Genetic diversity analysis, QTL mapping

and marker-assisted selection for shoot fly

resistance in sorghum [Sorghum bicolor
(L,)Moench]

SHIVAJI PANDURANG MEHTRE

M.Sc.(Agri.)

DISSERTATION

Submitted T o The
Marathwada Agricultural University, Parbhani
In Partial Fulfillment Of T h e Requirement
For T h e Award o f t h e Degree Of

DOCTOR OF PHILOSOPHY
IN
Agricultural Botany
(Genetics a n d Plant Breeding)

DEPARTMENT OF AGRICULTURAL BOTANY

MARATHWADA AGRICULTURAL UNIVERSITY
PARBHANI - 43 1 402 [M.S.] INDIA
 DEDICATED
TO
MY LATE BELOVED
PARENTS
 CANDIDATE'S DECLARATION

I hereby declare that the dissertahon or part thereof has not been prev~ously

submitted by me to any other University or Institution for a degree or diploma

Place: Parbhani

Date:1$January 2006
Dr. S.T.Borikar
M.Sc. (Agri.),Ph.D.
Director of Research,
Marathwada Agricultural University,
Parbhani - 43 1 402 [MS]India.

CERTIFICATE - I

This is to certify that the dissertation entitled "Genetic diversity

analysis, QTL mapping and marker-assisted selection for shoot fly

resistance in sorghum [Sorghum bicolor (L.) Moench]" submitted to

Marathwada Agricultural University, Parbhani in partial fulfillment of the

requirement for the award of degree of Doctor of Philosophy in Genetics and

Plant Breeding embodies the results of bonafide research carried out by Shri.

S.P. Mehtre under my guidance and supervision and that no part of this

dissertation has been submitted for any degree or diploma.

The assistance and help received during the course of this investigation

have been fully acknowledged.

<
-
* -i ---.
Place: Parbhani (Dr. S.T. Borikar)

Date: I?, January 2006 Research Guide

 CERTIFICATE - I1
This is to certify that the dissertation entitled "Genetic diversity

analysis, QTL mapping and marker-assisted selection for shoot fly

resistance in sorghum [Sorghum bicolor (L.) Moench]" submitted by Shri.

S.P. Mehtre to the Marathwada Agricultural University, Parbhani in partlal

fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY in

the subject of Genetics and Plant Breeding has been approved by the

student's advisory committee after oral examination in collaboration ~ 5 t hthe

external examiner.

(1>6.P2~bkd
_ --
Lf---.-

( Dr.S.T. Borikar)
External Examiner Research Guide

(D wd!
.C. Sharma)

Associate De ipal (P.G.)

College of Agriculture,
M.A.U., Parbhani 431 402
 ACKNOWLEDGEMENT

It is my pride and honour to expre.r.7 mvprofound .sense ofgratitude,for providing

tne un o p p o r t u n i ~to undertake tltis research projeer at Inlernational Crops Re.reurch
histitt~tefor the Semi-Arid Tropics (IC'RISA71. I intmeit.r.elj~rhrr~tkDr. S. T. Borikur,
Direcror of Resec~rch,Marathivadti Agricultural (inrversitj' (A4.A l ! ) , I1arhhurii,for hi5
constant encouragement, meticulous fittidnncc ond tmccusing support during n1.v studv tit
M A. (1,.Purhhuni and during the course o f uchiciling thr, final .sh(rl,e o f m,v lhesis ut
I(,'RISAT.
It is nvl profoundprivilege and deep .tcn.re o f rcvcrcncc ant1 gratitude to Llr. C'. T,
Iliish. I'rinciple Scienti.rt, Moleritlar Breeding. ICRISAI: co-ridvisor uf niv reseiirch
cor1iit711ti.e.fi,r provriling excellen/ re.secri~ch,fiicilitie.s, n7c,/iculou.r guiticr~ice and
conslructive cri/icisrn throrrgh out the period y i 1hi.s invesiigation aniiprcparc~/ioiiuf the
I?~NI~II.YC~'~I~.

regard.\ to Dr. I1 C. .Ylii~rnra.

I u~ishl o record mj' ~ i n c r r eg~atituiieund rc.sj~cc/firl
I'rincii,/c Scienli.rl, E n t o m o l u ~ ,I('RI.S.41: mci~iher i f 1i7j. aihv.\oi:i, coninzittee /or
providing ercellent research fucillties, upp port. i i , h i l ~ Iteurted co-i~/~eratioii(in(/
cf?coiirigement illiriiig the corrrse of i~iy.rrur!l~
I am .smcerel> ohliged and iniiehtt~rlto Dr 0. 1: S. Redi!,,, 1'riticil)l~~
Scientist,
Sorgh~~ni
Breeding, member of mny cithiaory contnirilee fi)r j~rovidir7h.rc~etrrchmateritrl.
cor?.strrrctii,csug,qe,stioii, melrcu1ou.c. picliince a i d pr~porirtio17o f the n?anir.\cri~)rdurilig
iiij course ofrnve.stigi~lion.
I iviih to uxpre.ss n?,i sincere and 1i hvle-hecrricd grarirtnle to Dr. 1.'. G. Ktilkarrti
Ijeud, Del~arlmento?f Genetics and Ploni Ureedin~.111, '4. I:. Porbhani. mciriher o f n7j.
orh~rsor?,committee, fir his conti~iuourencourogenie~it(itid fi7.e~rtS U ~ J J O Ito
. ~ car1:1,O N I ti!),

doclorrrl research at I('RISA1:

I .rincere!~)thank I'ice-chuncellc~r,'1.1. .4. U Porhhani, Dr. S S. Kcidrint, , f i r Ilia
kiiid s ~ p ~ ~to
o rcarry
t out my doctoral resecrrch a1 ICRI.C-1T
I ant also lhankful 10 Dr. M. 1' Dhnhle. Dean and Z)ire~,torof histruction.
.21 .4. U., Parbhnai and Dr. D. K. Shelke, ,4ssociate Deon and Principal, College (!j

,4,qricultu~e,Parbhorli for his valuable strg,qestions and encouragement.

I ani profoundly thankful to all the staff members o f Department q f Genetics and
Plant Breeding, Sorghum Research Station and Tissue Culture Center; A4 .4. 1,'.
Parbhuni,for their co-operation, encouragement and help rendered during the course q f
this sludy.
 I u~ishto acknowledge and express sincere thanks to Messrs. C Muralidhar.
Basheer Ahmed, Somaraju C. Ashok Kumar, and Mrs. Pochamma (RWF) ,for their kind
co-operation in the greenhouse, field and lab work ,for succes.r/ul cornperition of niy
docroral research project.
I am perronaliy obliged to Dr. S Chandra (Bionietriciun). Mr. 1i.P. Pra.sanfh, untl
Ms. Rupa.fi)r valuahle suggestions and help in .stoti.rricctl cma~,sc.s.
I wish to record my sincere gratiiude und re.rpectf11 regard^ to 1)r.v. J 11 Croucii
und D. 1~~)isington
ex- and,former- Global Theme Leaders Btotcchnology. ICRlSA1: for
extending support and ,facility, and Dr. Rolf Folkerstuma, Ex l'osf-docrorctl Fel/ou'.
I(.'RISAZ: for genuine guidance, construcrivt, suggesrions and suppori in inolecular
.srudie,s.
1,fccl privilege to render my hearr:felt worti.~o f ctppreciatron lo Entomolol3; uriii
stclff at ICRISAT: Mr. G. Pampapathy, Dr. Mukcsh Dhillo~i. .b4essr.s. Xtljci Rao,
~ . Kuniar and Ramakri.rhnu fir rhcir
Hrindranafh, Matihltsudhan Reddv, Peter h ~ a )Shivu
kind co-operation, hard work in recording observatron.~and all other ~)o.s.srhlehelp
durrng tile course of this rnve,stigatiun.
A4y sincere and .s/~ecra/riiarih goes lo Mr. I4jay Dalvi. l'h. D.Scholcir, and Dr.
S ~ m t o ~Deshpanile,
h C'onsultant, I('RIS.47: .fi,i. [heir untiring Iielp. ~i'holchcmrfcd co-
operufion and moral support all along the course of this sruc!),.
I express heart ,felt word of apl~rcciationto (111 ICRlSAT PMB crnd Sor,~/iuni
Breeding SrufJ specialli; Ah. Baskur Roi and Dr Kamesh,for their kind co-opertrlion and
hc,lp rendcred I pcrso~ial/ythank ,\lr.\. S.Devi,ji~rher help in the.sis drrrfi pogc, sc>fring.
and Mr. S.B. st an!^^ for hi~ikind undpron~l~t
co-operuiion.
I am grateful to the AGL scieizrjfic and rechnical sitrjf's~)eciull~~,
Dr Scnthil~cl
I:Ai E;ris/intr
Mr.r. Seethu b n n a n . Mes.~rs,iliarsi Redd~m,Br?an IL1o.s.s. Gqbor. Isii~rlcr~-.
and K.D. Prasad for their help in completing n7j /oh ivork ~/ficicnfli~.
cxl~ctlienr!,,rinrl
.smooth/~v.
I will fake this opportunir); to thank Dr. O.P. Rupela (co-orclinnlor). Dr. Balaji.
Program Leder, ,Mr. Prusod Rao and ali other stuff of Learrirng S),sicni Linit. Libra/.!'.
housing and food services, securip, farm and enyineerrng scn,iccs ond,iinance division
,for iheir excellent co-operation during my research work at ICRIS'4T Special Thanks to
Dr. C. N Reddy, Senior manager Jeld and medical unit, for hi.^ encouragement and
excellent medical facilities during my stay at ICRISAT.
 I express my co-ordial thanks to all my,fiiend.s hfr. Navghire, Toprope. Dr. Shete,
Dr. Godkc, Dr. Rizvi, Mr Girish Gunjorikar. Goyiji, Sreejeerh. Kassuhun, Arun, Mohan,
Devvarr, Pradeep, Ramu, Rajaram, Sripathi, Parikshit, Pranjan, Bhushan, G'ulia, Raghu.
Nepolian, Sathish, Gandhi, Veiu, Baskuran, Gomashe, Dcshmukh, I'alingc, Ms. Jyoliti.
Mrs. Rupasree a17d hfrs. Smiiho ,for giving ci nice company and making mv .staj,
romfurtablc both a/ ICRISAT and M. A. U , I'arhhani.
No word are enough lo express m.p regurd~ lo 111ylule helored jmrenrs u~ho
hlessed and pon'ered their love, affection crud s ~ ~ i rofi t cnlhusiusm for excelicnce in my
ucudcn2ic carrier. 1 an7 in deurih o f words to expr(,rc.cs177). love and t~flection/o n1,v .sisters.
brothers and relalives spcciaily .father-m-law'. n7othcr-i11-lou'atid their j?uniily tnembcr~
u~ho ,sparkled lifc wirh happiness, exciiemcnl, nsi)iration tmtl cncourcigcmenr ,for
nchieving excellence in my arcidemic carrier.
Dicfation is no1 enough to express my 1ove.fiom the core of m , heart
~ to my loving
ii~ifeSushmn, .son Aksha], and daughter Shrushti,fi,r !heir love i ~ n dunhi11dit7gafikctiun
(intipatience during thc course o f n ~ y1he.sis rc.cecrrc,h.
I grargfui1y ucknowiedge A.rian Dcve1of)menr 13i117k(inti I('R1,S.A l',fi)rj~roving,fiinci
and rc.serirch,felio~',s/7ip
lo carr). out my doctorcrl rcscorch.
 CONTENTS

Chapter
No.
Title I Page NO.

INTRODUCTION

REVIEW OF LITERATURE

MATERIALS AND METHODS

RESULTS

DISCUSSION 1 123

SUMMARY AND CONCLUSIONS I 169

LITERATURE CITED 1 177

APPENDICES I I-XXVIII
aTable

2.4

3.2

3.3
3.4
3.5

3.6
LIST OF TABLES

Title

Sources of resistance to sorghum shoot fly, Atherigona soccata

Inheritance of factors associated with resistance to sorghum shoot fly
Atherigona soccata
Characteristics of different sorghum genetic linkage maps published tc
date
Summary of qualitative and quantitative trait loci identified in sorghum
QTL mapping for insect resistance in cereals
Fluorescent-labeled sorghum SSR primers (Applied Biosystems) used in
sorghum diversity study.
PCR protocols used for amplification with labeled SSR primers ir
sorghum diversity study.
Salient features of parental lines of RIL mapping population
Aphid ratings in relations to percentage
Target genomic regions, linked SSR markers and associated shoot
resistance QTLs for marker-assisted selection
Fluorescent labeled sorghum SSR primers (Applied Biosystem) usrc
for foreground selection in marker-assisted breeding for shoot fl)
flj
In-
between
Pages
17-18
27-28

33-34

37
38-39
54-55

54-55

57
60
67

68-69

resistant.
3.7 Summary of SSR markers used to identify pure parental plant: 72
:mployed for FI hybrid production
3.8 Details of numbers of plant genotyped, markers used, number 01 73-74
selected heterozygous plants and number of backcrosses effected and
ldvance to the BClF, generation
3.9 Details of number of recurrent parental plants genotyped and number ol 73-74
Jure plants selected and used for pollination.
3.10 Details of genotyping of BClF, plants with linked marker (foreground 74-75
selection)
3.11 3etails of recurrent parental plants genotyped and number of parental 75
ype plants selected for (BC,Fl) backcrossing
3.12 3etails of discarded BCIFl plants 76

3.13 'arental data of allelic size (bp) polymorphic SSR markers determined 76-77
In the ABI 3700 used for background selection
 Table In-
Title between
No. Pages
-
3.14 Fluorescent labeled sorghum SSR primers (Applied Biosystems) used 76-77
for background selection in marker-assisted breeding for shoot fly
resistance
3.15 Details of background screening of BCIFIselected plants and advanced 77-78
to BC2Fl generation
3.16 Details of number of seed planted crosswise and set wise in BC2FI 78
foreground selection
3.17 Details of set-wise sowings of recurrent parents 79
3.18 Detail of foreground BC2FI populations numbers of plants genotyped, 80-8 1
and list of selected heterozygous plants
3.19 Detail of seed planted of each recurrent parent and plant used in 81
backcrossing to advance BCIFl generation
3.20 Detail list of selected BC2Flplants for background screening 82
4.1 Multiplex primer sets used for amplification of SSRs: allele sizes and 86-87
number obtained in diversity study
4.2 Means and standard errors for two parents of cross 2968 (susceptible) 92-93
X 1S 18551 (resistant) for different component of resistance to shoot
fly, and other traits in two screening environments at Patancheru, during
2002 - 2004
4.3 Mean, standard errors, and range of phenotypic values for RILs derived 94-95
from the cross 2968 (susceptible) x IS 18551 (resistant) for different
components of resistance to shoot fly and other traits in two screening
environments at Patancheru, during 2002 - 2004
4.4 Genotypic variance. G X E interaction and respective standard errors of 94-95
resistance to shoot fly and other traits in sorghum RIL mapping
population derived from cross 2968 (susceptible) X IS 1855 1 (resistant)
under individual and across screening environments
4.5 Means of parents, the RlLs population, their difference and proportion 96-97
,f RILs with the values outside the parental limits based on pooled
neans over two screening environments
4.6 3perational heritability estimates (broad-sense; entry mean basis) for 98-99
:omponents of resistance to shoot fly and other traits in sorghum RIL
~opulation derived from cross 2968 (susceptible) X IS 18551
:resistant) evaluated under two screening environments and across hvo
rcreening environments, at Patanchedru.
 Table In-
Title between
No.
-
4.7 Characteristics of QTLs associated with putative components of
Pages
100-101
resistance to shoot fly (in two screening environments, kharif and rabi)
based on Composite lnterval Mapping (PLABQTL, LOD>2.5) using
213 RILs derived from cross 296B (susceptible) x IS 1855 1 (resistant).
4.8 Characteristics of QTLs associated with putative components of 106-107
resistance to shoot fly (based on across-season averages) based on
Composite Interval Mapping (PLABQTL, LOD>2.5) using 213 RILs
derived from cross 296B (susceptible) x IS 18551 (resistant)
4.9a Target genomic regions, linked SSR markers and associated shoot fly Ill
resistance QTLs for marker-assisted selection
4.9b Parental polymorphism (allele size, bp) using eleven SSR markers that 112-1 13
were used for foreground selection in marker assisted breeding for shoot
fly resistant in this study.
4.10a Genotypic data of hybrid 28B (288) x RIL 189(3 12) for foreground 113-1 14
selection
4.10b Molecular data of BCIFl population [28B(288) x RIL 189(312)]x 113-1 14
28B(288), recurrent and donor parent for foreground selection
4.10~ List of introgressed plant with the targeted QTL and associated 114
characters
4.11a Genotypic data of hybrid KR 192 (304) x IS 18551 (267) for fore 114-1 15
ground selection
4.11b Molecular data of BCIFl population [KR192(304) x IS 18551(267)] x 114-115
KR192(304), recurrent and donor parent for fore ground selection
4.1 1c List of selected BClFl introgression heterozygotes with their targeted 115
QTLs and associated characters
4.12a Genotypic data of hybrid 2 0 8 (186) x RIL 252 (318) for foreground 115-116
selection
4.12b Molecular data of BCIFl population [20B(186) x RIL 252(318)] x 115-1 16
20B(186), recurrent and donor parent for fore ground selection
1.12~ List of introgressed plant with the targeted QTL and characters 116
associated
1.13a Genotypic data of hybrid 2 0 8 (179) x RIL 153(248) for foreground 116-117
selection
1.13b Molecular data of BCIFl population [20B(179) x RIL 153(248)] x 116-117
208(179), recurrent and donor parent for foreground selection
 Table In-
Title between
No.
-
4.13~ List of selected QTL introgression heterozygote plants with the targeted
Pages
117
QTLs and associated characters
4.14a Genotypic data of hybrid KR 192 (300) x RIL 252 (319) for foreground 117-118
selection
4.14b Molecular data of BClFl population [KRI 92 (300) x RIL 152(3 19)] x 117-118
KR192(300), recurrent and donor parent for fore ground selection
4.14~ List of introgressed plant with the targeted QTL and characters 118
associated
4.15 List of SSR marker loci that are monomorphic between pairs of parents 120
4.16 Genotyping data for background selection of foreground-selected BClFl 121-122
plants of [28B(288) x RIL 189(3 12)] x 28B(288)
4.17 Genotyping data for background selection of foreground-selected BCIFl 121-122
plants of KR192 (304) x IS 18551(267)] x KR192 (304)
4.18 Genotyping data for background selection of foreground-selected BC,Fl 121-122
plants of [208(186) x RIL 252(3 IS)] x 20B(186)
4.19 Genotyping data for background selection of foreground-selected BClFl 121-122
plants of [208(179) x RIL 153 (248)] x 20B(179)
4.20 Genotyping data for background selection of foreground-selected BCIF, 121-122
plants of {KRI 92 (300) x RIL 252 (3 19) x KR192 (300))
4.21 Details of background screening of selected BCIFl plants with targeted 121-122
QTL
4.22 Molecular data of BC2Fl population [28B(288) x RIL 189(3 12) x 121-122
28B(288) ] x 288, recurrent and Donor parent for Fore ground selection
1.23 Molecular data of B C ~ F I [ K R192(304) x IS 18551 (267) x KR 121-122
192(304)] x KR 192(304), recurrent and Donor parent for fore ground
selection
1.24 Molecular data of BC2Fl [20B(186) x RIL 252(318) x 2 0 8 (186)] x 121-122
20B(186), recurrent parent and donor parent for foreground selection
1.25 Molecular data of BC2Fl [20B(179) x RIL 153(248) x 20B (179)] x 121-122
20B(179), recurrent parent and donor parent for foreground selection
1.26 Molecular data of BC2F, [(KR 192(300) x RIL 252(319) x KR 121-122
192(300)] x KR 192(300), recurrent parent and donor parent for
foreground selection
1.27 Details foreground selection of five BC2F1 populations with targeted 121-122
shoot fly resistance QTL character associations
Table In-
Title between
No.
-
1.28 Details of BC2Fl foreground selected plants chosen for background
Pages
122
screening.
5.1 Recombinant inbred lines (RILs) with shoot fly resistance, measured in 153-154
terms of deadhearts (%) incidence, comparable to resistant parent 18551
and good agronomic adaptation
5.2 List of selected QTL-introgressed heterozygote plants and associated 162
characters of five backcross (BCIFI)populations.
5.3 Details of background screening of BCiFl selected individuals and of 163-164
B C ~ Findividuals
I to be genotyped, with targeted QTLs
5.4 Details of foreground selection of five BC2FI populations with targeted 166
shoot fly resistance QTL and character association
5.5 Details of foreground-selected BC2F, plants chosen for background 167
getlotyping from five BC2F, populations with targeted QTL and
character associations

-
 LIST OF FIGURES

Figurc In-
Title between
No.
- Schematic diagram of RIL development procedure
Pages
57-58
3.1
3.2 Genetic linkage map of sorghum based on 213 recombinant inbred lines 63-64
derived from cross 296B x IS 18551.
3.3 QTL positions of shoot fly resistance component traits for 252 68-69
recombinant inbred population derived from cross BTx 623 x IS 18551
across two screening environments
4.1 Dendrogram generated from data for 91 sorghum accessions using SSR 87-88
genotype data revealed by silver-stained PAGE of PCR products from
20 polymorphic loci distributed across 9 of I0 sorghum linkage groups.
4.2 Dendrogram generated from data for 91 sorghum accessions using SSR 89-90
genotype data revealed by capillary electrophoresis of PCR products
from 20 polymorphic loci distributed across 9 of 10 sorghum linkage
groups.
4.3 Frequency distribution of 213 RlLs derived from cross 296B X IS
95-96
18551 for components of shoot fly resistance and few agronomic traits
in two screening environments, viz., late khar(f(E1) and rabi (E2) at
Patancheru.
1.4 QTL positions of shoot fly resistance component traits for 213 100-101
recombinant inbred population derived from cross 2968 x IS 1855 1
under two screening environments, late khar$(indicated by purple
color) and rabi (indicated by pink color) at Patancheru during 2002 -
2004.
1.5 QTL position of shoot fly resistance component traits for 213 106-107
recombinant inbred population derived from cross 296B x' IS 1855 1
across two screening environments at Patancheru, during 2002-2004.
 LIST OF PLATES

In-
Plate
Title between
No.
Pages
1.1 Various life stages of the sorghum shoot fly, Atherigona soccata. 1-2
( 3.1 1 Interlard fish-meal technique used to field screen the test material for 1 58-59
resistance to sorghum shoot fly
Glossy leaf and trichomes traits associated with resistant to sorghum
shoot fly
Plumule and leaf sheath purple pigmentation scores of sorghum
genotypes at 5 Days to seedling emergence.
Recurrent and donor parents used in marker-assisted selection for shoot
/ / fly resistance and its component traits. I
Silver-stained gels showing some of the more polymorphic SSR
markers used in diversity study.
Genotyp~ngof five hybrid (FI) populations to confirm hybridity using
SSR locus Xtxp75.
Genotyping of five BC,F, populations for foreground screening using
SSR locus Xtxp75, Xtxp37 and Xtxp312, in three different sowing sets.
Graphical presentation (chromatograph) of BCIFI population screening
with SSR marker Xtxp65 analyzed using the ABI prism 3700 DNA
sequencer.
Background genotyping of selected foreground plants from five BC,F,
populations using SSR locus Xtxp265 and Xtxp 274.
Background genotyping of selected foreground plants from four BC,F,
populations using SSR locus Xtxp298 andXlxp 57
Background genotyping of selected foreground plants from three BC,F,
populations using SSR locus Xtxp289
Background genotyping of selected foreground plants from four BC,F,
populations using SSR locus Xtxp145
Background genotyping of selected foreground plants from three BC,F,
populations using SSR locus Xtxp3 I and Xtxp 21.
4.1 la-b Background genotyping of selected foreground plants from three BC,F,
populations using SSR locusXcupU2 and Xtxp258.
 In-
Plate
Title between
No.
Pages
- Background genotyping of selected foreground plants from three BC,F, 121-122
4.12
populations using SSR locus Xtxp230.
4.13 Graphical presentation (chromatograph) of selected BCIFl individuals 121-122
(background screening) with SSR marker Xtcup28 analyzed using the
ABI prism 3700 DNA sequencer.
1.14-16 Genotyping of five BC,F, populations for foreground screening using 121-1 22
SSR locus Xlxpi.5, Xlxp37andXlxp312, in three different plates.
1.17 Graphical presentation (chromatograph) of BC2FI population screening 121-122
with SSR marker Xtxpl5 analyzed using the ABI prism 3700 DNA
sequencer.
5.1 Agronomically desirable RILs with shoot fly resistance (measured in 153-154
terms of deadhearts incidence) comparable to resistant parent IS 18551.
 LIST OF APPENDICES
Appendix In-between
Title
No. Pages
I Particular characteristics of sorghum accessions used in sorghum I-IV
diversity study.
il Preparation of stock salutations V - VIII
111 Details on pedigree of 259 F7-F8RlLs derived from cross 2968 x IX -XI
IS 18551.
1 1V 1 Mean performance of RlLs ( I -259) and parents (2968 and IS 1 XI1 - XXI
1855 1) evaluated under the late kharifscreening environment
during 2002.
V Mean performance of RILs (1-259) and parents (296B and IS XXII - XXVllI

1 18551) evaluated under the late Rabi screening environment during

2002
INTRODUCTION
 CHAPTER I
INTRODUCTION

Sorghum (Sorghum bicolor (L.) Moench) is the fifth most important cereal crop
globally after rice, maize, wheat and barley (FAO, 2004). It is grown in about 86
countries covering an area of about 47 million hectares (ha) with a grain production of
69 million tons (t) and average productivity of 1.45 t ha" (ICRISAT, 1996 and FAO,
2004). It is grown mostly in tropical and subtropical areas. Sorghum occupies 14.1
million'ha in Asia. The major sorghum producing countries are Nigeria, Mali, Sudan,
India, China, Pakistan, USA, Australia, Argentina, and Mexico. The grain is used as
human food in various ways and both grain and stalk are used for animal feed. India is
major producer of sorghum with the crop occupying an area of 9.9 million ha and
yielding an annual production 8.00 million t during 2003104 (FAS, 2005).
Productivity of sorghum is highly variable from county to country. Several
constraints affect grain productivity. Among these, drought, pests (particularly
sorghum shoot fly, spotted stem borer, midge, aphids and head bugs) and diseases
(particularly anthracnose and grain molds) are the predominant ones. Grain yields in
farmers' fields in Asia and Africa are generally low (500 to 800 kg ha"). One of the
major factors causing these low sorghum grain yields is insect pest damage. Each year
nearly 32% of the actual produce is lost due to insect pests in India (Borad and Mittal,
1983). 20% in Africa and Latin America. and 9% in the USA (Wiseman and
Morrison, 1981). The annual loss of sorghum production due to shoot fly in India is
estimated at nearly US$ 200 million (ICRLSA'T, 1992).
Sorghum shoot fly (Atherigona socculu Rond.) is a key pest of sorghum in
many countries including India. Shoot fly female lays a cigar-shaped eggs on the
lower leaf surface of young sorghum plants in the 1-7 leaf stage, i.e. 5-25 days after
seedling emergence. Egg hatch in 1-2 days and first instar larvae move along the
shoot to the growing point of the seedling. The larva cuts the growing point resulting
in wilting and drying of the central leaf, causing the typical 'deadheart' symptom
(Sharma et al., 2003) that appear 1-4 weeks after seedling emergence. In order to
compensate for the loss of central shoot, damaged plant produces side tillers that may
subsequently be attacked by shoot fly. Larval development is completed in 8-10 days
after which the insect pupates in the soil (platel.1). Pupation lasts approximately eight
days and the entire life cycle is completed in 17-21 days depending on prevailing
weather conditions (Sharma et al., 2003). Shoot fly incidence is high in late sown
kharif(rainy season), early sown rabi (postrainy season) sorghum crops. The level of
infestation even may go up to 90-100 % (Usman, 1972) and the losses due to this pest
have been estimated to reach as high as 86% of grain and 45% of fodder yield
(Sukhani and Jotwani, 1980). Adoption of chemical methods for insect control in
staple food crops is not economically feasible for resource poor farmers of the semi-
arid tropics (SAT) as the low crop value per acre precludes the use of insecticides for
control of insects (Dhams, 1943). Therefore host plant resistance combined with
timely sowing is the most realistic approach to minimize grain and stover yield losses
due to insect pests such as sorghum shoot fly. Genetic variability for shoot fly
resistance in plant exists in sorghum germplasm. Many of the gemiplasm sources for
resistance to this pest have poor agronomic features and grain yield potential, and
sources with high levels of resistance are not available in the cultivated species.
Gcrrnplasm accessions with absolute resistance have been found in wild relatives of
sorghum (Sorgh~~rn
purpuroseric'arint. Snilidutn. S vevsicolur and S austruliense)
(Mote. 1984; ICRISAT. 1991): however, their utilization in sorghum breeding
programs is hindered by crossing barriers.
Screening procedures have been standardized and low to moderate levels of
resistance have been identified in several gem~plasmsource materials (Sharrna el al..
1992). Several mechanisms of resistance have been identified in these resistant lines
such as 'non-preference far oviposition' (components of which include trichomes.
glossiness. and restric~ed leaf surface wetness), 'antibiosis', and 'tolerance' or
'recovery' (Shamla and Nwanze, 1997). Some of these resistance sources have been
used in conventional breeding, but the levels of resistance available for selection
aluong the segregating progenies were not high. The selection of sorghum genotypes
for resistance to shoot tly by utilizing one or a few resistance parameters is inefticient
because several components are i~ivolvedin resistance and one or more genes govern
each of these resistance con~ponents. Further, expression of many of these
components is influenced by environmental variation; hence shoot fly resistance is a
quantitative trait and shows a large amount of genotype x environment interaction.
Marker-assisted selection has considerable potential to improve the efficiency of
selection for quantitative traits (Hash and Brarnel-Cox, 2000), such as shoot fly
resistance, for which expression is sensitive to the testing environment. As resistance
to shoot fly are mostly quantitative in nature, it is important to identify quantitative
trait loci (QTLs) from the viewpoint of genetics and breeding. The ultimate goal of
such QTL analysis is to develop tools that are useful for marker-assisted selection in a
practical breeding program aiming at increasing the level of resistance in
agronomically elite backgrounds through gene pyramiding for shoot fly resistance.
Traditional quantitative genetic studies on shoot fly resistance with different
sorghum genetic materials have been reported by many workers. Also recently QTL
analysis for shoot fly resistance component trait has been carried out using a set of
sorghurn recombinant inbred lines (RIL population) derived from cross B'fx623 x IS
1855 1 (Saijanar. 2002: Folkerstama et a/. 2005, unpublished). These studies revealed
the complex nature of shoot fly resistance and quantitative inheritance of resistance
for some of the coliiponent traits with possible genotype (G) x environment (E)
interaction. Quantitative genetic a~ialysisof shoot fly resistance requires replicated.
multi-environment testing under a wide spectrum of shoot fly pressure because of the
utipredictahility of field screening environments. This can be accomplished by
utilization of a RII, population. This allows measuring of the environmental (E)
contribution and G x E contribution to total phenotypic variance allowing less biased
estimates of genotypic ( G ) variance. In sorghum linkage maps have been developed
using a number of Restricted Fragment Length Polymorphism (RFLP) (Subudhi and
Nguyen. 2000). Amplified Fragment Length Polymorpliisrn (AFLP) (Bovin et ul..
1999) and Simple Sequence Repeat (SSK) markers (Bhattranakki er rtl., 2000).
Amolig the different types of ~noleculartnarker systems available. SSR markers best
satisfy the criteria of sufticie~itpolymorphism. repeatability and cost effectiveness
required for successful utilization in applied marker-based selection. In sorghum a
reasonably large liu~nberof SSR markers have been developed (Brown 41 01.. 1996:
Taraniino et al., 1997; Kong er (11.. 2000; Bhattra~nakkirr a/., 2000; Schloss rf 01..
2002). often using the elite breeding line BTx623 as a source. and these are suitable
for screening the existing sorghum RIL population to construct a genetic linkage map
and to identify QTLs for shoot fly resistance and its coniponent traits.
The analysis of genetic diversity and relatedness among individuals within a
species or among different species or populations is a central task for many
disciplines of biological science. Genetic diversity and phylogenetic studies were
initially conducted using quantitative and qualitative traits, which are mostly
morphological. using various statistical methods i.e. analysis of variance. covariance,
D* and Metroglyph analysis. These analyses are nlostly based on quantitative
traits that are highly influenced by environmental effects and require tedious
statistical procedures. Molecular markers are being widely used in various areas of
plant breeding as important tools for evaluating genetic diversity and determining
cullivar identity (molecular fingerprinting). Establishment of a molecular marker and
phenotypic assessment database of crop germplasm will help breeders to trace down
the origins and degrees of relatedness of many landraces and cultivars. Considering
the potential of molecular markers crop breeders can extend their hands to use these to
suppleme~ltother tools currently being used in their crop breeding program. In this
present study. we used SSR markers to estimate the level of allelic differences among
91 sorghum accessions collected from different parts of the world and previously
identified as resistant to one or more major insect pests of this crop. with the aitn of
assessilig their ge~ieticdiversity.
DNA markers that are tightly linked to agrononiically important genes can be
~ ~ s easd a molecular tool for marker-assisted selection (MAS) in plant breeding
(Ribaut a11d Hoisington. 1998). MAS involves using the presencelabsence of a marker
as s~~bstitute
for or to assist in phenotypic selection. in a way that makes it more
efficient, effective, reliable and cost effective compared to the phenotypic based
selection in co~ivcntionalplant breeding methodology. llost plant resistance can play
pivotal role in integrated pest management. Sources of resistance to insect pests have
long since been identified: however. these have not been uscd effectively in crop
iniprovcmcnt programs because the levels of resistance available are either too low or
i t is only rarely possible to develop optimum levels of insect infestation to screen the
test material. Use of biotechnological approaches can play a significant role in
developing cultivars with resistance to insects. There is an urgent need for innovation
in the iniprovenieot of phenotyping systems for assessing resistance to insect pests.
Once accurate and precise phenotyping systems for insect resistance have been
established, the n~olecular markers can be used in dissecting the genetic basis of
resistance, identifying the location of underlying genes and the nature of their gene
action. Such knowledge will significantly accelerate the introgression of insect
resistance genes into high yielding cultivars. The final outcome of marker-assisted
crop breeding will be the rapid production of iniproved varieties and at lower cost.
 With this background, a research program 'Genetic diversity analysis, QTL
mapping and marker-assisted selection for shoot fly resistance in sorghum [Sorghum
birolor (L.) Moench]'. was therefore attempted with following objectives:
1 . To assess genetic diversity by SSR markers in a set of insect resistant
lines.
2. Phenotyping a set of RlLs (296B x IS 18551) for components of resistance
to shoot fly over seasons.
3. Identification of QTLs for shoot fly resistant component traits using the
marker genotyping and phenotyping data of the RIL population derived
from 2968 x IS 1855 1 cross.
4. lntrogression of shoot fly resistant coniponent traits in agronomically
superior genotypes using moleculnr marker-assisted selection.
REVIEW OF LITERATURE
 CHAPTER I1
REVIEW OF LITERATURE

