. 1990. Sources of resistance to downy mildew and rust in pearl millet. Plan!
Dis., 74 : 87 1-874.
Singh, S.D. and R. Gopinath. 1985. A seedling inoculation procedure for detecting
downy mildew resistance. Planr Dis.. 69: 582-584.
Singh. S.D.. J.P. Wilson, S.S. Navi. B.S Talukdar. D.E. Hess and K.N. Reddy. 1997.
Screening technrques and sources of resistance to downy n~l lden and st In pearl
millet. Information Bullet~n. No. 48. ICRISAT: Patanchcru. Andra Pradesh. 502
324, Ind~a.
Singh, S.D., S.B. King and J . Werder. 1993. Downy mildew disease of pearl millet.
lnformation Bulletin no.37. International Crops Research InstituE for the Semi-
Arid 'Tropics: Patancheru, Andra Pradesh, 502 324. India. Pp. 36.
Srngh. S.D.. S.L. Ball and R.P. Thakur. 1987. Problems and strategies In the control of
downy mildew. In: Proceedings International Pearl Millet Workshop (Witcombe,
J.R. and S.R. Beckerman. Eds.). ICRISAT: Patancheru, Andra Pradesh, India.
Pp.161-172.
Soller, M. and J.S. Beckmann. 1983. Genetic polymorphism in varietal Identification and
genetic improvement. Theor. Appl. Genet.. 47: 179-190.
Soller, M. and J.S. Beckmann. 1990. Marker-based mapping of quantitative trait loci
using replicated progenies. Theor. Appl. Genet.. 80: 205-208.
Soller. .M., A. Genizi and T. Brody. 1976. On the power of experimental designs for the
detection of linkage between marker loci and quantitative loci in crosses between
~nbred lines. Theor. Appl. Genet.. 47: 35-39.
Spaner, D., B.G. Rossnagel. W.G. Legge. G.J. Scoles, P.E. Eckstein, G.A. Pemer, N.A.
Tinker, K.G. Briggs, D.E. Falk, J.C. Afele, P.M. Hayes and D.E. Mather. 1999.
Verification of quantitative trait locus affecting agronomic traits In two-row
barley. Crop Sci., 39: 248-252.
Stam. P. 1993. Construction of integrated genetic linkage maps by means of computer
package-Joinmap. The PIanr Jour.. 5: 739-744.
Stam. P. and J.W. Van Ooijen. 1995. JoinMap TM version 2.0:Software for the calculation
of genetic linkage maps CPRO-DLO: Wageningen, The Netherlands.
Stuber. C.W. 1992. Biochemical and molecular markers in plant breeding. In: Janick
(ed.) Plant Breeding Revrews. 9: 37-61.
Stuber. C.W.. M. Edwards and J . Wendel. 1987. Molecular marker-facllitaled
investigations of quant~tative trait loci in maize. 2. Factors influencing yleld and
its component traits. Crop Scr.. 7 : 639-648.
Stuber. C.W.. M. Goodman and R. Moll. 1982. Improvement of y~eld and ear number
resulting from selection at allozyme loci in a maize population. Crop. Sc~cr.. 22:
737-740.
Stuber. C.W., Polacco. M and M. Lynn Senior. 1999. Synergy of empirical breeding,
marker-asslsted selection. and genomlcs to increase crop yield potential. Crop
Scl., 39: 1571-1583.
Stuber, C.W., R. Moll. M. Goodman. H. Schaffer and B. Weir. 1980. Allozyme
frequency changes associated with selection for increased grain yield in maize
(Zea mays L.). Genetics. 95: 225-236.
Stuber, C.W., S.E. Lincoln, D.W. Wolff, T. Helenjaris and E.S. Lander. 1992.
Identification of genetic factors contributing to heterosis in a hybrid from elite
maize inbred lines using molecular marker. Generics, 132: 823-839.
Suiter. K.A., J.F. Wendel and J.S. Case. 1983. Linkage-I: A PASCAL computer program
for the detection and analysis of genetic linkage. J. Hered., 74: 203-204.
Tanksley, S.D. 1993. Mapping polygenes. Ann. Rev. Genet.. 27: 205-233.
Tanksley, S.D. and J . Hew~tt. 1988. Use of molecular markers In breeding for soluble
solids content in tomato- a re-exam~nation. Theor. Appl. Genet., 75: 81 1-823.
Tanksley, S.D., H. Medina-Filho and C.M. Rick. 1982. Use of naturally occuning
enzyme variation to detect and map gene controlling quantitative traits in an
interspecific backcross of tomato. Herediy. 49: 11-25.