2.1 Application of SSR markers in diversity analysis of sorghum insect

resistant germplasm accessions
The present review covers the assessment of genetic diversity at a molecular level and
its application in crop improvement in general and sorghum in particular.
Sorghum [Sorghum bicolor (L.)Moench] is the fifth most important cereal
crop of world providing food and fodder throughout the world (Dogget, 1988). It is a
crop with extreme genetic diversity. Its adaptation to harsh environments, specifically
its high levels of resistance to biotic stresses and tolerance to abiotic stresses, accounts
for its success throughout the semi-arid regions of the world. It has numerous
mechanisms that allow it to survive and still be productive in these conditions. Harlan
and Dewet (1972) subdivided the cultivated sorghum into five morphologically
distinct races: bicolor, guniea, caudatum, kafir, and durra. Intermediate races are
designated, for example, as kafir-csudatum, durra-bicolor, etc. They speculated that
the race durra and bicolor arose from the wild subspecies aethiopicum, that the kafirs
arose from ve. Ticill~'Joi-um,and that the guineas evolved from anmdinaceum.
Subraces or working groups (Murty and Arunachalan, 1967) describe some of the
variation within races and intermediate races and often refer to commonly used
groups by sorghum scientists as feteriatas, zera-zera, kaura, kaoliang, milo, sorgo,
sudangrass, etc. A refinement of the working groups as they fit with and complement
the Harlan and de Wet race clssification has aeen proposed by Dahlberg (2000).
The germplasm pool of the genus Sorghum is characterized by abundant
diversity. The immense morphological diversity of the cultivated races of sorghum
had resulted from variable climate and geographical exposure in which its wild
ancestors evolved, coupled with selection pressure imposed by the environment and
the man during and after domestication. Many sources of exotic and unique
germplasm have been discovered and utilized over the years for sorghum
improvement. Traits such as grain yield, resistance to shoot fly, stem borer, midge and
greenbug had been found and incorporated into current germplasr~land it has resulted
in tremendous improvement in crop adaptation, resistance to biotic stresses, tolerance
to abiotic stresses, and food and fodder productivity.
 Understanding and management of the natural variation within the
domesticated cultivars and their wild relatives of a plant species is very important in
the establishment of an efficient breeding program aimed at crop improvement.
Exploiting natural variation is very important for several reasons. Genetic uniformity
in crops is undesirable as it makes the crop vulnerable to epidemics and
environmental disasters resulting in yield loss. Many wild relatives of crop plants
contain genes conferring resistance to biotic stresses such as pests and diseases, and
tolerance to abiotic stresses such as drought. cold, and salinity. When these traits are
incorporated into economically important varieties, large losses in yields can be
acoided. A plant breeder also aims at iniproving certain desired characters such as
grain quality and yield for specific end use adaption. A pre-requisite for i~iiprovingthe
overall plant characteristics is an understanding of the germplas~navailable for use in
breeding, which in turn will allow a systematic sa~ilplingof the gennplasn~for
breeding and conservation purposes. DNA tilarkers have been used to quantify genetic
diversity and determine phylogenetic relationships in several plant species (Clegg,
1991; Lee. 1998). Cluster analysis is useful for studying the relationships among
closely related accessions while ordination (principal component analysis) provides a
more complete rcpresentatioti of the relationship among major groups. Such an
analysis is very useful for producing 'core' collections at the international centers
(Virk er ul.. 1995). which can represent most of the diversity in the germplasm
collcctio~iand allow one to extrapolate conclusions to the entire collection.
Following donlestication, genetic variation in crop plants has continuously
narrowed due to continuous selectioli pressure for specific target traits, i.e.. yield and
its attributes. This narrowing of genetic variation has render crops more vulnerable to
disease and insect epidemics and jeopardized the potential for sustained genetic
improvement over the long term (I-iarlan. 1989). This risk was brought sharply into
focus in 1970 with the outbreak of southern corn leaf blight. which drastically reduced
corn yields in USA, and was attributed to extensive use of a single system of
cytoplasmic-genetic male sterility (Texas type) for hybrid seed production, which was
unfortunately linked to disease susceptibility (Ulstrop, 1978). Thus, it is extremely
inlportant to study the genetic composition of the germplasm of existing modern day
cultivars in comparison with their ancestors and related species. This will not only
provide information on their phylogenetic relationship but also indicate where there
are chances of finding new and useful genes, as the accessions with most distinct
DNA profiles are likely to contain a greater number of novel alleles. DNA profiling to
make such sampling decisions is now underway in most crops. Many DNA markers,
both specific as well as arbitrary, have been used so far for DNA fingerprinting of
various classes of germplasm (Callow, 1997; Virk. et a/., 1997). AFLP markers are a
new class of molecular marker that has gained popularity for the study of genetic
polymorphism, especially in species where polymorphism is extremely rare using
other types of marker systems. Pakniyat et a/. (1997) used AFLP for studying
variation in wild barley with reference to salt tolerance and associated eco-geography.
More recently the discovery and application of several more readily
reproducible of polymorphism assays based on variation in the number of short
tandemly repeated DNA sequences (i.e.,SSRs) has increased ..le utility of PCR-based
molecular marker genotyping for genetic diversity and marker-assisted breeding at
least in crops where the necessary investment to develop appropriate primers can be
made. DNA simple sequence repeats are numerous and are highly polymorphic in
plants (Morgante and Olivieri, 1993; Wang et al., 1994; Rongwen et a!., 1995; Yang
et al., 1994). SSRs are a highly useful class of such PCR-based genetic markers.
Although costly to develop relative to some other classes of genetic markers, once
developed their analysis is both easy and inexpensive. They are co-dominant, occur in
high frequency, and can display a high level of polymorphism even among closely
related accessions. Their high information content and other favorable characteristics
make them excellent genetic markers for many types of investigation including
marker-assisted selection and fingerprinting of germplasm collections (Brown er a/.,
1996). SSR markers are detected utilizing the polymerase chain reaction with pairs of
unique DNA primer sequences flanking the repeated region. They have not only
revolutionized mammalian genome analysis (Hearne er a/., 1992). but have also
facilitated plant breeding and genetics. Recently, SSR marker technology has been
developed and used for genome mapping and DNA fingerprinting in crop plant
species such as rice (Wu and Tanksley, 1993), wheat (Roder et a/., 1998). barley
(Saghai Maroof et al., 1994). maize (Senior and Heun, 1993; Taramino and Tingy,
1996), sorghum (Brown et al., 1996; Taramino et a/., 1997; Dean et a/., 1999;
Bhattramakki et al., 2000; Dje et al., 1999, 2000; Kong er a/., 2000; Smith el a/.,
2000; Ghebru et al., 2002; Haussmann et al., 2002; Schloss e t a / . ,20021,
Genetically mapped markers tagging specific genes of interest to plant
breeders have been identified. Examples include resistance genes for blast and gall
midge @air el al., 1995a; 1996) using RFLP- and PCR-based approaches in rice, and
leaf rust resistance gene LR 28 in wheat (Naik er at., 1998); QTLs for protein content
in wheat (Prasad el at., 1999) heterosis in rice (Nair et at., 1995b). downy mildew
resistance (Jones et a!., 2002) and drought tolerance (Yadav er at., 2002, 2004) in
pearl millet.
Germplasm analysis to study genetic diversity is other important area in which
a lot of efforts have been put for fingerprinting of crops like rice (Ramakrishna et al.,
1994; Gupta et al., 1994), wheat (Sen et al., 1997; Pujar el a/., 1999), pearl millet
(Chowdari el at., 1998) etc. are being carried out extensively. This information has
potential in strategic planning of future crop breeding efforts to improve agricultural
sustainability in the SAT. Information on the genetic diversity available within a crop
species is important for selection of parental strains and in the prediction of hybrid
performance especially in crops such as rice, sorghum and m: ,ze in which hybrids are
commercially important. The various steps involved In hybrid breeding programs,
such as making several crosses and screening the combination for superior
performance and heterosis are very costly, laborious, and time consuming. Hence, if
heterosis can be predicted before making the crosses, then the number of crosses to be
performed and the progeny to be screened in field trials can be reduced considerably.
Various investigators are trying to correlate genetic diversity, as quantified by DNA
markers, to predict hybrid performance, in various hybrid-breeding programs because
the level of genetic diversity between the parents has been proposed as a possible
predictor of heterosis. Studies with maize (Godshalk el at., 1990; Melchinger el at.,
1992) revealed that molecular marker analysis is useful for assigning maize inbreds to
heterotic groups, but the RFLP based genetic distance cannot be used to predict
hybrid performance, while in oats, Moser and Lee (1994) have shown that molecular
marker based genetic distance could be a predictor of hybrid performance only for
those crosses where the parents belong to the same heterotic group and can not be
extended to crosses between different heterotic groups.
Smith et al. (1990) observed a significant relationship between parental
genetic distance and F, performance with a simultaneous increase in sample size as
well as the number of markers used for analysis. Stuber el al. (1992) reported a
significant relationship between parental heterozygosity and hybrid yield when the
number of parental inbred lines was increased. While Lanza et a/. (1997) observed
consistent correlation between grain yield and random amplified polymorphic DNA
(RAPD) marker-based genetic distance in maize, Martin el al. (1995) and Barbosa-
Neto et 01. (1996) were not able to establish any relationship between marker-based
genetic distance and hybrid performance in wheat. In rice, Zhang el al. (1994, 1995)
used eight lines representing a major portion of the elite rice germplasm used in the
hybrid rice breeding programs in China, to determine the relationship between marker
locus heterozygosity, performance, and heterosis. Their studies revealed that
correlations between mid-parent heterosis and specific heterozygosity (based on
positive markers) were large and may be useful for prediction of heterosis. If such
correlations are contirmed using a larger sample size, then it can certainly aid in
planning the most productive crosses in the hybrid-breeding program.
The results of Xiao er al. (1996a) involving crosses between four japonica and
six indica elite inbred rice lines had indicated that genetic distance measures based on
RAPDs and simple sequence repeats (SSRs) could be useful for predicting yield
potential and heterosis of intra-subspecific hybrids, but not of inter-subspecific
hybrids. Heterotic groups are not clearly defined in sorghum as in maize, studies
using molecular markers cluster Am-pairs and R-lines separately (Ahnert et al.,
1996); however the majority of RFLP patterns were coltimon to both groups
suggesting that Am-pairs and R-line groups have not diverged to an extreme degree
(Verling et al. 1994 and Ahnert er al. 1996). A close relationship between
morphological markers and molecular markers with respect to cluster formation has
been reported by many workers [e.g., Virk el al. (1995) in rice and Bhattacharjee er
al. (2002) in pearl millet]. Smith and Smith (1991) identified 47 maize hybrids using
80 RFLP probes. Wall er al. (1984) used zein protein markers to differentiate maize
inbreds. Dallas (1988) identified rice cultivars by using RFLP markers identified
using two human mini-satellites as probes.
Ghareyazie er al. (1995) assessed genetic diversity among 35 Iranian rice
varieties by comparing these with two typical indica and three typical japonica
varieties using PCR-based RFLP markers Virk el a/. (1995) used RAPD markers to
identify duplicate accessions in a rice germplasm bank. Taxonomists had traditionally
used morphological markers to classify genetic resources in sorghum. The
morphological traits used in the taxonomic classification of sorghum to different races
are conditioned by a relatively small number of genes. However, more economically
important traits, which are related to adaptation exhibiting enormous variability across
sorghum gemplasm, are often complex and quantitatively inherited. Hence
classifying germplasm accessions based on solely on a few discrete morphological
characters would not necessarily provide an accurate indication of genetic divergence
among the cultivated genotypes of sorghum (Menkir er al., 1997). In sorghum, Tao er
01. (1993) demonstrated the use of RFLP and RAPD markers to differentiate sorghum
accessions and obtained different clusters according to their sub-specific groupings
(i.e., Durra, Zera-zera, Caud-Nig and Caffrorum). The result also indicated that
individuals of similar taxonomic grouping but different geographic origin may be
genetically less identical than previously considered and similar frequencies of
polymorphism were obtained with RAPD with RFLP markers. Results of these
experiment indicated that a high level of genetic uniformity exists within Sorghum
bicolor. Deu el al. (1994) used RFLP markers and related allelic variation in these to
racial differentiation among 94 sorghum germplasm accessions and breeding lines
primarily of African origin. Oliveira er a[. (1996) used RFLP, RAPD and Inter
Simple Sequence Repeat (ISSR) markers in genetic diversity studies of 84 sorghum
lines, and found that both racial characterization and geographic origin correlated with
relatedness. Several workers have selected diverse parents fc crossing based on the
genetic diversity revealed by RFLP and/or RAPD markers, and obtained close
relationships between the levels of marker diversity observed and heterosis expressed
by the FI hybrids (Smith el al., 1990 and Xiao el al., 1996a. in sorghum and rice,
respectively). The advent of PCR-based molecular marker techniques such as RAPD
(Williams el al., 1990) has further facilitated analysis of the sorghum genome.
Pammi et al. (1994) identified conditions that allowed reproducible
amplification of RAPD markers and tested them on 32 different genotypes of
sorghum. Cui er al. (1995) compared the restriction fragment length patterns of 53
sorghum accessions from Africa, Asia, and USA and detected different levels of
polymorphism according to source continent. Deu er al. (1995) assayed mtDNA
variation using RFLP and showed a significant genetic differentiation among the
cultivated sorghum crop. The bicolor and guinea races exhibited the highest variation
while the kafir race had the least. The homogeneity of koJir may be due to its
relatively recent domestication (Harlan and Dewet, 1972). Dje el al. (2000) evaluated
the use of microsatellite markers to quantify genetic diversity within as well as among
the accessions solnplcd from the world gcr~nplas~llcollection of sorghum.
Considerable variation was found at 5 microsatellite loci analyzed, with an average
number of alleles per locus equal to 2.4 within accessions and 19.2 in the over all
sample of 25 accessions. Results shows that microsatellite data are useful in
identifying individual accessiotis with higher relative contribution to the overall
diversity of the collection.
Grenier et al. (2000) evaluated the genetic diversity three subsets of around
200 accessions each from the world sorghutil germplasm collection using 15
polymorphic microsatellite loci. The average allele richness of each subset was
equivalent to 16.1. 16.3 and 15.4 allelcs per loci~sfor the subsets PCS (selective
sampling based on quantitative characters). L (random sampling after stratification of
the entire landrace collection). 'T (selection based on tlie geographical origin of the
landlxces and the traits under farnlers selection). respectively. Average genetic
iliversity was estimated at 0.81 for the PCS subset. 0.77 for the L subset, and 0.80 for
the T subset.
Smith rt 01. (2000) evaluated the potential ability of SSK techriology for
research product de\elopment. seed pr(1duclion quality assurance. and genetic
resource conser\,ation management lor sorghum. Fifty getietically diverse elite
sorghum inbreds aitli k n o n n pedigrees mere used to compare the discrimination
abilities of 15 SSR nlarkers \\it11 104 RFLPs. KF1.P data a l l o ~ ~ eall
d lines to be
~tniquelyidentilied evcept t\bo litlcs that could 1101 he distinguished by the molecular
tlatfl. (As the set of lines used in the present study encompass a relatively broad array
of ger~nplasnidiversit!, representing diSferent geographical areas. maturity ranges.
gcrmplnsni groups and inter-group crosses: even the very snlall set of these 15 SSR
loci \\ere able to uniquel> identify these lines). f h e mean Polymorphism Information
Content (PIC) values were 0.62 and 0.58 for RFLl's and SSRs. respectively.
C'orrelations for pair-wise molrcc~larprotile rlistance witii pedigree distance among
the ~naintainer parents (D-lines) were 0.52 and 0.53 for RF1.P and SSR data,
respectivel!: and for male parents (R-lines) were 0.41 and 0.47 for KF1.P and SSR
clata. respecti\,ely. 'This set of 15 SSK markers could be used to help the genetic
resource conservation management in sorgbun~.
Ghebru et 01. (2002) used sorghi~m SSR n~arkersto characterize genetic
diversity in 28 Eritrean sorgllum landraces atid compared this diversity to a
representative sa~lipleof the wo1.1d sorghu~ncollection. Pools of SSR markers were
sized and score on autoniated DNA sizing gels. A high level of diversity was
observed anlong the Eritrean landraces compared to other sorghum gennplasm. in
both the nutrtber and size range of SSR markers, Individual laildraces were found to
carry a high level of within-population diversity and lleterozygosity, and between-
populations diversity was equally high. Most of the Eritrean sorghum landraces
evaluated clustered it1 a separate sub-group from the other sorghum germplasm
included in this study. These results indicate that a great deal of gertnplasm diversity
and genetic novelty are available in Eritreatl sorghum and that SSK markers can
contribute to the wise use of this diversity for sorghum improvement.
Jordan er 17l. (2003) investigated the value of molecular marker-based distance
infortnation to identify high yielding grain sorghum hybrid in Australia. Data from 48
trials were used to produce hybrid perfomlance estimates for four traits (grain yield.
height. maturity and stay green) for 162 hybrid combinations dcrived from 70 inbred
parent lines. Each line was screened \%it1111.3 mapped RF1.P markers. The Roger's
distances helaeen the parents of each hybriil were calculated fro111 the marker
inCurmation on a genome basis and individually Ihr each of the ten linkage groups of
sorghum. Some of the inhred parents were related so the hyhrids \+ere classified into
75 groups. with each group containing individual hybrids that showed a silnilar
p:tt~crn of Roger's distance across lilikage groups. Correlation between the hybrid
groi~pperfimnnnce and hybrid group Roger's distances were calculated. A signilicant
correlation \+as observed between whole genome based Roger's distance and grain
yielil (r=O.J2). This association is Iuo weak to be of value tbr identitying superior
hybrid combinations. One reason Tor the generally poor association between parental
genetic dilersity and yield Inay be that important QTLs influencing lieterosis are
located in particular chro~nosomeregions and not distributed evenly over the genome.
I>iversity on individual linkage groups \\as explored to predict hybrid perCormance
and [his detected trio linkage groups cxplailiing 38% of the total variation in hybrid
perl'or~iioncefor grain yield, ahile another nod el combining phenotypic trait data and
parental diversity on a particular linkage group explained 71% of [he total variation in
grain yield and had potential for use in the selection of lieterotic hqhrids.
Monica el 01. (2004) assessed the genetic diversity in elite sterility maintainer
(B-lines) and fertility restoring (R-lines) sorghum inbreds as compared with a group
of exotic and converted ger~nplasnifrom world collections. A set of 100 SSR markers
and 1357 AFLP marker with known map positions were utilized to determine genetic
si~liilarityin the groups of B-lines, R-lines. and US public inbreds. Cluster analysis of
genetic si~llilarityestimates revealed that tlle classification of sorghum inbreds is
based on the sorghum working groups, Zera-zera, Kafir, Kafir-Milo, Durra and
Feterita. Cluster analysis failed to give a clear differentiation between B- and R-lines.
suggesting that R- and B-lines do not represent well-defined heterotic groups in this
set of public lines. By comparing the different classes of molecular markers (SSRs.
AFLPs. co~ubinationof SSRs and AFLPs), it was detertilined that the distribution of
markers and the coverage of tlie genonies by the markers did not affect tlie
classification of genotypes.
Kan~ala et ui. (2005) studied genetic and phenotypic diversity atnong 36
randotnly selected downy-milde\v resistant snrylii~niaccessions, the former using 10
SSli marker loci and the latter using 20 phenotypic traits. The nu~nberof alleles (ai) at
individual loci varied from 5 to 14 uith an average of 8.8 alleles per locus. Nei's gene
diversity (I-1.i) wried fiom 0.50 to 0.92 ~ i t hall average of 0.81 per locus. I-iigh gene
dibersity and allelic richness were observed in races durra-caudatum (Hj=0.76,
;1.i=J.3) and guinea-caudatum (lli=0.76. aj= 3.8). and among accessions from east
Africa (I-11.-0.78. a,i=7.2). The regions were genetically more differentiated than the
races as indicated by Wright's F,,. The pattern of SSR-based clustering of accessions
\\as 111ore ill ~ccordiniceurith their geographic proximity than with their racial
likeness. This clustering pattern matched poorly uith that obtained Trom phenotypic
traits. 'The inter-accession genetic distance vc~riedfrom 0.30 to 1.0 with an average of
0.78. while inter-accession phenorypic dirtance varied from 0.1 to 0.55 with an
alerage 0.33. Eleven accession pairs had pllet~ot!pic distances more than 0.5 and
genetic distance more than 0.7. These could he u~etlas potential parents in sorghum
<lo\vny nild dew resistance breeding prograln (personal co~nmunicatiot~
fro111 Dr. S.
C'liandra. Principal Scientist. Statistics and Head. Bioinibr~naticsUnit. ICRISAI').
Casa er (11. (2005) iluantified atid cli;~ractcrized diversity in a panel of
culti\ated and \vild sorghum \\it11 98 SSK loci distributed through out the genome. In
a panel of' 104 accessio~iscomprising 73 landraces atid 31 wild sorghums. Evaluation
ot' SSli polymorpllism indicated that landraces retained 86% of the diversity observed
in the \rild sorgllum. The landraces and wild accessio~ls were moderately
differentiated. but there were a little evidences of population differentiation atrloilg
racial groups of cultivated sorgl~um.Neighbor-,joining analysis showed that wild
sorghums generally formed a distinct group and about half of the landraces tended to
cluster by race indicating a history of gene flow atnong the \larious cultivated type or
recent common ancestry.
2.2 Host plant resistance: mechanisms a n d inheritance of shoot fly resistance a n d
its component traits in sorgtrum
Sorghum [Sorgl?uri?bicolor (L.) Moencli] is an important cereal crop of semi-arid
regions of Asia, Africa. the Americas and Australia. Generally the lower yields in
Asia and Africa are associated with pest damage. Nearly 150 insects species liave
been reported as pests on sorglium (Reddy and Davis. 1979: Jotwani er al.. 1980).
Slioot fly is one of the ma.ior pests of sorghum in Asia and Africa. Adoption of
chemical control methods is not economically feasible for no st of the sorghum-
growing farmers. Therefore. utiliration of host-plant resistance is the most realistic
approach to reduce losses caused by sorglium insect-pests. E\en though genetic
Lariability for shoot fly resistance is available in the sorylu~mgermplasm. the le\,el of
resista~iceis not high and the a\ailable sources of resistance liave poor agronomic
features. 1'11~q~latntitativenature of resistsunce to this illsect and n large environmental
variation in its expression hinders genetic ~nanipulation of shoot f l y resistance by
co~lve~ntionnlplant breeding procedures. Resistance of plants to insects is the
coliscquence o f lleritable plant cllamcters that result in a plant being relalively less
damaged tlnan plants without those characters (Sharma. 1997). Man!, other studies
have also revealed that e number of component traits ore associated with shoot fly
resistance
'The Ipresent relieu co\,ers the areas of control of shoot fly, with main
enipl~asison host plant rrsistance. mechanisms and inheritance of resistance, and
breeding fiir resistance. It also sunimarizcs !reports on moleccllar markers. QTL
~ n a p j ~ i natid
g statistical techlliql~cs for mappings ill general and for sorghum in
particular.
2.2.1 Shoot fly control
Control of sorglium shoot Ily can be achieved by early and/or timely sowing.
increased seed rate. thinlling and destroying tlie seedlings with deadhearts, crop
rotations. fallowing and others methods like use of insecticides (Sing11 and Sliarma.
2002). However, tinlely sowing depends on several factors like cropping system.
rainfall, soil type and ~ n o i s t ~ status
~ r e at sowing time. tnatl) of which are out of control
of farmers. From previous studies (Jotwani er cri.. 1970) it has been established that in
khcrrif season shoot fly incidence and damage increases with delay in sowing date.
Planting time studies during khtrrYf season using high yielding cultivars showed that
early khor~fsowing with tlie onset of the soi~tl~west
mo~lsooneither avoided or
significantly reduced the incidence of damage by shoot fly (NRCS, 1998). This
finding proved to be extreiiiely useful for the widespread cultivation of high yielding
cultivars possessitig lesser levels of shoot fly resistance. In tlie case of rirhi sorghum.
advancing sowilig dates gives better yield potential and efficient use of residual
moisture under rainfed condition. Houever such plans for advancing rahi sorglium
sowing arc spoiled due to higher slioot fly i~icidence.For control of shoot fly in rcrhi
sorghum, iise of various insecticides like phorate. disulfaan and carbofuran have been
advocated. Iio\cever. under high shoot fly pressure such attempts at chemical control
hil. A s far as biological control is concerned, tilore than 15 species o r shoot fly
predators have been recorded. but their predation potetitial has not been assessed
ulider field conditions (Sing11 and Sharma. 2002).
2.2.2 Host plant resistance
'fhc usc of' resistant varie~iesmay offer tile best and pcrhaps the only econotnical
method of uontl.ol o f certain pests like sorghi~tnslioot lly. because the control of
insects on a crop of low value precludes the use of insecticides (Uhams. 1043).
Painter (1951) detined resista~icein plants to insect attack as the relative amount of
hcritahle qualities of the plant that i~ltluenccthe ulti~uatedegree of damage done by
tllc it~srct.While accordil~glo Smitl~(1989). resistance of plants to illsects etlables a
plant to avoid or inhibit host selection. inhibit o\~il~ositioii
and feeditig. reduce insect
survival and ilevelopment. and tolerate or recover froin i11,juryby insect populations
that W O L I I ~ cause greater darnagc to other l~lantsof the same species under similar
e~lviron~nental
conditions.
2.2.2.1 Sources of resistance
The existence of resistance ill sorghum to shoot fly mas first reported by Ponnaiya
(lr)S1a). who identitied resistant culti\,ars: must of then^ \\ere li.0111peninsular India.
Subserluentlj. Kao and Kao (1956) aiiil .fain and Bhatnagar (1962) evaluated 42 and
196 ci~ltivars.respectively ntid selected a few promising resistatice sources. 'rhe
search Tor sources of resistance to shoot fly continued through field evaluation of
thousands of entries of the World Sorghum Collection by the All India Coordinated
Sorghum Itiiprovement Prqject (AICSIP) during the 1960s (Sing11 et crl., 1968:
Pradhan, 1971; Young, 1972) a1111 by AICSIP and ICRISAT during tlie 1970s and
1980s (Jotwani. 1978; Rao et 01.. 1978; Jotwalii and Davies, 1980). Rao (1972)
remarked that most of these identified resista~icesources belong to the ri~oldar7dior
dtigdi types of Indian winter sorghums or the shollu types usually grown mixed with
ii~oldtrtldior dogdi types, which consequently survived in sniall populations.
Several workers had screened sorghum gerlnplasm for resistance to shoot fly
considering the needs of the local breeding programs and identified resistance sources
(Table 2.1 ). As the work on shoo1 fly resistance continues, several new sources are
being put at breeders' disposal e v e y year.
17requency distributions of slioot fly reactions among sorgllum gern~plastli
accesqions assessed for susceptibility to shoot fly re\ealed that out of 16694
accesstons evaluated, 133 accessions sho\\ed high le\els of resistance in the rainy
season. but only 18 accessions showed high levels of resistance in the postrainy
season (Sharma rt trl., 2003). As far as taxonomic distribution is concerned, out of
1290 accessions showing some degree of lrsistance to slloot fly in tllc rainy season.
niost oS the accessions belonged to the race Lllirrn (471) or C C I I I C ~ I I(185).
I ~ I ~ I I'he
~eogmphicdistribution of these sorghum gern~plasmaccessions pointed out India as
the main area of origin o r accessions sho\ving rcsistruice to shoot fly in the rainy
season. ti)llo\\ed by Sudan and Nigeria. For postrainy season conditions. most or the
resisla~~t
ticcessions 01-iginated Srom Inclia. tbllo\zed by Ethiopia. Sudan and Nigeria
(Sharnlc~el (11..2003).
2.2.2.2 Mechanisms of resistance
All tlie three nieclianisms of resistance suggestecl by Painter (1051) 1.i:. ovipositional
nm-]>reference (Soto. 1974). antibiosis (Kaina tat c11.. 1081 ). and tolerance/recovcry
resislance (Doggett t.1 ul.. 1070). are Lno~iln to cvist in sorghum Sor shoot fly
resistance. The priniary meclianis~nsof resistance to sorghum slloot 11) have been
ohsrrvcd to be ion-preference for oviposition and perliaps a low lecel of antibiosis to
tllc larvae (Young. 1072).
2.2.2.2.1 01ipositional non-preference / aatixenosis
Sain and Rhalnagar ( 1 962) fit-st reported ovipositional tlon-preference by shoot fly in
resistant sorghum cultivars. I.ater several workers considered it as the primary
mechanism of resistance in sorgliu~n (Blum. 1967: Krishnananda ct a/., !970;
Kangtlang et trl.. 1970: Jotwjani er (11..1971 : Pradhan. 1971 : Young. 1972: Soto. 1974;
Nara~ana.1975: Sharma et (11.. 1977: Singh and Narayana. 1978: Singh and Jotwani.
l980a: Singh e/ crl.. 1981; Sharma and Rana, 1983; Kana rt trl.. 1984: and Unnithan
and Rrddy. 1985).
conr ...
Genotypes Season(8)of Resistant Genotype(s) Reference
screened (Nos) screening

IS 2 122, IS 2 123, IS 4660, IS 5092, IS 5480, and IS Uniithan and Reddy (1985)
18551
9 Rainy E 20 1 to E 208, and E 303 Kishore (1986)
20 Rainy IS 1082, IS 2146, IS 2312, IS 5470, IS 5622 and IS Mote et al. (1986)
5633
8 Late kharif IS5604, IS 5490, and IS 2146 Nimbalkar and Bapat (1987)
67 Rainy and post- IS 1456, IS 7094, and IS 1261 1 Jadhav et al. (1988)
rainy
20 Late kharif IS 1054, 1s 2123, IS 2312, IS 2146, IS 18551 etc. Omori et al. (1988)
IS 1054, IS 18551, IS 2123, and IS 5469 Singh and Verma (1988)
20 Rainy, late IS 2205, IS 1054, IS 5469, IS 5619, IS 18557, IS Pate1 et al. (1985);Pate1 and
kharif and 8320, S 386, and SPV 102 Sukhani (1990)
summer
Rainy and post- P 24, E 302, 370 x 3660A, IS 1199 etc. Dalavi et al. (1990)
rainy
205 Postrainy IS 2312, IS 2191, IS 4516, IS 17596, IS 33714, and Balikai et al. (1998)
IS 33843
39 Rainy PGN 1, PGN 8 , PGN 19, PGN 20, PFGS 2, PFGS 8, Kishore (2001)
PFGS 27 etc.
16694 KhariJ rabi IS 1034, IS 2146, IS 2205, IS 2312, IS 4664, IS Sharma et al. (2003)
5604, IS 22121, IS 22144, IS 22145, IS 22148, IS
22149, IS 22196, and IS 18551
 Jail1 and Bliatnagar (1962) screened 196 sorglium varieties from the World
Collection to assess varietal resistance to shoot fly and reported significantly less
oviposition on resistant varieties as compared with susceptible ones. Similar results
were also reported by Blum (1969b). Jotwani et 01. (1971) and Jotwani and Srivastava
(1970). 'They also reported that the efficacy of this mechanism was not stable and that
i t breaks down under no choice co~iditionsor under lieavy shoot tly pressure. When
geographic distribution was considered, degree of slioot fly preference was foutid to
be more (55%) in tetilperate and comparatively less (33%) in India11 varieties (Singli
L'/ l i l . . I981 ).
Ueha~ziouralresponses of shoot fly slio\ved that initial choice of a susceptible
culti\rar. CSll 1 was random. but that the duration of female stay on resistant
gemiplasm accessions IS 2146. IS 3602 and IS 5613 was brief (Sliarma and Kana.
19x3). In addition. adlilt females laid eggs on non-preferred cultivars only after laying
se1:ernl eggs on alternate susceptible CSII 1 seedlings.
Koina c/ 111. (1984) reported that it1 single choice tests, signilicant non-
preference for cwiposition was obserled on IS 2146. IS 3962 and IS 5613. 111another
eupcri~iientwhere rctiiales were gi\en no clloice fix an oviposition substrate but could
escape into an outer cage, ovipositional non-preference was evident Sir five tlie seven
test cultivars. IS 2 \ 4 6 atid IS 3962 byere consiste~itlynon-preferred for oviposition in
both ofthesc tests.
Singli and Jotinani (19SOa) and Horikar t7t 111. ( I 9x21) indicated that efficiency
o f this mechat~ismof resistance is not stable and it tends to breakdown under 110

clloice conditions and under h c a q slioot fly population pressure. Mote er 01. (1986)
reported that the leaces of tlie some sorglium cultivars resistant to sl~oottly were pale
grcen coniparecl to dark grern colour of the susceptible cultivars. Texture and widtli of
tlie leaf \yere also inlportant factors in selectio~iof the oviposition substrate by the
fetilale fib. Narrowness and erectness o r the l e a ~ e sreduce oviposition substrate
resulting ill less egg laying and lower deadliearts incidence compared to plants having
broad and droopi~igleaves. Genotypes ICSV 705. IS 1054. IS 2146, lS2206, IS 4663.
IS 5613. PB 15881-3, IS 18551, and IS 2312 have been reported to displaq high levels
of atitixenosis for oviposition ( < I 7 eggs seeding") as compare to susceptible check
Swarna (18.8 eggs seedling-') across Indian slioot fly screening locations (AICSIP,
2003). Ka~iiatarand Salimatli (2003) suggested that plants with eggs contributed
directly to deadhearts incidence (%) and could be used as a criteria to select sorghum
resistant to shoot fly. while leaf colour. seedling vigor. glossiness, leaf width and
seedling height colltributed indirectly towards plant resistance. Wild sorghuIll
gertnplasn~ accessions belong to Pala-sorghum and Stiposorghum sections were
imlnune to shoot fly damage, while IIeterosorg11~111~
and Chaetosorghum accessions
showed negligible damage and the test accession of section Sorghum exhibited
silsceptibility to shoot fly under !multi-choice conditions (Venkatesh\rran, 2003).
2.2.2.2.2 Antibiosis
Antibiosis to shoot fly was reported hy Jotbani and Srivastava (1970). Rlum (1972).
Soto (1974) and Sharma et tri. (1977). Survibnl and develop~nentwere adversel>
affrcted when slloot fly larvae \\ere renretl on resistant varieties (Jotwani and
Srivastava. 1970; Narayana. 197.5: Raina c l (11.. 1981; Llnnitlian atid Reddy. 1985)
coinpared wit11 susceptible genotlpes (Singh and Nurayana. 1978). Growth and
development were retarded. ant1 the larval and pupal pcriods \\err extended bq 8-15
days o n resistant varieties (Sing11 aritl Jotnani. IOXOb). Survival ri~ldfi.ctttldity \\ere
also better on Iiighl) s~rsceptiblevarieties (Singh and Karayana. 1078). but adversely
affccted on resistant varieties ( I'aneja and L.euqchnei. IC)85).Survival ruid longevity
01' Scmales rund fecundity \\.ere adversely aR'ectetl \\llcn the larvae \rere reared 011

sl~oot11) r e ~ i s t ~ genotypes
nt (Raina e/ 111.. 108 I). l.ar\ a1 and total gro\\tli indices were
signilicnnll\ lowered on resistant conipared \\it11 susceptihlc \.arieties. The percentage
p ~ ~ p a t i o non resistant karieties mas signilicantly loner compared with that 011

susceptible varieties (Ilha\,on e/ (11.. lOo3).

Ilain;~o (11. ( 1981) suggeatcd that tricho~nelessculti\ars nccumulate more dew
atid stay met lot~ger.'l'llis situatiuti woitld facilitate the luo\cment of' ti.cshly llatched
larvae to the base of central shoot. On the other hand, trichomed culti\ars tend to dry
faster, making the doutlward journey of lar\ae more difficult. The earliest work tlial
reported to antibiosis as a possible ~iiechanisinof shoot fly 1,esistance in sorghu~n\\,as
tliilt of I'onnuiya (1951a.) He attributed this to earl) deposition oi' irregular shaped
silica crystals in the resistant culti\ars. ullich \\as confirmed by Blur11 (1968).
Railla (1985) reported that three different factors. individually or in
combination. may contribute to the expression of alitibiosis to shoot fly in sorghutn:
( i ) trichomed cultivars hinder the movement of newly hatched lar\ae to the base of the
whorl: (ii) resistant cultivars had greater silica deposits and lig~lificatio~l
of cells,
which nay restrict larval penetration to the base of the whorl: (iii) biochemical
deticie~lciesor presence of chemical factors in resistant cultivars may adversely affect
the development and survival of larvae and reduce the fecundity of the resulting
adults.
Stability parameters for IS 8315 and IS 2123 revealed that the level of
oviposition will differ on these two resistant lines under different levels of infestation
pressure but there will be relatively less mortality in these resistance sources than in
more susceptible sorghum genotypes. This is probably indirect evidence of antibiosis
resistance mechanisms present in these two genotypes (Borikar and Chopde, 1982).
Some cultivars are preferred for oviposition; however, levels of infestation as
measured by deadhearts production are low mainly due to antibiosis (Mote et al.,
1986). Lower larval survival on resistant genotypes as compared to a susceptible one
has also been reported by Jadhav el al., (1986). The mortality of the first instar was
highest (96%) in the first 24 hours (Mowafi, 1967; Bushara, 1972; Zein el Abdin,
1981), which depends not only on the ability of the female to select a suitable
oviposition site, but also difficulty in penetrating the leaf sheath, and covering the
distance between the egg deposition site and the seedling growing point (Delobel,
1982). The larva growing on a resistant variety is typically sickly in appearance and
smaller compared to that grown on susceptible varieties. No larval survival was
observed on accessions of Stiposorghum and Heterosorghum (Sorghum laxiflorum)
and Para-sorghum had relatively higher levels of deadhearts incidence, but there was
no fly emergence (Venkateswran, 2003). These results indicated that along with the
non-preference mechanism of resistance to shoot fly, a high degree of antibiosis is
also present in different groups of wild Sorghum accessions. The resistance of
sorghum to the sorghum shoot fly is largely a cumulative effect of non-preference and
antibiosis mechanisms (Raina et a/. 1981).
2.2.2.2.3 Tolerance 1 recovery resistance
Five shoot fly resistant and 2 shoot fly susceptible sorghum varieties were studied in
order to evaluate the association between several plant traits and tiller survival both
under field and stimulated conditions (Blum, 1969a). In both experiments, tillers of all
resistant varieties grew faster than tillers of the susceptible ones and also infestation
by shoot fly was delayed by 2 days in resistant varieties as compared with susceptible
ones. This form of resistance has been referred to as tiller survival, while Doggett el
al. (1970) referred to this phenomenon as recovery resistance. Similar results were
also observed by Blum (1972).
 Doggett (1972) pointed that synchronized tillering after the main shoot is
killed. is a potential for111of recovery resistance. In Africa, farmers actually prefer an
initial infestation of their sorghu~nby shoot fly that led to profuse tillering and
subsequently a good harvest. Howe\rcr. Indian sorghums were known to be non-
tillering and any basal tillering was a consequence of failure of the main shoot to
grow due to deadheart formation. IIowever. the tillers of susceptible varieties
continue to be attacked by so~.glwrn slioot fly ~ ~ n d eoutbreak
r conditions, thus
resulting in failure to yield the harvestable heads (Sharmn et 81.. 1977).
Raina (1985) opined that tolerance can be greatly intluenced by grow~li
conditions and thus may not always be predictable at c a r i o ~ ~locatiuns.
s particularly
tliose wit11 irregular rainfall patterns. Further. recovery resistance/tolerance does not
appear to be an useful mecllanism particularly when slioot fly populations
progressibely increase as tlie rainy seasol1 continues (Singh ct (11.. 1981.Singh and
Rana. 1986).
Mote er trl. ( 1 985) observed that SPlI 196 and SPH 325 were leas susce]>tible
to 11. socccrrti at the initial stapes of seedling gronth and expressed the highest
frequency o r recovery resistance and Ilence glxin yield anlong I4 sorghum hybrids
tested. Tiller development consecjucnt to dcadlleart fc7rnlation in the main shoot and
tllc subsecl~~ent
s~11.\4\'nland recover> of the sorghum plant depends in part on tlie
l e ~ e lof primary resistance. Varieties with Iiigh recover?. trf resistance appeared to
!.ield more under shoot fly infestation (Rani1 c / t r i . . 1985).
2.2.2.3 Factors associated with resistallce
Some seedling (pliysico-morpl~ol~rgical)
characters (Btum. 1068: h'laiti and Didinper,
1979: Raina. 1981: Maiti el tri.. 1984). as \%ell as sonle biocllemical factors. are
associated with shoot fly resistance in sorgllu~n(Sing11 and .lotwani. 19XOc: Patel and
Sukliani. 1990). Resistant cultivars are ~~sually
tall with thin stems having lotig
internodes and short peduncles. Also they typically have narrow glossy and
yellowish-green leaves. These leaves possess trichomes on tlie abavial surfhce, which
act as physical barriers to movenient of y o ~ ~ nn~aggots
g (Kishore er <I/., 1985; Mote er
01.. 1986). Colour of leaves. glossiness of leaves and presence of trichomes are
prominent attributes conferring resistance to shoot Ily in sorghum (ladhav er a/.,
1986). These factors liave been studied in detail and hence are revie\ved individuall}
below.
2.2.2.3.1 Glossiness
The glossy trait, a characteristic of most of the winter ( ~ w b i )sorghum varieties of
India (Blum, 1972; Rao rt 01.. 1978). is reported to be associated with shoot fly
resistance (Blum, 1972: Bapat er crl.. 1975: Maiti and Bidinger. 1979; Taneja and
Leuscliner. 1985; Omori el 01.. 1988). Tarlin~oto(1980) reported a simple screening
technique for identification ot' glossy cultivars atilot~glarge gerrnlllasm sets. The
difference between glossiness and non-glossiness can be detected by whether or not
sprayed water adheres on leaf blades.
Maiti and Bidinger (1979) screened approximately 8000 lines li-om the \vorld
sorghum germplasm collection for resistance to sl~ootfly and observed that lilies with
trichomes on their abaxial surface were more resistant to slioot fly than lines lacking
such trichomes. These resistant lines also had otlier clistinctive cllalncteristics. \\liich
\\ere evident only in first 3 weeks of seedling growth: leaves tended to he more erect
and narrower. with yellowish-green glossy appeatxnce. \cllich is tern~cdas tlie 'glossy
trait'. rZ systematic survey of the ucirld gernlplasm collection indicated a low
frequency of accessioiis \\,it11 tlie glossy trait (only 495 of 17.536 ger~nl>laam
accession.; screened) and 84"/0 of these lines where of' Indian orifill. While glossiness

is cleaily manifested in tlic scedling stage. it graduall) disappears as the scedli~lg

g~.o\\silnd soil fertilit!, does not affect its expression (blaiti er c r i . . 1984).
Taneja and I.euschner (1985) identified 42 lines that were consistelltly
rcsista~~t
ti1 sl~ootfly. and out of tliese 42 lines. 37 were glossy. Furtllcr evaluation of
these lines for shoot fly reaction in rainy and postrainy senqtins rexe'11t.d that shoot fly
incidence w i ~ shighel. in non-glossy lines than glossy ones in tllc post rainy season.
Ilonever, glossiness contributed less to slioot 1ly resistance during tlie rainy season. .
Glossiness of seedling leaves may possibly all'ect the cjuality of light reflected
from leaves and influence the orientation ol'o\ripositting slioot flies to\wrds their host
plant. Also glossy leaves might also intluence llost selection by rneans of chemicals
present in tlie surface waxes or by altered permeability of such waxes to chemicais
present it1 the leaves (Sharnia. 1993). Most of the lines resistant to shoot fly exhibit
the glossy leaf characteristic during the seedling stage. The intensity of leaf glossiness
at the seedling stage is positively associated with level of resistance to shoot fly
(Sliarnla and Nwanze. 1997). Both A- and A-line components of pairs SPSFR 9406.
SI'SFR 94034, SPSFR 94036 and SP 55301 were significantly less susceptible to
shoot fly than susceptible check CSH I and had glossiness scores of <3 (ICRISAT,
1999).
Kamtar and Salimath (2003) observed highly significant inverse relationships
between seedling glossiness score and both deadhearts incidence and oviposition
levels. The level of resistance to shoot fly was higher when both glossy and trichomes
traits occurred together (Agawal and House, 1982). The presence of trichomes and
glossiness have independent and apparently additive effects in reducing the incidence
of damage by shoot fly (Maiti et ol., 1980).
2.2.2.3.2 Trichomes
Levin (1973) described the role of trichomes in plant defense and pointed out that in
numerous species there were negative correlation between trichome densities and
insect feeding and oviposition responses, including nutrition of larvae.
Maiti and Bidinger (1979) identitied 32 lines from 8000 sorghum germplasm
lines with trichomes on abaxial surface of the seedling leaf blade. These accessions
had fewer plants with deadhearts and lower ratios of plants with deadhearts to plants
with eggs than 35 lines without trichomes. Maiti el a/. (1980) observed that the
presence of trichomes on the seedling leaf surface resulted in a lower frequency both
of oviposition by shoot fly and subsequent larval damage. Resistant accessions IS
2146, IS 3962 and 1s 5613 had high densities of trichomes on the abaxial leaf surface
while susceptible hybrid CSH I was found to lack trichomes. However, under heavy
shoot fly infestations, the density of trichomes appeared not to make any difference
between preference and non-preference for a sorghum cultivar as a substrate for
oviposition.
Three wild Sorghum species (Sorghum versicolor, S. purpureosericeum, and
an unidentified wild genotype) amongst 57 entries covering different species were
found to be immune to shoot fly (Bapat and Mote, 1982b). It vas observed that these
immune entries all had high densities of trichomes on the lowers of their leaf blades,
which contribute to their resistance.
Maiti and Gibson (1983) suggested that trichomes might be less effective
during the rainy season than during the postrainy season, possibly because of
physiological factors or more severe shoot fly attacks during late rainy season
plantings. Biradar et ol. (1986) reported that the intensity of trichomes on the adaxial
surface was 2 to 6 times more than abaxial leaf surface. Although, the trichome
density on the abaxial surface of the leaf have significant and negative correlation
with deadhearts, it has indirect effect on oviposition by sorghum shoot fly (Dhillon,
2004). Role of plant trichomes in insect resistance is through physical barrier in the
movement of insects on the plant surface (Peter et al., 1995). Trichomes in sorghum
deter penetration of the young shoot fly larvae in the whorl (Maiti et al., 1980).
Jayanthi et al. (1999) observed that the expression of trichomes in hybrids depended
on the type of parents involved and the seasion of testing. If the postrainyseason-
adapted resistant male-sterile lines were involved, trichome expression in hybrids was
lower in the rainy season than in the post rainy season.
Trichomes can act as an insect resistance mechan~smby limiting the insects'
contact with the plant. Such trichome can act as a physical barrier to insect movement.
In addition, glandular trichomes can contribute to insect resistance by producing toxic
compounds, which poison the insect through contact, ingestion, and/or inhalation, and
by producing gummy, sticky or polyme~izingchemical exudates, which impede the
insect movement (Duffey 1986, David and Moorthy 1988).
2.2.2.3.3 Interaction of glossiness and trichomes
A study of four combinations-glossy leaf and trichomes, glossy leaf only, trichomes
only, and neither - revealed that the mean deadhearts percentages were 60.7, 70.9,
83.5 and 91.3, respectively (Maiti and Bidinger, 1979). The glossy trait alone (mean
of 71% deadhearts) seemed to be more effective in reducing deadhearts incidence
than trichomes alone (84% deadhearts). However, the combination of both characters
(61% deadhearts) was significantly superior to the mean of the two resistance
component traits taken individually. Similarly Aganval and House (1982) also
reported that the level of resistance was greater when both the glossy and trichome
traits occur together.
2.2.2.3.4 Seedling vigor
Blum (1972) reported that shoot fly-resistant lines grow faster than susceptible ones,
while Dhillon (2004) found that shoot fly-susceptible lines initially grow faster and
attracted by shoot fly for oviposition, resulting the early deadheart formation,
however the resistant lines delays oviposition, resulting in less deadhearts. Singh and
Jowani (1980d) indicated that longer and narrow leaves and faster seedling growth as
indicated by leaf sheath length (8.36 cm in CSH I compared to 12.36 cm in IS 5469)
and seedling height (29.13 cm in CSH 1 compared to 39.33 cm in IS 5469), coupled
with hardness of the leaf sheaths may be contributing towards resistance to shoot fly.
 Khurana and Verliia (1985) studied plant characters of nine sorghum lines (6
resistant to shoot fly and 3 susceptible) and concluded that faster growing resistant
plants may remain in the favorable height for relatively lesser period as compared to
the slow growing susceptible plants. 'l'aneja and Leuschner (1985) observed that in
tlie postrainy season. shoot fly incidence was higher in sorgllum lines that were less
vigorous at seedling stage: ho\cever. tlie same trend was not observed in tlie rainy
season. Also, it was observed that fast seedling gro~vtllnliglit prevent the first instar
larva frorn reaching the seedling g r o ~ i n gtip. although leaf margins may be cut
withont causing deadheart symptoms.
Jadhav 't (11. (IOXh) studied morphological plant character^ in 158 sorgliun~
entries for interaction wit11 response to slioot fly measured in terms of deadhearts
incidence and concluded tiiat apart from the glossy lrait and presence ol'trichomcs.
initial faster plant growth rate confers resistance to slloot fly in sorglii1n1.
Kar:ui.ikarer 111. (1992) observed positive relationsiiips between vigor of the
plant and its cscape fro111 shout fly attack (The seedlitig vigor score was recorded
i~nde! moderate level of slioot fly infestation). Sing11 (19V8) concluded that rapid
seedling grontli and long, thin seedling leaves make plants less susceptible to shoot
tl!. Seedling vigor \vas signilicantly and negzztively associ:ited \\it11 deadhearts ant1
u\.iposition ('laneja and Leuschner. 1985). (I'he rapid seedling growth o f
toleranllresistant genotypes acts as escape mecllanism against shoot fly infestatioll.
hence less oviposition and in turn less deadliearts incidence). Regression analysis
indicated inverse associationi between seedling vigor score and deadhearts incidence
and direct associations with ovipositio~i incidence and esg count (Kamtar and
Salitnath 2003).
2.2.2.4 Inheritance of resistance
Hluni (196917) developed 8 hybrids (made from 2 slioot t11 susceptible and 4 resistant
sorglium lines) and their Fz progenies. The parental lines and all F2 populations \\ere
evaluated under three levels of shoot 11) infestation. The F? data indicated that
resistance was ~artiallydominant when evaluated tinder low shoot fly population
pressure, v+hile when evaluated under liigli shoot fly population pressure.
susceptibility appeared to be dominant.
Balakotaiah rt 01. (1975) co~iducted a genetic analysis of resistance to
sorghum shoot fly based on large F2 populations from a diallel mating system
involving exotic. Indian, atid derived lines as parents. Gene effects estimated fro111
generation means analysis exhibited predonlina~ice of additive gene effects for the
inheritance of shoot fly resistance.
Slia~liiaet ai. (I 977) conducted a diallel analysis involving four agronomically
superior dwarf and four resistanr varieties of sorghu~ii to study inheritance of
resistance to sorghum shoot fly (It is a 8x8 diallel (without reciprc1cals) consisting 4
agronomically superior lines + 4 resistant lines). Inheritance of resistance was
reported to be quantitative as indicated by the prevalence of cotitinuous variation in
different generations and the ititernlediate resistance levels expressed in liybrids of
resistant and susceptihlr parents. Resistance was 11iainlyaclditive in nature. It was also
ohserved that FI hybrids of susceptible and resist;ltit parents were slightly more
susceptible than the ~nid-parental value and thus. susceptibility appeared to be
partially dominant.
l3orikar and C'liopde (1980) evaluated an TI diallel cross wit11 4 rcsistant uncl 4
susceptible parental lilies, under 3 tlistinct levels of shoot fly inrestation ill rainy
seiision The magnitude of additive components of variance. as compared to
doniinance colnponents, increased with increases in shoot fly population pressure.
Sorgllutn susceptibility to slioot fly appears to be recessive under low insect
pop~~lation
pressure hut exhibits donlitlance under high shoot fly population pressure.
Ilane e/ (11. (1081) studied the bchaviour of shoot fly resistance in the F I ,Fl. FI
~uiiiadvanced generations of crosses between resistant and susceptible parental lines.
111this study the mas observed to be alnlost intermediate bet\\een the two parents
with an added lieterotic advantage of lo\vcr deadliearts percentage. I<esistetice showed
partial dominance under lo\$ to moderate shoot fly population hut this relationsilip
may sliifi under heavy infestation conditions. I'lie resistance is polggenic in nature
and governed by additive genes.
Halnlli et ul. (1982) repcvted that ill a seven parent diallel cross. comprising
Ibur high-yielding varieties and tluee cultivars with varying levels of shoot fly
resistance crossed in all possible con~binationsand evaluated in post rai~lyseasion.
The inheritance of slioot fly resistance was found to be controlled by both additive
and lion-additive genes effects.
Halalli ef al. (1983) evaluated advanced generations during kharif to estimate
extent of variability, heritability arid genetic advance for shoot fly resistance. Five
BCIF3 progenies, one Fi progeny. and 3 F4 progenies were found to be significantly
more resistant than the most highly resistant parent, IS 5604. suggestilig transgressive
inheritance of the character.
Patel et 01. (1984) studied combining ability for shoot fly resistance in an 8-
parelit diallel cross without reciprocals. They reported negative general co~iibi~iitig
ability (OCA) erfects in resistant paretit for percent deadliearts both in normal and late
sowings, suggesting preponderance of additive genetic variance. Patel cr al. (1985)
observed tliat in both normal atid late sown conditions: additive (D) as well as non-
additive (Hi. 112) components of genetic variance were significant for resistance to
shoot fly.
Nimbalkar a ~ i dBapat (1992) evaluated an 8-parent diallel cross under thl-ee
levels of shoot fly infestation to study combining ability and genetic coliipolietits (3s

variation for shoot fly resistance. 'fhey observed additive gcnc action for shoot fly
resistance.
2.2.2.4.1 Heritability estimates
liano ct (11. ( 1 975) reported that ililf'rrelices between slioot Ily susccl3tihlc antl resistant
progenies are establislied from tlie t', generation and tlie 1iel.itability esti~iiatefor shoot
11) resistatice mas about 25%. Dorikar and Cliopde (1081a) analyzecl an 8-parent
dinllel cross in the T Iand I;? generations to s t ~ ~ dtlie
y genetic architecture of shoot fly
resistance atid indicated that licritability for slioot fly resistance appears to be around
23 to 25 percent.
IIalnlli 1.r (11. (1 983) screelied advanced generation ~nalerialsand reported tliat
broad sense heritability was around 30%, indicating a large influence of envirotllue~it
on shoot fly resistance. A sunilliary of inheritance studies Tor different shoot fly
resistance traits and their getietics alotig with lieritabilit> values reported by different
workers are presented in Table 2.2.
2.2.2.5 Breeding fur r e s i s t a ~ ~ c e
Although tlie work on sorglium shoot fly resistance \\as initiated in tlie early 1050s.
11~1attempts \rere made to incorporate resistance into a variety with good agronomic
base. Only since the late 1960s have various sorghum research workess [including
Ulum (1965. 1967, 1968. 1969a. b) from Israel. Doggett er 01. (1970) and Starks er (11.
(1 970) from East Africa. Harwood et crl. ( 1 972) from Thailand. Rao et trl., (1974).
Balakotaiah et (11. (1975) Rana ef LII. (1975. 1981, 1985). and Agrawal and Ilouse
(1982) from India liiade significant co~itributionsin breeding for shoot fly resistance.
 i!aql pue slalopai
aslamp Z I pue Saul1 aluajs-aleur paldepe
(6661) ' I D la !qlue-cer jueu!urop llle!)~ad- .G!suap awoqs~ij,. ~ o q -uoseas .6u!e1pod pue .Cu!ei a , q a ~ ~
a>ueu!urop
x a.\!l!ppe pue a n ~ ~ r p p-eaaejins jeal [e!xeqy
(6861) asueu!wop c3E
ie.wnprreq3 pur? ~ e y ! ~ o g x asueu!ruop pue an!pppe - a q i n s jeal le!xepy ' 1 3 8 '76 'Id '(alq!ida3sns)rd '(lue~s!sai)I d

asuels~saiu! panlonu! osle ale p a p o d a ~ sauq leluaied

.i[sno!aaid jeqq ueql lay30 !lo[ Iauo!)!ppe OM$ ssalamoyJ!ix pue pauroqsul Bu!ssoi~
( ~ 8 6 1U) O S ~ !pue
~) !)!en ]seal l e :asuels!salloj 1 0 1 3 ~
ioleur
~ e - samoq3uL .iq padola~apsuo!leiauaB Bu!leBai%as
sauq [ e $ u a ~ ~ d
ssalauoq3u~pue pauroqsul %u!ssoi~
( ~ 8 6 1!l!)=W P m u'Jsq!t) auaB an!ssa>ai al%u!s - sauroqs!ll 30 asuasald .iq padola~apsuo!$wauaB Bu!~eBaiBas
mnqBiog
( 6 ~ 6 1~~SU!P!E
1 P U !IWW
~ aua%an!ssa3ai a[Zu!s - sauroq3!.11 JO asuasaq 30 uo1l~a1103P I ~ O M
~ 0 1 sa-1
3 0008
dq!suap amoqqq pus samoqopq JO aouasqs l o aouasaq
SIJ n a q l
suoseas ssoise alqels pue sia1o)sai asian!p ZI pue sau!l aluals
(6661) 'la la !qlue.Cer uo!ssa~dxa 1!si1 pue iueu!mop ssau!ssolZ-uo~ -aleur paiq h'u!ei-lsod prre ?(u!e~a a 1 a u
(~861)
weqwqy pue pmeiBy
(I&) isso12 an11 p m 06) .(sso12 '(15)kso18
( ~ 8 6 1o)l o u r n i e ~ sa[al[a a[d!qnn -uou Buoure sassois jo suo!lelndod rd
(0861) 0lo'JJNe.L an!ssasai a[dw!s id "8 '(ASSO@)7d ' ( r k ~ ~ [ Z
Id - ~ ~ ~ )
8samssol%-uoupus ssaupsolt]
aouarajay panIoau! uolaoa a u w pasn w a ~ s m
%u!paua SUOJJB~
(Table 2.2 cont ...)
Factom Breeding material used Gene action involved Reference
Seedling height
PI (resistant), P, (susceptible),FI, F,, BCI, Addltlve Sharma et al. (1977)
BC.I
PI (resistant),PI (susceptible),FI, F,, BCl, Predominantly non-additive Borikar and Chopde
BC2 (1981b)

Deadhearts incidence
Large FI population from a diallel mating Predominantly additive Balakotaiah et al. (1975)
system involving exotic, Indian and derived
lines
PI (4 resistant), PJ (4 susceptible), FI, F d , Predominantly additlve Sharma et al. (1977)
BCI, BC2
Seven parent diallel (susceptible x Both additive and dominance effects responsible Hallall et al. (1982)
resistant) along with FI for resistance; two recessive genes govern
resistance
8 X 8 diallel (resistant x susceptible) Additive Pate1 et al. (1984)
PI (resistant). Pd (susceptible),FI, F,, BCI, Both additive and non-additive (dominance, Biradar and Borikar
BC2 additive xadditive, dominance X dominance) (1985)
16 F2 progenies from crosses between Two duplicate recessive genes govern the Rana et a/. (1985)
susceptible x resistant parent resistance
PI (resistant), P2 (susceptible),F I , Both by additive and non-additive components Biradar et al. (1986)
backcross involving susceptible line a s
recurrent parent
Eight parent diallet (3 resistant and 5 Additive and additive x additive gene interaction Nimbalkar and Bapat
susceptible parents) (1987)
Seven ~ a r e n half-diallel
t analvsis Both additive and non-additive Dhabholkar et al. (1989)
 Bluni (1965. 1967) inipro\,ed the resistance of M 35-1 by t\vo cycles of mass
selection and successfully incorporated the resistance of the selected line into an
adapted line with good agrononiic characteristics using the pedigree method. Doggett
et 111. (1970) utilizing the recovery resistance ~nechanism available in the cultivar
Namatera in crosses to elite line Serena de\,eloped high yiclding lines possessing
recovery resistance, by adopting the backcross metliod. Based on the largely additive
genetic variance for recovery resistance, Doggett el 111. (1970) established random
mating populations for a long-term recurrent selection program
In Thailand. llarwood e l < I / . (1972) tried to in~pro\~e
tlie shoot Ily resistance in
matcrial adapted to local conditions using resistancc sources like IS 5604. IS 5383 and
IS 4567. I'llrce different approaclles \%eretaken: i) crossing these resistance sources
nith locally adapted varieties; ii) crossing these resistance sources with male-sterile
sccd parents of released hybrids, and i i i ) inter~nat~ng
tlie resistancc sources. There \&,IS
limited success because of various problems cncountered due to undesirable
characters of the resistance sources, which werc ph<~tosensiti\,e.
tall and susceptible to
mold and rust.
In 1068. breeding for resistance to shoot Ily started in India (Vidyabh~~shanaiii.
1972).Initial11 the shoot f l ~resistant and popular local postrainy senwn variety
M 35-1 was crosscd with five susceptible lines. viz.. CK 60B. 22108. IS 84. IS 3691
and 367513 (parents ofreleased rainy seasoll-adapted li! brids). Since the recover! of
desirable F? segregates was very lo\v, the F I 11) brids \\ere hackcrossed to duarf shoot
fly s~lsceptiblelilies. Selection of plants Srom tile bacltcross progenies was done for
t\\o generations. Wile11tested ltl~dcrIienvy shoot Ily inli.slation, the selected progenies
proved to be highly susceptible.
Kao el ol. (1974) recoml~~ended
tllal due to superiority of hybrids of over
parents and additive nature of inheritance. it could advantageously be capitalized in
hybrids and line de\.elopment programs. It \\as also opined (Balakotaiah r.1 01.. 1975
and Shanna er (11.. 1977) that the resistant x resistant crosses did not exhibit an
improvement over the parents indicating no diversity anlorig resistant lines. They have
also concluded that resistance is due to gradual accumulation of desirable alleles
rather than due to the presence of one or ma.ior genes.
Balakotaiah el ai. (1975) observed that the characteristic way in \vIlich the
seedling mortalities due to shoot fly gradually decreased from 65% to 23% in the
order exotics, exotic x exotic, exotic x derivative, exotic x Indian. derivative Indian.
Indian x India11 and Indian continns that shoot fly resistance was due to gradual
accuniulation of desirable alleles rather than due to one or two major genes.
Rana el trl. (1975) opined tliat transfer of resistance to shoot fly. which is
due to ovipositional non-preference, from tlie tall and generally late Indian
varieties to dwarf, semi-dwarf and early-iiiaturing fornis is apparently feasible since
inheritance appears to be largely additive. It was also suggested that the selection of
resistant progenies, which exhibit seedling mortalities one standard deviation belo\\
tlie population mean il~iderreasonable levels of il~festation.
Sliarmn et 01. (1977) also reported that resistance to shoot tly was due to the
gr;~dualaccuniulation of resistant genes of sillall effect. rather than hci~iglargely due
to one or two major genes. 'Shey also reported positive associations orresistance witli
seedling height and pcrfbrniancc per. se of ~esistant barieties for ovipositiotl
incidences, seedling height, effective tillers percentage. plant recover) and yield per
plant This necessitated selection of dwarf and liigh yielding plants fro111 resistant
limilies of susceptible x resistant crosses. It was opined that under such
where absolute resistance is lacking and threshold levels of rcsistance
~ircu~iistances.
depend on shoot fly population, Ion intensity selection pressure should be applied
under reasonable levels of infestation (50-80% shoot fly deadhearts on susceptiblc
controls). Selection for effective tillers or plant recovery />ci. .sc seems to be
unnecessary. being a functio~iof deadlieart formation in the main shoot. tiulkarni er
111. (1078) proposed tliat a shoot fly resistance 11ybl.icl breeding program should
include dwarf female parents having some degree ol'resistance combined with Indian
tali local resistant parents.
In a study of an eight-parent diallel cross (\+it113 resistant and 5 susceptible
lines) Roi.ikar and Cliopde (1981a. b) observed that additi>e variance a sizeable
proportion of total gcnetic variance for shoot fly rcsistance. despite the presence of
non-additive components for traits like plant recovery and eggslplnnt. It was opined
tliat a varietal breeding programme. through tlie exploitation of this tixable
component by adopting a biparental cross approach. uill he rewarding.
Rana et 01. (1985) opined that in tlie absence of an immune source of
resistance. a moderate level of resistance could be build-up in high yielding
backpound. It was further inferred that breeding for resistance to shoot fly is a slow
process. which requires several cycles of crossing to combine higher levels of
resistance with yield potential and grain quality.
 Singh and Rana (1986) observed that the behavior of resistance in F l , F2 a ~ i d
F3 and advanced generations suggested the possibiliky of gradual improvement in
resistiltit x intermediate and intermediate x inter~uediatecrosses, where intermediate
represents the high-yielding deri~ativesof resistant x susceptible crosses. By adopting
such selection criteria in telilperate x tropical crosses. wliich signifies susceptible x

resistant crosses, it was possible to improve the level of resistance and develop a
n ~ ~ m bof
e r high-yielding varieties with adequate levels o f slioot fly resistance.
2.2.2.6 Stability of resistance
Ilosl plant immunity to slioot tly attack being absent. the lcvcl of 'deadhearts'
sylnptolns in susceptible and resistant varieties varies \\it11 seasons. years and
ilil'estation Icvcls (Sitigh and Ra~ia,1986). 'I'llib 11as made it difficult to identify stable
sources ol'resistance amongst the available pool of resistance sources.
Singli er a/.(1978) co~iducted a stabilit! stud) on I5 promising resistant
\nrietics. identified on basis of preliminary screening of the world collection of
sorghuni, in six en\ironments rcprcscnting three crop growing seasons and t\vo
loc:ltions. It was noticed that ~ilostof the genotypes tested were consistent in theit.
slioot fly reactions, but IS 1054. IS 5460 and IS 5400 were found to be tlie most
stable. Borikar and Ciiopde (I 981 a) also reported that IS 5490. IS 4 6 9 and IS 5400 *
IS 5604 exhibited 11igIi degrees of' resistance and greater phencitqpic stability under
tlircc different shoot fly populations.
Chundurwar and Borikar (1983) evaluated 0 fi, dcri\atives of shoo1 fly
resistant x susceptible crosses under four levels of infestation to stt~d! thcir stabilitb
for resistance. O ~ i l yfive genotypes revealed dcadheart ~ncidencelevels (%) at par
~vitli resistant control entry IS 168. regression coeflicients less than unit) and
nonsignilicant deviation from regression; indicating superiority of tliese genotypes
o \ e r this control in respect to the stability of their sliout fly resistance. A~iiongthe
yerniplasm lines tested in Indian Coordinated trials. IS 1082, IS 2146. IS 4664, IS
5470. IS 5566. PS 144454. PS 18061-3. PS 18822-4. PS 21318. 1% 1-2121 and SPV
401 showed greater stability of resistance to shoot fly than IS 1054 (AICSIP. 1984).
Cliundurwar et '11. (1992) evaluated 32 sorghum ge~iotypesto study genotype
x e ~ i v i ~ . o ~ l ~interaction
lle~lt for shoot fly reaction in 4 different sowing dates. The high
magnitude of en>ironmental variance indicated that tlie level of slioot fly population
played a major role and genotypes like IS 2146 and IS 5566 exhibited a high degree
of stability for slioot fly resistance.
2.3 Molecular marker studies
2.3.1 Sorghum SSR markers
Simple sequence repeat- (SSR-) containing clones isolated from both bacterial
artificial cliromosome (BAC) and enriched genomic-DNA (gDNA) libraries and
database secluences that contait~SSRs were the sources for the sorgllum SSRs mapped
by Bliattratnkki er ul. (2000). Targeted isolation of SSR loci using BAC clones as
proposed by Crega11 r t 01. (1999) is likely to be the most efficient mclhod for placing
SSR loci in specific target genomic regions. BTx623 (Frcdcriksen and Miller. 1972) is
the reference genot)pe ilsed for sorghum molecular marker genotyping and it u a s the
source of DNA used to construct [lie enriched libraries and 111c two sorghum BAC
libraries that are currently available (Bliattramakki er (11.. 2000). 1'('1< primers for the
a~nplificationof DNA fragments containing SSRs frc1111~ o r g l ~ i ~\\ere
n i successfullp
dc\eloped through three difl2rent ;~pproachesb> Brown ~r '11 (1006) and it mas
reported that sorghi~mfragluenls can be amplified sing it1 lrasl s o n ~ enliiize SSR
primers (13ro\\n e/ trl.. 1996).
blap locations have been published for nearly 300 sorgl~u~il
SSR loci having
primer sequences in the public domain (CT Hash. pers. comm.). Bhattratuakki er ol..
(2000) rel~orletlmap location o r 46 SSR loci based on previously reported primer
sequences ( ' I aramino r r (11.. 1997: Tao e l trl., IOOXn; K o n g ui ~rl..2000) and 113 SSR
loci (includi~lgTour SSR-containing gene loci) hased on novel primer sequences.
'l'licsc SSR marker loci were incorporated into pre-existing KFLP-haced maps of Xu
FI (11. (1004) (Kong er t11., 1997) and Peng rr L I I . (1999) (Bhattmtnakki el 111.. 2000).
The number of SSR loci available per sorghum linkage group ranged fron~8 to 30.
Eight SSR loci that. although monomorphic among the 1 8 survcy accessions. have
high degree of homology to known genes (Bhattramakki er (11..2000) remained to be
mapped. Tlie average number of alleles detected per locus at the poly~llorphicloci mas
3.88. (AG/'I'C)N and (ACITCi)Nrepeats colnprised the nlajority of t11csc SSRs (529'n)
and 91% of the dinucleotide SSRs at these loci (Bhattramakki er (ti., 2000). The
estimated average probability that two accessions in a working group. would have
different alleles at a locus ranged from 0.88 to 0.67 depending up011 the working
group to which the accessions belong (Kong er LII..2000). In addition, the number o f
alleles per locus was positively correlated (r = 0.68. which is significant at the 1%
level of probability) with the nulilbcr of repeated units at the locus in BTx623, the
strain from which the SSRs were originally isolated (Kong e l (71.. 2000). This
cotifirms that many Sorght/ni biculor SSR loci are sufficiently polymorphic to be
useful in marker-assisted breeding programs (Kotig e t 01.. 2000). First complete
genetic linkage map of sorghi~~ii,co~ii]~risedof tell linkage group putatively
corresponding to the ten gametic choroniosome of Sorghum bicolor and Sorghum
propinquurn.The map i~icludes276 KFLP loci, predo~ninately detected by pstl-
digested Sorgliu~iibicolor genomic probes. segregating in 56 F2 progetly of a cross
betueen Sorghum bicolor and Sorghum propincluuni. The remarkable level of DNA
polyniorphism between these species will facilitate development of a high densit)
genetic iiiap (Chittenden et al 1994) Scliloss (2002) Reported. the RFI-P probes
sequence were evaluated for presence of simple sequence repeat (SSKs) and 60SSI<s
( S C ~series)
I~I were developed atid assayed in all array of sarghutn germplasm
co~nprisinginbreed. land races atid wild relatives. 'She sequence information and SSR
loci generated tllro~~gli
this study will be vnluable in gene disco\,ety. marker assisted
selection, diversity and pedigree analysis.
2.3.2 Linkage maps in sorghum
Genetic studies of morphological traits in sorghum began early this past cetltiiry and
Iloggett (1988) summarized genctic linkage of morphological and physiological
mutants involving 49 loci. To date over 200 mc~rphological and agronomically
itilportitnt markers have been identified (Berlian et (11.. 1993); however, otlly nine
linkage groups could be establislied will1 these markers and these consisted of only 2-
10 loci (Pereira et ~ r l . .1994). 'l'lie biggest linkage group consisted of tell linked
morphological marker loci (Doggett, 1988). Sorghum genome mapping based on
DNA markers began in the earl) 1990s and since then several genetic Innps of
sorglii~mhave been developed with large nitmbers of DhA-based niat.kcrs including
R121.Ps. AFLPs and SSRs. Where opportunities have permitted. morpliological marker
loci have beeti integrated into these ~nolecularmarker-based getictic linkage maps.
'Tlie~e[naps will be i~sefulin advanced breeding and genetic studies.
The constructioci of the first DNA-based sorghum linkage niap was done using
rhe KFLP technique with heterologous maize probes (Hulbert et (11.. 1990). Later
several tilore RFLP-based lilikage maps of S hicolor have been co~istructed(Binelli et
01.. 1992: Wllitkus e t ol., 1992: Berlian et a/., 1993; Chittenden c / 01.. 1994; Pereira er
(11.. 1994; Ragab et 01.. 1994; Xu er 01.. 1994: Dufour e t 01.. 1997: l'ao er (11.. 1998a:
Peng et "I., 1999, Haussmann er 01.. 2002: Bowers et al., 2003). Similarly, the RFLP
maps of Xu e t a / . (1994) and Peng e t irl. (1999) have been improved with addition of
over 100 SSR markers (Kong er LI~., 1997: Bhattramakki e l (rl.. 2000). while that of
Dufour rt (11. (1997) has been augtnented with AF1.P markers (Hoivin er a/.. 1999).
Rece~ltlyhigh-density genetic lliaps using AI'LP. RF1.P and SSK markers (Menz er
111.. 2002) and KI'LP probes (Bowers tr (11.. 2003) have heen reported. These liigll-
density integrated maps will accelerate gelio~nemappiilg and comparative napping
activity in sorghum and other related gross species. Tlie characteristics of differeilt
sorghum genetic maps are given in 'Table 2.3.
2.3.3 Marker-trait : ~ s s o c i e t i o ~ ~ s
Qiiantitative characters have bee11 a major arca of genetic study for u \ e r a century
hecause they are a colnmon feature of natural varint~on in populations of' all
euliaryotes (Kearsey and Falqullar. 1998). First attempts at studyii~gthen1 stemmed
froin the nark of Gillton (1x89) on nian bcforc the rediscovery of Mendelian
inheritance of quantitativc characters through the pioneering work of Fischer (1918).
\\.hicli Ins been followcd up by \+'right ( 19341. Metlies (19-19) and Falconer (1989) to
the new cra. Despitc these studies, the number of gencs and their illteractive effects
controlling the expression ol'quantitative traits are poorly understo~~d.
The basic co~iccptof associating gc~iutic~uarkersu i t h quantitati\e traits was
lirst proposed hy Sax (1923). Since then there has been great inifrest in genetic
dissection of cluantiative variation. Geneticists have ~.ecogniredthe potential use of
li~lkagesbetween quantitative genes and Q'L'L for studying the nature of quantitative
genetic variation (Sax. 1923: Lindstro~n.1926. 193 1 : Waxclso~l.1933: licerson and
Schaller. 1955:and Thoday. 1061) Ilnfurtunately the relativel> s~nollnulnbers and
nature ofclualitati\e marker genes \vas cxtrernel) li~nitingfor
sometimes-deleterii~t~s
li~thagestudies \+ith quantitative genetic variation (Hul~eckPI crl., 1003).
Anal!sis of biochemical and DNA markers in crorses betneen parelits that
differ for a quantitative trait can be used to find ~narkerslinked to getics controlling
the q~~antitative
traits or QTLs (Gale and Witco~nhe.1992). In plants the lirst attempts
lo use markers to perform genome-\vide analbsis of rl~iantitative variation used
allozyti~es(Tanksley et (11.. 1982: Edwards e/ 01.. 1087). Later RFLPs were used as
DNA markers (Beckman!? and Soller. 1983: 1,ander and Uotstein. 1989), but these
u e r e followed by PCR markers such as IIAPDs. nlicrosatellites and AFLPs that were
cheaper. safer and provided more nlarker data per unit of DNA (Westman and
Kresovich. 1997). These polylnorphic markers pro\,ided the framework maps with
which the polygenesiQTLs could be located (Kearsey and Farquhar, 1998).
cont.. . - -- -

Size and Genome

Reference Parents type of Markers Length LG Probe sources
~oiulation -
Bovin et al., 1999 IS 2807 X 3793 110 F, RILs 2 9 8 RFLPs, 137 AFLPs Sorghum, crreals
Crasta el al., 1999 B35 X RTx430 5 6 Fb - RILs 142 RFLPs Sorgl~unl,ccrcals

Peng el of., 1 9 9 5 BTx623 X IS 3620C' 137 Fh-a RlLs 3 2 3 RFLPs Sorghum, cereals
BTx623 X S 2395 locl based o n 1525
Bowers el al., 2 0 0 0 propcr~quum' 6 5 F2 RFLPS Sorghum, cereals
Kong et al., 2 0 0 0 BTx623 x IS 3620C3 I 3 7 Fo a RlLs 11 RFLPs, 3 3 SSRs Sorghum, cereals
Bhattramakki ef al.,
2000 354 RFLPs, 1 4 3 SSRs Sorghum, cereals
Sorghum, ccreals,
Tao et al., 2 0 0 0 QL39 X QL411 152 F5 RlLs 281 RFLPs. 2 5 SSRs
sugarcane
835 X Tx700 9 8 Fi RlLs Sorghum. maize
Framework m a p d c r ~ v e dfrom comparison

of the m a p s of I<ong ef al. (20001, Peng el 154 RFLPs. 3 4 SSRs, 1 0

Sorghum. cercals
al. (1999). Pereria el 01. (1994) and Berhan morphological markers
el ul (1593)
44 SSRs, 8 5 AFLPs, I
RTx x Sureno 1 2 5 F5 RlLs Sorghum
morpholog~caimarkcr
125 AFLPs. 4 5 SSRs. 14
Haussmann el ul., 2002 IS 9830 X E 36-1 2 2 5 F3; RlLs Sorghum
KFLPs, 3 fC\PDs
158 AFLPs, 5 4 SSRs, 16
Sorghum
RFLPs
3 3 5 AFLPs, SSRs, RFLPs
Cornposrte map of thr two pupulatlons 1424H II Sorghum
and R-2PDs
 Several statistical approaches have been developed for detectillg alld
the strength of these associations between markers a ~ l dtraits (Soller a ~ ~ d
Urody. 1976; Edwards et al.. 1987; Lander and Botstein, 1989: Knapp. 1989). The
ability to detect a QTL with a marker is a function of the tllagnitude of Q fL's effect
on the character. the size of !napping population being studied and the recombination
frequency betjveen the nlarker and the Q f L ('l'anksley r t ill.. 1989). I'he realized QTI,
effect is a function of how large an efSect the Q 1.1. has and how tightly it is li~lkedto
the niarker or flatikitlg markers (Ciale and Witcornhe. 1902). There arc. ho\vever
dangers associated with the establishment of breeding progranls based on correlations
of marker genotypes with quantitative traits bcfore the identified factors (QTLs) have
been tested in several genetic backgrounds and evaluate for associated effects on other
characters of agronomic or economic importance (Tanhsleq and Ile\vitt. 1988).
It is well understood b! plant breeders that genotype x enLirotlment ( G x E)
interactions exist for many quantitative traits. suggesting that general conclusions
ahout Q'l'Ls. particularly those with s~nallei'fects Jctcctcil on the hasis of single
cnvirotit~~etits
and single populations could lead to crroncous clecisions. 'The use of
0 I'L itlcntiiicatiotl by breeders also will be influe11ct.d hy the consistency of QTL
regions acl.oss the gernlplasm (Bubech r l (11.. 1993). One challenge of plant breeding
is (a take advantage of favorable direct eSfects of QT1.s. \\l~ile~ n a s i ~ n i r i nfavorable
g
oncs (Bubech el 01.. 1993).
environmental ititeractions and minimizing unft~\~olable
A greatly abbreviated list of agronomic traits sut>jectetI to marker-based
mapping and Q'fL nnal>sis includes drought tolerance (hlartitl et rrl.. 1989). seed
htirdness (Keim el trl.. 1990). seed size (Tatoku~lcr 01.. 1992). rnatt~rilyand plant
I~cight (I.in c / <I/.,1995). disease resistance (revie\ved hy Yott~lg. 1996). oil and
protein content (Diers er trl.. 1992). solublc solids (Tanksley and Hewitt. 1988) and
yield (Stuber rr L I ~ 1987).
.,
2.3.4 Statistical techniques for QTL analysis
Q'fL analysis is predicated on lookitlg for associatioils beticeen the trait

and the marker alleles segregating in the 11ia17pi1lgp o p t ~ l a t i ~It~has

~ . two esse~ltial
stages: the nlapping of the markers and association of the trait with the markers. Both
of these require accurate data and statistical software ( K e a r s e ~and Farqullar. 1908).
The basic tlleol>r underlying marker ti~appit~g
has been available since the I920s
(Mather, 1938). but has to be extellded to handle l~ttndredsof markers silnultaneously.
The availability of computer software packages has made this niuch easier (Young,
2001).
The traditional approach (Soller and Brody. 1976: fanksley rt 01.. 1982;
Edwards rt crl., 1987) for detecting a Q'I'L in the vicinity of a marker involves
studying single genetic markers one at a time. Ho\vever. il'the Q'fL does not lie at tlie
marker locus, its phenotypic effect diniinishes relative to the true effect of the Q'TT, as
the distance (recombination frequency) increases het\veeli the marker locus and the
QTI. (Edwards el 01.. 1987: I,atider and Botsteitl, 19x9). To olrrcome this, Knapp
(1'189) developed an approacli that utilizes pairs of markers in a sequential manner
and estimates the phenotypic eFect of tlie QTI. ant1 its signitic:unce in the region
bracketed by tlie two markers i n each pair. Lander and Botstein (1989) rcportcd
d r \ e l o p n ~ e n to f such a method of mapping QTl-s. interval ~nappitig using [.OD
scores. Intervals between adjacent pairs of markers along a chromoso~iieare scanned
and the likelihood profile o f ' a QTL bcing at any particular point in each interval is
Jctcl.niined: or to be more precise. tlic log oC the mtio of the lihelihoods (LOD) of'
there being one QI'L 1,s no VTL at a particular lpoint is determined (1,nnder and
Botstein. 1089). An alternative approach using multiple regresqion \+as de\,eloped h)
I laley and Knott ( 1 992). It often produces ver! siniilar results to L01) tilapping both
in terms of accuracy and precision. but has the advantages of spced arid simplicity of
p~ogl.a~nming.Tests of si$nificance and confidence intcrvals can he obtained.
l'anksley and Nelson (IYY6) ad\ise that the sl;~tistical detection o f QTLs is likcly to
depend not only on the type of populatio~iutilized. but is also like]? to depend on the
intra-locus and inter-locus interactions of the segregating QTLs.
For most mapping projects the most widely used genetic mapping softwarc is
MAI'MAKER (Lander rt (11.. 1987). MAPh,lAKIIR is based on the co~iccptof the
L.OD scorc. ..the log of odds mtio" (Morton, 1')iS). The popularit!. o f MAPMAKER
is based on the ease n i t h \vl~ichit perl'ornls multipoint analysis of Inan) linked loci
(Young, 2001). Tlie computer program .IOWMAP is especiall) suited to relate one's
map to tliose derived from other mapping populations (Stam, 1993).
T o apply linkage maps to QTIJ analysis. MAPL\4AKER1Q'TLhas been written
to carny out si~iipleinterval mapping (SIM) QTL. analysis using matl~ematicalniodels
and interfaces very ~iiuch like tlie original MAPPVIAKER program (Lander and
Rotstein. 1989). Other programs like QTL Cartographer (Basten et (11.. 1998) provide
very nlllch the satme type of analysis. QTL analysis can also be perforlned by using
co~nposite interval mapping (CIM) with the PI-ABQTL software as described by
Ranii et ul. (1998) or \*it11 QTL Cartographer. For large-scale use of linkage
illformation ill a marker-assisted breeding. a prograni like Map Manager (Manley and
Cudmore, 1998) helps to keep track o r marker data in the population of interest.
Hypergene (Young and Tanlisley. 1989) or Graphical (;enotypcr (GGT) can lielp to
display graphical genotypes. '('lie prograni qGENE seeks to bring all of these
i~iiportantDNA ~iiarkertools together into single package (Nelson, 1007).
2.3.5 QTL mapping in sorghum
Numerous studies to identify Q'fLs for agrononiically important trait\ liave been
conducted in sorglium and QTLs liave been itle~itiliedl i ~ ra wide arl.ay of important
traits (Table 2.4). 'This work has bee11 important ill impro\ing our utiderstanding of
tlie genetic inheritance of specific traits and tlie best breeding approaches for tlie~ii
(Rooney. 2004). Aduption of uthcr molecular tcchnulogics is important und is being
tested. Markers detected for sinlply inherited traits sucll as maturity. height and
fertility restoration have been identilied and tested fix tile applicability of MAS
schenies. 'l'liese tests liave had varying depees o f success. Q 1'Ls h a ~ cbeen idcntitied
for drought stress (pre- and post-flowering), grain mold resistance. grain yield, and
grain quality. Coulibalq (2002) \+as i~nsucccssii~l
ill using C)'f1. marhcrs to iiitrogress

post-llowering dl-ought stress fro111donor parent U35 to several elite iiibreds. Fraiilts
(2003) had liniited success in using marltcrs flanking Q'1'L.s for grain mold resistance
to enllaoce grain ~noliiresisracicc: they mere effecti~cill progenies \r.ith the exact
saliic pedigrec in which the Q f L s \vcrc ]napped. but they \vere no1 errective in all>
otlier population. Tlie potential remains for the use of ~~iarkers
fur siniply inherited
traits tbr introgression rir pyran~iding of' traits. but there have heen 110 reports
published to document their use in sorghum (Roo~le!. 2004 1.
Table 2.4 Summary of qualitative and quantitative trait loci identified in
sorghum
Trait Reference
Drought tolerance (pre- and post- Tuinstra et a/. (1996, 1997), ~ r a s t ael a/.
anthesis) (1999), Subudhi et JI. (2000), Tao et al.
(2000), Xu et al. (2000), Coulibaly (2002);
and Haussmann et at. (2003)
Anthracnose resistance Boora el al. (1998) and Mehta (2002)
Rust resistance Tao et al. (1998b)
Head smut resistance Oh el a/. (1994)
Downy mildew resistance Gowda er al. (1995) and Oh el al. (1996)
Maturity Lin et al. (1995) and Childs et al. (1997)
Height [.in el a/. (1995) and Pereria and Lee (1995);
Klein et al. (200 la)
Yield and components Pereria el al. (1995), Tuinstra et al. (1997),
Rami el al. (1998) and, Sanchez-Gomez
(2002),
Grain quality and mold resistance Rami et a/. (L998), KIein et a[. (2001a), and
Franks (2003)
Leaf blight resistance Boora er al. (1 999)
Fertility restoration Klein er al. (2001b)
Pre-harvest sprouting resistance Lijavetzky et aL (2000)
SIriga resistance Haussmann el al. (2004)
Greenbug resistance Agrama el al. (2002),Katsar el al. (2002),
Nagaraj el al. (2005)-
Midge resistance Tao et al. (2003)
Shoot fly resistance Folkertsma et a1.(2005) unpublished;
Sajjanar (2002). Deshpande (2005)
Tillering Paterson el al. (1995)
Seed size and dispersal Paterson er al. (1 995)
2.3.6 QTL mapping for insect resistance in cereals
Like other cjuantitative traits, inheritance of resistance to a nutnber of insects ill

cereals is polygenic (Khusli and Brar. 1991). Phenotypic selection for such traits is
difficult. Selection based on markers could theoretically ease the matiipulation of such
traits without affecting other agro~iomictraits. blolecular tilapping experinients for
qllantitative insect resistance in maize. sorghum. ricc. \\heat and barley have heen
conducted and tlie details are 11resented in l'ablc 2.5. The mapping population types
generally used were F23. RlLs and doubled Iiaploid lines (DH1.s). The sire of
popul~ilions used varies between 71 (RII q) and 475 Signilicant ()XI:
i~itelactionwas observed for rcsialance to corn borers in terms of lcar feeding ratcs
(.lampatong tZ/ol., 2002; Uohn pr 01.. 1096: Rollti ei (11.. I997 ant1 Groh ci rrl.. 1998).
This indicates tlie influence ot'environment on the expression of resistance traits.
Taking cognizance ol'the low po\ber of (]'I'L dctcction for small sample sizes
(*,300) found in simulation studies (I!tz and 'vlclchinger, 1994). sever;^! reasonably
large sized RII. !napping popl~latiotisIiat'e been rlevelope<l in sorgliom at ICRISA I'.
Patancheru. 'l'llrse are being screened for resistance to rlioot fly, midge iuld sten1
borer.
Among the cereals. cxtensi\'e QTL ~llappingsxperiments \\ere done in niaize
for resistance to dillkrent species of corn borers. A cotn~iionl>held view is that rncli~c
n:lturc. A sufticiiritl>
is exceptionally poiymorpliic. due to its highly cro~s-pollin:~ted
large number of polynlorphic RI:LP loci call be found Sor m a i ~ ein intraspecilic
crosses in colitrast to many othcr crops \\hcrc interspecific crosses are llsetl to
obercome lack of nlarker pulytilorphisnl \\ithi11 thc cul~igen. 111 addition. lnrgc
nu~nbersof RFLPs that llavc already beell mapped in tnaiZe genome are publicly
available (Bohn et '11.. 1996). In case of sorghum. sorghutn RF1.l' li~tkagem;lps
(Subudhi and Ngu)e~i. 2000) and an integratetl SSR illid RF1.P linkage ]nap
(Bhattramakki e/ trl., 2000) are available (Ilnussnlantl r / cil.. 2002). These in turn 11me
been suppleniented by AFLP markers (Menz t,t c11.. 2002) and III-LP ~iiarkersfrom n
wide array of gralninaceous crop species (13owers et al.. 2003) to provide higli density
base tilaps for sorghutii. The markers on these maps are of potential use in mapping
sorgliulll genome regions associated with resistance to shoot fly, sten1 borer alld
midge.
 ,ping ...for insect resistance in cereals No. of
Size of No. of
crop I Pest cross

873 B52
population
F23
mapping
population
300
endronmnts
evaluated
Two locat~ons
~~a~~~~~
Tunnel length
idez:;ed
7
QxE
I,,temNion
Non-
Reference

Schon et aL

1
X
corn borer ECB tunneling significant (1993)
(Ostnnia 873 X B52 RILs 200 Pour Leaf feedlng 9 Cardinal et al.
nubilalis rates (2001)
HCibner) B73Ht x F2.3 244 Three Tunnel length 9 Sign~ficant J m p a t q ef
Mo47 Tunnel length al. (2002)
Three Tunnel length 7 Sign~ficant Krakowsky ef
al. (2002)
De811x F23 147 Three Krakowsky et
B73 I aL (2004)
DeSll X RlLs 191 Three / 10 Sign~ficant Cardinal and
Lee (2005)
973
j
CML131 X F23 171 TWO Leaf feeding 10 Significant Bohn et al.
corn borer CML67 rates / (1996);Bohn et
al. (1997)
Southwestern CML131 x F23 17 1 Tu.0 seasons Leaf feeding 6 S~gn~ficant Bohn et al.
corn borer CMLO7 Three seasons k a f feeding 9 (1997)

1
(Dzalrea Protein
grandiosella CML131 X RlLs 187 One season ronc?ntratlon 5 Significant Groh et al.
CML67 Leal (1998)
I ! 145 ! One season tou~hness 7 , -
~ e aferding
i
rates
Leaf feeding
Two seasons Leaf
toughness
One season
(Table 2.5 contd ...)