Tanksley, S.D., M.W. Ganal and G.B. Martin. 1995. Chromosomal landing: a paradigm
for map-based gene cloning in plants with large genomes. Trends. Genet.. 11:
63-68.
Tar'an, B., E.T. E. Michales and K.P. Pauls. 2002. Genetic mapping of agronomic traits
in common bean. Crop Scl.. 42: 544-556.
Teulat. B.. 0. Merah. 1. Souyr~s and D. -Th~s. 2001. QTLs for agronomic tram from a
Mediterranean barely progeny grown In several environments. Theor. :lppl.
Genet.. 103: 774-787
Thoday, J.M. 1961. Location of polygenes. Nature, 191: 368-370.
Tinker, N.A.. D.E. Mather. B.G. Rossnagel. K.J. Kasha, A. Kleinhofs. P.M. Hayes, D.E.
Falk, ,T. Ferguson. L.P. Shugar. W.G. Legge. R.B. Irvine. T.M.Choo. K.G.Briggs,
S.E. Ullrich. J.D. Franckow~ak. T.K. Blake. R.J. Graf. S.M. Dosing, M.A. Saghai
Maroof. G.J. Stoles. D. Hoffman. L.S. Dahleen. A. Kilian. F. Chen, R.M.
Biyashev. D.A. Kudma and R . J . Steffenson. 1996. Reg~ons of the gcnomic that
affect agronomic performance In two-row barley. Crop Scr., 36: 1053-1062.
Tiret, L., L.Abel and R.Rakotovao. 1993. Effect of ignoring genotype-environment
Interaction on segregation analysis of quantitative traits. Generic Epldemiologv.
10: 581-586.
Ullstnrp, A.J. 1973. An overview of the downy mildew of corn and sorghum. Report of a
workshop on the downy mildew of sorghum and corn, Technical Report, Texas
Agricultural Experimental Station. Department of Plant Sciences. 74-1: 5-12.
Utz, H.F. and A.E. Melchinger. 1995. PLABQTL: A computer program to map QTL,
Version 1.0, Univers~ty of Hohenheim, Germany.
Utz, H.F., A.E. Melchinger and C.C. Schon. 2000. Bias ahd sampling error of the
estimated proportion of genotypic variance expla~ned by QTL determined from
experimental data in maize using cross validation and validation with independent
samples. Genetics, 154: 639-849.
Van Ooijen, J.W. 1999. LOD significance thresholds for QTL analysis In experimental
populations of diploid species. Heredity. 83: 613-624.
Van Ooijen, J.W. and C. Maliepaard. 1996. M ~ ~ Q T L ~ ~ version 3.0: Sofiware for the
calculation of QTL posittons on genetic maps. CPRO-DLO, Wageningen, The
Netherlands.
Veldboom, L R . M Lee and W L Woodman 1994 Molecular-mdrker-fac~l~tated studled
of morpholog~cal tralts In malze I L~nkage analvsls and determlnat~on of QTLs
for morpholog~cal tralts Theor lppl Gener 92 230-214
Via. S and R Lande 1987 Evolutron of genetlc \anablllty In a spat~ally heterogeneous
env~ronment Effects of genotype-env~ronment Interaction Gener Res -19 147- 156
Vos. P . R Hopers. M Bleeker. M Re~jans T Van de Lee. M ~ o m e s : A Fanjters, J
Pot, J Peleman. M Ku~per and M Zabeau 1995 AFLP A new techn~que for
DNA fingerpr~nt~ng Nltcl 4crds Res 23 4407-4414
Wang, G L , D J Macklll. J M Bonman. S R McCouch. M C Champoux and R J
Nelson 1994 RFLP mapplng of genes confemng complete and partla1 resistance
to blast In a durably resistant nce cultlvar Genetrcs 136 1421-1434
Weber, D and T Helentjans 1989 Mapping RFLP locl In malze uslng B-A
translocat~ons Genet~cs 121 583-590
Weller, J , M Soller and T Brody 1988 Linkage analys~s of quantltatlve tralts In an
~nterspec~fic cross of tomato (Lvcopersrcon esculentum x L prmprnellrfollum) by
means of genetlc markers Genetrcs 118 329-339
Weller, J I 1987 Mapplng and analysls of quant~tat~ve tralt l o c ~ In Lvcopersrcon (tomato)
w ~ t h the a ~ d of genetlc markers uslng approximate maxlmum l~kellhood methods
Heredrrv 59 4 13-42 1
Welsh, J , and M McClelland 1990 F~ngerpnntlng genornes uslng PCR w ~ t h arbitrary
pnrners Nucl Acids Res . 8 7213-721 8
Weltzlen R E and S B K ~ n g 1995 Recurrent select~on for downy m~ldew res~stance In
pearl m~l le t PI Breed 114 308-3 12
Wlll~ams, J G K , A R Kubellk, K J L~vak, J A Rafalskl and S V Tlngey 1990 DNA
polyrnorph~sms ampl~fied by arb~trary pnmers are useful as genet~c markers
Nucl Acrds Res 18 653 1-6535
W ~ l l ~ a m s , M N V . N Pande. S Na~r , M Mohan and J Bennett 1991 Restnct~on
fragment length polymorph~sm analyses of polymerase cham reactlon products
amplified from mapped loci of rice (On,:o scltr~u L.) genomlc DNA. Tlteor .4ppl.
Getter.. 82: 489-498.
M:~lliams. R.J.. S.D. Singh and M.N. Pawar. 1981. hn improved field screening technique
for downy mildew resistance in pearl millet. Plrzrrt Dis . 65: 239-241.
Wilson. J.P. 1999. Pearl millet diseases: A compilation of information on the known
pathogens o f pearl millet. Penrtisertrm gluucum (L.) R . Br. Agriculture Handbook
No.716. U.S. Department of Agnculture. Agnculture Research Servtce. National
Techn~cal Information Service: Springfield. VA. USA.
Winter, P. and G. Kahl. 1995. Molecular marker technologies for plant ~mprovement.
WorldJ. Microbiol. Biotech.,ll: 438-448.
Witcornbe, J.R. and C.T. Hash. 2000. Resistance gene deployment strategies in cereal
hybrids using marker-assisted selection: Gene pyramiding, three-way hybrids and
synthetic parent populations. Euphvrica, 112: 175-186.
Wu, W.R. and W.M. Li. 1994. A new approach for mapping quantitative trait loci using
complete genetic marker linkage maps. Theor. Appl. Genet., 89: 535-539.
Xiao. J., J. Li. L. Yuan and S.D. Tanskley. 1996. Identification of QTLs affecting traits of
agronomic importance in a recombinant inbred population derived from
sub-specific nce cross. Theor. Appl. Genet.. 92: 230-244.
Yadav, R.S., C.T.Hash, F.R. Bidinger and G.P. Cavan. 2002a. Quantitative trait loci
associated with traits determining grain and stover yield in pearl millet under
terminal drought- stress conditions. Theor. Appl. Genet., 104: 67-83.
Yadav, R.S., F.R. Bidinger, C.T. Hash, Y. P. Yadav, S.K. Bhatnagar and C.J. Howarth.
2002b. Mapping and characterization of QTL x E interactions for traits
determining grain and stover yield in pearl millet. Paper accepted in Theor. Appl.
Gener.).
Yamamoto, T., Y. Kuboki. S.Y. Lin. T. Sasaki and M. Yano. 1998. Fine mapping of
,quantitative trait loci Hd-I, Hd-2 and Hd-3, controlling heading date of rice, as
single Mendelian factors. Theor. Appl. Gener. 83: 813-820.
Ynn. J . . J . Zhu, C. He. M. Benmoussa and P. Wu. 1999. Molecular marker-ass~sted
dissection of genotype x cnvlronment interaction for plant type tralts In rice
IOn,ztr surrvci L. ) . Crop .$,,I . 39. 38-544 .
l'ang, H., M. Shankar. B.J. Bu~rchell. M.W. sweetingham. C. Camlnero and P.M. C.
Smith. 2002. Development of molecular markers using MFLP linked to a gene
confening resistance to Diaporthe to.xicu in narrow-leafed tupin (Luprnus
ungtistifolius L.). Theor :lppl. Genet.. 105: 265-270
Yano. M. and T. Sasaki. 1997 Genetlc and niolecular dissection of quant~tative malts In
nce. Plaltt Mol Rfol.. 35. 145-153,
Young, N.D. 1994. Constructing a plant genetic linkage map with DNA markers. In:
DNA-Based Markers in Plants. (Phillips, R.I. and I.K. Vasil. Eds.). Pp.39-57.
Young, N.D. 1996. QTL mapping and quantitative disease resistance In plants. Annu.
Rev. Ph.vtoputho1.. 34: 479-501.
Young, N.D. 1999. A cautiously optimistic vision for marker-assisted breeding. Mol.