Crop Pest Cross Mapping

population
Size of
mapping
population
No. of
OTLs
identified
PXE
interaction
1 Reference
Rice Brown plant
hopper
(Nilaparuata
lugents)
7-
Minghui 63
250 l Mechanisms
of resistance
(antixenosis,
antibiosis
and
2
al. (2001)

tolerance)
Lemont X RlLs
Teqing

Yellow stem
borer
lR64 x
Azucena
DHLs

and white-
1 Soundarars
Janet al.
(2004)
Selvi et al.

II
incertulas

Green
leafhopper
(Nephoteniu
1 uirescens
I Distant)
/
I
Taichung65

1
FIORils
1
/
/
Antibiosis

Barley Cereal aphids Harrington DHLs

X TR306 pour et al.
 Correlating a genetic nlap to tlie physical map \vould be highly valuable to
plant geneticists for map based cloning of genes responsible for a particular QTI-.
Recently. an attempt has been made to locate molecular markers (r,r~icl05aon the
sholt alyn of cliromoso~ne9, c.si1145n on the long arm) that flank QTLs for resistance
to sugarcane corn borer (SCB) and southwestern corn borer (SWCB) in niaize (Sadder
and \Veber, 2002). It was suggested that further polymorphic DNA sequences have to
be identified before attenipting to isolate these QTLs.
2.3.6.1 Slioot fly resistance coniponent traits QTL mapping in sorghum
Scijanar (2002) and 1:olkertsnia t.1 trl. (2005 unpublished) genotyped 252 recotiibinallt
inbred lines (RILs) of a (RTx623 x IS 1855 I)-derived ~iiappingpopulation using 109
SSII markers, l'he genetic linkage map \\.as constructecl using JOINbl~II'versions 2.0
and 3.0. resulting in tlie formation of 10 linkage grtiups \\it11 a total map length of
1468 cbl. OIL analysis using I'labQTI, revealed tlie prcsence of 28 QTLs detected at
least in two of three screening environnients (four QTLs for seeclling glossiness score.
two Q r L s for seedling vigor I. t i w Q I Ls for seedling ~ i g o IrI , two Q'1Ls for ahaxial
leaf' surface trichomc density. three QSLs for adasial leaf surfi~lcetricliome density,
two QTLs for slioot fly ovipositio~i incidence 14 days after seedling elnergeticc
(DIII;). one QTL for shoot fly oviposition incide~lcellDAE. four QTLs for shoot tly
cleadhearts incidence 21 LlAE. three QTLs for shoot fly dec~dheartsincidence 28 1)AL
and cine QTL for seedling height I). Markers have heel1 identified closel! linked to the
h u r dcadhearts resistance QTLs. Tlley will be used in ~narker-assistedhackcrossing
programs at ICRIStYI' and biAll-Parhhani.
Ileslipande (2005) gellotyping 213 KI1.s of Z96B X IS 185.51 mapping
~xqx~lation
using 114 SSR niarkers. 'l'lie genetic linkage map lias been constructed
using MapmakerIEXP 3.0 with tlie LOD threshold value at 3.0 and linkage distance
(ch4 units) calculated using tlie Ikildane (1919) niapping fi~nction.Markers \+ere
nlapped in 10 linkage groups with a total map length of 2165.8 cM. QTL analyses
perhrmed using composite interval ~iiapping(I'lahQTL version I . I ) revealed the
presence of 13 QTLs detected across two environments for important shoot Ily
resistant traits including seedling glossiness score ( 4 QI'Ls), seedling vigor score 1 (2
QTL). seedling vigor score 11 (1 QTL), deadhearts incidence (%) 28 DAE (1 QTL),
seedling lieiglit I (1 QTL). seedling height I1 (1 Q-fL). tricliome density of upper leaf
blade surface (1 QTL). tricliome density of lower leaf surface (2 QTLs).
2.1 Marker-assisted selection (MAS)
This section gives a detailed literature overvie\+ of diffirent topics that deal with the
study of marker-assisted selection in general. and for disease and insect resistance in
crops in particular.
Marker-assisted selectio~i(also referred as 'marker-assisted breeding') may
greatly illcrease the efficiency and effectibeness of plant breeding compared to
co~lventio~lal
breeding methods. Once markers tliat are tightly linked to genes or
QTLs of interest have been identified. prior to field evilluation of large numbers of
plants, breeders may use specific DNA marker alleles as diagnostic tools to identifq
plants carrying the target genes or QTLs (Michelmore. 1995: Kibaut er (ti.. 1997:
Young. 1996). The advantages of MAS i~icludea) substituting for coniplex lield
trials (tlial need to be conducted at particular tinies of year or at ~pecificlocations. or
arc technically co~iiplicated)with ~ilolcculartests helps in sabing tiiiie and elimillating
unreliable phenotypic evaluation associated \bit11 held trials due to environmental
effects: b) selecting genotypes at seedling stage: c) gene 'pyramiding' or comhi~~ing
multiple genes simultaneously: d ) avoiding the Iransrer of i~iidesirableor deleterio~~s
genes ('negative linkage drag': this is of particular relevance Sor introgression of
genes fiom uild species): e ) selecting for traits with low heritability: alid I) testing I'or
specific traits \rliere phenotypic evaluatio~iis not feasible (e.g.. cl~~arantine
restrictio~is
Iney prevent euotic pathogens to bc used for screening).
In MAS the liglit linkage of ~iiarkersto a gene if interest is exploited for.
indirect selectioli of traits in a breeding programlne. Two pre-rcqi~isitesfor adopting
MAS in plait breeding progranls arc:
I. one or more lliarker loci tightly linked to the gene of interest. and
2. a population that is polylnorphic for the marlter(s) and gene of interest. which are
in extreme linkage disequlibrium.
There are at least three possible approaches to applying MAS in plant breeding:
(a) selection based on ~narkersalone with no measurelnents of phenotype:
(b) simultaneous s e l e c t i o ~ l omarkers
~~ and phenotype: and
(c) two-stage selection with the first stage involling use of markers to select alnolig
the genotypes and second involvi~igphenotypic selection among the previously
selected genotypes.
 The potential efficiency of marker-aided selection depends on tlie heritability of
tlie trait, tlie proportion of genetic variance explained by tlie markers, atid the
selection method used.
MAS an important plant breeding tool in mhicli molecular biology can be
applied to transfer traits from donor parents to recurrent parents. MAS lias been a
usefill tool for facilitating rapid generation advancement in case of application of
Q'1'l.s in breeding progranis (Lande and 'l'liompson. 1000: Knapp. 1994 and 1998).
Ginelfarb and Lande (1995) presented detailed analysis of the relationship between
genetic markers and Q'TLs in the process of MAS. Molian er 111. (1997) concluded that
MAS could be used to pyramid major genes, including disease and ir~sectresistance
genes, with the ultiniate goal 01' producing crop cultivars with larger numbers o r
desirable traits. A study conducted by Eatingtoti el (11. (1997) assessed tlie usefilltiess
o f marker-assisted effects estimated r r ~ t i early
i generation testcross data for predicting
later generations testcross pcrfortnance.
MAS can be used to pyramid several segregating resistance genes into singlc
host cultivars where hybrids are possible M'itcombe and Ilesh (2000) l~avedescribed
how pmcticnlly to strategically deploy resistance genes in a potentially more durable
manner has been previously been pmcticcd. by exploiting tlie ability of MAS to
introgress niultiple resistance genes into a common hybrid seed parent background.
atid then intermating the products to produce agrononiically uniform cultivars that
segregate Cos multplir resistance genes. The frequent! of genotypes liaving resistance
alleles at sever:~l loci increases greatly in both the seed parent and its hybrids when
the o\erall Srecluency oFresista~iccalleles in (lie maintainer li~ie(s)increases.
The ability to manipulate genes responsible for quantitati~e traits is a
~xerequisitefor sustained improvement in crop plants. MAS in pedigree. backcross
and population improvement breeding methods is especiall! l~sefulIhr traits tliat arc
other\cise difficult or impossible to deal with by conventional tiiealls alone (Hash and
Bramel-Cox, 2000). 'fliere lias been an implicit expectation tliat marker-based Q I'L
analyses will make it easier and faster for breeders to manipulate these traits (Soller
and Beckmann. 1983: Tntiksley. 1983). but this expectation has often not been
realized-in large part because of tlie empliasis in researcll on niodel systems and
subsequent difficulties in extrapolating fro111 such lnodels to more complex (and less
well understood) applications.
 The development of linkage maps with abundant markers in a wide range of
crop species was accelerated by development of newer and simpler DNA marker
systems like RAPDs (Williams e l al., 1990). AFLPs (Vos ei al., 1995) and SSRs, also
known as microsatellites (Akkaya el al., 1992). Scientists soon began to believe that
the promise of MAS originally proposed by Sax (1923) and Thoday (1961) might
soon become a reality (Young, 1999). Analyzing plants at the seedling stage,
screening multiple characters that would normally be epistatic with one another,
drastically minimizing linkage drag, and rapidly recovering a recurrent parent's
genotype in genomic regions distant from genes that are the target of introgression
were some of the potential advantages of MAS (Tanksley el al., 1989).
In order to tag any gene of interest with selection fidelity of 99%. Tanksley
(1983) observed that it would be necessary to have marker loci spaced at 20-cM
intervals throughout the genome. Selection can be exerted for a number of markers
simultaneously, which will have the effect of selecting for QTLs with positive effects
on the quantitative trait of interest (Paterson er al., 1988). However, one of the major
drawbacks is that when the linked marker used for selection is some distance away
from the gene of interest, this permits crossovers to occur between the marker and the
target gene. This produces a small percentage of false positives/negatives in the
screening process (Mohan el al., 1997). Therefore, in the final analysis, the success of
MAS will depend on identifying highly polymorphic marker(s) as close to the target
gene as possible to ensure itdtheir utility across many breeding populations (Mohan el
al., 1997).
For efficient MAS some additional QTL mapping steps have been suggested
by Young (1999):
1) repetition of phenotyping over several years and locations,
2) repetition of combined genotyping and phenotyping in a larger sibling population,
3) repetition in genetically unrelated populations, and
4) detailed analysis in marker-generated near-isogenic lines (NILS) that isolate the
effects of individual QTLs.
Marker-aided selection has been well demonstrated in traits that are largely controlled
by major genes, such as blast resistance (Hittalmani ei al., 1995). gall midge
resistance (Nair e l a/. 199Sa) and semi dwarfism (Cho ei al., 1994) in rice. However,
the utility of MAS in manipulating quantitative traits was presented by Dudley (1993)
in his paper on the potential of molecular markers in manipulation of genes affecting
quantitative traits. Stuber el al., (1987) reported the exploitation of MAS in
quantitative traits manipulations for maize improvement, demonstrating the
effectiveness of marker-based techniques for identifying and locating QTLs and for
detailed genetic investigation of quantitative trait variation. He reported more precise
mapping of QTLs in several plant populations and multiple trait associations within
specific genomic regions. Stuber (1994) demonstrated the transfer of QTLs using
MAS for improving the yield level of maize inbred lines.
Breeding of insect and disease resistance and tolerance to abiotic stresses has
become a worldwide issue for crop improvement. To identity the insectidisease
reaction of breeding materials, plants must be inoculated artificially or naturally or in
specific environments where the biotic stress exists. Artificial inoculation may be
impractical when the insect pest or disease is under quarantine control. However,
evaluation of plant response to different insects or diseases or different
biotypes/strains/races of the same stress agents is often very difficult. Using molecular
markers associated with each of the stress responses will help select for resistance to
multiple insect pests, plant diseases, or variants of these without inoculation or
creating the specific screening environment required for conventional phenotypic
screening. Similarly, plant response to multiple biotic stresses can be predicted
simultaneously using molecular markers associated with tolerance or sensitivity to
these stresses. There are several successful examples of u 'ng MAS to select for
resistance to biotic stresses in rice. For example, Hittalmani er al. (2000) used marker-
assisted selection to combine three rice blast resistance genes (PII, Pi2-5, PI-TA) into
a single genotype. For PR-5 a single marker was used where as flanking markers
were used for the other two targeted host plant resistance genes. MAS was effective in
developing a resistance gene pyramid in line containing all three resistance genes. The
product breeding line with this resistance gene pyramid had a broader resistance
spectrum than lines with only one of the three resistance genes. Huang el al. (1997)
pyramided four bacterial blight resistance genes (Xa4, Xa5, Xu13 and Xa2l) using
PCR-based markers. Sanchez et a/. (2000) transferred three bacterial blight resistance
genes into a susceptible rice line possessing desirable agronomic characters. Ribaut el
al. (1999) identified five QTLs for drought tolerance that were stable over across
several drought stress environments, and successfully transferred these to an elite but
drought-sensitive line in maize. Shen er 01. (2000) at lRR1 reported that after QTLs
affecting root parameters were identified using a rice double haploid population
derived from cross IR 64 x Azucena, a marker-assisted backcrossing program was
started to transfer the alleles of Azucena (a drought tolerant upland rice variety) at
four QTLs for deep roots (mapping on rice chromosomes 1,2, 7,9) from selected DH
lines into IR 64. The resulting breeding products showed significant improvement of
root mass and root length. Marker-assisted selection for QTLs controlling the stay-
green trait (a component of terminal drought tolerance) in sorghum is in progress at
ICRISAT-Patancheru (Hash el al., 2003). Six QTLs of relatively large effect from
donor parent B35, which have been independently mapped by two or more groups of
earlier workers, are targeted in this program, with agronomically elite and genetically
diverse sorghum varieties R16, ICSV I I I, IRAT 204 and ISlAP Dorado as recurrent
parents.
Molecular marker based QTL analysis in tomato demonstrated that QTLs
isolated from wild germplasm can improve phenotype of c mmercial varieties for
many economic characters (de Vicente and Tanksley, 1993; Eshed and Zamir, 1994;
Zamir and Eshed, 1998) as a result of which specific QTLs for increased yield and
soluble solids were transferred to cultivated tomato varieties. Xiao er al. (1996b)
demonstrated that wild rice species 0. rufipogon carries favorable alleles at two
QTLs, which increase grain numbers per plant and thus have potential to substantially
increase yield of rice. A slow growing wild relative of the cultivated tomato,
Lycopersicon pennellii, has been observed to have genes for increased rate of dry
matter accumulation and 'soluble solids' concentration.
2.4.1 Efficiency of marker-assisted selection
The analytical approached of Lande and Thompson (1990) focused on first generation
selection. Succeeding studies have focused on the efficiency of MAS over several
successive generations using computer ;imulations (Zhang and Smith, 1992, 1993;
Gimelfrab and Lande, 1994a,b, 1995; Wittaker et al., 1995). Results from these
studies showed that MAS could be more efficient than purely phenotypic selection in
quite large populations and for traits with relatively low heritabilities. The simulations
also showed that additional genetic gain provided by MAS, when c ~ m p a r e dwith
purely phenotypic selection, rapidly decreased when several successive cycles of
selection had occurred, and that MAS may become less efficient than phenotypic
selection in the long term. This situation becomes more acute when the effects
associated with markers are not reevaluated at each generation. MAS was as effective
as phenotypic selection for developing populations with diverging grain yield (Stuber
and Edwards.1986). MAS with an index of 34 sweet corn traits was as effective as
phenotypic selection (Edwards and Johnson. 1994). I11 common bean, Schneider et al.
(1097) i~sedtive RAPD markers for MAS of ?ield in a drought-stressed eilvirolunent.
MAS improved yield performance by 1 1 percent and 8 percent under stress and lion-
stress conditions, respectively. hut tlie conventionnl selection for yield failed to
improve performance under stress.
The efficiency of the backcross method in transferring QTLs will be govenied
by tlie niagnitude of linkage drag and correct identification o f QTL-marker
associatioils during the process of Q'rI mapping. The current Q'l-I. mapping
techiiology maps Q'1'l.s witllin gelloniic segments of 15-20 cM. \vllicli increases tlie
probability of linkage drag. Further, large numbers of t'alsc positive associatioils
('l'ype I error) are lnore important than failure to itlentify the real associations (Type I1
error), because tlie marker-bascd selection in the former case becomes an exercise in
rutility (Dudley. 1993). Another errw (Tlpe 111) emerging from detection of
significant association of a QTL. with a wrong marker is even more serious. QT1.s
~nnpped\%it11stringent levels of signiticance and high threshold values that arc based
on the size of the genome. using fully-saturated genetic liiaps \\ill be more expensive
to gelleratc. but are expected to improve the efficiency of ~i~i~rker-assisted
QTL
tl.ansfer tllrough backcrossing.
Hospital et (11. (1997) used computer simuI;~tionsto study the efficiency of
M,4S based oli an index combining the phenotypic value arid molecular score o i
intli\'iduals. Tlicy obher~edtIi;~tin tlle first generation t11c relatibe erficienc! (RE) of
expected efficiency of MAS over tile expected el'liciency of purel> phenol!pic
selectio~igenrlally incrrases wit11 I ) larger llopulatio~i size, 2) lower lieritability
\'dues of tlie target trait, and 3 ) high type-I error risk. Their studies showed that
higher efficiency of' MAS for lixation of favorable alleles at QTLs \\it11 large effects
in early generations is balanced over successi\,e generations by a higl~errare of
fixation of ~nlfavorablealleles at Q'fLs with small effects in later generations. This
explains why MAS niay become less efficient tlia~iphenotypic selection in the long-
term. MAS efficiency therefore depends. at least in part. on the genetic determination
of that trait.
The efticielicy of MAS generally reduced with increasing distance between
markers. So, the optimal distance recommended between two adjacent markers
flanking a particular target QTL is about 5-10 cM (Hospital er ai.. 1997). However.
the efficiency of marker-assisted selection is less efficient than the phenotypic
selection in the long-term (Hospital er al., 1997) if there is linkage between favorable
alleles of large effect and unfavorable alleles of small effect in the genomic region(s)
subject to marker-based foreground selection.
Knapp (1998) presented estimates of the probability of selecting one or more
superior genotypes by MAS to estimate its cost efficiency relative to phenotypic
selection. The frequency of superior genotypes among the selected progeny increases
as the selection intensity increases. Van Berloo and Stam (1998) assessed the
effectiveness of MAS compared to phenotypic selection, showing that MAS appears
particularly promising when dominant marker alleles are present at the QTL and
linked in coupling phase. Uncertainty in estimated QTL map positions reduces the
benefits of MAS.
Young (1999) pointed out that despite innovations like better marker systems
and improved genetic mapping strategies, most marker associations are not
sufficiently robust for successful MAS. Charmet el al. (1999) studied the accuracy of
QTL location determination, showing that it greatly affects st :ction efficiency. MAS
for QTLs have recently started to be applied to the genetic improvement of
quantitative characters in several crops such as tomato (Lawson el al., 1997;
Bernacchi el a!., 1998). maize (Graham el a[., 1997). and barley (Han el al., 1997;
Toojinda el al., 1998).
Hospital and Charcosset (1997) provided a general framework for the
optimization of the use of molecular markers in backcross breeding programs aimed
at introducing one to several superior QTL into a recipient line. Using at least three
markers per QTL allows a good control of the donor chromosome segment over
several generations. When several target alleles are monitored simultaneously.
background selection among the limited number of individuals resulting from the
foreground selection step accelerates the increase in genomic similarity with the
recurrent parent with only limited increase in the cost. Frisch el al. (1999b)
determined the number of marker data points (MDP) required in background
selection, the size of the population to be used and compared a two-stage selection
procedure (one background and one foreground selection step), with alternative
selection procedures (one foreground and two or three background selection steps).
They concluded that as the number of selection processes increases, the number of
MDP required decreases.
 Moreau et 01. (2000) evaluated tlie relative efficiency of MAS in the first cycle
of selection tlirough an analytical approach taking into account tlle effect of
experimental design (population size. ~ i i ~ m b of
e r trials atid replicationitrial) on QTI.
detection. They concluded that expected economic returns of MAS compared to tlie
phenotypic selection decrease with increases in the cost of genotyping..B~~nya~iiin
er

rrl (2003) repotted MAS for colilplex traits in conuiion bean using an index based
on QTL-linked ~iiarkers and ultramtric gerietic distnnce(from a cluster analysis)
between lilies and a target parent. A compariso~iof the mean seed yield of the top five
lilies selected by different schemes denionstrated tliat tlie highest yiclding group \\as
selected on the basis of a combination of phenotypic perfor~nanceand high V r L -
l~asedindex, followed by groups identified b) liigh QIL-based index. conventional
selection, and low Q'TL-brised index. respectively. The study showed that use of a
VI'L-based index in conjugation \+it11 the ultra metric genetic distance to tile targeted
parcnt would enable a plant breeder to select lines tliat retain itiiportntit QTI,s in a
desirable genetic background. 'I'herefore this type of blAS would be cspected to be
superior to plie~iotypicselection.
2.4.2 General consideration for all trait categuries in rnnrlter assisted selection
2.4.2.1 Gene introgression
Gene introgression involves the introduction of a target gene into a producti\'e
recipie~~t
line or culti\,ar. (iene introgression can be used in botli bnckcrossing and
intercrossing programs. By using DNA niarkcrs to identify reco~nbinalits.introgressed
cl~romosomesegments might bc "tri~iimed" to minimal s i x , rcdi~ci~ig
tile extent to
\vhicli the recurrent genotype is disrupted hy undesirable alleles closely linked to
genes controlling the target trait (Tanksley and Rick. 1980). It is often critical in plant
breeding that allelic substitutio~ibe precise sci that ollly the target gene and the
shortest possible segment of tlie linked chromosotiie are transferred from the donul-
parent to the recipient parent. the latter of whicli is usually a cultivar or inbred line
with very good combining ability. To reduce false positives in MAS. markers used fol-
foreground selection must be tightly linked to the geneiQ.I'L co~itrollingtlie target
trait. atid flanking tilarkers or multiple ~narkers around the region call be used
simultaneously. A three-marker system. with three markers located on a short
cliromosomal block of a few (4)cM. will be desirable in such cases (Zliang and
Huang. 1998). The marker in the middle. preferably intragenic or co-segregating with
the target gene. will be used in foreground selectio~ito indicate the presence of the
target gene in the selection process. 'l'lie marker on each side will be used to indicate
tlie absence of the cliromosoriie segtiient from the donor parent (negative selection).
tliat is. selection for recombination beween the target gene locus and the ttiarker
locus. For genes tliat have bccli cloned. the marker in tlie middle can be developed
from tlie cloned gene or gene sequence. l'liis system will bc very useful when the
target gene is only available in wild species and litlkage drag is proten to be
associated with tlie chromosome seyliient to be inlrogresscd.
The first such study employed a cross between the wild rice relt~~ive
O,:,,zcr
n the Chinese ittr!iccr hybrid 'V20':'Ce64'
r ~ f i l ~ o g oand (Xiao et rrl.. 1998). Although
the 0. rri/i~~ogori
accession was phenotypicall) inferior tbr all 12 trails studied.
t~.ansgressivesegregation was observed for all traits. and 5 1% of the QTL detected
had hc~icticialalleles from O rifi~popoft.By MAS and tield sclectioli. an excellent
CMS restorer line ('0661') carrying one of the niajor Q'TLs for yicld components was
developed. Its hybrid. 'J23A'I'Q661'. out-!ielded the clieck hybrid by 3 5 O h in a
replicated trial fix the second rice crop in 2001 (Yuan. 2002).
2.4.2.2 Wholc genolae selection
MAS can also he practiced at the whole genome I c ~ e l .DNA nlnrher-based \vhole
genome selection or "background selection" can be used to accelerate recovery of
recurrent paretit genotype in the backcrossi~lgprocess Ihr breeding irnpro\cd p~~rentnl
lines. Coniprired to a conventional backcross program that usually takes five to seven
generalions lo reco\.er riioct of the recurrent parental backgmund. MAS may savc t ~ v o
to Ibi~rbackcross generations in tlie transfer of a single tarfct allele (Tanksley et (11..
1489; Hospital er 01.. 1992: Fisch er ol., 1099). Combined with selection for Inryct
traits. whole genome selection allo~vstlie breeder to si~iiultnneouslq transcer targeted
traits througli backcrossing. (Combined firreground ancl background selection allo\vs
a single QTL. hut simultaneous
the breeder to save a few generations for transferri~~g
transfer of multiple traits remains a ver! difficult and expensive exercise) using MAS.
It is 17robahly lnorc cost-effective to transfer multiple QTLs in parallel single-V'I'L
backcrossilig programs and then use a cornplex series of crosses of the single-QTL
introgression lines to pyramid the QTLs).
As genetic mapping infomiation accumulates fioni different mapping
populations. it will be possible to establisll a colilplete profile fbr all the genes
associated with a specific trait or trait categor).. Whole genoliie selection can he used
to select the best traitigene combit~ationsbased on selection for each of the target loci
for which position in the genome is known. It is possible to select the best cassette of
marker alleles for any trait andlor trait combination.
TO transfer the bacterial blight resistance gene Xa21, 128 RFLP markers,
evenly distributed across the 12 rice chr~mosomes,were used to recover the genetic
background of 'Minghui 63', a widely used parent (Chens el a/., 2000). MAS was
also be used by the same group to improve '6078', an elite restorer line with high
yield potential by transferring Xu21 from IRBB21 (Chens er ol., 2001).
2.4.2.3 Selection for multiple genesltraits
MAS provides opportunities for simultaneous selection of multiple traitslgenes. In
some cases, multiple pathogen races or insect biotypes must be used to identify plants
for multiple resistances, but in practice this may be difficult or impossible because
different genes may produce similar phenotypes that cannot be distinguished from
each other. Marker-trait associations can be used to simultaneously select multiple
resistances for different disease races andlor insect biotypes, and pyramid them into a
single line through MAS. To find a CMS restorer in rice through testcrossing and
progeny testing, a candidate male parent has to be testcrossed with a CMS line and
male fertility of the resulting hybrid progeny assessed to find out if the candidate male
parent has fertility restoration ability. However, sterility in the testcross hybrid could
result from the absence of either fertility restoration genes or \bide compatibility genes
or both when an intersubspecific cross is involved. MAS can be used to distinguish
between these two different causes of sterility. Hybrid rice provides an advantage
over inbred cultivars because dominant genes and/or QTLs with favorable effects
from both parents that can be integrated into one hybrid. An integrated breeding
program including MAS was initiated in China to improve elite hybrid rice.
2.4.2.4 Integrated genetic mapping and MAS
In many cases, genetic mapping results obtained from specific crosses cannot be used
for MAS for the same traits in different crosses. There are three reasons for this
phenomenon. First, quantitative traits are usually controlled by many genes. Genes are
only segregating at the loci where two parents are genetically different and thus can
be mapped using the population from these two parents. For a randomly selected
mapping population, the parents will have a strong chance to share identical alleles at
some of the genetic loci. There is a high probability that segregating genes already
mapped in one mapping population are not segregating in a second mapping
population. Second, mapping population parents could have alleles that are different
from those of elite breeding populations. Interactions among these multiple alleles
will modify marker-trait associations when different allele combinations are
considered. Third, GxE interaction could make the establishment of marker-trait
association depend on specific environments. One of the best ways to avoid these
limitations is to integrate the genetic mapping for a trait with improvement of that trait
in an elite background, i.e., identify the marker-trait associations from a breeding
population so that they can easily be used for MAS of the same population. This is
critical for quantitative traits, which are genetically controlled by many genes and
interact with environments. Advanced backcross QTL analysis proposed by Tanksley
and Nelson (1996) to accelerate the process of molecular breeding is one of the
approaches that can be used for this purpose.
2.4.2.5 Cost benefit analysis of MAS
The cost of using the 'tools' of MAS in applied plant breeding programs is a major
consideration. The cost of using MAS compared to conventional plant breeding varies
considerably between studies. Dreher el al. (2003) indicates that the cost effectiveness
of MAS needs to be considered on a casc-by-case basis. Factors that influence the
cost of utilizing markers include: inheritance of the trait, method of phenotypic
evaluation, fieldlglasshouse and labour costs, and the cost of resources.
In some cases, phenotypic screening is cheaper comj ?red to marker-assisted
selection (Bohn el al., 2001; Dreher et a/.,2003). However, in other cases, phenotypic
screening may require time-consuming and expensive assays, and the use of markers
will then be preferable-especially in private-sector breeding programmes where
reducing the time required to achieve a unit gain in varietal performance can help
make a company's products more competitive in the market. Some studies involving
markers for disease resistance have shown that once markers have been developed for
MAS, it is cheaper than conventional methods (Yu el al., 2000). In other situations,
phenotypic evaluation may be time-consuming and/or difficult and therefore using
markers may be cheaper and preferable (Dreher et al., 2003; Young, 1999; Yu el a/.,
2000). An important consideration for MAS, often not reported, is that while markers
may be cheaper to use, there is a large initial cost in their development. An estimate
for the cost to develop a single marker was AUD $1 00,040 (Langeridge el al., 2001).
MATERIALS AND METHODS
 CHAPTER 111
MATERIALS AND METHODS

3.1 Application of SSR markers in diversity analysis of sorghum insect resistant

germplasm accessions
3.1.1 Plant material:
Ninety-one sorghum genotypes were selected for the present study. These include
elite open-pollinated varieties, hybrid parental lines, recurrent parents used in marker-
assisted backcrossing programs at international Crops Research Institute fur the Semi-
Arid Tropics (ICRISAT) Patancheru for the stay-green trait (a component of post-
flowering drought tolerance), and germplasm accessions exhibiting resistance to
sorghum shoot fly, spotted stem borer, and sorghum midge. rhey are currently used in
breeding programs at the ICRISAT, Patancheru andlor at the National Research
Center for Sorghum (NRCS), Rajendranagar. Hyderabad and breeding programs of
state agricultural universities in India. Seeds of thesc accessions have been maintained
by the ICRISAT germplaslii unit (Appendix I).
3.1.2 Ninety six-well plate mini-prep genomic DNA extraction:
Details on the preparation of solutions and buffers used in DNA extraction are
presented in Appendix 11.
Lysis buffer (3% CTAB) was preheated to 65'C in a water bath before the start of
tissue sample collection.
Steel balls (2 per tube), pre-chilled in the freezer at -20°C for about 30 minutes, were
added to plastic extraction tubes.
Leaf strips 6-cm long were collected (final weight 30 mg) from one-week-old
seedlings of each germplasm accession, cut in pieces (1 mm length) and transferred to
tubes.
A. Grinding and extraction:
1. 450 p1 of preheated (65°C) CTAB buffer was added to each tube containing
leaf tissue samples.
2. Tissue sample grinding was conducted using the Sigma Geno-grinder at 500
strokes per minute for 2 minutes. Grinding was repeated until the color of the
sample solution became pale green.
 3. After grinding, the tube box was fixed in a locking device and incubated at
65°C in a water bath for 10 minutes with nianual shaking at regular intervals.
B. Solvent extraction:
1. 450 y1 of mixed chloroform:iso-amyl alcohol (C:IAA=24:1) was added to
each tube and centrifuged at 6200 rpm for 10 minutes.
2. After centrifugation, the aqueous layer (approximately 300 111) u a s transferred
to a fresh tube.
C. Initial DNA precipitation:
To each tube 0.7 volume (approximately 210 y1) of cold isopropanol was added and
the samples were kept at -20°C for 10 minutes.
The box of 06 tubes was then centrifuged at 6200 rpm for 15 minutes using the box
centrifuge.
Supernatant was decanted from each tube and crude DNA pellets were allowed to air
dry (minimum 20 minutes).
D. RNAse treatment:
200 y1 of low salt buffer (Tris-HCL 1: EDTA 0.5) and 3 yl of RNAse (stock 10
nigipl) were added, mixed properly and incubated at 37'C for 30 minutes (can be kept
overnight at room temperature).
E. Solvent Extraction:
200 PI of phenol:chloroform:iso-amyl alcohol (P:C:IAA=25:24:1) was added, mixed
well, and centriruged at 5000 rpm for 5 minutes.
After transferring the aqueous layer to a fresh tube. this step was repeated with the
chlorofor~n:iso-amylalcohol mixture (C:IAA=24:1).
F. DNA precipitation:
To the tubes containing aqueous layer (1110th of its total volume approxin~ately)
sodium acetate (from 3M stock) and 2 volumes (300 PI) of 100% ethanol were added,
mixed, and the tubes subsequently kept at -20°C for 5 minutes.
Following this brief incubation the box of tubes was centrifuged at 6200 rpm for 10
minutes.
G . Ethanol Wash:
After centrifugation the supernatant was carefully decanted. In order to remove excess
salts, 200 yl of 70% ethanol was added to the pellet followed by centrifugation at
6200 rpm for 5 minutes.
H. Final reauspension:
Supernatant was decanted and the pellets were allowed to air dry for one hour.
Dried pellets were re-suspended in 100 to 150 pl of TE buffer and kept at room
temperature to dissolve completely (approximately one hour).
Dissolved DNA samples were kept in a refrigerator at 4OC.
3.1.3 Checking DNA quality and DNA concentration
The DNA quality was checked using 1.2% ready-made agarose gels (Amersham
Biosciences). For this, 1 pl of DNA solution was mixed with 1 p1 of orange dye and 8
pl of distilled water and the mixture loaded into a well on the 1.2?/0 ready to run
agarose gel. The gel was run for 10 minutes, after which the quality was checked
under W. A smear of DNA indicated poor quality whereas a clear band indicated
good quality. Samples of poor quality were re-extracted.
The DNA concentration was assessed using a Spectrafluor Plus
spectrophotometer after staining the DNA with PicogreenTM(11200 dilution). Based
on the Relative Fluorescence Units (RFU) values and using the standard curve, DNA
concentrations were calculated. The DNA was diluted to a final concentration of
2.5ngipl. Figure presents a calibration curve where
DNA concentration = -2.78273+0.002019*RFU.
Standard curve showing the linear relationship between RFU and DNA
concentration.

DNA concentration in pg
3.1.4 Primer selection:
For assessment of genetic diversity of the 91 sorghu~ngenotypes includes in this
study. 21 SSR primer pairs were used including pairs from the Xctcl? series (15 primer
pairs), the Xrxp series (4 primer pairs), and for both Kt$ and Xgtrp84. These primer
pairs detect 21 SSR marker loci that had carefully selected based on the following
criteria:
- Markers should be mapped at different loci on different sorghum linkage
groups.
- The markers should display a range of allele sizes in prior publications.
Primer sequences for the markers used in this study have been described in the
following publications: the X/xp markers by Bhattramakki L.! trl. (2000) and Kong cr
ti/. (2000). the .Xgap 84 markers by Urown el ill. ( I 996). the ktrf'marker b), Taratnino
et ol. (1997). and the ,licup markers hy Scliloss el (11. (2002). Seven groups of three
primer pairs here formed. Each group contained three pairs of pri~nerswith the
forward primer of the first pair labeled with 4.7.2',4'.5'.7'-hexachloro-6-
carboxyfluorescein (HEX), the forward primer of the second pair labeled with 6-
carboxyfluorescein (6-FAM), and one primer of the third pair labeled with 7'3'-
benzo. 5'-fluoro-2',4,7-trichloro-3-carboxyflourescei
(NED) (Table 3.1).
3.1.5 PCR Amplification:
Polymeric chain reaction (PCK) a~npliticationof each SSR loci was performed in a
total reaction mixture volunle of 5 111containing sorghum genomic DNA. PCR buffer
(Applied Biosystems), dNTPs, MgClz (Applied Biosystems), forward primer (Applied
Biosystems) labeled with HEX, NED, or FAM dyc phosporamidites (Applied Bio-
systems), reverse primer (MWG), and Ampli Toy Gold DNA polyn~erase(Applied
Bio-systems) in an Applied Bio-systems Gene Amp PCR system 9700 ther~no-cycler
using a "Touch Down" PCR technique. PCR conditions were previously optimized
for each primer pair using a grid of nine reactions. Three different sets of PCR
conditions were used for PCR amplification ('Table 3.2).
The initial DNA denaturation at 94OC for 15 min, to activate the Tuq
polymerase, was followed by 10 cycles with the following profile: denaturation for 15
sec at 94OC, annealing for 20 sec at 61°C (the annealing temperature was decreased
by 1°C for each cycle) and extension for 30 sec at 72'C. This was followed by 31
cycles with the following profile: denaturation at 94'C for 10 sec, annealing at 54'C
Table 3.1 cont...

SR locus I Label 1 LG' l ~ e ~ e a t s l~rimerSequence References

I I I b. A A A rTTGCACTTGTC Kong et al. (2000)
.rnGGCACTAG
I I
ACTGTGAGCAGC l~chlossel al. (2002)
3GTTGTTGTGCC

kcup02
-
I I
NU) G ~CCA)6 E.
:GACGCAGC'ITTGCTCCTATC
,,
, A , , ,
R. u 1 LL-LLAACCCACGTATC
Schloss el al. (2002)

F: CTAGAGGATTGCTGGAAGCG Schloss el a1. (2002)

Xcup07 FAM I (CAA) 8
7
R: CTGCTCTGCTTGTCGTTGAG
F: CTCCTCGCCGTCATCATC Schloss el a!. (2002)
Xcup52 FAM J (AATT),
R:TAAAGAGAAACGCAGGCAGG
F: AGCATCTTACAACAACCAAT Tamarino et al. (1997)
XsbKafGKl FAM J (AAC)9
R: CTAGTGCACTGAGTGATGC

bpl5 I-- /
FAM J /(Tc) 16
: CACAAACACTAGTGCCTTATC P(ong et al. (2000)
:CATAGACACCTAGGCCATC
1-
Sorghum llnkage group designations following the system of Peng et al. (1999), Subudhi and Nguyen (2000), and Menz el al. I !002), which
have the following relationships with sorghum chromosome designations assigned by Kim el al. (2005):
A= SBI-01, B= SBI-02, C= SBI-03, D= SBI-04, E= SBI-07, F= SBI-09, G= SBI-10, H= SBI-08, I= SBI-06, J= SBI-05
 Table 3.2: PCR protocols uscd for amplification with labeled SSK primers i n sorghum divcrsily stud).

Protocol No: 7

SSR locus Primer (1 pM)/pl MgCl, (1 mM)/pI dNTP ( 0.375 mM)/pl DNA (1.25 ng)/pl Enzyme (0.2 U)/pI Buffer (1X)Ipl k a t e r (pl)
XcupO7 97 97 36.375 0.5 19.4 48.5 138 225
Xcup I 4 97 97 36.375 0.5 19.4 48.5 138.225
Xtxp114 97 97 36.375 0.5 19.4 _ 48.5 138.225
Fgap84 97 97 36.375 0.5 19.4 48.5 138.225
Protocol No: 4

R locus brimer (0.5 pM)/pI )Mgcl2(0.75 mM)/pI JdN~p(0.5mM)/pI ~ N (1.25

A ng)/pl J ~ . n z ~(0.25
m e U)I p l IBuffer (lX)/pl b a r e r (pl)
I 48.5 72.75 48.5 1 0.5 24.25 1 48.5 1 194
Protocol No: 5

I I I I I I I I I
for 20 sec, extension at 72'C for 30 sec. After these 31 reaction cycles the extension
at 72°C was prolonged for 20 nlin. Subsequently, the PCR product samples were
stored at 4OC. Then 1 pl (for FAM- and HEX-labeled PCR products) to 1.5 p1 (NED-
labeled PCR products) was transferred to a 96-well ABI plate containing 7 pl
formamide, 0.3 p1 ROX size standards, and 4.2 pl Double Distilled Water. 'The
remainder was PCR-amplified for an additional 6 cycles with the following pmtile
(denaturation at 94OC for 10 sec. annealing at 54°C for 20 sec. extension at 7?OC for
30 sec). San~pleswere then stored in -20°C until further use.
3.1.6 Electrophoresis
a) Non-denaturing polyacrylamide gels
One 111 of loading buffer was added to 3.0 to 3.5 p1 of each PCR sample. Two pI of
this buffered PCR product was then loaded on each lane of a 96-track 6% non-
denaturing polyacrylamide gel containing 29: 1 acrylaniideibisacryla~nide,10X 'TBE.
and water. In addition, four wells were loaded with a 100 bp size standard to ensure
proper sizing of the amplified fragments. The gel was run at 600 V of constant power
in O.5X 'I'BE for 3 h, using a RioKAD gel sequencing apparatus.
3.1.7 Silver staining
After PCR product separation by PAGE, the gel was placed in water [or 5 min.
soaked in 0.1% CTAB for 20 minutes with gentle shaking, incubated in 0.3%
ammonia for 15 min, and placed in silver nitrate solution (0.1% silver nitrate, IM
NaOH and 25% ammonia) for 15 min with gentle shaking. After incubation in this
silver nitrate solution. the gel was placed in developer (30 g sodiuni carbonate and 0.4
ml formaldehyde in 2 liters of water) with gentle shaking until bands became visible,
rinsed in water for 1 min to stop the staining reaction, and placed in fixer (30 ml
glycerol in 2 liters of water) for a few seconds.
After silver staining the PAGE gels, the size(s) (base pairs) of the most
intensely amplified specific bands or alleles for each SSR marker were estimated
based on migration relative to the 100 base pair (bp) DNA ladder (consisting of
fragments ranging from 100 to 1000 bp). The presence (1) or absence (0) of each PCR
fragment was scored for each of the 91 genotypes.
3.1.8 ABI Prism 3100 genetic analyzer
PCR products of each group of 3 primer pairs were pooled post-PCR. Because of the
different signal intensities of the fluorophores, 1 pI (in case of FAM- and HEX-
labeled PCR products) to 1.5 p1 (in case of a NED-labeled PCR products) was added
to a mix of 7 p1 formamide, 0.3 p1 ROX size standards, and 4.2 111 Double Distilled
Sterilized water (total volume 15 pi). The samples were denatured for 5 min at 94°C
and cooled on ice. The plate with the samples was then centrifuged 1 n ~ i nat 760 rpm
(Eppendorf) and stored at -20°C until separation on the ABI 3100 or AD1 3700
capillaq electrophoresis DNA sequencing machines.
96-well plates (96 genotypes x tllree primers) were placed in the AB1 3100 or ABI
3700 machine. The san~pleswere separated using the following protocols:
ABI 3100: dye-set "Dm. run module "SSR 20 minutes". and analysis
module "GSHD Analysis". 'The fragments were separated in a 36-cm
capillary array, using POP4 as a carrier.
ABI 3700: dye-set "D", run module "GeneScan2-
POP6DefaultModule". and analysis module "GSHD Analysis". 'The
fragments were separated in a 50-cm capillary array, using POPh as a
carrier.
After completion of the run, the peak patterns were sized using Gene Scan. Presence
or absence of allelic fragments were scored using the Genotyper software.
3.1.9 Data Analysis
SSR data was analyzed for both PAGE and AD1 PCR product separation methods.
Clear and distinct amplification products were scored as ' I ' for presence and a '0' for
absence of bands. The N'SSYS (Numerical Taxononly and Multivariate Analysis
System) program was used for cluster analyses. 'The data was used to generate
Jaccard's similarity coefficients based on SSK bands. The Jaccard's coefficients
between each pair of accessions were then used to construct a dendrogram using the
un-weighted pair group method with arithmetic averages (UPGMA).
3.2 Phenotyping of RILs 296B x IS 18551 for components of resistance to
sorghum shoot fly
3.2.1 Material
The experimental material consisted of a set of 259 Recombinant Inbred Lines (RILs)
(F7.8),derived from a cross between two sorghum-inbred lines, viz., 296B (susceptible
to shoot fly) and IS 18551 (resistant to shoot fly). Table 3.3 elaborates salient features
of these two parental lines. The RIL population progenies along with both parents
were used for phenotyping and genotyping.
Table 3.3 Salient features of parental lines of RIL mapping population
Parents Salient features
29GB Derived from Aisptrri. Semi-compact earhead. white
grain, foliage tan coloured. Leaves of seedling are non-
glossy with no trichornes. Highly susceptible to shoot
fly.
Originates from Ethiopia, race Dzrrru. Earhead uith
straw coloured grain and glun~eslarger tho112960
Leaves of seedling are light green. glossy, narrou and
pointed upward with densc trichomes on both sides of
the leaf balde. Resistant to shoot fly. Very tall at
maturity.
3.2.2 Development of mapping population
'The RlLs were produced at ICRISAT, Patancheru. Alier the initial cross between
2968 and IS 18551, a single FI plant was selfed. The resulting F2 seeds were sown
and Fz plants were selfed. 'The Fj seeds were sown head-to-row, each 173 plant was
selfed and from each head-to-row a single plant was randomly chosen to provide the
seeds for the next generation. This, modified single-seed-descent method, where each
line is maintained through selfing a single randomly selected plant. was repeated for 3
to 4 generations, up to F7. During RIL development the plant material, recommended
protection measures were taken to protect the plants against shoot fly and other
insects. Bulked seed was harvested from randomly selected F h plants to produce 259
F7 recombinant inbred lines (RILs). Each F, line represents the individual F2 plant
from which it is derived. The details on pedigrees of 259 F7 g RlLs of cross 2960 X

IS I855 1 is given in Appendix 111 (Fig. 3.1).

3.2.3 Evaluation of RlLs for resistance to sorghum shoot fly,Atherigono soccata
Screening of the RIL for shoot fly resistance was carried out at the International Crops
Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra
Pradesh. A total of 300 lines (259 N L s + 14 repeated checks of each of 296B and IS
18551, and a standard check, CSH 9 repeated 13 times). were sown on 16'~
August
during the 2002 khargseason (El). For early rabt season (E*), a total of 270 entries
Fig 3.1. Schematic diagram of RBI, develuprnent procedure
(259 KlLs 4-4 tir~irsrepeated ciiucks ul'eacii ~>1'796f3;tiid IS 1853 I ;-sti:~iJ;irdiiicch
CSt-I 9 repeated 3 times), \irere sown oil 16'" Octvber 100-! I'iic i ~ l os ~ r c c ~ l i ~ i g
cr~vironn?entsare referred as El (late itl~<:iif)arid 1:; (earl) n,b!). I'lre teat rn;lti.r i,li !i,i\
planted in balanced a dcsign, witli 75 cni ~ t i d 10 ciii iiitcr- ailti inir't-ro\\~ sp:icirig.
rerpccrively. In the late khiirfurid vtibi seasons, cacli clitrj !\.I:, gr-oirri itl ti\(,-r~~\s
plots o f 2 tn length in f o i ~ and
r tiiree rcpiicarrotir, t~especticc!~
Shooi fly inYcstatiun \has quite liiyii di>ri!igtile :,/:iii-;/ scaioii. illiring 111~. l _ O i ! i

iuhi scason, the shuot fly irrfsstation w'ls ielutrvcly iow. l o ciisutc uniforrii , ~ , ~ i i
o p t i n ~ u ~shoot
n 11) inibstation under field cotiditio~i:,,t l ~ ciriii.rl,~~d-ti>li
!iie,il ii:~llrircjuc
(Sharma i.1 (11. 1991) was fol!o\ved to scrceri i i ~ resistn:ice
r to ilioo~ti! (1)Iaie .i.
/ ).
3.2.3.1 Obse~.vationrs
C)hser\atio~is were recordeii uti leaf glossincs\, iricliome dcnsitq ori tiL>.o.rai . ~ n d
adaxial surfacc of ieaf, pcrcent plants \+ir!i eggs arid desdiic.,i~?\, iiinc to 50%
Ilowcririg, piant irciyl~t,recovery I-esiatdticc, apliid d;~rrwgc\cute, n~ndgr;iiri >iclil 111

cacti plot during the 2002 Mzurf and 21304 ?.,lhi Leasons Obsl-ivatiurr\ \icrc aim
recorded or1 Icai'pignient;itii~ti.hccdli~~g
vii:t)r, secdling hciglii. iiiidpr da~iingt:< c t > r ~ .
and ;~gronon;icperfonliarice during il:e 2!)1)4 ?.<I/II sciison.
3.2.3.1.1 Cdlossiilrss
1.caf glossiricss was rscurdzd drrtiiig carly liioriirtig liuurs v,'l,!r~ ~.etiectionuf Irglit ii

masin~urn (i'late 3.21). )nterisity of leiif glossinrss wii\ rccoriicd ~ i s u a l l yoil a scaic ill
1 to 5 ( 1 = pale green, shiny, marrow, and leaves poiritcd trpirards. dnd 5 ~ L~IIII
IXO~:LI,
i , cirooping leaves) ai 7 days aiizr etiicrgcr1;e jDAL3j (Sliarlna ci a).. 1'107).
g ~ e e ~atid
3.2.3.1.2 Seedling vigor
Seedling vigor (a combinatioi~of heigtit, leat'gr.owtl!, ur~drobiisiries>) was t-v,liua~ed
on a 1 to 5 scale at 9 DAE ( 1 = plants \\,it11 muairtiunl iicighi, leaf Cl;p:illSloli :inti
robusiircsS, arid 5 - platits with niinirnuri~ grn\btii. little IuaS exjia~?sio~r.
nlic! poor
;idaptation) (Sharma et a\., i997).
3.2.3.1.3 Seedling l~cigiil
Seedling lreight ( c ~ n was
) measured frorn thc base ul' the plant to lllc ti!) u f tiis lor-
most completely opened leaf on tliree iarrdorr~lyselcclcd pI;bii~sIrorn c;icIi p l ~ ? l ;it LO
DAE.
3.2.3.1.4 Plarat starld
a1 70 1);\1.
l.he total nlrlnber ofplanls in each pic( w;is dztzrrni~~ed 'Iliib nililiber i\a\
used to calculate the percentage of oviposition ;iird deadirearts iiicidc~icc.
Plate 3.3: PIumule and leaf siieath purple pigmentation scores of sorghum gellotypes at
5 days after seedling emergllce. A = 1 (dark pink). 8 = 2 (fair pink). C' = 3 (light pink).
1) = 4 (very light pink). li: = 5 (green).
tillers and panicles similar to the main stem, and 9 = less than 10% plant with uniform
height tiller and panicles similar to the main stem panicles of non-damaged plants.
3.2.3.1.12 Agronomic score
Agronomic performance of the test material was evaluated at crop maturity on a scale
of 1 to 5 based on panicle length, and production potential, where I = good productive
potential, and 5 = poor agronomic performance, poor productive potential and poor
adaptation to agro-climatic conditions.
3.2.3.1.13 Midge damage
Midge damage was evaluated at crop maturity on I to 9 scale, where, 1 = < 10%
midge damaged spikelets, 2 = l l to 20% midge damaged spikelets, 3 = 21 to 30%
midge damaged spikelets, 4 = 31 to 40% midge damaged spikelets, 5 = 41 to 50%
midge damaged spikelets, 6 = 51 to 60% midge damaged spikelets, 7 = 61 to 70%
midge damaged spikelets, 8 = 71 to 80% midge damaged spikelets, and 9 = > 81%
midge damaged spikelets.
3.2.3.1.14 Aphid damage
Aphid damage was evaluated at crop maturity on a I to 9 scale, where, I = few aphids
present with no apparent damage to the leaves and 9= heavy aphid density on infested
leaves (Table 3.4)

Table 3.4 Aphid ratings i n relation to percentage

Aphid densitylinjury Aphid densitylinjury
rating YO
1 1-10
3.2.3.1.15 Grain yield
,411 the mature panicles from each plot were harvested in bulk and threshed together,
and expressed the yield as gram per plot.
3.2.4 Statistical analysis
3.2.4.1 Pheaotyping data analysis
3.2.4.1.1 Analysis of variance (ANOVA)
The analyses o f variance for phenotypic data sets were performed using the residual
maximum likelihood algorithm (ReML). which provides the best linear unbiased
predictors (BLUPs) of the performance o f tested genotypes (Patterson and Thompson,
1971). ReML estimates the components o f variance by maximizing the likelihood o f
all contrasts with zero expectation. For each trait and for each entry, the predicted
means were calculated with entries as fixed effects for both individual environment
(season) and across screening environments (seasons) analyses; replications, error,
and entry x replication interactions as random effects in individual screening
environment analyses; and replication, error, entry X replication, and entry x
environment interactions as random effects in the across screening environments
analysis.
Entry means were estimated by generalized least squares with weights
depending on the estimated variance components according to Paterson (1997). The
data was analyzed using the GenStat (6Ihedition) package (Payne, 2002).
3.2.4.1.2 Estimates o f broad-sense heritability (h') on of pr ,geny-mean basis
Broad-sense heritability h2 (progeny-mean basis) was estimated across RlLs in each
of the two screening environments for all candidate resistance component traits as
well as for the traits measured at crop maturity, It is the ratio o f total genotypic
variance to phenotypic variance and was calculated following Falconer (1989) for the
data recorded in individual environments, E l and E2:
Broad-sense heritability estimates across the screening environments were coniputed
by the formula,
v,
h2 =

v, + V,, +v c

where.
h2 = broad-sense heritability
V, = genotypic variance
Vp = phenotypic variance
V, = environmental variance
V,, =G x E interaction variance
3.2.4.1.3 Superiority of RILs over the parents (transgressive segregation)
The calculation of superiority of RILs over parents for shoot fly resistance and other
traits were worked out using following formula:
Sl = (RIL-P I)/Pl

S2 = (RIL-P2)/P2
where,
Sl = superiority to P1 (2968)
S2 = superiority to P2 (IS 1855 1 )
P1 = Mean of parent 1 (2968)
P2 = Mean of parent 2 (IS 1855 I )
The information obtained was used to estimate the proportion of transgressive
segregants in the RIL population (based on nleans across the two screening
environments); RILs showing phenotypic characteristics with values lying outside the
parental limits for shoot fly resistance components as %ell as other traits were
considered transgressive segregants.
3.2.4.1.4 Test of significance of means

-
TO test whether the difference between means of each parent and mean of RILs is
small enough to accept the null hypothesis, i.e., XI =
-
X2, the t-test was applied and
calculations were made using the formula given by Singh and Chaudhary ( I 996):
 -
(XI1 - x1l2
where, s = -
nl -l

The calculated value of 't' was compared with the table value of 't' to test its
significance at (111 + nz) - 2 degrees of freedom.
3.2.5 Marker data analysis
3.2.5.1 Occurrence of non-parental alleles
'This populatioti was found to contain unexpectedly large number of progeny with
non-parental alleles. Around 18% of the total population (46 RILs out of the 259-
entry RIL population) was detected to be carrying non-parental alleles. These entries
were discarded from the dataset used as input for the linkage and QTL mapping
analyses.
3.2.5.2 Information on genetic linkage map used for QTL mapping
The QTL mapping was done by utilizing the skeleton map developed by Deshpande
(2005). The linkage map of 296B X IS 18551 based RIL population had 11 1 SSR
marker loci mapped over 1 1 linkage groups (Fig 3.2). The linkage groups varied in
length from 22.9 cM to 385.2 cM with a total linkage map length of 2165.8 cM.
3.2.5.3 QTL analysis
A total number of 213 RIL progenies from the cross 296B x IS 18551 were used for
marker trait associations. The BLUPs of these 213 RILs derived from the cross 2968
x IS 18551 were used for QTL analyses. QTL analyses were performed using
Figure 3.2. Genetic linkage map of sorghum based on 213 recombinant
inbred lines derived from cross 2968 X IS 18551. Numbers on the left of
each linkage group indicate map distance. relative to the firet marker. in
d U determined using Haldane function. The loci indicated in red have
rcamangements in their order compared to the sorghum linkage maps
produced by BhattmmakH et aL (2002) and Polkertsma et aL (2005).
composite interval mapping (CIM) (Jansen and Stam, 1994; Zeng. 1994). Required
computations were performed using PlabQTL Version 1.1 (Utz and Melchinger.
2000). which performs CIM by employing interval mapping using a regression
approach (Haley and klott, 1992) with selected markers as cofactors. Markers to
serve as cofactors were identified using stepwise regression with an F-to-enter and an
F-to-delete threshold value of 3.5. The presence of a putative QTL in an interval was
tested using a critical LOD threshold as determined by PlabQTL using the Bonferroni
x2 approximation (Zeng, 1994) corresponding to a genome-wise type 1 error of 0.25.
Since the mapping population used in the present study was constituted of RILs. the
additive model 'AA' was employed for analyses in which additive X additive
epistatic effects were included. The point at which the LOD score had the maximum
value in the interval was taken as the estiliiated QTL position. QT1.s detected in
different environments were treated as common if their estimated positions were
within 20 cM of each other and their estimated effects had identical sign. The
proportion of phenotypic variance explained by a single QTL was estitnated as the
square of the partial correlation coefficient. Estimates of the additive effect of each
detected QTL, the total LOD score, and the total pl.oportion of phenotypic variance
explained jointly by all detected Q'fL were obtained by litting a tnultiple linear
regression model that simultaneously included all detected QTL for the trait in
question. QTL x environment interaction was analysed over all three environments as
described by Utz and Melchinger (2000). The proportion of genetic variance
explained by the QTL was adjusted for Q'fL x environment interactions to avoid
overestimation. After the QTL analysis with PlabQTL, the QTl..s identified for
components of resistance were assigned to the linkage groups based on linkage
positions of markers on the linkage map developed by Bhattra~nakkiet (11. (2000).
3.2.5.4 QTL analysis for a single environment
To localize and characterize QTLs controlling components of resistance to shoot fly.
the combined phenotypic and n~oleculardata were analyzed with PlabQTL (Utz and
Melchinger, 1996). Interval mapping using multiple regression approach with
flanking markers (CIM i.e. composite interval mapping) was followed according to
the procedure described by Haley and Knott (1992). Since the mapping population
used in the present study constitutes RILs, the additive model AA was chosen for
analysis in which additive x additive effects were included.
 The LOD score was calculated from the F-value for the multiple-regression
(Haley and Knott, 1992) as
LOD = n In (I + p*F /DFres)*0.2171
where,
p = number of parameters fitted;
F ratio = SSR(full) - SSR (red) 1 p MSE(ful1)
where;
SSR(hl1) = sum of squares for regression with full model, i.e. with QTL and
cofactors
SSR(red) = sum of squares for regression with reduced model. i.e, without the QTl.
MSE(full) = SSEIDFE = residual mean square (full model)
P MSE = number of estimated QTL effects
DFres = number of degrees of freedom for residual sum of squares in multiple
regression;
The percentage of phenotypic variance explained by a putative QTL (R2%)
was calculated. This is based on the partial correlation of the putative QTL with the
observed variable, ad.justed for colbctors (Kendall and Stuart, 1961). In the
simultaneous fit. the cofactors are ignored and only the putative QTLs initially
detected and their estimated positions werc used in multiple regressions to obtain the
final estimate of the additive effects and percentage of phenotypic variation for a
particular trait that could be explained by thc Q'CL(s). 'The adjusted R ' ~ % (adjR2%),
the finally explained portion of the phenotypic variance, was estimatcd according to
Hospital et u1. (1997). The additive erect was calculated as half the differences
between genotypic values of two homozygotes (Falconer, 1989):
Additive effect = (Parent P2 - Parent P1)/2
3.2.5.5 QTL analysis across the environments and Q x E interaction
The analysis was done with PLABQTI, (following the same procedure described
above) to identify QTLs for the traits using BLUPs across the two screening
environments and for each of the two individual screening environments. The
occurrence of additive x additive interaction was tested for significance by adding
digenic epistatic effects to the additive effects in the model. The Q x E interaction for
shoot fly resistance was estimated by a fitted model to the adjusted entry means of
each environment as described by Bohn er ul. (1996). A simultaneous analysis with
all detected putative QTLs was performed for each screening environment. The
results were obtained in the form of tables showing ANOVA and the estimated
effects.
The additive effects were obtained for all detected putative Q T L ~for each
environment as well as across the environments. The estimated MS (Q x E) were
calculated from the difference of the fits of the data from individual environnlents and
across environments. These values were tested for significance with a Sequentially
Rejective Bonfenoni F-test (SRBF).
3.3. Marker-assisted selection for shoot fly resistance traits in sorghum
3.3.1. Background of marker-assisted selection for shoot fly resistance and
component traits in sorghum
Screening for the shoot fly resistance and component traits in three environnients of
252 recombinant inbreds lines (RILs) of BTx623 X IS 18551 mapping population
was done, while genotyping using 109 SSR markers was undertaken to explore the
genomic regions associated with shoot fly resistance (Sajjanar, 2002. Folkertsama ei
ul 2005, unpublished). The genetic linkage map was constructed using Joinlilap
version 2.0 and 3.0 resulting in the formation of 10 linkage groups, with total map
length of 1468 cM. QTL analysis using PlabQTL version 1.1 revealed the presence of
28 QTL detected at least in 2 of the 3 environments (4 for leaf glossiness. 2 fhr
seedling vigor I, 5 for seedling vigor 11, 2 for abaxial leaf surface tricholiie density. 3
for adaxial leaf surface trichome density, 2 for oviposition at 14 DAE, I for
oviposition at 21 DAE, 4 for deadhearts at 21 DAE. 3 for deadhearts at 28 DAE, and
2 for seedling height I). Closely linked markers were identified for the four deadhearts
QTLs, which can be used in a marker-assisted backcrossing program. In the present
study efforts are being made to transfer the deadhearts QTLs into elite sorghum
breeding lines developed at Sorghum Research Station (SKS), M.A.U., Parbhani by
marker-aided selection using the closely linked markers. The markers associated with
shoot fly resistant traits are listed below (Table 3.5)
 Table 3.5 Target genomic regions, linked SSR markers and associated shoot fly
resistance QTLs for marker-assisted selection (Folkertsma era/., unpublished).
Linkage Associated SSR markers QTLs co-localized with genomic regions
group
A Xtxp75, Xtxp37 Deadhearts I. Ovlposition 1
E Xtxp40, Xtxp3 12 Deadliearts I. Oviposition I
G .Yixp263,XgapOl,Xtxp141 Glossiness, Tricliome density upper and
lower leaf surface. Seedling vigor 11.
Oviposition I and 11, and Deadhearts 1 and I1
J Xi.~p258,X1xp65,Xlxp15 Glossitless, Seedling vigor 11. Oviposition I
and 11. Deadhearts 1 and 11

3.3.2 Plant material

Elite sorghum breeding lines were obtained froni SRS. M.A.lJ., Parbhani (personal
communication with Dr S S Ambekar. Sorghum Breeder. SRS. M.A.U.. Parbliani).
while donor parents were obtained from 1CRISA'f. Patancheru, A.P..lndia (personal
communication with Dr R V S Reddy. Principal Scicntist, Sorghum Breeder.
ICRISAT).
Recurrent oarents: PMS 288. PMS 200, KR 192
Donor parents (ICRISAT): IS 18551 (shoot tly re~istantparent). KIL 153, RIL 189,
and RIL 252 derived from the BTx623 x IS 18551 shoot fly resistance mapping
population.
3.3.3 Salient features of'parental line used in backcross program
Recurrent parents
PMS 28B: Maintainer line developed from ICSB 940400 x MS 2968. The special
feature of this line is its long panicle with a large number of pri~~laries
and
secondaries. It is kharif adapted, tan type, good combining ability, juicy white midrib,
compact panicle, medium seed size and is susceptible to shoot fly (Plate 3.4).
PMS 20B: This A> cytoplasm maintainer line, is derived from a cross between MS
296B (A2) and SPV 900. It has bold seed with yellow endosperm. The panicle is
oblong with a large number of secondaries. It is rabi-adapted with non-tan plant type
and medium height. It is agronomically superior, has good combining ability, juicy
white midrib, awns and is susceptible to shoot fly (Plate 3.4).
 192: This is a mid-tall restorer line derived from cross SPV 544 x ~ p 462,v has
good combining ability for yield and yield-contributing characters. I t can be grown in
kharif; rabi and s m m e r seasons. I t has broad leaves, white juicy midrib, and pearly
white grain color. Panicles are compact and awnless with medium-bold grains having
50 Percent of the grain surface covered by glumes. It is grain mold resistant and shoot
fly susceptible (Plate 3.4).
Donor parents
I S 18551 (resistant parent): Origin from Ethiopia, race durra, panicles with straw-
colored grain and large glumes. Leaves o f scedlings are light green, shiny, narrow and
pointed upwards with dense trichomes. Resistant to shoot fly, very tall at maturity.
agronomically poor, four shoot fly resistance QTLs mapped (Plate 3.4).
RIL 189: This RIL was selected from the shoot fly resistant population (BTx623 x

IS 18551). I t is mid tall, has non-tan foliage, juicy white midrib, glumes partially
covered, narrow upward pointing leaves. awnless, high number o f trichomes and
glossy, three shoot fly resistance QTLs mapped.
RIL 252: Derived from shoot fly resistant mapping population BTx623 x IS 1855 1 ,
mid tall, narrow upward leaves, juicy white midrib, panicle like IS 1855 1, white grain.
awnless, high trichome density and glossy, three shoot fly resistance QTLs mapped.
R I L 153: A mid tall, juicy white midrib, medium to high trichome density, medium
glossy, and three shoot fly resistance QTLs mapped.
3.3.4 Parental genotyping with SSR markers
Approximately ten to twelve seeds/set of each o f the parents were sown individually
in small pot. Staggered sowing was employed in two sets to ensure nicking of
flowering period between donor and recurrent parents. The sowing was done at
ICRISAT, Patancheru during second and fourth week o f August 2003.
3.3.5 D N A extraction
DNA from parental plants was extracted from individual one-week-old seedlings by
using a modified CTAB method (Mace el ol.,2003) as described in section 3.1.2.
3.3.6 Selection o f the markers
A set of eleven sorghum SSR markers linked to targeted shoot fly QTLs (Table 3.6)
from four linkage group (Fig 3.3) was used for PCR amplification using DNA from
recurrent and donor parents as templates in order to identify ~ o l ~ m o r ~ SSR
hic
markers among recurrent and donor Parents.
9 '2
LGi dad
Table3.6 Fluorescent labeled sorghum SSR primers (Applied Biosystem) used for foreground selection
in markerassisted breeding for shoot fly resistance
SSR locus Linkage Label Primer sequence
group F- or R- Forward (F) Tm Reverse ( R) Tm SSR repeat motif
Xtxp37 A , F-Hex AACCTAAGAGGCCTATTTAACC 56.5 ACGGCGACTCTGTAACTCATAG 58.4 (TC)23
Xtxp75 A F-Fam CGATGCCTCGA.MWAAACG 55.9 CCGATCAGAGCGTGGCAGG 63.1 (TG) 10
Xtxp312 E F-Ned CAGGAAAATACGATCCGTGCCl63.0 GTGAACTATTCGGAAGAAGTTTGGAC 64.0 (CAA)26
Xtxp40 E F-Fam CAGCAACTTGCACllGTC 53.7 GGGAGCAATTTGGCACTAG 56.7 (GGA)7
Xi10362 E F-Hex CCTTCGTGTTTGGAAAGTT - CCGGTTGGATGAGAAGTA ATlCTlGT
Xtxp141 G F-Ned TGTATGGCCTAGCTTATCT 55.0 CAACAAGCCAACCTAAA 47.9 (GA) 23
Xisp10263 G F-Fam TATCTTCTCCGCCCTTTC - TAAGNGCCAAGGGAATG CACTG
TCCTGTllGACAAGCGCTTATA AAACATCATACGAGCTCATCAATG
1pa&l G F-Fam (AG16
Xtxpl5 J F-Fam CACAAACACTAGTGCCTTATC 55.9 CATAGACACCTAGGCCATC 56.7 (TC)16
Xtxp65 J F.Hex CACGTCGTCACCAACCAA 56.0 GTTAAACGAAAGGGAAATGGC 55.9 (ACC)4(CCA)3CG(CT)8
Xisp10258 J F-Ned GCAGGACCGGATAGAGAT - ATCCCGGAATGATGAAGT CAAlCCG
 $ -
-8 -
-= (I)

-
5- =
D
5

#sm#a8$i ;g# 8, 8;
.s%Be2,se
XXXXXXXX 8:X rX
X 38
X X 2
X
gj
I I
0
I 1
0NV)rrnV)lc
N 3 R 8$ ! r o
kt

\I
II
mbm ammo
00
-,a mV)V)(Ds8c~a~$!$

-
-- 8
- - .c
(I)
6-
5 88P.S
,P 4@$
X X X X X
3.3.7Amplifieation of SSR markers
PCR reactiotls were conducted in 384 wells plates in a PE 9700 Perkin Elmer
(Norwalk Conn.) DNA thermocycler. 'The reactions were performed in volumes of
511 using four different PCR protocols and a touchdowll PCR program. Reaction
mixture contains 10 mM Tris-HCI (pH 8.3). 50 mM KCI. 1.25-2.5 ng of DNA. 2pm
of forward and reverse primer, ImM MgC12. 80-100 14M of each dNTP and 0.1 units
of Tug DNA polymerase. PCR conditions mere previously optimized for each primer
using a grid of nine reactions Smith el ul., (1995). Three different protocols were used
for PCR amplification i.e. protocol 5, 7, and 4.
The touch down PCR program consisted of initial denaturation for 15 min at
94°C and then 10 cycles of denaturation for 10 sec at 94OC, annealing at 61-52'C for
20 sec, the annealing temperature for each cycle is reduced with 1'C. and extension at
72'C for 30 sec; 35 cycles (for SSRs screened using PAGE) and 31 cycles (for SSRs
screened using the ABI sequencer) of denaturation for 10 sec at 94'C. annealing at
54°C for 20 sec and extension at 72'C tbr 30 sec. The last I'CR cycle was followed by
a 20 min extension at 72'C to ensure an~plificarionto equal lengths of both DNA
strands.
If the parents showed PCR product polymorphism of more than 5 bp. then
PCR products were separated on 6% non-denaturing PAGE (Poly Acrylaniide Gel
Electrophoresis) gels and silver stained using the modified procedure of Tegelstrom
(1998). If the polymorphisnl between the parents is less than 5 bp, then PCK products
were separated by capillary electrophoresis using a ABI Pris~il3700 (Perkin Elmer)
sequencing. For this purpose fluorescent-labeled primes were used. All ABI primers
were screened with the help of Dr. Rolf Folkertsma, PDF. ICRISA'I'.
3.3.8 Non-denaturing PAGE (Poly Acrylamide Gel Electrophoresis)
1 p1 of loading dye (orange red + EDTA + NaCl + Glycerol) was added ro 3 p1 of
PCR product. From this mixture. 2pl of the sample was loaded into a well of the 6%
non-denaturing PAGE gel.
The gel was prepared using
52.5 ml of double distilled water; 7.5 ml of IOX TBE buffer; 15.0 ml of
Acry1amide:Bis-acrylamide (29:l) solution 450 p1 of Ammonium Per Sulphate
(APS); and 100 p1 of TEMED.
Along with the samples, the 100 bp marker ladder (50 ngipl) was also loaded in the
first and last lanes of the gel to ensure proper sizing of amplified PCR fragments.
Most of the markers used allowed clear differentiation of donor and recurrent parent
alleles. The gel was run at 550V of constant power in 0 . 5 ~TBE buffer for 3 hours
using a BioRad sequencing gel apparatus.
3.3.9 Silver staining
After running of PAGE gels for the required lime, the gels were developed by silver
staining.
Seauential stem involved in silver staining
The gel was placed in
i water for 5 min
i 0.1% CTAB solution for 20 nlin (2 g in 2 lit of water)
i 0.3% ammonia solution for 15 min (26 nil of 25% a~nmoniasolution in 2 lit or
water)
0.1% silver nitrate solution for 15 min ( 2 g of silver nitrate t 8 ml of 1 M
NaOH in 2 lit of water and add ammonia solution up to the solution becomes
colorless)
i water for few sec
i Developer (30 g of sodiu~ilcarbonate + 400 p1 of formaldehyde in 2 lit of
water)
After developing the bands gels were rinsed in water for 1 minute and placed it in
fixer (30 ml glycerol in 2 lit of water) for a few seconds.
Continuous shaking is required throughout the silver staining procedure.
After silver staining of the PAGE gels, the size (base pair) of the intensely
amplified specific bands or alleles for each SSR marker was estimated based on its
migration relative to the 100 bp DNA ladder (fragments ranging from 100 bp to 1000
bp) and presence or absence of parental alleles were scored.
3.3.10 Data collection and analysis
Scoring of the gels
The silver stained bands of amplified PCR products in the gels were scored as A. B,

II, OFF and "-" based on their pattern compared with those of the parents. "A" was

defined as the homozygote for the allele from the recurrent parent (28Bi20BIKR 192),

"B" was defined as the homozygote for the allele from donor parent IS 18551iRIL

189lRIL 252lRIL 153, "H" was defined as the 1ieteror.qgote (presence o r both

recurrent and donor parent alleles), "OFF" h a s defined as an allele observed from

unknoun source, and "-" u a s a ~iiissing sample. During parental polymorphism

studies. amplified PCR products were scored as "A" and "U", wh~le for the

backcrossing program amplified PCR products were scored as "A". "B" and "11".

In parental genotyping the PCR product was run on llic autoniatic DNA

sequencer (AH1 3700) and the a~ilplifiedallele size was scored. Polymorphic SSK

markers, that flanked targeted QTL of each of four linkage group, were identified.

Based on parental allelic size the pure ho~nozygousplants were selected and tagged at

seedling stage and these plants adcanced for crossing to produce FI hybrids (plant to

plant crosses). A detailed list of the numbers of plant yenotyped and numbers of pure

plants selected for crossing is presented below (Table 3.7).

 Table 3.7 Summary of SSR markers used to identify pure parental plants
employed for FI hybrid production
Nameof Plant No. of No, of SSR markers used No. of parental
parent numbers plants type plants
geno- selected for
typed crossing
208 SPI-12 12 Xl~p75.~Y/xp37,Xl.rp15.X/xp225.?rixp6S.09
SP 177-186 10 09
,I(i~pl0258.Mxpl4l,Sgu1~l,Xi~p10263.
Xlxp3 12, ,Y,~p40,Xi.s/)10362
28B SP133-135 03 03
X/xp75,Xlxp37.?tix/~15.X1,~p225,Xix/~65.
SP287-296 10 ,I(i.v1110258. Xlxpl41. Xgul1l.,Yis/)10263. 08
,Yt,xp3 12. ,Ytxp40, ,Yisp10362
KR 192 SP 141-152 12 X1xp75. Mxp37. Xlxpl5. S1xp225.Xl~p65. 05
9'297-306 10 Xi~s~110258,
X/.ryl4l. Xgopl,Xi.sp10263. 06
Xr.up312, Xlxl140. ,YiYi>pl
0362
IS 18551 SP 109-120 12 ,Y/,rp75. X1xp37, Mxpl 5. X/xj1225. .i'/x1165, 11
SP267-176 10 XisplOZS8,X/xp141.,Ygc1pl,~Yi~rpl0263.08
?i',.xp3 12, Xlxp40, Xisp 10362
R1L 189 SP 153-164 12 Xxp75. Xi.uy37,.Y1xp15,~Y1xp225,X1x~~65,
05
SP 307-316 10 Xisj110258, X1xp141,,Ygopl,Xi~,1>10263, 06
X l ~ p 312, ,YIX/J~O.
Xi.T/110362
R11, 153 SP 85-96 12 X/xp75.,Y/x1137, ,Y/xp15, Xl.~p2?5,Stx/165. 04
SP 247-256 10 X/x~~14l,Xgupl,.~splO263.04
d%\i;.~p10258,
Xxp312, Xtxp40, XisplO362
RIL 252 SP 165-176 12 06
Xtxp75,X~xp37,X1xp15.,Y~xp225,S/.u/165.
SP 317-326 10 ,Yisp10258,,Y&p141,Xgc1pl,Xisj1l0263. 05
X1xp3 12, X1xp40. Xispl0362
In the present study we have tried to produce seven hybrids by using three recurrent
parents and four donor parents, as listed below.
I. 28B (288) x RIL 189 (312)
2. KR 192 (304) x IS 18551 (267)
3. 20B(186) xRIL252(318)
4. 20B (179) x RIL 153 (248)
5. KR192(3OO)xRIL252(319)
6. 28B (293) x IS 18551 (268)
7. 28B (292) x RlL 153 (252)
3.3.1 1 Testing of hybridity with SSR markers
Staggered sowing was employed to ensure nicking of flowering period. The sowing
was done in two sets between the second and fourth weeks of April 2004. Five and
fifteen seeds each of seven hybrids and three recurrent parents, respectively, were
sown individually in pots for each set of sowing. DNA was extracted from one-week-
old seedling leaf tissue and genotyping of each FI plant was undertaken with four SSR
markers flanking the targeted QTLs in each linkage group. Heterozygous plants in the
five hybrid populations were chosen for each targeted QTLs at the seedling stage and
these plants were advanced by crossing to produce BCIFl seed. A detailed list of
numbers of plant genotyped, markers used, numbers of selected heterozygous plants,
and numbers of backcrosses effected and advanced to the BCIFl generation is
presented Table 3.8.
3.3.12 Testing of recurrent parent purity with SSR markers
DNA was extracted from each of the recurrent parent plants and genotyping of each
recurrent parental plants was done with four SSR markers to test their purity. The
plants found homozygous at all loci were used in the backcrossing program. The
details of numbers of plants genotyped and numbers of pure plants selected and used
for pollination are presented in Table 3.9.
3.3.13 Cenotyping BClF, populations with SSR markers for foreground selection
Ten seeds of each of five BCIFl populations, eight seeds of each of three recurrent
parents and one seed of each of four donor parents were sown for each sowing of a
total of three sets of staggered sowings to ensure nicking of the flowering period.
Sowing of first set was done during the fourth week of August 2004, the second set in
Table3.8. Details of numbers of plant genotyped, markers used, numbers of selected heterozygous plants and numbers of backcrosses
effected and advanced to the BClFl generation
Name of hybnds No.of Details of SSR markers LG No. of Plants No. of Detail of crosses Crosses advanced to
pbots plants used heterozy ous selected backcrosses BC,F, generation
genotyped plants sekcted efrected
2 8 8 288) x PJL 05 SP541-545 Xtxp75, A 02 SP 541, 02 SP 541 x SP405 SP 541 xSP405
189 &12) Xmp37 G SP 542 SP 542 x SP 406 SP 542 xSP 406
xff4p1bzss
xtx I J

KR 192 (300) x 09 SP 561-570 Xrxp75, A 08 SP561, 08 SP561 xSP465 SP561 SP465

RIL 252 (3 19) Xtxp37Xtrp41 G SP 563, SP 563 x SP465 SP568 SP465
,i%pl0258 J SP 565, SP565 x SP462 SP566 x SP466
SP 566, SP 566 SP 466
SP 567, SP 567 x SP 463
SP 568, SP 568 SP 465
SP 569, SP569x SP464
SP 570 SP570 SP464
288 293) x IS I0 SPSOI-510 X1~p75,~Ylxp A 0
18555 (268) 37x1~41 G
~;~~l&58' J
288 292) x RIL Nu1 -
153 6.52) gemmated --
Table3.9. Details of numbers of recurrent parental plants genotyped and number of pure plants selected and used for pollination

Name of No. of plants Details of SSR markers used LG No. of Plants used in Details of
recurrent genotyped plants homozygous backcrossing plants
parent plants selected
28B (288) 30 SP 401-430 Xap75,Xap37 A 25 02 SP 405,SP 406
Xap4 1 G
XISP10258 J
September first week, and the third set in the second week of September 2004 at
Patancheru.
Individual plant DNA of each BClF, plant, recurrent parent plant and donor
parent plant was extracted from one-week-old seedling tissues. Marker genotyping of
individual BClFl plants, recurrent and donor parent plants was done with eleven SSR
markers flanking the targeted shoot fly resistance QT1.s (foreground selection) on
four linkage groups. A total of 69 heterozygous plants having an appropriate allelic
constitution were selected (Table 3.10) before floweri~ig from five BClFl
populations. These selected plants were crossed with their respective recurrent parents
(plant x plant cross) to produce BCzFi seed.
Each recurrent parental plant was genotyped with eleven SSR markers and
plants conforming to parental type allelic constitution were selected (Table 3.1 1) and
used as pollinators.
Table3.10 Details of genotyping ofBC,F, plants with linked marker (loreground selection)
Name of BC,F,- populntwn
. Number of Detail of SSK markers used LC Heterozygous plants selected Total
plant plants selected
ecnotyped ~lants
BC,F, 288 (288) x RIL 189 (312) x 28B (288)
IF, (SP 541)l x SP 405 10 SP610-619 Xlx~75.Xrxp37. A SP612,SP613,SP617,SP619
Xlxp 40, Xixp 10362, Xrxp 3 12, E
Xup 10263, Xgap I. Xlxp 4 1 G
xup10258,xixp65, xlxp IS J
[F, (SP 541)] x SP405 5 SP710-714 SP710,SP711

F,(SP 542)] x SP 406 5 SP710-714 SP 719

[F, (SP 542)] x SP 406

[F, (SP 51 111 x SP 477 I0 SP 629- 638 ,Y&p75. XLrp37, A SP 629. SP 630, SP 633, SP 636
XIxp40, Xixp 10362. Xrxp 3 12. E
Xixp 10263, Xgap I. XLrp41 G
XLrp 10258, Xhp65, X&p I5 J
[F, (SP 513)] x SP 476 10 SP729-738 SP 729, SP 73 1, SP 732. SP 736, SP 737

[F, (SP 518) x SP 482 10 SP 829- 838 SP 830, SP 832, SP 833, SP 834, SP 835,
SP 036

BC,F, 20B (186) x RIL 252 (318) x 208 (186)

[F, (SP 525)] x SP431 10 SP 648- 657 Xlxp 75, XLTP37, A SP 649, SP 650, SP 651, SP 655, SP 656, SP 657
Xrxp40, Xrxp 10362, Xmp3 12, E
Xirp 10263, Xgup I , XIxp41 G
Xcxp 10258, Xlxp 65, <U.YPI5 J