Breeding, 5 : 505-5 1 .
Young, N.D. and S.D. Tanksley. 1989a. Graphics-based whole genome selection using
RFLPs. 111: Current Communlcntions in Molecular Biology: Development and
Application of Molecular Markers to Problems In Plant Genetics, (Helentjans, T. and
B. Burr. Eds.). Cold Spring Harbor Press: Cold Spring Harbor. NY. Pp.123-129.
Young, N.D. and S.D. Tanksley. 1989b. Restriction fragment length polymorphism maps
and the concept of graphical genotypes. Theor. Appl. Genet.. 77: 95-101.
Young, N.D., J. Miller and S D. Tanksley. 1987. Rapid chromosomal assignment of
multiple genomic clones in tomato using primary trisomics. Nucl Acids Res.. IS:
9339-9348.
Yu, Z.H., D.J. Mackill, J.M. Bonman and S.D. Tanksley. 1991. Tagging genes for blast
resistance in nce via linkage to RFLP markers. Theor. Appl. Genet., 81: 471-476.
Zeng, Z.B. 1993. Theoretical basis of separation of multiple linked gene effects on
mapping quantitative tralt loci. Proc. Narl. Acud. Sci. USA. 90: 10972- 10976.
Zeng, Z.B. 1994. Precision mapping of quantitat~ve tralt loc~. Ge~ier~cs. 136: 1457-1468
Zhang, X.Q.. K. Ross and J . P . Gustat'son. 2000. Phys~cal locat~on of homeologous groups
5 and 6 molecular markers mapped in Trirrcu111 ot,srii ,~i~~l I.. Ge~lerlcs. 91:
441 -445.
Zhu. H., G. Briceno, R. Dovel, P.M. Hayes, B.H. Liu, C.T. Liu and S.E. Ullrich. 1999.
Molecular breeding for ga in y~eld In barley: an evaluation of QTL effects in a
sprlng barley cross. Theor :lppl Getler.. 98: 772-779.
Zhuang. J .Y. , H.X. Lin. J . Lu. H.R. Qian. S. Hittarnani. N. Huang and K.L. Zheng. 1997.
Analysis of QTL x environment interaction for yield components and plant height
i l l rice. lheor. Appl Genet.. 95: 799-808.
Zietkiewiez, E., A. Rafalsk~ and D. Labuda. 1994. Genomic fingerprinting by simple
sequence repeat (SSR)-anchored polymerase chain reaction amplification.
Genornzcs. 20: 176-1 83.
1 2 m m m m m m m m m m u m m m m n ~ 1 ~ ~ n ~ r z z ~ z r z z 1 u
I , - i 1 ' 1 z m a < a , a m m m m m ; ,;u; , ; = c b
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , 8 . : 1 ' ' 9.. I...... ,
i a r m m m ~ a . m a m m m m o ; z c c < l r ~
I , , , , , , , , ~ / = n m m a m s ~ n n a a n m m a ~ ~ s ~ = ~ = r = u z u z C i z i i
1 l ~ m m ~ m m s ~ s m m m m a r r n m ~
1
~ . . , , , , , , , , , , , , , , , , < . ' , < I .....a a ! . . m;m . n . u u ~
r u z r r z z u I ' 1
m = i = ; z o z . z n z u
, ; p s h ; ~ ~ h ; g a s < ~ ~ n m ~ ; ; ~ h g z 0 9 ~ ~ E : g ? - ? : ; ~ ; ; ; ; ? : 2 > ? ; 2 2 ; ? ? , $ 2 < ~ ~ > ~ ~ B f i ;4$z$$$$$$$$$$$$$$&$$$$$$&$$s$$
- I 5 m ~ z ~ ~ = ~ s = ~ ~ m m = r ~ r ~ = ~ r n r n ~ r ~ a s m m r n I
< ~ ~ = T S ~ Z Z T Z ~ Z Z <
4 4 4 m m m m m m 4 4 4 4 4 4
c" cr
1
l a 1 2
I < a < < <
r o m m m
s ' < d ~ d < i ‘ , F 4 4 1 ~ T T T L T 4 4 d 4 4 4 1
L J = ~ 4
4 4 4 4 4 6 4
3 C T m Z
I g 44.444. 4 4 4 = z = = = m z = = = 4 4 = a < < < i
r n b r d i - c e < < < r r r = = = z r z r z r / < o < < z I
4 < 4 ~ 4 4 = 3 = ~ ~ = ~ 4 4 4 z = z < <