[F, (SP 525)) x SP 431 10 SP748-757 SP 754, SP 757

[F, (SP 525)] x SP 43 1 10 SP 848- 857 SP 848, SP 849, SP 850, SP 853 12
a a
V) rn
-- V)
m-
R P
rn % %
m- w r-- ID P-- ID
W P - ID C VI P-
W I D r-"-0000
a 4 a n 4 n.
V)
V) mvlcnrn

4 w o -

rn
Table 3.11 Details of recurrent parental plants genotyped and number of
parental type plants selected for (BCIFI)backcrossing
Name of No. of Details of Parental type plant used in backcrossing
recurrent plants plants
parent genotyped*
288 (288)
SP 405 8 SP 601-608 SP 601. SP 602. SP 604. SP 605. SP 606. SP 607. SF
SP405 8 SP 701-708 SP 707, SP 708
SP406 8 SP 801-808 Nil

S1'476 8 SP 820-827 SI' 821, SI' 822. SP 823. SP 826

* Eleven SSR markers used to evaluate purity of recurrent parental plants: Xr?p75,
Xlxp37 (LC: A), Xtxp40, Xtxp312, Xsp10362 (LG: E), Xgupl, Xtxpl41. Xisp10263
(LG: G), Xisp10258, Xtxp65, Xrxpl5 (LC: J)
 Out of 69 heterozygous BCIFlplants selected on the basis of foreground
molecular data. 8 plants were discarded (Table 3.12) due by infection of maize streak
virus at seedling stage.
Table 3.12 Details of discarded BCIFl plants
Name of backcross No, of Details of plant removed
plants
rcniobed
[28B (288) x RIL 189 (312)l x 28B (288) 3 SP 612, SP 613, SP 619
[KR 192 (304) x IS 18551 (267)j x KR 192 (304) 1 SP 636
[20B (186) x RIL 252 (318)] x 208 (186) 1 SP 650
[20B (179) x R11. 153 (248)l 208 (179) 3 SP 672. SP 771. SP 775
[KR 192 (300) x RIL 252 (31911 KR 192 (300) 0

3.3.14 Parental polymorphism using SSR markers (for background selection)

Initial parental screening with 38 SSR tnarkers was carried out before actual
genotyping of selected BClFl plants. Tlie main objective for screening parental plants
with these 38 SSR markers (Table 3.13) was to detect polymorphism among the
parents. I'CR amplification of each SSR marker was perfornied in a total volume of 5
ul reaction mixture containing parental genomic DNA. PCR buffer. DNTPs, IvlgC12.
foruard primer labeled with HEX. NED or FAM dye phosporaniidities, reverse
primers and AnipliTaq gold DNA polymerase. PCR reactions were conducted in 96-
well plates in a PE 9700 Perklin-Elimer. DNA thermocycler using a touch down PCK
technique. Tlie PCR products were run on an autotnatic DNA sequencer (i.e. capillary
electrophoresis using an AB1 3700). After completion of the sequencer run, presence
or absence of allelic fragments was scored using the Genotyper software. When the
parents show polymorphism of more than 5 base pairs, the PCR products of the
respective backcrossing population plants were separated on 6% non-denaturing
PAGE (Poly Acrylamide Gel Electrophoresis) gels. If the polymorphisni betueen the
parents was less than 5 base pairs the PCR products were separated by capillary
electrophoresis using an ABI Prism 3700 (Perkin-Elimer) DNA sequencer. For this
purpose fluorescent labeled primers used. The details regarding fluorescent labeled
sorghum SSR primers used in parental polymorphisn~of the backcrossing populations
presented in Table 3.13 and 3.14.
-'=not amlified, Monomorphic parent pairs highlited in bold font
Table3.14 Fluorescent labeled sorghum SSR primers (Applied Biosystems) used for background selection in marker-assisted breeding for shoot fly resistance

SSR locus Llnkaae Label

~ m u p F- or R- Forward (F) primer sequence Tm Reverse ( R) primer sequence Tm SSR mow
XbrP 2s B F-Fam CCATTGAGCTTCTGCTATCTC 57.9 CATTTGTCACCACTAGAACCC 57.9 (cT)12
X&D 298 B F-Fam GCATGTGTCAGATGATCTGGTGA 60.6 GCTGTTAGCTTCTTCTAATCGTCGGT 63.2 (AGA)23
~ L x296
p B F-Hex CAGAAATAACATATAATGATGGGGTGAA 59 3 ATGCTGTTATGAmAGAGCCTGTAGAGn 62 7 (CAJ18
Xcup 63 B F-Ned GTAAAGGGCAAGGCAACAAG GCCCTACAAAATCTGCAAGC 573 (GGATGCp
xhp 283 CGCCCGAACTCTTCTTAAATCT ATTATGCCCTAACTGCCTTTGA
xcup 11 TACCGCCATGTCATCATCAG CGTATCGCAAGCTGTGTTTG
Xbrp 228 ACAGGTTGGCGATGTTTCTCT TTCTTnTCGAATTCArTCCTTTr
xtxp31 TGCGAGGCTGCCCTACTAG TGGACGTACCTATTGGTGC
X w p 32 ACTACCACCAGGCACCACTC GTACTTTlTCCCTGCCCTCC
x w p 61 TTAGCATGTCCACCACAACC AAAGCAACTCGTCTGATCCC
Xhpll4 CGTClTCTACCGCGTCCT CATAATCCCACTCAACAATCC
Xgap 236 GCCAAGAGAAACACAAACAA AGCAATGTAnTAGGCAACACA
Xcup 14 TACATCACAGCAGGGACAGG CTGGAAAGCCGAGCAGTATG
Xbrp 59 GAAATCCACGATAGGGTAAGG GACCCAGAATAGAAGAGAGG
XgsPlO GTGCCGCTTTGCTCGCA TGCTATGTTGTTTGCTTGTCCCTTCTC
Xhp2l GAGCTGCCATAGATTTGGTCG ACCTCGTCCCACCTTTGTTG
X w p 28 GGTGTGAGACTGTGAGCAGC TATAGCACGGTTGTTGTGCC
X&p 27 AACCTTGCCCTATCCACCTC TATGATGAATCAAGGGAGAGG
Gpsb 050 GGCTTCTTCCTCTCC GAGrTCTTTrATGrnGTGT
m p343 CGATTGGACATAAGTGTrC TATAAACATCAGCAGAGGTG
Xgap 206 ATTCATCATCCTCATCCTCGTAGAA AAAAACCAACCCGACCCACTC
XLxp 289 AAGTGGGGTGAAGAGATA CTGCCTTTCCGACTC
xcup 02 GACGCAGCmGCTCCTATC GTCCAACCAACCCACGTATC
Xbrp 258 CACCAAGTGTCGCGAACTGAA GCTTAGTGTGAGCGCTGACCAG
X h p 230 GCTACCGCTGCTGCTCT AGGGGGCATCCAAGAAAT
Xho 321
7 - TAACCCAAGCCTGAGCATAAGA CCCATTCACACATGAGACGAG
mp47 H F-Ned CAATGGCTTGCACATGTCCTA 57 9 GGTGCGAGCTAGTTAAGTGGG 61 8 IGT)B(GCJ5+(Gn6
X ~ 273D H F Fam GTACCCAmAAATTGmGCAGTAG CAGAGGAGGAGGAAGAGAAGG (TTGJPO
XhploS H TGGTATGGGACTGGACGG - TtiTTGACGAAGCAACTCCAAl
XtxpPlO H CGCTmCTGAAAATAnAAGGAC 55 9 GATGAGCGATGGAGGAGAG
m p 354 H TGGGCAGGGTATCTAACTGA 57 3 GCCTTI IICTGAGCClTGA
m p 57 I GGAACTTTTGACGCGTAGTGC 59 8 CGATCGTGATGTCCCAATC
Xt?p 145 I GTTCCTCCTGCCATTACT 53 7 CTTCCGCACAICCAC
Xbrp317 I CCTCCl-rrTCCTCCTCCTCCC 63 7 TCAGAATCCTAGCCACCGTTG 59 6 (CCT)5(CAT)11
Xtxp 265 I GTCTACAGGCGTGCAAATAAAA 56 5 TTACCATGCTACCCCTAAAAGTGC 61 (GAA)l9
Xtxp 274 I GAAATTACAATGCTACCCCTAAAAGT - ACTCTACTCCTTCCGTCCACAT wc)lg
Xtxp 6 I ATCGGATCCGTCAGATC 52.8 TCTAGGGAGGTTGCCAC 52.8 (cT)33
xtxp 17 I CGGACCAACGACGATTATC - ACTCGTCTCACTGCAATACTG (TC)16+(AG)12
3.3.15 Screening of BCIFIselected foreground plants for background selection
A total of 61 BCIFI plants selected through foreground screening from 5 backcross
populations were genotyped with 38 SSR marker loci covering the entire genome
except the region harboring targeted QTLs (i.e. regions covered in foreground
screening). Approximately 2 SSR marker loci were selected to cover the top, middle
and bottom portion of the six non-target linkage groups. The main objective of
background genotyping was to ascertain recovery of the recurrent parent genome.
Out of 61 BClFl plants, 12 plants were selected carrying recurrent parental
alleles at most of the SSR loci used for background screening along nith a few
heterozygous loci. Those individuals homozygous for donor parent allele "B" type
were rejected as they could only be derived from selfing of their FI female parent, The
backcross selected 12 individuals advanced Ibr planting the BC# generation, A
detailed list of background genotyping i.e. number of plants genotyped. number of
markers used, plants selected with each targeted QTL, and advanced to the BCzFl
generation is presented in Table 3.1 5.
3.3.16 Genotyping of BC2FI population with SSR marker for fore ground
selection
3.3.16.1 Planting of BC2FI population
Numbers of plants to be sown from each of five BCzFl populations to recover a given
number of plants possessing the target QTLs were estimated following Sedcole
(1977). 'The sowing of five BCzFI populations and four-donor parents (Table 3.16)
was done in two sets. Staggered sowing was employed to ensure nicking of flowering
period. The first set was sown in the first week of April 2005 and the second was
sown in the 2""eek of April 2005. Sowing of the recurrent parent done in four sets
(Table 3.17). with the first set sown in the last week of March, the second set sown in
the first week of April, the third set was sown in the second week of April, and the
fourth set was sown in the third week of April 2005. One seed was sonn in each pot,
with pots filled with a mixture of sand, vertisol. and FYM (1:l : I , V: V: V).
 Table 3.16 Details of number of seed planted crosswise and set wise in BCIFI
foreground selection
No. of seed sown Total no.
Name of the cross of plants
Set 1 Set I1
BC2Fl 288 (288) x RIL 189 (312) x 28B (288)
(BCIFI (719)) x SP 606 07 07 14
(BCIFI (814)) x SP 607 07 07 14
(BCIFI (818)) x SP 608 14 14 28
BC2Fl KR 192 (304) x IS 18551 (267) x KR 192 (304)
(BC 1F1 (620)) x SP 62 1 07 07 14
(RCI Fl (830)) x SP 823 14 I4 28
BC2Fl 208 (186) x RIL 252 (318) x 20U (186)
(ACIFI (757)) x SP 740 07 07 14
BC2Fl 20B (179) x RIL 153 (248) x 208 (179)
(UCIFI (669)) x SP 661 07 07 14
(BCIFI (773)) x SP 759 14 14 28
( B C l r l (871)) x SP 865 07 07 14
(BCIFI (874)) x SP 860 07 07 14
BC2FI KR 192 (300) x RIL 252 (319) x KK 192 (300)
(BCIFI (889)) x SP 877 14 14 28
(13C 1T 1 (895)) x SI' 878 07 07 14
Donor parents
IS I8551 (267) SP 828 02 02 04
RIL 252 (3 18) SP 847 02 02 04
RIL 252 (319) SP 885 02 02 04
RIL 153 (248) SP 866 02 02 04
RIL 189 (3 12) SP 809 02 02 04
The sowing of three recurrent parents (Table 3.17) was done in four sets to ensure the
sufficient pollen availability during the peak crossing period.
Table 3.17 Details of set-wise sowings of recurrent parents
Namelplant no. of No. of seeds sown
Total
recurrent parent Set 1 Set I1 Set I l l Set l V

3.3.17 Genotyping of the BC2FI populations with SSR markers for foreground
selection
DNA of individual plants of each BC212,population was extracted from one week old
leaf tissue. In case of recurrent parents leaf tissue from one representative plant from
each set was used for DNA extraction, and DNA of four plants was pooled in one
well. The same procedure was applied for donor parent plant DNA extraction.
Genotyping of 224 BC2FLplants was accomplished with 11 SSR marker loci linked to
targeted shoot fly resistance QTLs in four linkage groups (A, E, G and J). Around 100
heterozygous BC2FIplants having appropriate allelic constitution were selected before
flowering and crossed with their respective recurrent parents (plant to plant crosses) to
produce B C ~ F Iseeds. Details of numbers of plants genotyped, markers used and
numbers of heterozygous plants selected for each of the targeted QTLs are presented
in Table 3.18.
For recurrent parent plant populations, self-seed of selected genotyped plants
used in backcrossing the Fl plants and selected for their purity were sown Tor
backcrossing to develop the B C ~ F populations.
I The details of recurrent parent seed
planted and plants used in backcrossing to develop the BC2Fl generation are listed in
Table 3.19.
 Table 3.19 Detail of seed planted of each recurrent parent and plant used in
backcrossing to advance BC2FIgeneration
Name of No. of Plant nos. and No. of Detail of plants used as pollinators
recurrent seeds source plants used
parent sown in as
Isets pollinators
288 (288) 60
3.3.18 Genotyping BC2F1 selected foreground plant for back ground selection
68 B C ~ F Iselected on the basis foreground genotyping from five backcrossing
populations will be genotyped with 38 SSR marker loci (Table 3.20) covering the
entire genome except the region harboring targeted QTLs. This background screening
will be restricted to the loci that were heterozygous in the previous generation. Also
genotyping will be done for loci that were not amplified in BClFl background
screening.
Table 3.20 Detail list of selected BClFl plants for background screening
Backcrossing population Number of Details of Targeted
plants plants QTL
selected selected
BC2FI [28B x (288) x RIL 189 (312)l x 28B I8 1009 A
(288) x 28B (288) 1010 A
1012 A
1013 A
1017 E
1019 E+G
1026 J
1027 G
1207 A+J
1212 A
1215 G
1216 E
1218 E+G
1219 E
1221 G
1224 G
1226 E
1227 E+G
BC2FI [KR 192 (304) x IS 18551 (267)] x KR 12 1031 A+G
192 (304)] x KR 192 (304) 1033 A+G
1034 G
1039 A
BC2Fl [20B (186) x RIL 252 (318)l x 04
20B(186) x 20B (186)

BC2Fl [20B (179) RIL 153 (248)l x 200 24

(186) x 20B (1 86)
 1271 A+J
1277 A+J
1285 E
BCzFl [KR 192 (300) x RIL 252 (319)l x KR 10 1092 A
192 (300) x KR 192 (300) 1094 J
1101 A
1103 J
1105 A
1106 G
1109 G
1296 A
1298 A+J
1301 J
RESULTS
 CHAPTER IV
RESULTS
Sorgliitm [Sorghirrn hicoior (L.) Moencli] is an important cereal crop of semi-arid
regions of Asia, Africa. tlie Americas and Australia. Generally tlie lo\%eryields in
Asia and Africa are associated witli pest damage. Shoot fly is one of the major pests
of cultivated sorghum in Asia and Africa. Adaptation of chemical control tiietliod is
not economically feasible for most of the resource poor sorghum-growing rartners of
Asia and Africa, Therefore utilization of host-plant resistance is the most realistic
approach to reduce losses caused by sorghum insect pest. The quantitative nature of
resistance to this insect, and large environ~iictitalvariation it1 its expression hinders
genetic ~iia~iip~tlatio~i
of slloot fly resistance by coti\e1itiona1 plant breeding
procedures. Therefore efforts are being ~iiadeto this end in current Pli. D. research
project on ' genetic diversity analysis, QTL mapping ant1 marker assisted selectio~ifor
shoot fly resistance in sorghum.' 'rhe results obtaincd fionl these studies tire presented
belo\\ objective\vise.
4.1 Application of SSR markers in diversity analysis of sorgl~uniinsect resistant
ger~nplasmaccessions
Assessn~enlof gciirtic diversity based on SSK marker genotypes was conducted in
sorghum gemiplasm accessio~isresistant lo sorgliiim shoot fly. spotted stetii borcr.
and sorghum lilidge following separation of P('R products by polyacryla~nide gel
electropliorrsis (PAGE) and capillary electropl~oresis(ARI). Oul of 21 SSR primel.
pairs used for ditersity analysis. that for ,YCi~pl4\\as iiot itsed for PAGE and that for
.\i.111)32 \vas not ~ ~ s eond the AH! because of tlieir poor amplificatio~i.Data obtained
from 70 SSR loci lbr eacli separation tnetliod (19 common loci across tlie trio
tnetliods) \\.ere used to esti~iiatetlie genetic iliversit~ o r tlie 91 sorghum genotypes
studied.
4.1.1 Polyacrylamide gel electrophoresis (PAGE)
Twenty primer pairs \\ere used for separation of PCK products using DNA samples
from 91 sorghum genotypes. The allelic cor~ipositiori of eacli genotype was
deteniiined and scored individually from tlie amplified products separated on 6%
denaturing polyacr?.la~iiidegels. Only 11 o r 20 SSR primer pairs revealed high levels
of polymorphism on silver-stained PAGE gels. A total of 69 alleles were detected by
silver staining usi~igthese 20 SSR pri~ners.On an average 3.45 fragments were
anlplified per SSR locus for the 91 sorghum genotypes studied. Gel image of tlie most
polymorphic SSR markers are presented in (Plate. 4.1).
The polymorphism infor~natioticontent (PIC) value of these 20 SSR markers
were calculated from the 91 sorgl~umgenotypes evaluated. These 20 SSR markers
revealed high levels of polymorphist~i:I I of 20 (i.e., 55%) of tlie SSR primer pairs
used detected high levels of polymorpliis~n with PIC values 20.5. The PIC balue
range o b s e ~ ~ w
was
l 0.13 to 0.83. Tlie highest level of polymorphism was found with
primer pair Sh6-84 (0.83). which detects locus Xg'k.rlp84.followed by those for .l'~xpIS
(0.82). and S/.~l1320(0.77). Tlie lowest poly~liorpliis~n
was found with tlie primer pail
for .\i-1/1,62 (0.13) (Table 4.1 ).
4.1.2 Capill;iry electropl~oresis(ABI)
'fhc genotypes studied using separation of PCK products PAGE were also assessed
for their polymorphism using ~iutomatedcapillary electrophoresis (A131 3100 or ABI
3700 DNA sequencing machines). A tolal of 118 alleles generated by 20 SSR primers
a c r e detected using the automated capillary electrclpliorcsis system. O n average. 5.1
Sragments were amplilied per SSR locus. The I'IC value of each of the 20 SSR
markers was calculated from the ADI-generated data set for tile 91 sorgllum
geuutypes. In this data set I3120 (i.e.. 65%) of primer pairs detected high levels o f
~pol!~norphisn~
with PIC \ d u e s 10.5. The PIC \,slues observed mngcd lixlrn 0.21 to
0.81. 'l'he liiglies~level of pol!morpliis~n was found wit11 the prinier pairs li>r,Yt.~1~320
(0.81) and .Yts1115 (0.81 1 li>llo\vrd b! pri~ilerpair Sb6-84 (0.79). and tile lo\cest
IXII! morpl~isni(-fable4. I ) b a s liiund \\it11 the primer pair for .\i.i11160 (0.21 ). Thus the
~liostpolymo~pliicgroup of sorgll~1111
SSR ~ilari\ersdid not show substantial changes
across tlie two I'CR product sepal.stion and visualization systems (silver-stained gels
from PAGE and electronically captured fluorescence measurements during capillary
electrophoresis).
4.1.3 PAGE dendrogram
Tlie dendogram for genetic similarit! between genotypes based on the PAGE-
generated data set showed clustering for geographical origins. sorgllum races. raw
germl~lasnlvs. elitelimproved breeding lines (including stay-green recurrent parents).
and specific traits sucli as insect resistance. The accessions studied were broadly
grot~pedinto clusters representing four of the five sorglii~mraces [dtirro, cul~dtrt~riu
(

elite lines derived from zera-zera landmces). hicolor., and griinetr] according to their
molecular diversity. The correlation coefficient bet~veen the cophenetic matrix
I'late 4.1: Silver-stained gels showing write of the more polymor[hic SSR [markers
used in divenity study.
Table 4.1: Multiplex primer sets used for amplification of SSRs: allele size, No of alleles and Polymorphic information content (%)
obtained in diversity study.
PAGE ABI
Expected Expected Observed allele size Observed
allele size Number (bp) number of Observed
SSR locus (bp) of alleles alleles PIC*) Observed allele size (bp) number of alleles PIC
Xcup62 200 2 32012951190 3 0.13 1861189 2 0.49

22512221220121812 1011 22511 261224122212201218121

Xtxp I5 240 3 6 0.82 12 0.81
80 6121412 12121012081198
Table 4.1 Cont.---

Xcupl l 185 3 1851180 2 0.48 122112111681163/162 5 0.53

Xcup 7 271/2681267/2591256/253/19
2001280 5 290128012701190 4 0.55 8 0.55
31190

Xcupl4 200 6 --- --- ---

Total
number of
69 118
alleles
computed fro111 tlie dendrogra~nand the original similarity matrix was r = 0.74 (t -
16.7, P=l). The results suggested good fit o f the tree generated from the rougll data.
The coefficient of similarity values ranged from 0.28 to 1.00 (Fig. 4.1).
Ninetyolie sorghum genotypes could be separated into two super clusters at
the 30% level of similarity. The first super cluster consisted largely of ilr~rr~ilandrace
genotypes resistant to sorghum slloot fly andlor spotted ste~iiborer. The second super
cluster consisted largely of genotypes having resistance to sorghum ~iiidgeand elite
zera-7era derivatives being used as recurrent parents in a marker-assisted backcross
programme for tlie stay-green trait. These 91 sorghum genotypes diverged into 20
clusters at approximately the 50% level of similarity. /\mong these 20 clusters, the
largest was cluster 4 (18 genotypes): follo\hed by cluster 12 (13 genotypes); cluster
I? (12 genotypes): and. cluster 11( 9 genotypes). l4owever. some of the clusters (e.g..
clusters 5. 7. 9. 10. 14. 19 and 20) accommodated only single genotype (IS 17948.
ICSH 4.57. IS 2269, IS 4756. Sup1lanbur.i 160. IS 19476, and SSG 59-2, respectivcly).
Clusters 1. 3. and 6 acconi~ilodatedonly two genotypes each.
Cluster 1 had two getlotypes 296B and HC 260. This cluster --as well
sul~purled\\it11 an operational bootstrap value of 61.4%. LIlite hybrid niaintainer line
296B is susceptible insect pests hut is a potent combiner t'or higli grain yield. It has
been ilscd cutensi\el> ill hybritl development programs in India. Tliis elite line is used
as a susceptible pal.ent in shoot 111 resistance OTL mapping studies reported else
where ill this thesis.
Cluster 4 included 18 genotypes. most o r \ c h i c l ~exhibits resistance to sorglluni
shoo( tly andlor spotted stem borcr. All sorghi~mgenotypes in the cluster originated
from the race. This group contains gcnotypc IS 18551. wliich lias been used as
~iiii.~.ii

the resistant parent in development of two ICI<ISA'I' s o r g l l ~ ~mapping

ti~ populations
targeting shoot fly resistance. Most of tlie genot>pes found in this cluster possess late
maturit! and ruediuni grain yield potential but lia\e relatively high degrec if insect
resistance. IS 22121. IS 2265. IS 2312. IS 2 \ 9 5 . ant1 IS 2123 have been identified as
sources of resistance against both sorghum shoot fly and spotted stem borer. Most of
the genotype pairs in this cluster exhibited operatio~ialbootstrap values greater than
5046. wl~ichprovides confidence about their clustering. In pal-ticular. genotype pairs
involving the group of IS 2122. IS 2123. and IS 5470. aiid tlie group of IS 18573. and
IS 18577 actually exhibited 100% operational bootstrap values, indicating that they
l d be distinguished based on silver-stained PAGE
are perhaps identical as they c o ~ ~ not
Fim 4.1. Dendragram generated from data for 91 sorghum accessions using SSR genotype
data revealed by silver-stained PAGE of PCR products from 20 polymorphic loci distributed
across 9 of 10 sorehum Linkwe emups.

0.46 0.64 0.82

Jaccard' s similarity coeff~cient
gel banding patterns for PCR products of 20 SSR primer pairs previo~lsly
demonstrated to detect polymorphism in cultivated sorglium
Cluster 6 i~icludes two genotypes. BTx623 and Suphanburi 11. The
operational bootstrap value 82% for these two genotypes strongly supports this
cluster. Both genotypes are susceptible to sorghum shoot fly and 13Tx623 was used as
susceptible parent of tlie first ICIIISA'I' sorghum RIL population developed for Q'TL
mapping for shoot t1y resistance.
Cluster 8 consists of four genotypes representing an intermediate population
developed from crosses of t111rr.rrand co~ititrrtl~ll
materials. l'liese are all elite breeding
lines atid known for their combiliation sl~ootfly resistance witli better agronomic
performance. The genotypes li-om this grot111 c o ~ ~ l he
t l used as a source tor tlie
developlnent of a mapping popillation for grain yield and shoot fly resistance, but it
appears that a single representative of this group would he s~iffice.at least initially.
Cluster I I includes nine genotypes: a sub-clurter of two germplasm lincs
exhibiting higli levels of midge resistance and a loosc sub-clurter of tlie remaining
seven agronomically elite genotypes ( i t . . diverse recurrent parents in ICRISAT's
backcross programme for marker-assisted introgression of stay-green QTLs). Eitlicr
Al. 28 or IS 22806 could be used as thc resistant parent of a nen mapping population
if we choose to niap midge resistance. but since these t\\o entries cluster closely
(o]~el.atio~ial
boot strap value of 82.9%) based 011 tlieir n~olecularmarkcr genotypes. it
is likely that tlie bases of their midge resistance(s) arc similar. 'rlie remainder of tile
genotllxs in this cluster are elite breeding lines and released vat.ieties from several
difkrcnt countries, and all are largely be hased (In crosses of rera-zera derivatives.
'The specially of clilster 12 is tliat except for converted zcra-zera line CS 354 1
(C'SV 4). all of its constituent genotypes were bred at 1C'R1SAT-PatancIieru in a
Iprograni to conibine insect resistance (to sorghum slioot fly or sorghum tnidge) witli
superior agronomic perfor~nance and excellelit grain q~cality. These ICRISA'T
genotypes were developed from crosses involving converted zera-zera landraces.
Twelve genotypes are found in cluster 13 is another elite group of materials.
consisting largely of recul~entparents for tlie stay-preen backcrossing programme.
Most of these genotypes are agronomically elite caudatum-type breeding lines or
improved cultivars adapted to tropical sorghum production zones of Latin America.
Africa or Asia. Some of them also exhibits resistance to sorghum midge.
 LS 1, LS 2 and Malisor 84-7 fonn a separate cluster (cluster 16) of improved
genotypes. with a moderate operational bootstrap value of 52%. The first two
genotypes originated from the People's Republic of China and tlie third was
developed fro111 guinea -caudatum materials in ICRISAT's sorglium breeding
prograninie in Mali. All these lines are potential recurrent parents Ibr the stay-green
marker-assisted backcrossing progranime. but any one can be used as they are si~liilar
to each other.
Cluster 18 contains four genotypes having various degrees of eliteness. midge
resistance, and shoot tly resistance. Single genotype clusters 19 and 20 appear to
represent tlie grassy bicolor race of sorghum.
4.1.4 ABI dendrogram
The UPCiMA nietllod was used for generating a dendrogram as we did for the PAGE-
generated marker data set. 'l'lie results suggested a good fit for generation of tlie
tree.For the SSR data set generated using the A131 serli~enciiig machines. the
coefficie~itsof similarity ranged from 0.21 to 1.00 .
Marker alleles detected in the AB1-generated data set grouped 91 so~ghum
genot>pes into 28 clusters at tlie 50% level of similarit). When conlpared to the
dendrogram iiom the I'AQF-generated dnta set. tlie number of c l t ~ s t e rdetected
~ with
llir ABI-generated data was conlparatively higher (Fig. 4.2). This niay he due to the
greater sensitivit? of the autoniated sequellcer. wl~icllallo\\s it to dctect SSR alleles
differing bq smaller numhers of repeat units so that it can erfecti\.ely detect liigller
levels of polymorpliisni.
Accordiiig to the operational hootstrap \'slues obtained, only a. re\\ clusters
\\ere suppc)rted at tlle 50% level. T\\o pairs of test entries (CSV 14R and IS 1034.
and. Macia and ICSV-LM 89.522) appeared to be almost identical in their genetic
niake-up based on the ABI-generated nlarlter data sets be. as they had 100% levels of
sinlilarily. These t\\o tight clusters were eacli supported by strong operationni
bootstrap values (1009'0). It is possible tliat one of tlie genotypes in such a pair might
llave been de\,eloped as an improved line fsom the other genotype. or the two of them
shared genetic material at loci that were assessed in this study due to one or more
common aiicestors. However, if it is not the case. it will be necessary to reconfir111the
lack of marker polymorphism between such pairs of lines. starting with extraction of
DNA from seedlings 60m selfed true-to-type plants of eacli line in order to
co~iipletelyeliminate tlie possibility of a mix-up in samples having identical marker
genotype data sets.
The largest cluster according to the ABI-generated SSR marker data set was
cluster 17, which consisted of 14 genotypes. Many of the genotypes in this cluster
show ~uidgefly resistance andlor are agronomically elite lines selected as potential
recurrent parents for the ICRISAT marker-assisted backcrossing progratntile f o r the
stay-green component of temiinal drought tolerance. All these lines are ayrotlolnically
elite caudatum-type breeding lines and released varieties.
Another 12 genotypes were grouped together to form cluster 3. Gellotypes in
this cli~sterorigi~iatcdfrom the durra race and possess moderate levels of resistance to
sorghum shoot fly andfor spotted stem borer. Seven additional durra genotypes werc
grouped in adjacent cluster 4. These genotypes also have sorghilm shoot fly and/or
spotted sten1 borer resistance. Actually these two clusters based on tlie ABI-generated
data set (3 and 4) formed in a single cli~sterin the PAGE-gene~xteddata set (4). The
use of single reprcsentativc from ARI-generated clustcr 3 and another fiolu ARI-
generated cluster 4. as parents in mapping populations targeting shoot fly and/or
spotted stem borer resistance. would seem to be a reasonable starting point. .4s the
resistant parent (IS 18551) of boil1 cllrrently available siioot Ily resistant mapping
populations falls in AB1-generated cluster 3, an! li~lurcshoot fly resistance mapping
population should I i a e its resistant parent fi.0111 ARl-generated clustel. 4. Several
genotype pairs (i.e.. IS 18551 and IS 2265 in cluster .3. and all possible combinations
invol\ing IS 18573. 15 18577. IS 221r5. IS 2195. and IS 5490 in cluster 4) exhibited
operational bootstrap \ d u e s greater than 50%. indicating good fit of the genotypes in
these clusters.
Cluster 14 contained sik iniproved genotypes, hlvst of these have sorghum
shoot fly resistance and some of tllrni have sorglium midge resistance, all in
agrono~nicall!: superior zera-zera-deri\.rd genetic backgrounds similar to that of insect
pest susceptible CS 3541. Cluster 15. which inclu~led tive genotypes could be
designated as a cluster of agronomically superior midge resistant breeding lines.
Nearly all genotypes fiom clusters 14 and I5 \rere developed at IC1<1SAT-Patanclleri1
ti.otii crosses designed to introgressed insect resistance into elite zera-zera
backgrounds having superior agrononiic characteristics and excelle~itgrain qualit). A
single selected genotype from these two clusters could be used for developnient of a
mapping pop~~lation
targeting for grain yield. grain quality, and insect resista~ice.
 Compared to the clusteril~gpatterns obtained from the PAGE-generated data
set. many genotypes formed single-genotype clusters at the 50% level of similarity
when the ABI-generated niarker data set was used. 'These genotypes included ICSV
700, I-IC 260. IS 17948, IS 4756, IS 18581, SDSL 88'128. ICSV 757. DJ 6514.
Godahunian. ICSV 197, PB 12779-2. Suphanbur 11, IS 2367, and SSG 59-2. Among
these single-genotype clusters, many of them originated from difrerent countries; e.:..
Suphanbnr came from Thailand. ( ~ o d a r n l ~ ~ ~originated
man from Sudan. and IS 18581
and IS 26367 are Nigerian breeding lines.
By and large. niost of the clusters tliat appeared from the PAGE-generated
SSR marker data set werc separated further and their positions relative to other
clusters cllanged moderately in the dendrogram based upon the ABI-generated SSR
~narkerdata set. 1-his is expected as the ABI should give a more accurate picture than
PAGE because of its superior ability to detect small polymorphisms hetween tile
genotypes. For example except for a very feup large clusters of rclated breeding
pl.oducts or insect resistance germplas~naccession. ail clusters detected based on the
('AGE-generated marker data sets \vere separated into disti~ictsub-groups by the ABI-
generated ~narkerdata set. If we look at around the 40% level of similarity. hoth the
PAGE- and ABI-generated data sets detect 12 clusters. Hut positions of the senotypes
uitiiin these clusters were sliglitl) modified by die wperior sensitivity of tlie PCK
prc~ductseparation 011 the AH1 ~nncliines.At tlie same titlie if we look the clustering
pattern at around the 70% level of simil;~rity. hotli of the systems classified the
accessions into a larger numher of clusters. uhich indicates that the 91 studied
genotypes \\ere \\,ell diverged in their genetic make-up. Finall!. according to both tlie
AUI- atid PAGE-generated data sets tliere arc very fe\v genotypes rormitig single-
genotype clusters (e.g.. genotype SSG 59-2) at ecen the 30O/o level of similarity.
revealing their distant relatedness to tlie remainder of the cultivated sorghums
s i ~ n ~ p l e din this study. and their distinctness \\as hell supported by v e 5 l o u
operational bootstrap values (0.6'Yo).
All of tlie genotypes tliat were well disti~iguishedfiom each other at tlie 50?%
level of siniilarity had divergent geographical origins. Tliough materials of dissimilar
geographic origins have sonietimes fallen in different clusters. ~iiost materials of
diverse origins intermingled wit11 each other within clusters irrespective of their
origin. For example breeding lines IS 18573. IS 18577 and IS 18579 from Nigeria Pall
in cluster 4 (for both PAGE- and ABI-generated dendograms). But the other
genotypes in this group were predominantly of Indian origiti. Another breeding line
from Nigeria, IS 18581, was loosely associated \vith several unrelated genotypes in
cluster 1 5 of the PAGE detidrogram, but was round in single-genotype cluster 10 of
the AD1 dendrogrru~i. Modest shifts of this type, where poor ability of PAGE to
discriminate among silnilar alleles was overcome by the superior sensitivity of the
ABI sequencers. were commonly found it1 our detidrogratiis.
Many reasons may be suggested for unexpected associations of genotypes
thought to be of distinctly different geographic origin. Most o r the genotypes are
cultivated. One well-known possible reason for similarities is due to widespread
exchange of genotypes across rcgiotis: especially of ~naterials liabing superior
performance for traits (e.g., picld pote~itialand grain quality) that may he of corntiion
interest across regions. In addition. lines arc introducetl fro111 other countries for
spccilic purposes like debelopment of male-sterile li~ies.or as sources of insect and
disease resistance. Though they are thought to hale come from otie state. their origins
could he quite different so that tllough they are now dcrited from disti~ictgeographic
regions. origin;~llythey appear to ha\e been deribcd l'rom a common gene pool andior
to share comllion origins it1 the distant past.
4.2 P l ~ e n o t ? p i a gRILs from cross 2968 x IS 18551 for components of resistance
to sorghum slloot fly
An experimental s t ~ ~ i l\\as
y catried out to characterize the recombinant inhred lilies
(RILs) de\eloped from cross 20OH (stisceptiblc) IS 18551 (resistant) in order to
itl~proveu~lderstanditlgof the genetic tnakeuli of sliuot t l ~ lresista~~t
components in
sorglii~m.The experin~entswere coiiclucteil under two environments \ i r , late klior.(l
2002 ( E l )and r.crhi 2004/2005 (E2)at I1ataticlieri~.Tlie ~hser\~atiotis
were recorded o n
diffet-ent cumponetits of resistaiice to slloot fly and other traits. Tlie entry mean
performance of shoot Ily resist;l~lceand other traits for inditidual RIT. and its parents.
evaluated under late khtrrif and rob; seasons. are prcsented in Appendix I\ and V
respectively. The results combined from pl~enotj,picancl combined plienotypic and
genotypic data are presented utlder suitable headings.
4.2.1 Estimates of pl~enotypicr~ndgenotypic variation
4.2.1.1 Mean performance of parents
Tlie tnean perforniance of the t\+o parents revealed significant phenotypic differences
(Table 4.2) for the slioot fly resistance and other agronomic traits. except for titlie to
50% flowering (days) in E l : and overall recovery score, aphid da~nagescore. and
grain yield in E?. The parental perfor~nanceunder different environtiients varied for
all tlie traits, except for glossitless (El, E2). recovery resistance (E2). and aphid
damage (E2).
The resistant parent IS 18551 (P:) showed n~aximumleaf glossitless (score 1.0
and 1.1 in El and E 2, respectively). ~uoderateseedling vigour 1 (3.0 in Ez). high
trichome density [189.7 in Ez. and 152.2 inEl) on upper surface of leaf blade: and
83.1 in E2 and 73.0 in El no./ microscopic field on lower surface of leaf blade, and
high seedling heiglit (11.6 cni in E2). For titlie to 50% flowering. there were no
significant differences among the parents in El. \*,liile in L2, significant differences
were observed bet\+een tlie parents. Thc resistant parent IS 18551 was tall [244.0 cm
( E l ) and 161.5 (Ez) cm]. and liad better shoot fly resistance recobery score 12.9 in E l
anrl 3.4 in Ez]. and loner apliid damage score 14.2 and 5.0 in E2 and El, respectively]
than its susceplible counterpart 296B. The resistant parent also recorded high midge
damage (8.6 in E2) atid sliowed poor agronomic desirability (3.8 in E?). Oviposition
incidetice (O4) and dcadheart incidetice (%) were significantly lo\\er. in P2 than P I . in
b<~th
screening environments. Ilo\vever. the range of phencitypic values for these traits
varied sig~iificantly in tlie two screening environments. 'Tlic phenotypic ~ a l u e sfor
oviposition Ibr IS 1855 I were lower in environment E: [I 5.5% and 33.1% plarits \vitli
eggs at I4 and 21 IIAE. respectively/ than i n E t [45.0°h platits with eggs at 14 DAE.
ontl 79.0% plants with eggs at ?I DAE]. Similarly tile deadheart fortilation in IS
I8551 was lo~vcr in environ~ncnt E2 (17% deadliearts at 2ll)hE. mid 25.8%
cleadhearts at 28 DAE) than ell\ iro~ililrntEl (51.5% and 68.5% deadliearts at 21 and
1 8 DAE. respectivel!,). IS 185.51 also showed Iligli pigmentation (1.1). i.e. dark pink
col~irationat tlie seedling stage (non-tan type).
The susceptible parent. 296R ( P I ) was non-glossy. \\it11 lo\\ trichome densit!
on both leaf snrtices in both s c r e e n i ~ ~environments.
p Ilin\ever. 296R slio\+ed poor
seedling vigor (4.3 in I:?).and least seedling height (7.9 in E:). 296B liad significatirlq
higher oviposition (68.7% atid 95.5% plants with eggs at 14 and 21 D.4E.
respectively) it1 environment 1'1 than in E* (55.2% and 81.2% plants witli eggs at 14
and 21 DAE, respectively). Similarly, tlie deadlieart incidence in 2968 \%as
signiiicantly Iiiglier in E l (75.7% and 96.1% deadliea~ts at 21 and 28 DAE.
respectively) than in E2 (60.0% and 77.4% deadliearts at 21 and 28 DAE.
respectively). This parent also sliowed a liigll pigmentation score (2.9, tan type) in E2.

The two parental lines flowered almost at same time (89 days and 88 days for 2968
atid IS 18551, respectively) in El. but 2 9 6 8 flowered late (82 days) than tlle parent IS
18551 (75 days) in E2. Parent 2 9 6 8 was shorter (I 10 cnl atid 105 cnl in El and E2,
respectively) tllan IS 1855 1. 296B sho\ved ]>oar recovery (8.1 and 4.0 in E; and Ez,
respectively) and high aphid dan~age(8.0 and 5.3 in El and respectively) tha11 IS
1855 1 i t i both screeni~lgenvironments.
4.2.1.2 Mean and ranges of R l L s
The mean and ranges of RIL population for different shoot fly resistance components
and agrononiic traits (Table 4.3) varied between the environments. except for the
glossiness (score 3.6 and 3.5 in P I and Ez. respectively). The mean \slue for trichome
density on the upper surface of leaf blade v n s greater in E2 (87.3 per microscopic
field) than in I;! (80.6 per n~icl-oscopicfield). Similarlc. tile mean values for tricliome
density on the lo\%erleaf surface were iligher in L:-> (43.7 per microscopic field) than
in C I (32.0 per microscopic ficld). The KII. Ineiln vitlues \<ere lo\\er for o~iposition
incidence (35.0% and 57.8% plants with eggs at I4 and 21 DAE. re~pectively)and
dcadhearts incidence (35.0% and 50.691 plants with deadhearts at21 and 28 DAE.
rcspectively) in C r than in El ((11.2% and X').7°h plants icitll eggs at 14 and 21 DAE.
a ~ l d7?.3'/0 and 87.7% plants \\it11 deadhearts a121 and 28 DAC. respectively). Tlie
obser\>ed ranges of I<LL means \sere larger 1111.oviposition incidence and dearlhearts
incidence in E: than in I < I .For other agronomic triiits. KII. means for time lo 50%
flo\\ering and plant height (73 days and 106 CIII.repecti\elq) \vere lower in L? than
in E l ( X i days and 195 cni. resliecti\el!). R11 lllcn~lsfbr o\,erall reco\.el-j score (4.2)
and aphid danlage scorc (4.0) in F2 \\ere louer than in I-'](5.7 nild 6.4. respectivel\).
The nlcan perli,rnloilcc of tlie I<ll.s for grain !icld \\as better in El (219 1 g plot-1)
than in Ez (470.0 g plol-l).The mean performance of the IULs for midge damage was
ltig11 (7.2) and agronon~icscorc \\as fairl) govd (3.8) i l l E?.
4.2.1.3 Alinlysis of variance
'l'l~e anallsis of variance for different shoot tl!. resistance components (Table 4.4)
sllowed that variances due to genot!pes (RILsl were sigliificanr for all the traits
studied based on performances in individual screening environlnents. as ~ v e l las
averages across the t\vo screenil~gen\'irontnents, For shoot ily resistance component
traits such as glossiness score. ovipositio~i incidence. deadhearts incidence. and
trichome density on the lo\\er surfilce of the leaf blade. genotypic variances were
greater in environment Er than in environment El. IIowever. for trichotue densit! on
the upper surface of the leaf blade. time to 50% Ho\vering. plant height, overall
recovely score, aphid daniage score, and grai~iyield. the genotypic variances were
greater in environ~nent El than i n environme~it E2. The across-season analysis
revealed that variances due to genotypes ( G ) and G x E interaction \yere significant
for all tlie traits observed. The genetic variance values were more for resistance
components such as glossiness. trichome density (both upper and lower surfaces of
seedling leaf blades). time to 50% tloxvering, aphid damage score, and plant height
more than double tlie G x E ititeraction variance. For other traits such as O\~ipusition.
deadhearts. overall recovely resi?tance score. and grain yield. the genotypic variances
were niore than the variance due to G x I: interaction.
4.2.1.4 Frequency distribution
The \,ariation for shoot ily resistance and other traits ivas represented grnpllically
tising li-equency distributions of entry nieans in the t%o screening en\*iroruments.The
n~easurements grouprd into ecl~~ally
spaced classes on t l ~ eX-axis. and the
frequency of individuals falling in e:~cliclass was plotted on Ilie Y-axis. The resulting
histograms showed normal curves (Fig 4.3 A-(2). In general. frequency distrib~1tion.i
for most o f the traits under stutly approximated a nornial curve in the rabi screening
en\.ironment. 13ut ill case of tlie 2002 kharif c~i\,il.onnicnt.the Frequency distributions
Ibr a I'cw resistance traits Lve1.e s k e ~ t e d Thc
. values for means and ranges of these
characters varied. and tlie pei~ks\\ere seen at ilift'ercnt points Tor each of tlie traits in
thc t\\o testing envirul~rnents.111 tile 2004 rabi environment. the distribution cur\es
\%ere nor~nal. except fur glossiness. 1rich(11nc tlensi~! (hot11 on upper and lower
s ~ i r ~ ~ l cofe sseedli~ig leaf blades). seedling vigor I. pigmentation score. and midge
clomage score. In case of tlie 9002 Lliarif enviromiicnt. ~ h efrequent> curves \\ere
11. trichoine density (both on
~iormnl.except leafgiossiness. o\iposition 11. deadhet11-t~
upper aalid lower surfaces of seedling leaf blades). plant height. ancl aphid clamage. For
the trait leal' glossiness. altliough the character varied continuousl). it sho\ued a A~nd
o r bi~nodaldistribution. ~ l ~ i c\\-as
l i evident in each a f t h e screening enviro~inients.Tor
triclionie densities oti tlie upper nnd lower surfaces of the leaf, the histograms dra\\n
sho\\ed discontinuous distributions. i.e.. bimodal distributions tilet \\.ere skened
tolvards trichomelessness in botli the kharif and the rabi enviro~iments. For
ovipositio~iI1 and deadliearts 11. tlie histogranis showed discontinuous distribution.
which were skewed towards shoot fly preference for egg la! inp and high deadheart
incidence in the 2002 kharif screening environn~e~it.
'I'lie l~istogramsfor plant height
and aphid damage also slio\ved discontinuous distribution. which were ske\ved
 Figure 4.3 (A - Q). Frequency distribution of 213 RILs derived from cross 296B X IS
18551 for components of shoot fly resistance and few agronomic traits in two screening
environments, viz., late kharif(E1) and rabi (E2) at Patancheru. On the X-axis groups
of concerned trait arc plotted and on the Y-axis frequencies of each group (i.e. the
number of RILs in each group) are plotted.

Glossiness intensity

hhanf
IRab~
-.

0 1.5 ? 2.5 3 3. 4 4.5 5

Glossiness score (1-5 scale)
-~ ~ ---- -- - --- -

'!.I
- -
@ kharif
IRabi
--

, 0 4- -
10
r-

20 30 40 50 60 70 SO 90
Ovipsition incidence I (%)
-- - - - -- -. - -- . --
1
El
-
Oviposition incidence I1

120

p so,
: 60 ~~harif::
I--R ~ ~ I -
I
I

0 --- I

- - - -

Deadhearts incidence I (%)

10 20 30 40 50 60 70 80 90
I
1 Deadheartt incidence l (%)

Deadhearts ~ n c ~ d e

1 140
I 120 I
I El00 1 .-
01 Khar~f
I r a b ~ -1 '
-
-

0 4-- 7 7- T

30 40 50 60 70 80 90 100
Deadhearts incidence I1 (%)
- -- --A - --- -- .-
 Trichome density (upper leaf blade surface)
1
1

0 20 40 60 80 100 120 140 160 180 200 220 240

Trichome density (no. 1 microscopic fieid)

-
. -. - .- --
Trichome density (lower leaf blade surface)
I

-
~ h G f
II rabi I
.. -I
I

I
0 10 20 30 40 50 60 70 SO 90 100110120130140
Tricbome density (no. / microscopic field)
-- -- - - - -

Plant height (cm)

1

120 140 160 180 200 220 240 260 280 300

Plant height (cm)

-
1 1 I Time to 50% flowering (dl
I
80 -
,x 60 -
c
hKhafri
4 40-
.1
-
rabi
- -

65 70 75 80 85 90 95 100
Time to 50% flowering (days)

Overall recovery score I

7 --
Khnfl
II
Irabl
- J.-
I

--

I Recowry scale (1-9)

Aphid damage scorp

I I2O? WEE Im m
IN Grain yield (glplot)

100 200 300 400 500 600 700 800 900 1000 1100 1200
Grain yield (g)

Seedling vigor I 1!

1
i
Seedling vigor score (1-5) ~
I
Seedling pigmentation

m
1-

40
20 I

0 1.5 2 2.5 3 3.5 4 4.5 5

Pigmentationscore (1-5)
!
 -- -
.

Agronomic score
I

Agronoms score (Id)

L .
- -- _ -- . _ _ -I

I
Midge damage score

5 6 7 8 9
Midge score (1-10)

;-.---
-- - -- -

;
Seedling height (cm)

1 140
120-
I p100.
: 80-

E L- .

r - I
3 6 9 12 I5

~ Seedling height (em)

--- -- A
I
 96
towards greater plant height and lower apliid daniage score in the 2002 kharif
screening etivirotinient. For seedling vigor I, pigmentation score. and midge damage.
tlie histogratlls showed discontinuous distributions in tlie 2004 rnbi screening
etivironrnetit. Seedling vigor I showed a bitllodal distribution. while pigmentation was
skewed towards dark-pink pigtiletlt (non-tan foliage). nhile midge damage score was
skewed towards susceptibility.
4 .2.1.5 Transgressive segregation
'The R1l.s lying outside the parental limits were identified based on trial entry means
across the two screening en\ironments. T11e RIL population mean atid individual
parent means were subjected to T-test to assess tlie significance of dift'ercnces
between the means (Table 4.5). The analysis revealed that RIL means differed
signilicantly fiom hot11 the parents for shoot fl> resistance traits such as glossiness.
oviposition incidence. deadhearts incidence. trichonle density (both 011 upper a r ~ d
loner surfaces of seedling leaf blades). plant height. oxerall recovery score. aphid
clat~iageand glxi~iyield. For oviposition at 21 lli\E. the RII. mean did not difrer
signilicantl!, irom that of resistant parent IS 1855 1. 'Transgres?ive segregants nith
phetiotypic \alucs oittside the parental limits were obse17,ed for most of the traits.
except for leaf glossiness (both parents). deadhearts I1 ( 2 9 6 0 ) . tricho~nedensity (both
on upper and lo\ver surfi~cesof seedling leaf blades) (29613). plant height (2968). and
grain yield (IS 18551 1. For o\.iposition I and 11. dead11eal.t~1 and 11. ol.erall recover!
score. aphid damage score. attd grain yicld. the R11. populnti~>tlmean \\as less than thc
11iid-parental \nlue. In contrast to plossiness. the trichomc densit? (both o t ~upper and
Icr\\'el. surlhces of seedling lcai hlndes) and plant heigl~t.RIL populatioti tnean xulue
\\ere greater than the mid-parental value. Ihe prtoportio~lo r KIL outside the parental
limits were greater tilr those outside thc lo\\-scoring parents. and lo~vcrTor oittside
Iiigli-scoring parents.
4.2.1.5.1 Glossiness
The de;,i;ttion uf RI1. mean Trom mid-parental value \\as positive. but no transgressi\e
segregwnt RIL \bere observed \\it11 phetiot>pic values outside tlie high- and lo\\-
scoring parent.
4.2.1.5.2 <hiposition incidence (Oh)

The KIL population mean de\,iated from mid-parental ~'nlue to\vards that of tlie
resistant parent. Favourable transgressive segregauts were observed at 14 and 21
DAE. I-lowever. the proportion of RIC lying outside the mean of the low-scoring
parent (IS 18551) was comparatively higher for oviposition I1 ( at 21 DAE) than for
oviposition 1 (14 DAE).
Table 4.5 Means of parents, the RIL population, their difference and proportion of RILs with values outside the parental limits based on pooled means over two
screening environments
Character PI P2 Midparenta RIL Test of significance Proportions outside
2968 IS 18551 value population of means the parental limits
mean PlIRIL P2/W P1 P2

Glossinessintensity 5.0 1.1 3.0 3.6 +* ** 0.000

Oviposition 1 (%) 66.3 40.9 53.6 49.9 be ** 0.015
Ovipositim U (%) 93.0 72.6 82.8 76.1 tt ns 0.003
Deadhearts I (%) 72.9 46.8 59.9 56.9 *t " 0.015
Deadhearts I1 (%) 92.6 62.6 77.6 71.8 t* " 0.000
Trichomc density (upper surface)(no./microscopicfield) 0.0 157.4 78.7 83.6 t* ** 0.000
Trichome density (lower surface) (no.lmiaoscopicfield) 0.0 74.4 37.2 37.7 at " 0.000
Plant height (an) 109.5 232.3 170.9 182.3 t* ** 0.000
Overall ecovcry score (scale) 7.4 2.9 5.2 5.1 ** " 0.019
Aphid damage score (scale) 7.6 4.8 6.2 5.8 t* " 0.003
Grah yield (glplot) 228.7 999.6 614.2 502.1 t* 0.027
--

N : nonsignificant * significant atp= 0.05 ** significant atp= 0.01

 97
4.2.1.5.3 Deadhearts incidence (%)
The RIL population mean deviated from mid-parental values for tliis trait.
Transgressive segregants were observed at botli tlie observation stages. No
transgressive segregatit RIL was observed wit11 phenotypic values outside the high-
scoring susceptible parent 296B for deadliearts 11. However, tlie proportion of RIL
lying out-side the low-scoring resistant parent. IS 1855 1 \vas coniparatively higher for
deadllearts I1 (at 28 DAE) tlian fin deadhearts I (at 21 DAE).
4.2.1.5.4 T r i c l ~ o m edensity
Mean triclionie density in the RIL population was on par wit11 the mid-parcntal balue
Sor trichonie density on tlie lower surface of leaf blades; however. favourable
transgressive segregants were obscrved for tliis trait. v,hicIi were relatively high in
frecluency with values outside thr high-scoring parent IS I855 1 (KIL nos. 39, 50. 80.
1 10. 11 7. 129. 140. 208. 242, and 258) For thc tricliome density on the upper surface
of' leaf blades. the mean of the l<lI. popi~lationdeviated positibely froni the rnid-
parental value and the proportion of favournhle transgressive segregants Mas e\en
Iliglicr than that observed for trichome densit on the lo\ber surface o r leaf blades.
Transgressive scgregants with phenotypic values lying outside the high-scoring IS
1855 1 included KIL nos. 33. 42. 50. 80. 1 1 7. 129. 174. 208. 24 1 . 242. and 258 for
trichonle density on the upper surface of the leaf
4.2.1.5.5 Plant Ileight
'She two parents differed significnnlly for plant height. The meat1 \slue of the RIL.
was higller than tlie mid-parental value. Ira~isgressi\e ssgregants \\ere obsened
exhibiting greater plant heiglit than taller parent IS 185.5 I . altliouyli their proportion
was low.
4.2.1.5.6 Overall recovery score
Mean recovery resistance score of RIL popi~latio~l
\bas at pi~r\\it11 tile mid-parental
value for this trait. A low proportion (IT transgressive segregants \\ere observed. \\it11
a few RlLs outside the lo\\ scoring parental values. The l>roportion uf transgressive
I<IIJs was higller for individuals \\it11 \slues I!ing outside tlie Iligll-scoring susceptible
parent 206B.
4.2.1.5.7 Aphid damage score
The RIL pop~tlatio~imean deviated from the mid-parental value for this trait.
Transgressive segregants were observed outside limits of both the high and low-
scoring parents: however. the proportion of R1I.s I!ing outside the lo\\-scoring (mure
resistant) parent IS 18551 was comparatively higher.
 98
4.2.1.5.8 G r a i n yield
T11e two parents of the RIL population differed significantly for grain yield under
conditions of moderate to severe shoot fly pressure. l'he mean value of the RIL
poplilation was lower than the mid-pare~~tal
value. A low proportion of transgressive
segregants were observed for this trait. No favourable transgressive segregate were
observed. All observed transgressive segregates for this trait had grain ~ i e l dvalues
lying outside the lower-scoring shoot fly susceptible parent 2 9 6 8 .
5.2.1.6 Inheritance of resistance
4.2.1.6.1 Broad-sense heritability
Esti~natesof broad-sense (operational) Ileritability (entry mean basis) fol- shoot fly
resistance components and other traits at maturity were obtained from the data
collected in individual rnvironlnents 1,;:. El (kharil' 2002) and E2 (rabi 2004).
tleritability estimates were also obtained based 011 acerage performance c>ver these
t\vo e ~ ~ v i r o t l m e(Table
~ ~ t s 4.6)
4.2.1.6.l.l Glossiness
I lcritability esti~ilates\\ere c<~nsislentlyhigh for leaf glossinecs (17',0.85) in thc t\\o
screening environments. but wcre moderatc (/i2>0.64)acrosi environments.
4.2.1.6.1.2 Oviposition incidence (I%)

Operational heritability \vas lo\\. to nlodcrate. hut consiste~ltlor crviposition in the t ~ u

inilividual scieening envirol11nents. and also at dit'tkrent ohser~ation inter~als.In
kharif 2002. the operational heritability estimates for o\ iposition incidence (%) \\ere
lo\\ at both observational sleges (1,' = 0.20 at 14 DAE. and 17' = 0.38 at 21 IIAE).
\vhile ill rabi 2004. operational Ile~itnbilities\\el-c i~igller(I?'= 0.67 and h.'= 0.70 at 14
and 21 1)Atl. respecti\ely). Ilo\\e\er. operatiollal Ilerilnbility across seasons for both
stages was lo\\. (I?-'= 0.17 and 0.02). indicating a signilicant influence of the screening
e ~ ~ v i r ~ ~ ~ andlor
~ t n e ngenotype
t. * en! il.c~nmei~t
intcractio~~.
4.2.1.6.1.3 1)eadhearts incidence (%)
Operntio~lallieritabilit). estimates for deadhearts were lo\\- to ~noderclte in the t\\o
screening environments. In kharif 2002. the lierit;~bilityestimates \\ere moderate. hut
consistent at hoth the obsercation stages (17' = 0.62 and 0.71 for deadhearts at ? I and
28 DAt?. respectively): wliile in rtrbi 2004. the estimates \\ere n~oderate(17' = 0.63
and 0.68 for deadhearts at 2 1 and 38 DAIi. respecti\,ely). However. the operational
heritability estimates across sensolls n e r e quite lo\v (0.17aitd 0.09 for deadllearts at ?I
and 28 DAE, respectively). iiidicatiiig signilicant influence of the screening
environ~nent,and genotype x environment i~iteraction.
4.2.1.6.1.4 Seedling vigor and leaf sheath pigmentation
The operational lieritability estimates for these traits were high in the rabi 2004. (h' =

0.79 and 0.85 for seedling vigor. and pigmentation score, respectively.
4.2.1.6.1.5 Seedling lleight
Seedling height showed nioderate operational heritability in 2004 rcrbi season (17' =

0.63)..
4.2.1.6.1.6 Tricl~omedensity
Tricliome density showed high operational heritability estimates (17' >0.95) in both the
environments for upper and lower leaf blade surfaces. However. tlie cornhined season
i~nalyses,lower estiniates of operational heritability \\ere seen for both upper (11' =

0.51) and lower (k' = 0.49) leal surfaces, indicating tlie genotype x environiiient
interaction.
4.2.1.6.1.7 Time to flowering
111 both scrcc~iing environments. rinic to flowering sl~o\ved high and consistent
operational lleritability (h' = 0.87 and 0.85 in the 2002 khurif and 2004 rtrbi.
respectively). Across seasons. this trait sliuwed nioderate heritability (/I
= 0.60).

probably as a result of dil'itrences ill ~lliotoperiod sellsitivit? or [lie parental lilies

contributing to genotype x envit.onment interaction.
4.2.1.6.1.8 Overall recoven score
Moderate operational lleritabilities were recorded Ibr overall recovery score in
iildibidual seasons (11 '= 0.76 and 0.71 for 2002 k11to.if and 2004 ~.trbi seasoils,
respecti\,ely). IIome~er. lou uperational Iieritahilit~ esti~uattrs \\ere obtained in
acrtiss-season analysis (I? = 0.38).
4.2.1.6.1.9 Aphid datnsge score
Moderate operational lleritab~lities were recorded for aphid damage score in
individual seasons (h' = 0.71 and 0.72 for kh(ri.!f' 200'2 and rtrbi 2004 seasons.
res],ectively). and a moderate operational lieritability estimate was obtained in the
across-season analysis (I,?= 0.59).
4.2.1.6.1.10 Midge damage score and agronotnic score
High and moderate operational lieritabilities estimates were recorded in the 2004 robi
for midge damage score (h' = 0.90) and for agronomic deasirahility (17' = 0.75).

respectively.
4.2.1.6.1.11 Plant height
Plant height sliowed high operational heritability estimates (h2> 0.80) in both
screening en\~ironments,as well as in the cotnbined analysis across environments (h2
= 0.84).
4.2.1.6.1.12 Grain yield
bladerate heritability was observed for grain yield under conditions of moderate to
severe shoot fly pressure in indibidual seasons (I,'= 0.75 and 0.53 for kharif 2002 and
rabi 2004, respectively), and a moderate to low operational lieri(ability estimate was
obtained in the across-season analysis (/i'=0.45).
4.3 QTL mapping
]:or QTL mapping and Q x E interaction analysis. the li~ihagcmap constructed using
the population of 213 RILs derived from cross 2Y6R x IS 18551 was used. I'he
sortware package Plab QTI. \\,as used to analyze [lie data by conlpositc interval
mappi~lg(CIZVI) procedures. The CIbI method \vas implemented using a I.OD value of
2.5 as the threshold for QTL significance. Tlle genetic nod el chosen \+as additke r

additive interaction fc~rthis Fi s r e c o n ~ b i ~ ~ apopulation.

nt The software calculates
additive el'fecls and estimates the poltion of phenotypic varintiun explained by each
intlividual VTL. T l ~ eresults fiom CIbl analysis for ide~itificatio~iof OTLs with
sig~lificanteffects for shoot fl! ~csistanceconlponents are described heion for two
indi~idual plie~lotypic en\,iro~l~llentsand across tllese plienotypic rn\ironments.
Anlong shoot 11) resista~lceand gritill jield or cornpc)t~enttraits studied. Q'fLs Mere
identified fill. all traits in botll the screening en1 irontnents e x e p t in casc of
oviposition I and deadhearts I i n the hharif 2002 environ~lient.
4.3.1 QTL analysis in single et~vil.onnic~~t
The phenotypic data from two pl~enot!ying en\ iron~ue~lts
and genotypic data for 21.3
KlLs \yere subjected to QTL analysis. 'The results of the two single-en\8ironmri~t
analyse are presented in Table 4.7 and Figure 4.4. The results of this QTI. a~lalysisfor
shoot fly resistance components are described belo\\.
4.3.1.1 Glossiness (score)
Based on the CIM analysis of phenot>pic data from (he ~ u I'atal~clieru
o screeliing
environments (kliarif 2002 and rabi 2004), three QTLs were detected using data from
the kliarif environment and t\\o y'fLs using data from the rabi en\,ironrne~lt.
accounting for 51.1% and 28.7% of the obser1,ed phenotypic variances in these
environments. respectively. Of the Q'fLs detected, one mapped on LG 'J' in both
W e 4.7 Characteristics of QTLs associated with putative components of resistance to shoot fly (in two
ncreening environments, khcvif and rabi) based on Composite Interval Mapping (PLABQTL. WDz2.5) using
213 RILs derived from cross 296B (susceptible) x IS 18551 (resistant)
Environment/Mt -
Linkage Position Marker i n t e n d --
Suomrt Peak WD' *% ~ffect*
group interval (cM) (additive)
CHouIneu Intensity
Kharif, Patancheru (E,) E 24 Xtxp 40-Xtxp 159 0-40 3.55 7.6 -0.264
H 86 XSbAGD02-Xtxp294 84-94 3.14 6.6 -0.199
J 28 Xisp215-Xisp258 22-54 17.99 36.9 -0.601
Sum: 3 QTLs 51.1
Final simultaneous fit LOD = 20.08 Adj. R' = 33.5%
Adj.genotypic var. ex.J = 48.5%

Rabi, Patancheru (E2)

J 32 Xisp2 15 - Xisp258 20-62 9.35 21.4 -0.465
Sum: 2 QTLs 28.7
Final simultaneous fit LOD = 10.34 Adj. R
' = 18.6%
Adj.genotypic var. ex. = 25.1%

Seedling vigor I (scale) A 72 x-75 - x-37 52-106 2.51 5.3 0.203

Rabi , Patancheru (E2) B 258 xtrpol- XW348 250-272 5.27 10.8 -0.300

- - - ---
Final simultaneous fit LOD = 7.53 Adj. R' = 12.6%
Adj-genotypicvar. ex. = 22.4%
Kharif, Patancheru ( E l ) Phenotypic observation not recorded
Wposition I (%)
Kharif, Patancheru (E11 QTLs not found
Rabi, Patancheru (E2) F 2 XtxplO -Xisp318 0.14 3.30 7.2 2.448
Sum: 1 QTL 7.2
Final simultaneous fit U)D = 1-72 Adj. R' = 2.8%
Adj.genotypic var. ex. = 6.4%
Enviro-ent/trPlt Linkage Position Marker interval Support Peak LOD R4% Effect
m"UP interval (cM) (additive)
Osl~itlon
II (%) E 22 Xtxp40 - Xtrp 159 0-42 2.86 6.1 -0.558
Kharif, Patancheru (El) J 26 Xisp2 15 - Xisp258 8-64 3.25 8.0 -0.623
Sum: 2 QTLs 14.1
Final simultaneous fit LOD = 3.99 Adj. R' = 6.5%
Adj.genotypic var. ex. = 46.7%
Rabi. Patanchem (E2) F 10 Xtxp 10- Xisp318 0-32 3.06 6.7 2.623
G 104 XgapO1-Xcup67 78.126 3.40 7.3 -3.751
Sum: 2 QTLs 14
Final simultaneous fit LOD = 5.06 Adj. R2 = 8.7%
Adi.eenotv~icvar. ex. = 19.0%
Deadhearts I ( O h )
Khanif, Patanchem (E1) QTL. not found

Rabi, Patanchem (E2) F 2 XtxplO - Xisp318 0.14 3.90 8.4 2.475

Sum: 1 QTL 8.4
Final simultaneous fit LOD = 2.31 Adj. R2 = 4.5%
Adi.eenotv~icvar. ex. = 10.0%
Deadhearta II (%)
Kharif. Patanchem ( E l ) F 12 XtrplO-Xisp318 4-14 3.02 6.6 -1.437

Sum: 1 QTL 6.6

Final simultaneous fit LOD = 0.55 Adj. R' = 0.2%
Adj.genotypic var. ex. = 0.6%
Rabi, Patanchem (E2) F 24 Xisp318 - Xtxp230 12-38 3.14 7 2.518
G 104 XgapOl - X a p 6 7 88- 122 5.29 11.2 -4.395

Sum: 2 QTLs 18.2

Final simultaneous fit ~ 0 ~ = 5 . 9 Adj.R2
8 = 10.5%
Adj.genotypic var. ex. = 23.9%
Table 4.7 cont.---
Environment/trait Linkage Position Marker interval Support Peak LOD' %
'
R Effect4
group interval [cM) (additive)
f3ee-g height I ( ~ m ) B 264 Xtxp348 - Xtxp207 254-272 3.63 7.6 0.258
Rabi, P a t a n c h e ~
(E2) I 62 Xtxp 17 - Xisp 347 60-66 2.88 6.7 0.203
Sum: 2 QTLs 14.3
Final simultaneous fit ~ 0 ~ = 4 . 1A
3 d j . ~ ' =6.8%
Adj.genotypic var. ex. = 18.3%
Kharif, F'atanchem (E 1) Phenotypic observation not recroded

Trichome density (upper leaf blade surface)(no./microscopic Held)

Kharif, Patanchem (El) G 112 XgapOl - X C Z J P ~ ~ 104-122 15.53 29 48.798

Sum: 1 QTLs 29
Final simultaneous fit LOD = 13.35 Adj. R' = 24.4%
Adj.genotypic var. ex. = 25.4%
Rabi, Patancheru (E2) G 124 XgapOl - Xcup67 110-126 7.31 15 22.367
Sum: 1 QTL 15
Final simultaneous fit LOD = 8.39 Adj. R2 = 15.8%
Adj.genotypic vat-. ex. = 16.8%
Mchome density (lower leaf bdde surface) (no./microscopic Held)
Khnrif. Patanchru (E1) G 118 XgapOI - Xcup67 108-136 15 28.4 16.839
Sum: 1 QTL 28.4
Final simultaneous fit LOD = 15.2 Adj. R' = 27.3%
Adj.genotypic var. ex. = 29.1%
Rabi, Patanchem (E2) C 26 Xtxp69 - Xtxp34 22-32 2.92 6.2 -7.482
G 112 x g a ~ o -l X C Z J P ~ ~106-126 7.27 14.9 13.127
Sum: 2 QTL 21.1
Final simultaneous fit LOD = 9.67 Adj. R' = 17.3%
Adj.genotypic var. ex. = 18.8%
hble 4.7 c0nt.---
Environment/trait Linkage Position Marker interval Support Peak LOD RZ% Effect
-UP interval (cM) (additive)
Aphid damage score
Kharih F'atanchem (El) E 34 X-40 - XlXp 159 22-46 6.94 14.3 -0.392
J 150 Xtxp 15 -X-283 140-160 7.70 17.3 -0.349
Sum: 2 QTLs 31.5
Final simultaneous fit LOD = 12-73 Adj. R' = 22.7%
Adj-genotypic var. ex. = 59.7%
Rabi, Patanchem (E2) J 144 Xtxp 15 - Xtxp283 132-156 4.47 10.4 -0.229
Sum: 1 Q T L 10.4
Final simultaneous fit LOD = 3.26 Adj. R2 = 6.0%
Adj-genotypicvar. ex. = 13.2%
Envlronment/trPlt - Position
Linkage Marker interval Support Peak LOD RZ% Effect
group interval (cM) (additive)
F'lgmentation score
Khorih Patanchem (El) Phenotypic o b s e r w t i o ~not recorded
Rabi. Patanchem IE2) A-
. 52 X-75 - XtXp37 30-66 3.51 7.13 -0.225
I 28 X1sp264 - Xcup 12 20 -40 5.9 1 12.00 -0.309
Sum: 2 QTLs 19.13
Final simultaneous fit LOD = 9.43 Adj. R
' = 16.9%
Adj.genotypic var. ex. = 25.6%
Mldge damage score
Kharix Patancheru ( E l ) Phenotypic observations not recorded
Rabi, Patancheru (E2) C 100 XlXp 114 - Xtxp218 92-126 2.63 5.5 0.310
I 2 Xtxp 145 - Xcup36 0.4 6.69 15.5 0.575
Sum: 2 Q T L s 21
Final simultaneous fit LOD = 5-86 Adj. R
' = 10.2%
Adj.genotypic var. ex. = 13.4%

Table 4.7 cont.---

Environment/trait Linkage Position Marker interval Support Peak LQD R~% Effect
POUP interval (cM) (additive)
Agronomic score
Khan3 Patancheru ( E l ) Phenotypic observations not recorded
Rabi, P a t a n c h e ~(E2) A 76 Xtxp75 - X-37 58-94 2.57 5.5 0.135

Sum: 1 QTL 5.5

Final simultaneous fit LOD = 2.79 Adj. R' = 5.0%
Adj.genotypicvar. ex. = 9.9%
Qlin Yield (glplot)
Kharif; Patancheru ( E l ) E 56 Xtxp40 - Xtxp 159 34-62 3.40 7.4 39.6
G 114 XgapOl - Xcup67 100 -126 5.74 12.1 73.4
1 hfi Xixu 17 - Xis0347 60.66 2.81 6.7 41.1
Sum: 3 QTL 26.2
~ i n simultaneous
d fit LOD = 6.96 Adj. R
' = 11.5%
Adj.genotypic var. ex. = 26.0%
Rabi, Patancheru (E2) C 98 XtxP114-XtxP218 90-130 3.08 6.4 -21.792
Sum: 1 QTL 6.4
Final simultaneous fit LOD = 2.55 Adj. R' = 4.55%
Adj.genotypic var. ex. = 16.6%

Glossiness (1 - 5 scale) : 1 = high intensity of glossiness, 5 = non-glossy

Seedling vigor I (1 - 5 scale) : 1 = high vigor. 5 = low vigor
' : Log 10 likelihood
: Percentage of adjusted phenotypic variance explained
a .. Percentage of adjusted genotypic variance explained

: + sign indicates that the homozygous IS 18551 allele genotype has a numerically greater value for the trait than does the
homozygous 296B allele genotype; while - sign indicates that the homozygous 296B allele genotype has a numerically greater
 Dhll
D

Dhll
I

-
Dhl
I

-
hlll

hll
-
- cry1

OR5

-
MI0
I

rmo
2,"
- ZP
Dhll

Ovill
mdge

-
.-
Plht
ms
Flo
PLht

-
Seht
screening environments. The QTL on 1.G 'J' is a major Q'TL explaining 36.9%
(kharit) atid 21.4% (rabi) of tlie observed phenotypic variance for glossiness intensity.
The fi~lalsimultaneous fits revealed peak LOD values of 20.08 (kharif) and 10.34
(rabi). explaining 33.5% and 18.6% of tlie adjusted phenotypic variances.
respectively. I'aretit IS 18551 contributed glossiness alleles for all of tlie detected
QTLs, with additive effects ranging from -0.199 to -0.601 (negative sign indicated
that greater degree of glossiness is fro111parent IS 1855 1 ).
4.3.1.2 Seedling vigor I (score)
For seedling vigor I. phenotypic observations were not recorded in the late kharil' ( E l )
environment. while in the rabi (E?)environ~nent tlie trait was nieasured and
phenotypic data from El detected three QTLs Tor seedling vigor score. Theses
putative QTLs were niapped 011 LGs 'A'. 'R' and 'D'. and together explain 24.2% of
the observed ]?henot!,pic variance. l'lle final sinlultaneoi~sfit o f thece three QTLs
using data from the E! en\~il.onmentre\ealed a peak a L O D ~ a l u e7.53 and 12.6% of
the adiusted plienotypic variance was explained by tliese tllree QT1.s. Out of tliese
three detected QTLs for seedling \igor 1. two QT1.s (one each on L.G 'A' and -U').
exhibited favorable additive genetic effects contributed by IS 18551 alleles (-0.203 for
tlie 01'L on LC; A and -0.300 Tor the QTL on LG B).\~.hilethe QTI. mapped on LG
'D'exhibited favorable additive genetic ellects fiom the 296B allele (0.280). The
0.1I. ~iinppcidon LC; 'B' appears to be tlie most important Q I'l., as it explained 10.8%
of observed phenotypic variance tbr seedling vigor I.
4.3.1.3 Oviposition incidence 1
screening environment.
For oviposition I. no QTI. was detected in tlie late kl?c~~.!/'(E~)
The most probable reason for this could be tlic ver! little difference bet\\t.en tlie RIL
parents and hence very limited variation for o\ipvsitinn incidence 1 in the KII.
population due to high shoot fly pressure in late kIltrr,if screening environ~neiit.In the
iuhi screetii~ig(E2) en\'ironnient. one QTI. was detected on LG 'P' explaining 7.20'0
of tlie obse~vedphenotypic variance. T l ~ efinal simultaneous fit analysis recealed that
only 2.8% of the ad,justed phenotypic \,ariance for oviposition I was explained by this
single QTL for tlie \vliicli peak LOD value was just 1.72. It was obser\,ed that this
single putative QTL exhibited favourable additive effects contributed by the 296B
alleles (susceptible parent).
4.3.1.4 Oviposition incidence 11
For oviposition 11, the QTL analysis detected two QTLs in each of the two screening
environments. For late khoi.if(E~)the two QT1.s detected were mapped on LG 'E' and
LG 'J' and together explained 14.1% o f tlie observed phenotypic variance for this
trait. Based on the phenotypic values collected in screening environment Ez (rahi).
two QTLs were mapped, one each on LC; 'F' and LG 'G', and together these
explained 14.0% of the observed phenotypic variance. Tlie final simultaneous tit
analysis revealed that only 6.5% of the adjusted pi~cnotqpicvariance was explained
by two QTLs for which peak LOL) value was 3.99 in E l environment. While in El.
final silii~tltaneoustit of the t\vo Q l 1.s detected explained 8.7% of tlie adjusted
phenotypic variance. ror which a peak LOD value of 5.06 x a s observed. The
fa\ourable additive genetic effects for tlie 0TI.s detected in E l and 1: were mostly
contributed by tlie IS 18551 (resistant parent) alleles. The exception was one Q'TL
mapped on LG 'F' ill E2 for which filvourablc a d d i t i ~ eel'rects were contributed b)
alleles from 200B.
4.3.1.5 Dendhenrts incidence I
For deadliearts I. no VI I. was filitnd using phenotypic data collected in tlic late khorif'
( t T l )screeliing environnirlit. Ho\ve\er. froni data collected in the rrrbr ( E l )
enr,ironment. one Q I'L was derected on LG '1''. accounting 8.4'/0 of the observed
pllenotypic variance. I'he final simultaneoils tit analysis revealcd that only 3.5% 01'
tlle adjusted phenotypic variatice was explained b\ iliis single QTL for which the peak
LOD \,slue was 2.31. It was observed tliat fa\,otlrahle additive genetic effects for this
()TI, on L(; 'I" \\err contributed by alleles from susceptible parent 296B.
4.3.1.6 Deatlhearts i ~ ~ c i d e n cI1e
QTL analysis for this trait revealed one Q'TL and trio 07 Ls tliat \\ere detected in
screening environments E l (late k/7ar.if'2002)and E: (i.rihi 2004). respectively. One of
these Q'f1.s was ~iiappedon I-G ' T ' in hoth screening seasons and accounted for 6.6%
and 7.0% of the observed pllenotypic variances in El and E?. respectively. The total
observed plienotypic variance explained by thc tmo detected QTLs in E2 was 18.2%.
Tlie second UTl2 detected for deadhearts 11 in Ez \bas mapped on LG 'G' and
accounted for 1 1 . 2 O ? of obser\sed plleliotypic wriance. This is the most important
Ql'L for this trait detected in the rrrhi 2004 screening environment. Tlie final
silnultaneous tit analysis for this trait in this environment revealed that only 10.5% ol
the adjusted pl~enotypicvariance was explained by these two QTLs. for which peak
LOD value was 5.98. Favourable additive genetic effects were contributed by IS
18551 (resistance parent) alleles at the QTL on LG 'F' in the late kitorif screening
environment and for tlie QTL mapped 011 LG 'G' in the r(r11i scree~ii~ig
environment.
111contrast. for the QTL niappeil on I.G 'F' in the rcrhi screening environment, allele
from susceptible parent 296B contributed the favoi~rableadditive genetic effects.
4.3.1.7 Seedling height I (cm)
Seedling height I was not recorded in tlie khoi,i/ screening environment (El). For the
rzrhi season screen (E2) two Q I'Ls for seedling height I were niappcd, one each oti LG
'B' and LG '1'. l'ogetlier tliese QTLs explni~ied about 14.3'% ooT tile observed
phenotypic 5ariance. Final simultaneous fit analy5is 1~vea1t.dthat only 6.8% of the
adjusted phenotypic variance was explained by tliesc two Q'SLs. \villi a peak LOL)
score of 4.13. 'The favo~trableadditive genetic effects for tliese tmo QTLs for seedling
height I were contributed by IS 18551 (resistant parent) alleles.
4.3.1.8 Trichome derisity ( u p p e r leaf blade surface) (no./microscopic field)
For triclionie density on tlie upper surfi~ceof tile seedli~igleaf blade. QTL analysis
detected one QTI. each in both (lie kh[rrif2002 ( E l ) and rtrhr 2004 (Ez) screening
cn\.irontiients. Interestii~gly.tlie QTI. tletccted \<as napped at tlie same position on
I,(i 'G' i ~ iboth screening environment?. The detectcd Q'l'L explained 24.4% of the
ail,jl~stedphenotypic variance \\it11 n peak IAID score of 13.35 in Fl and 1 . 8 % of the
adjusted phenotypic variance with a peak LOD score of 8.39 in El. Tlie allele for Iiigh
tricliome density on tlie upper curface of tlie seedling Ieai hladc was i~ilieritedTroni
resistatit parent (IS 18551) for the QTL detected it1 both of tliese screening
r t ~ \ , i r u t ~ r ~ ~'ef l~~~i st 111ajor
s. QTL detected on LG 'G' mapped to marher i~itcrvai
. ~ g l l / l-~, ~l c i i ] ~ 6 7 .
4.3.1.9 Trirliorne densit?. (lolvcr leaf bliade surface) (no./n~icroscopicfieltl)
t o r tricliome density on the lo\\er surface o r tlie seedling leaf blade. one QTL in tlie
kl~to,if'20U2 screening ( E l ) and t\b,o QTLs in the rrrhi 2004 screenillp (E2) were
detected. One QTL %as tnapped at same position on 1.G 'G' in both screening
enviro~imentsand an additional QTL was mapped 011 LG 'C' in El.The detected QTL
on LG .G' explained 27.3% of tlie acijusted phenotypic variance in El and had a peak
LOD score of 15.2. While in Ez the two QTLs detected on LG 'G'and 1.6 C explained
17.3% of tlie adjusted phenotypic variance and had a peak 1,OD score 9.7. For the
QTL mapped on LG G governing tricliome density on the lower surface of the leaf
blade. IS 18551 contributed the favourable alleles. Flowever. the favourable additive
 104
genetic effect for tlie QTL n~appedon LG 'C' \%as contributed by alleles from
susceptible parent 2968. Based on analysis of pooled means across E l and E2, one
major QTL governing trichonie density on the lower surface of the leaf blades
tilapped LG 'G' between ~iiarkers,Ygo1]?01 and Xi.111167.
4.3.1.10 Plant height (cm)
One QTL for plant lieiglit \%as detected in each of tlie screening environment.
interestingly. tlie QTLs detected were mapped on to common position on LC 'I' in
these environments. The QTL detected in Elexplained 8.7% of the adjusted
phenotypic variance and had n peak LOD score of 4.64. In E2 the Q1-I, detected
explained 10.0% of tlie adjusted phenotypic vatiilnce and had a peak LOD score of
5.32. 'I'lie additive genetic effects for increased plant height were contributed by
alleles from taller parent IS 18551 in both screening environments. 'I'he UI'L detected
in both environments mapped very near to marker locus k;,11l136.
4.3.1.11 T i m e to 5 0 % ~flowering (d)
'I'\co QT1.s for time to 50% flowering were iletected in the kliarif 2002 screening
environnient (I:]) and one QTI. \\as detected in tile sliortcr dn! length rabi 2004
screening environment (Ez). One C)I'IL mappcd un LG 'A' \\'as nlapped in tlie same
position in both E l and E? and tlic other QTI. mapped on LC; 'I' in longer day length
screening environliient (El). l'lie QTLs detected o n I.(? A. explained 11.8% of the
e flowering time in El\\it11 a peak LOD vali~enf 6.7
ad,ji~stedpl~enotypicv a r i n ~ ~ cfor
and explained 6.5% o f the adjusted phenotypic mriancc for this trait in Ez with a peak
LO11 value of 3.56. l'lie f6l\,ourable additive genetic effects (earl! flil\\ering) \yere
contributed by 1S 18551 alleles for the Q I L detected on LG 'A' iii both screening
t.n\ironnlents. \\liilc fvr the Q f L detected in E l on LC; '1'. 2 9 6 8 allele contl.ihuted
Ihvourahle additive genetic el'fects.
4.3.1.12 Overall r e c o v e n score
For overall recovery score, the analysis ilctected four QT1.s expressed iii the kharif
2002 screening environment ( E l )but 110QI'L was found for this trait in rabi 2004 (Ez)
screeniug environment. In E l one QT1. was niapped on each of LC; 'B'. LG 'C'. LG
'E' and LG 'G'. Tliese four QTLs explained .3?.-Iu% of tlie obser\,ed phenotbpic
val.iance. Final simultaneous analysis re\zraled that. 0111) 17.1% of the adjusted
phenotypic variance was explained by these fi711r QTLs. ~ l l i c l itogether Inanaged a
peak LOD value of 10.44. Tlie favourable additive genetic effects for the QI'L-s
detected on LG '13'. LG 'E' and LG 'Cj' were contributed by IS 18551 alleles. while
for the QTI. detected on LG 'C' the favourable effects were contrihuted by alleles
from susceptible parent 2963.
4.3.1.13 Ahid damage score
For aphid damage score, two Q1'Ls were detected based on plienotypic evaluation in
tlie kharif 2002 screening environment and one QTI, was detected based on screening
in E2. One of the QTLs detected mapped to satlie position of LG 'J' for botli screening
environments and one QTL mapped to LG 'E' based on screening in El. The two
QTLs together explained 31.5% of tlie observed phenotypic variance for tliis trait in
E l . Final simultaneous anal>sis revealed tliat 22.7% of adjusted phenotypic variance
was explained by tliese two QTLs. which liad a combined peak LOD score of 12.7 for
E l . Tlie single QTL detected in E2 explained 10.4% or observed pllenotypic variance.
I:inal simultaneous fit analysis revealetl tliat only 6.0% of tlie aclji~sted phenotqpic
val-iance was explained by this single QTL. Lvitli a peclk I.OD score of 3 26.
Favourable additive genetic effects for aphid lo\\ inciilence was cantrihuted by allcles
from shoot fl? resistant parent IS 18551 in both screening environments. A major
()TI, for aphid resistance was mapped on LG '.I' in tile n~arkerinterval Xr.x/115-
.\5;1~283.
4.3.1.14 Pigmentation score
Pigmentation \\as not ~recol-dedin tlie 2002 k/i<ir.ifscreelii~lgen\i~-onment( E l ) From
tlie 2004 /.(/hi screening environment data t\\o 0'fI.s li)r li>liagc color \vcre detected.
one each on 1.G 'A' and LG '1'. \vliicli esplaincd 7.196 and 12.0% of tlie uhscrved
plienotcpic variance. respectively. For botli of tliese 0 1I.s positive additive genetic
effects tliat is darker foliage color were contributed b! alleles from IS 18551. The
tinill siniultaneous tit analysis revealed that on11 16.9% ol' the ac!iusted phe~lot!pic
variance for pigmclltatioll score \+as esplained by tliese two Q1'l.s. \\liicli liad a
comhined peak 120D value o r 9.43 tliat was signilicaiitly better tlia~lthe best single
QI'L model for tliis trait.
4.3.1.15 Miilge ilamage score
The midge damage score \bas not recorded in the 2002 khrrrif' season screenillg
environnient (El). For the 2004 rrrhi screening environment (E2). t\v11 Q-rL.s \\ere
detected. one each on l.G 'C' and 1.0 '1'. explaining 5.5% and 15.5% of tlic obser\cd
phenotypic variance. respectively. 'Hie simultaneous lit analysis revealed that only
10.2% of tlie adjusted plienotypic variance was explained by these two QT1.s. wliicll
liad a combined peak I.OD score of 5.86. Tlie fa\,ourable additive getietic effects for
low midge damage score were contributed by 2968 alleles.
 106
4.3.1.16 Agronomic score
Agronomic score of the RlLs and their parents was not recorded in the 2002 kharif
screening environment (El). For tlie 2004 rohi screening environment ('>) one QTL
was detected on LG 'A' explaining 5.5% o r the observed phenotypic variance for
agronomic score. The final simultaneous fit analysis revealed tliat only 5.0% of the
phe~iotypicvariances was explained by this single QTL. nit11 a peak [.OD score of
2.8. The additive effects for desirable agronomic score were contributed by 296R
alleles in this envirolmient.
4.3.1.17 G r a i n yield (glplot)
For grain yield under moderate to severe shoot fly pressure. the QTL analysis detected
three QTLs in E I and one QTL in Ez. In 2002 X.ir[rl.if scree~iingenvironment E l . one
QTL \%asmapped on each of L-Ci 'I:'. 1.G 'Ci' and 1.G '1'. I liese three Ql1.s together
explained 26.2% of tlie observed phenotypic variancc Tor grain yield ntider severe
$hoot fly pressure. while in the 2004 i.~ibiscreening environment (El) one (>TI. \\as
mapped o n I.(; ' C ' . which explained 6.4% of observed p11enot)pic variance for this
trait. I'inal simultaneous fit a~ialysisfur the data li.om E l rcvealcd tliat a total of' 11.5%
of tile acljusted variance could be eslilained 114 the thl-ee detected QTLs. \~liicli
together had a ]leak I.OD value of 6.96. Favo~lrableadditive genetic effects Mere
contributed b> tlie IS 1855 1 parental alleles in the kharif screening cn~ironmcnt.
wllile 20613 alleles contributed l'a\,ourable effects in thc i.[ihi screening en\ il-onmenr.
4.3.2 QTL anillysis across tlie t \ ~ oscreening ear.ironrnents
In order to detrrtni~lccliron~oson~nl
regions that are i~nportantfor the e'xpression of
the traits under different en\ironmental conditions and dlso to detect the Q * C
interaction eftkcts. (>'I-L analysis \\as done based on pi>olecl means of the phcnot!pic
valnes averaged over tile two screening enviroliments Tlle resi~ltsare presented in
'I'nhle 4.8 and Figure 4.5.
4.3.2.1 Glossiness score
For glossiness two QTLs \yere deteclrd in (lie across-e1lviro1111ietits analysis, together
explaining about 41.6% of tlie total observed pllenot!pic variance ill pooled entry
nieans. These Q'l'L,s \\ere mapped on LC; 'G' and 1.G 'J'. I'lie QTL ~nnppedon L.(i ' J '
was a ma.jor QTL explaining about 33.6% of total phenotypic \,nriance n i t h a LOD
peak value 15.92. For both of the QTLs esllibitcd non-sig~lificantQ E interaction.
Both of the Q'TLs the favourable additiw genetic effects were contributed b! IS
18551 alleles. After adjustment of the phenotypic bariance for Q E interaction alld
A x A epistatic interaction. tlie two QTL explained 3 1. I % of the phenotypic variance
with a combined peak LOD value of 18.0.
 TrlL

-
I
Dhl
I
Ovlll

Ovll
I

ORS Aphid
I
-
Flo
I
-
Ld
10 0

::
B
3% X? 5;:
z N

2P 5P ggO
x x x x z x x x
 -
- -
OvlIl

Gls
Aphld
ORS
I

Plht
.
CrYl

9 m " 2 R
O 8 3 2 S

- GrYl

TrlL
I

-
-
TrlU
I
Dhll

OVlll

CIS
 107
4.3.2.2 Oviposition incidence I
For ovipositio~i I two QTLs were detected in tlie across-environments analysis,
together explaining about 15.1% of the total phenotypic variance for pooled entry
means. These QTLs were niapped on LG 'C' and LG 'F'. The Q T L for oviposition I
iiiapped on LG 'F' exhibited significant Q x E interaction, while the other QTL
mapped on LG 'C' exhibited non-significant Q x E interaction. Favourable additive
genetic effects were co~itributedby alleles Ao~nresistant parent IS 18551 for the QTL
mapped on LG 'C'. For tlie Q'I'L mapped on 1.G 'F'. favourable additive genetic
effects were contributed by alleles f'rom susceptible parent 296B. 1:inal simulta~ieous
lit analysis revealed that two Ql'l,s togetlier explained only 6.4% of adjusted
phenotypic variance in pooled RIL means Tor ovipositiorl I with a peak 1,OD value of
3.8.
4.3.2.3 O~ipositionincidence I 1
For oviposition I1 four QTLs were detected by the across-environtllents analysis.
explaining together about 30.5% of thc observed plienotypic variance. These four
VI'Ls were tiiapped. one each, on 1.G 'C'. L.G 'F'. 1,Ci 'G',
and LC 'J'. The QTLs
mapped on LC; 'I:' and 1.G 'G' explained about 6.8% and 10.1% of observed
phenotypic variance. respectively; and these two QT1.s exhibited significant C) .* E
interactir~n.\<'bile QTLs mappcd on LG 'C' and LG '.I'exliibited non-significant C) x
11. interaction and explail~cd 6.0?4 and 7.6%. rcspectivcly of obscr\ed p1icnot)pic
variance. Final simultaneous fit annl)sis revealed that only 1.5. 1% of adjusted
phenotypic variance could be explained by these rour Q'f1.s together mith a peak
L.OD \'nlue 9.26. 'lhe ihvourahle adilitivc gcnelic cfkcts were contributed by IS
15551 alleles ibr rlic QTLs mapped 011 LG 'ti'. L t i 'J', and L t i 'C'. \vllile for the
QTL 11iapped on LC; 't.' alleles lion1 susceptible parent 796B contributed favourable
additive genetic eit'ects.
4.3.2.4 Deadhearts incidence 11
For deadhearts 1 no Q'fI, was found across analysis. Failure to detect the significant
Q'I'Ls for deadliearts I using the pooled inran across El. E2 the nlost probable reason
for this could be the \.cry little iiifCerences b e h e e n the KIL parents and hence vet.).
limited variatioti for deadhearts I in the RIL population due to high shoot fly pressure
in late kllarif screening e~ivironment.In tlie rrrbi screening initially the deadhearts per
cent was ]ow due to low population of shoot fly and diff'erences was limited.
 QTL analysis using the plienotypic mean values for the individual RIL
progenies. averaged across two screening environments. detected two QTLs for
deadliearts incidence at tlie second observed stage a i d these mapped on 1,G 'F' and
LG 'G'. Final simultaneous fit of tliese two Q'l'Ls together explained 10.6% of the
ad.iusted plienotypic \'ariance for tliis trait witli a peak LOD value of 6.01. They both
exliibited significant Q x E i~iteraction.For the QTL on LG 'G'. favourable additive
genetic effects werc contributed by alleles Erom resistant parent IS 18551. However.
for the QTL 011 LG 'P'. alleles from susceptible parent 296B contributed the
favourable genetic effects.
4.3.2.5 'Trichome density (upper leaf blacle surface) (no./microscopic field)
Across en\,ironmetit analysis found one major QTL for tricliome density of the upper
surface of tlie leaf blade. 'fliir QTI. mallpeti on 1,G .Ci' atid explained about 30.1% of
tlie o b s e n e d phenotypic variance for this trait witli a peak LOD value of 16.38. Tlie Q
E interaction for this trait was non-significant and 2Y.6u/'0 oftlie adjusted plienotypic
X~ariancefor poolcd RIL c11tl.y iiieans \%asexplained by tliis major QTL. The alleles
for increased tricliome density on the upper leaf blade surface a putative shoot fly
resistance component mcrc inherited kom resistant parent IS 1855 1.
4.3.2.6 T r i c l ~ o ~ ndensity
e (lower leaf blirde surface) (no./niicroscopic field)
Tv.0 QTLs were detected for this trait it1 tlie across-sea~onanalysis, l'lie Q 1.1. mapped
on 1.G 'Ci' ,-;is ;I 11i:ijor Q'l'L explai~lingabout 27.5O/~ of the observed pilenotypic
variance with a peak LOU \slue of 14.4. Tlie second QTL ]mapped on LG 'F'
explaining 9.3% of obsrr\.ed plielic~typic wriance. Both of QTLs exhibited non-
significant Q * E interactioti. These two QTLs together explained 32.4% of tlie
adjusted plienotypic \.:lriatice for pooled RII, means of this trait. Favourable additive
genetic elf'ects (illcreased tricliome dcnsity) for both QT1.s were contributed by IS
1855 1 alleles.
4.3.2.7 Plant lieiglit ( c n ~ )
One QTL was detected for plant height in tlie across-seasoti analysis. This Q'TL
mapped on LG ' I ' and explainetl about 15.Gn/oof observed plienotypic variance witli a
peak LOD value of 7.38. This QI'L exhibited non-signiticant Q x E interaction. This
si~igleQTL explained only 9.8°/b of adjusted p1ienot)pic variance in pooled RIL
means o f tliis trait in the filial sim~~ltaneous
tit analysis witli had ~ e a klod values 5.21.
Favourable additive genetic effects for increased height were contributed by alleles
from shoot fly resistant parent IS 18551.
4.3.2.8 Time to 50% flowering (d)
Two QTLs were detected for flowering time in the across-season analysis. These
QTLs were ~ilapped on LG 'A' and 1-G 'E'; and together explained 26.0% of
observed phenotypic variance for pooled mean tlowering time. One maior QTL for
this trait on LG 'A' explained about 18.2% of observed phenotypic variance wit11 a
peak LOD value o r 9.15. 50th QTLs exhibited non-significant Q x E interaction.
Final simi~ltaneousfit analysis revealed that only 15.9% of adjusted phenotypic
variance in pooled RIL means for this trait could he explained by these two QTLs
together. The QTL on LG 'A' l1ad fi~vtl~~rahle
additive genetic effects for early
tlowering contributed by IS I8551 alleles. while the Q 1'L for this trait on LO 'E' had
favourable additive effkcts contributed b> 296B alleles.
4.3.2.90vcrall slloot fly recovery score
f u o QTLs were detected for this trait ill the across-season analysis. These two QTLs
were mapped on LG 'E' and 1.G '.I.. In tlie final silnultaneous fit analysis these two
Of1.s together explained onl) 8.5% of tlie adjusted pllenot!pic ~ a r i a n c efor pooled
RIL means for this trait with a peak LOD \8aluc of 4.95, Both tliese QT1.s exhibited
non-significant (.) x F inter;lction fat. this trait. Favourable additive genetic effects
(better overall recovery) usere co11t1ihuted by allcles from resistant parent IS 1855 I
4.3.2.10 A p l ~ i dd t l n ~ a g escore
I'\vo Q'1'l.s \vcre detected in the across-seasons anol!sis for this [I-ait. fliese mapped
on I.(? '1:' and I S 'S'. Final s i n ~ u l t a ~ ~ ctit
u ~anal!sis
is of the Q r1.s together explained
otlly 1 R.GO/~ of adjusted phenotypic variance in pooled RIL means for this trait with a
peak LOD value of 10.35. The (>TI- mnppetl on 1.G '.I' for this trait was a major one.
explaining 20.4% of observed l>l~eiiotypicvariance with a peak LOD value 9.26. The
QTL tilapped on LC; '1'' eshihiletl significant Q E interaction. \vhile tlie QTL
mapped on 1.G '.I' sllowed no~i-significant x E interaction. Tlie facourable additive
genetic effects rar both Q-1l.s \\ere contributed by IS I855 1 alleles.
4.3.2.1 1 G r a i n yield (glplot)
Three QT1.s bvere detected for this trait i~ndcrconditions of noder rate to severe shoot
fly pressure in tlie across-season analysis. These napped on 1.G 'A'. LG 'G'. and I.G
'I'. These three QTLs together explained about 12.3% of adjusted phenotypic
variance with a peak LOD value of 7.38. All three QTLs exhibited non-significant Q
x E interaction, The tkvourable additibr genetic effects for these three QTLs were
contributed by alleles from resistant parent IS 18551.
4.4 Marker- assisted selection for shoot fly resistance traits in sorghum
A backcross breeding progralii is aimed at gene introgression from a donor line into
the genomic background of a recipient line. Tlie potential utilization of molecular
markers in such programs has received considerable attention in the recent past.
Markers can be used to assess the presence of the i~itrogressed gene (foreground
selection) when direct phenotypic evaluatioli is not possible or too expensive or only
possible late in development. Markers can also be used to accelerate the return to the
recipient parent genotype at other loci (backgroitnd selection). Tlie use of molecular
markers for background selection in backcross prograni has been tested
experimentally and proved to be ver) efficient. In the present study. the target for
introgression of a QTI. (Quantitative Trait Locus). that is a gene or gene block \vliuse
position is nol known witli certainty. but on11 estimated. In fact intropressing tlie
fa\'ourable allele of a QTI. bb recurrent backcrossing can be a powerful Inran to
improve the economic \slue of elite lines provided the expression of the QTL is not
reduced in tlie recipient genomic background.
4.4.1 Marker-assistetl breeding for shout fly resistance and component traits
Conventional breed~ngfor shoot fly resistance and its compunenl traits is often an
extremely slo\+ and laborious process and hecause of significant genotype x

ell\ iron~nentinteractions. the results tcnd to be locatio~ispecilic. Tlie application of

DNA markers a i d Q'1'1. mapping tecli~iologyis expected to facilitate breeding for
complex traits such as shoot tly resistance. After mapping Q'rL(s) for shoot tl)
resistance and its colnponcnt traits in a do1101.pitrent. nifirkers linked to the QTL(s) bc
emplo!ed to transfer these Q'fl.(s) from that doiior (resistant parent) to n
agronolnicall! elite but no re susceptible parent (recurrent parent). This process is
reSer~edto as foreground selection ill a marker-Assiste~iBreeding (MAI3) program. In
addition. selection Ihr rccltrrcnt p;trctit alleles at 111arkcr(s)unlinked to the QTIA can
he used during the M/\H program to hasten recovery of recurrent parent genotypes in
peno~ilicregions that are not in\ o l ~ e d\\it11tlie target QTL(s) (background selection).
Sa,jjanar and Folkertsa~iiaet 31. (2005 unpnhlislied) evaluated that 252 KILs
of a BTx623 x IS 18551 derived napping population for shoot fly resistant
~ i t in three environments. Same set of RILs mas genotyped using 109
c o ~ i i p o ~ i etraits
SSR niarker loci and Q'SL analysis was perfomled witli the aim of identifying the
genomic regions associated \kith shoot tly resistance. QTL analysis using l'lab QTI.
Version 1.1 revealed tlie presence of 28 QI'Ls detected at least in tu-o of the three
screening environments. Closely linked markers were identified for four QTLs for
deadhearts incidence. In the present study efforts are being made to transfer these four
deadhearts QTLs, by marker-aided selection, into three elite hybrids parental lines
developed at SRS, MAU, Parbhani. The markers associated with shoot fly resistance
traits are listed below.
Table 4.9n Target genolnic regions, linked SSR markers and associated shoot fly
resistallce QTLs for marker-assisted selection
--
Linkage Associated S K a r k e r s QTLs co-localized with genornic regions

A=SBI-01 Xtq175.St.rp37 Deadheorts 1, Oviposition 1

E=SBI-07 cpI%
,i,' 0362. ,Yt.11,40. ,Ytsl,3 12 Lleadhearts i. Oviposition I
G=SRI-I 0 Sispi0263. ,~qtrl,Ol.,Ytrpl4l Glossiness. Trichome density of upper and
lover leaf blade surfaces. Seedling \igor 11,
Oviposition I and 11. md Deadhearts 1 and 11
J=SBI-05 Xi.\pi(l258. ,S1vl,65, AYr.~j~15 (ilossiness. Seedling vigor 11. Ovipoaition I
and 11. [leadhearts I and 11

4.4.2 Checking the DNA conce~~tration

After isolatitig the DNA sa~iiplesofparel~talplants. all effected crosses (FI I,,I,,,~,)and
their backcross populations were loaded i n to 1 . ~ O / O agrose gels along \\.it11 standard
for testilig the ilNA qualltit!. and quality. If the hands \\ere clear. this indicated good
quality. DNA collcentlxiticins \\ere also assessed \\.it11 a Spectrafluorplus
spec~ropllolo~iieterusing the green fluore~cent tl!c ~ i c o g r e e ~ i ~ I.ike\vise
"'. DYA
quality anel ~ l ~ a ~ i t iwt !i ~ sassessed for all generatiolis and dilutions \\ere ~mi~dr
a DNA
accordingly to p r o d ~ ~ c e\vo~hillg sc>lutiotis for eacll s a ~ i ~ p l lla~~itig
e
collce~ilratiotlo f 2 . 5 og/l~l
PCR was done \\it11 selected SSli pri~nersfor both foreground and background
selectio~ifor all backcross generations. After each I'CR reaction \\as conlpleted. PCR
products mere electrophoreticall! separated on 6% ~non-denaturing PAGE gels and
the? were scored for parental band after sil\,er stai~~irlg.I'arental and beckcrossing
populatio~l samples liad PCK products for some primer pairs separated on an
autolnatic DNA sequencer (ABI 3700) and a~nplifiedproducts \\ere tllrn scored using
the Genotyper sofhvare
4.4.3 Parental genotyping with SSR markers
Twelve SSR marker loci linked to targeted shoot fly resistance trait of QTLs were
used for genotyping recurrent and donor parent plants to detect polyrnorphisni
between the three recurrent and four donor parents. 'The results of parent
polymorphism screenig, (Table 4.9b) sliowed that the flanking markers Xtxp75-
.Ytxp37 (LG A). ,Ytx]1312 (LG E). XIxp141 (12G G), and Sisp102.58 (LG J ) exhibited
allele size differences greater than 5 bp betmeen all twelve (3 recurrent u 4 donor
parents) cross con~binations.
The remaining flanking SSR markers viz. [,Ytxp,.lO (LG E), Xgc1/)01 (LG G).
,\i.17165 (LC; .I). and ,Yr.rplS (LG J ) ] esliibited allele sire differences less than 5 bp
between recurrent and donor parent pairs. except that ;Y.Y~JO(1.G E) was
nionomorpliic betrveen all three recurrent parents (280. 2OB. KR 192 and the KIL 252
ilonor parent, and for nlarker X~CIIJOI(1.G G ) recurrent parent 20B was monotnorphic
wit11 all four donor parents. I\larker locus ,\'ir/,lO.362 (L.G E ) esliibited allele size
difference more than 5 bp het\veen all recurrent and dunor parent pairs escept that K K
192 was monomorpliic \vith donors IiII. 189. IS 18551 and RII. 153:. marker
.\'i\plO263 (LG G ) \\.as ~~icinomclrpliic
\\it11 recurrent pal-ent 20B and donor parents IS
18551 and Ill1 252.hut pol!morphic \\it11 donor parent RIL li3.siniilarl! for this
ma1ht.r KII 102 was muno~norl~hic
with RlL 153 btit j ) o I ! n i ~ r p l ~\\it11
i ~ dt1110rparent
IS 18551 and RII. 252. hlarher .\ivjlc)4 (LC; .I) \\as m~inon?orpIiic for all three
recurrent parents \\it11 all four donor parents.
Alirr detectii~g pol!.~liol.pI>isln bct\veen recurrent and donor parents.
hon~oj.ygousparenlt~l-typelplanrs of the t\\o parmlts \\ere s e l e c ~ c ~atl seedling stage
Tor subaequelit crosser (plant to plant crosses) Finall! \ \ e si~ccecdcdto de\elop 7 F I
11) hrids. \vl~ichare listed he lo\^.
28R(288) KII. 189(.312!
KIi lc)2(104) IS IX55l(267)
?OR( 186) * RII. 252(.718)
2013(179) x I<IL 153(248)
KR 192(300) * RII. 252(3lC))
288(293) IS 18551(268)
288(292) RIL 153(252)
Out of the seven F I hybrids listcd above. the "hybrid " prclgen: for cross 28B(29.3) x
IS18551 (268) showed only the recurrent parent alleles (indicating Illat c r o s s i ~ ~had
g
Table 4.9b. Parental polymorphism (allele size, bp) using eleven SSR markersthat were used for
foreground selection in markerassisted breeding for shoot fly resistance in thisstudy
SSR Marker6 LG 288 200 KR 192 RIL 189 IS 18551 RIL 252 RIL153
Xtxp 75 A 165 167 1 6 9 150 150 150 150
Xtxp 37 A 177 184 188 167 167 167 167
Xisp10362 E 370 350 360 360 360 365 360
Xtxp 40 E 135 135 135 138 138 135 138
Xtxp312 E 165 170 230 185 180 195 180
Xisp10263 G 320 340 320 320 340
Xgap 01 G 252 254 252 254 254 254 254
Xtxp141 G 156 150 150 162 156 156 156
Xiqr10258 J 190 190 195 185 185 185 185
Xtxp65 J 130 130 130 132 132 132 132
Xtxp 94 J 211 211 211 211 211 211 211
Xtxp 15 J 225 223 224 223 22 1 221 221
failed and the progeny were in fact selfed of recurrent parent 288(293)) based on
molecular data scored with SSR markers. Siniilarly hyb~idprogeny of cross 28B(292)
x RIL 153(252) fail to germinate. Thus. only five crosses (FI hybrid) were advanced
for generating backcross populations.
4.4.4 Screening of F I and B C I F I populations based on recurrent parent 2 8 0 with
SSR markers
4.4.4.1 Testing hybridity with SSR markers
Five plants putatively produced from cross 28H (288) x RIL 189 were genotyped at
seedling stage sing four SSR loci linked with targetcd QTLs (Table 4.10a). Two
lieterozygous F1 hybrid plants were selected (plate4.2) and crossed with recurrent
parent 28B to produce R C I F Iseed.
4.4.4.2 C;enoQ-ping BCnFl population [2HH(288) x KIL 189(312)1 x 28B(288),

recurrent parent and donor parent \\it11 SSR marker for foreground selection
I hirty plants of this backcross populatioii \\ere screened at the seedling stage at loci
detected by eleven SSR primer pairs that targeted shoot fly resistance (2'fL.s ('Tablc
4.10b). Based on scoring SSR niolecular data fourteen lleterozygous plants having
appropriate allelic constitutions were selected (plate 4.3.4.4.4.5.4.6) at seedling stage
and crossed \\it11 recurrent parent 2813 (plant to plant cross) to penerate BCzFl
pop~~lation.Iletails of fourteen-i~itrop~~sscd plant \\.it11 targeted O.1-r. and its
associated shoot 115 sesistiince rl.aits prcsenteil belo\\ ('fable 4.1 Oc).
I Pop l
Pop II Pop Il! Pop iv Pop V

Pop l
-
w
ki

a p w
Pop 11
*
=
rrsl

*a?
sl i
Pop V
e
i*
. -
r y e

I **
.*lr**"
D

A 5 H A B n
A
*. ..
B W
.
A 5
. . H A 8
* -
W
1
a " r t
I

- -.- -
*
--*ar
Pop I1 Pop 181
OUIL - I s -

B
* **
?
 Table 4 . 1 0 ~List of introgressed plant with the targeted Q T L and associated
characters
No. of Targeted Shoot fly resistance trait association
heterozy QTL linkage
gnus group
plant
selected
--
3 LGA Deadliearts 1 and Ovipositiun I
1 LGE Deadliearts I and O~ipositionI
2 LG ALE Deadliearts I and Oviposition I
1 LG A+S Glossiness. Seedling vigor 11. Ovipositioli I and 11. Deadliearts I and 11
3 I.(; E+G+J Glossiness. Triclionie density (upper atid lolver leaf blade surfaces).
Seedling vigor. Ovipositioti I and 11. and Deadliearts 1 and I1
3 LG AtE-LJ Glossiness. Secdling vigor 11. Ovipositio~iI and 11. Deadhearts 1 and I1
I LG A t G t J Glossi~less.Trichotlie density (upper and lo\\er leaf blade surfaces).
Seedling vigor. Ovipositioii I and 11. and Deadliearts I and I1

4.4.5 Screening of FI and BCIFI populations basctl on recurrent parent K R

192(304) nit11 SSR mnrlter
4.4.5.1 Testing Iiybridih wit11 SSR 1o;lrkers
l e n platit putati\ely producetl froni cross [KR 192(304) x IS 18551(267)] \\ere
genotyprd 11silig four SSR loci lirihed with targeted Q'TLs (Table 4.1 la). All ten
plallts \\ere idelitifietl at seedlilig stage (plate 4.2) as lieterozygous and crossed \vitli
selted prclgeiiy of recurrent parent KK 192(304) to generate BCIFI seeds
4.4.5.2 Getlotyping UCIFl population IKR 192(304) x IS 18551(267)] x KR
192(304), recurrent parent and donor parent with SSll markers for foreground
selection
Tliilty plants of this backcl.oss population were genotyped at seedling stage \+ith
eleven SSR marker loci linked to targeted QTLs for slloot fl) resistance traits (Table
4.1 l b). Fifteen 11lants having appropriate allelic constitutions were identified and used
as females for backcrossiiig (plate 4.3.4.4.4.5,4.6) \\-it11 recurrent parent KR 192(304)
to generate BC2FI seeds. Tlie details of selected BClFt QTL introgression
heterozygotes with their targeted QT1.s and associated characters are presented
(Table 4.1 1c).
Table 4 . 1 1 ~List of selected BClFl introgressioli heterozygotes with their targeted
QTLs and associated characters
No, of Targeted Shoot tly resistance trait associations
selected Q T L linkage
heterozygous group
plants
4 LG A Deadhearts 1 and Oviposition I
2 LG G Tt.ichome density (upper and lo\cer leaf blade
surfaces). Glossiness, Seedling vigor 11. Ociposition 1
and 11. Deadhearts I and 11
5 t.GA+G Trichome density (upper and lower leaf blade
surfaces). Glossiness. Seedling vigor 11. Oviposition I
and 11. Deadliearts 1 and 11
3 LG A+J Glossiness. Seedling vigor 11. O\,iposition I and 11.
I)eadlieal-ts I :lnd I1
I LG A+E+J Glossiness. Seedling vigor 11. Oviposition I and 11.
Ileadhearts I and 11

4.4.6 Screening of F l and UCtFl population based on recurrent parent 20B(186)

\r ith SSK marker
4.4.6.1 Testing hybritlity $1 ith SSll a ~ a r k e r s
Tell plants putatively produced fiom cross 20H(186) KIL 252(318) were genotyped
at seedling stage with four SSR marker loci linked \!it11 resistance trait Q'I'Ls (Tahle
4.12a). Only two heterozygous (FI h>hrid) plants 1vel.e (plate 4.2) identified and
crossed as fenlale with pollen from selfed progen) of recurrent parent 20B(179) to
generate BC,Fl populations.
4.4.6.2 Genotyping B C I F l populations [20B(186) x RIL 252(318)1 x 20B(186),
recurrent parent and donor parent with SSR markers for foreground selection
Thirty plants were screened at seedling stage with ele\,en SSR loci from four linkage
groups associated with shoot fly resistance traits (Table 4.17-b). Twelve heterozygous
plants for one more targeted QTL introgressions were selected and used (plate
4.3.4.4.4.5,4.6) in crossing to advance to the B C ~ F Igeneration. The details of
introgressed QTLs and their associated resistance characters are listed below for these
12 selected BCiFl plants (Table 4 . 1 2 ~ ) .
Table 4 . 1 2 ~List of introgressed plant with the targeted Q T L and characters
associated
No. of Targeted Shoot fly resistance trait associations
lieterozygous Q T L linkage
plants group
selected
--
6 LG A Oviposition I, Deadhearts I
3 LG E Oviposition I, Deadliearts I
2 LG A+E O\iposition I. Deadliearts 1
1 LG A+J Glossiness, Seedling vlgor 11, Deadliearts 1 and 11.
Oviposition I and I1

4.4.7 Screening of F1 and 13CIFIpopulations hased on recurrent parent ZOB(179)

\r ith SSR niarker
4.4.7.1 Testing of hybritlity with SSR markers
'I en plants putatively produced from cross 20B(179) x RIL 153(248) were gellotyped
at seedling stage with four SSI< niarlters from three linkage groups (Table 4.13a).
Eight heterozygous plants mere selected and used for crossing (plate 4.2) as females
with pollen fro111selfed progen? of recurrent parent 20B(179) to advance to the B C \ F I
gcnernlion.
4.4.7.2 Genotyping RCIFl populations [209(179) x RIL 153(248)] x 209(179),
recurrent parent, and dunor parent with SSR markers for foreground selection
Thirty plants of this backcross population were screened at seedling stage with eleven
SSR marker loci linked with targeted shout tly resistance QTLs. On the basis of
molecular marker data, 19 plants having appropriate allelic constitutions
(heterozygous for one or more target QTL intogressions) were identified (Table
4.13b) and crossed as female with pollen from selfed progeny of recurrent parent
20B(179) to advance to the BC2Fl generation (plate 4.3.4.4.4.5.4.6). The details of
these 19 selected plants heterozygous with various targeted QTLs (and associated
characters) are presented below (Table 4 . 1 3 ~ ) .
Table 4 . 1 3 ~List of selected Q T L introgression heterozygote plants with the
targeted QTLs and associated characters
-
No. of hetero- Targeted Shoot fly resistance trait associations
n g o u s plants QTL
selected linkage
grou~(s)
3 LG A Deadhearts I and Oviposition 1
5 LG G Trichomes (upper and lower leaf blade surfaces). Seedling vigor 11,
Glossiness. Oviposition I and 11. Deadliearts I and 11
I LG E Dcadhearts I and Ov~positionI
1 LG A t G 'Tricho~~les
(upper and lower leaf blade surfaces). Seedling vigor 11,
Glossiness, Ovipositioil I and 11, Deadhearts I and I1
2 I,G E t S Glossiness. Seedling vigor 11. Oviposition I and 11. Deadliearts I and I1
I LG A+J Glossiness. Sccdling vigor 11. O\-iposition I and 11. Deadhearts I and I1
1 I Tricho~nes(upper and lo\zer leaf blade surfaces). Seedling vigor 11,
(?lossiness. Obipositioli l and 11. Decidllearrs I and I1
1 1.G Tricliumcs (upper and lo\ver leaf blade surFaces). Seedli~ipvigor 11,
A+t.+Ci O\.iposition I aid 11. Deadllearts I and I1
Cilossi~~ess.
I LG (ilossiness. Tricllo~lledensit!. (upper and lo\\er leaf hladc surfaces).
AtQtJ Seedling vigor II. O\ipositio~lland 11. Deadllcarts I and I1
I LG (;lossiness. Seedling vigor 11. Oviposition I and 11. Deadllearts 1 and I1

4.4.8 Screening of FI and HCIFl populations based on) recurrent parent KR

IYZ(311U) aitli SSR markers
4.4.8.1 Testing hybridity wit11 SSR ~ n a r k e r s
Nine plants tllought to have bee11 produced fro111cross KR 192(300) x RIL 252219)
were screened at seedling stage uith four SSR nlarker loci associated \\it11 targeted
slloot fly resistance QTLs (Table 4.14a). Eiglit heterozygous plants nere identified
(plate 4.2) and crossed as feniales with pollen tiom selfed progeny of recurrent parent
KR 192(300) to generate seed of tlie BClFl generation.
4.4.8.2 Genotyping BCIFI population [KR 192(300) x RIL 252(319)1 x KR
192(300), recurrent parent, and donor parent with SSK marker for foreground
selection
Screening at seedling stage of thirt) BCtFl individual with eleven SSR marker loci
linked to targeted shoot fly resistance Q'I'Ls upasperfor~netl(Table 4.14b). The PAGE
separated SSR data revealed nine Iieterozygous plants (plate4.3,4.4,4.5,4.6)
appropriate allelic constitutions. 'These were crossed as female with pollen fro111selfed
progeny of recurrent parent KR 192(300) to generate 13CrF1 seed. The details of
selected QTL introgress heterozygote plants are presented ('Table 4 . 1 4 ~ )
Table 4 . 1 4 ~List of introgressed plant nith the t:~rgetedQTL and cl~aracters
associated
No. of Targeted Shoot fly resistance trait a*sociations
hrterozygous QTL
plants linkage
selected groups
4 LC; i\ Oviposition 1. 1)endhetil.t~I
1.G G Trichon~e dens~ly (vpper and lo~rer leaf blade
3 surfaces). Seedling vigor 11. O\ipclsition I and 11.
Deadl~eartsI and I1
1 I+GA+J Glossiness. Seedling \igcv 11. O\iposition I ai~d 11.

I LC; G-i-J Glossii~ess.Trichome densit! (upper and lo~ver leaf

vigor 11. Oviposition I and 11.
blade sul.fi~ccs).Secdlt~~g
Deadhearts I and I1

4.4.9 Parental genotyping with SSR tnsrker primel. pairs used for hackground
selection
Inilial parental screening with 38 SSR marker loci ~o\~eritlg
the entire genonle except
the four regions harboring targeted shoot fly resistance QTLs \\as carried out before
actual genotyping of selected foreground BCIFl plants for backgrouiid screening. The
main ob,jective for screening of parental plants with these 38 SSR marker loci was to
detect polymorphism among the parents. That could be used background selection to
speed recovery of recurrent parent genotype in genornic regions distant from the four
targeted QTLs. The parental genotypi~igresults (Table 3.13) revealed that SSR
markers pairs viz. Xcup63 (LGR). Xtxp283 (1,G B) Xtxp59 (LG C) and Xtxpl7 (LGI).
failed to detect polynlorphism among the pairs of recurrent and donor parents. The
remaining 34 SSR ]marker loci exhibited polymorphism alllong at least some of the
pairs of parents. However twenty-two SSR marker loci exhibited monomorphism
between one to four pairs of parents as given in 'Table 4.15
 Table 4.15 List of SSR marker loci that are monomorphic behveen pairs of
parents
Name of Lc Monomorphic parental line pairs
SSR
marker
locus
,Y/.rp25 KR 192(304). IS 1855 l(267)
.Yt,~p296 KR 192(304). IS 18551(267); 20B(179), R11.153(248)
Sc11pl l 208(186). RII. 252(3 18)
Xc11p6l 20B(186). KIL 252(318)
.S/.Y/I~I4 KR 192(300). KII, 2521319)
,ikrr/1236 KR 192(304). IS 18551(267); 20R(186). KIL 252(318): 20B(179). RIL 153(248)
.rcll]l 14 208(186), RIL 252(3 18); 200(179), RIL 153(248)
.Y?A/~10 280(288). RIL 1 XY(3 12); 20B( 186). KIL 252(3 18): 20R(179). RIL 153(248)
.\i,11/~28 288(288). KIL 189(312)
.Sg1)sb50 28B(288). RJI. 189(312); 20B(186), KIL 252(318); 2013(179), RIL 153(218)
.Y/YIJ~~~ KR 192(304), IS 18551(267)
,Y/.Y/l289 20B(186). RIL 252(318)
'Yl.lylo2 20B(186). RIL 252(3 18); ?08(179). RIL l53(?48)
.\i.1/1?30 KK 19?(300). RIL 25?(319)
.1'/.~{1??3 288(288). Rll. 189(31?)
.li.Y/'47 lOB(186). Rll. 25?(318): 20B(179). RII, 153(248)
.\k/lI05 2i?B(788). Rll. 189(31?): ?OB( 186). KIL 152(318):20B(179), KIL 153(?48)
.\ix11354 208( 186). RII. 252(318)
.Yl.Y/l I 7 2881288). RII.189(311-): 208(186). RIL !52(318): 20B(179). RIL 153(248): KR
192(3001. RII. 25?(319)
?OB(IX6). RII. 152(3 18)
288(288). RIL 189(3l?)
288(288). RIL 189(.3l?): 20B(186). RIL 252(318)
4.4.10 Background genotyping of B C I F I individuals selected on the basis of
foreground selection
A total of 61 BCIFI plants selected through foreground screeliing and forming five
back cross populations. were genotyped with a set of polymorphic SSR markers
(Table 4.15) covering the entire genome except the genomic regions harboring
targeted shoot fly resistance QTLs ( i t the regions covered in foreground screening).
Approximately two SSR marker loci were selected to cover the top, middle, and
bottom portion of each of these six non-target linkage groups. The main objective of
background selection was to confirm (and hasten) recovery ooT the recurrent parent
genome. Twelve plants were selected from five backcrossing populations, These
twelve plants each carly homozygous recurrent parental alleles (A genotype) at most
of the SSR loci and have a few heterozygous loci (1-1 genotype) used for background
screening (Tables 4.1 6, 4.1 7. 4.1 8. 4.19. 4.20). Those individuals homozygous Ibr any
donor parent allele ( B genotype) were rejected as they could only have resulted from
failure of backcrossing (i.e.. selfing) in the previous generation. 'l'he 12 selected
form five populations.
individuals (plate 4.7.4.8a.b.4.9a.b.4.10aab,4.11a.b.4.12.4.13)
had been crossed with their respective recurrent parents to advance this marher-
assisted Ql'L introgression programme to BC2FI generation. Detail regarding selected
individuals including targeted QTLs and character associations are presented in Table
4.21.
4.4.1 I Genowping of the five BCZFI populations, recurrent parent, and donor
parent with SSR markers for foreground selection
Genotyping of 124 BCzFl plants from five backc~.osspopulations \vitl? 10 SSR marker
loci linked to targeted QTLs associated with shoot fly resistance traits in four linkage
groups was performed (Tables 4.22. 4.23. 4.24. 4.25. 4.26).0ne hundred lieterozygous
BC:Fl plants with appropriate allelic constitutions were selected at the seedling stage
(plate 4.14.4.1 5.4.16.4.17) and crossed (as female parent)\vith their respective
recurrent parents to generate I3C3Ft populations. Details the nunlber of plants
genotyped. number of introgressed plant selected with its targeted QTL are presented
in table 4.27.
4.4.12 Genotyping selected BCIFl fore ground plants for background selection
Out of 100 BC2FI selected plants in foreground screening only 68 plants from five
back cross populations will be genotyped with a set of polyn~orphicSSR loci (Table
4.28) covering the entire genome except the region harboring targeted QTLs. This
background screening will be restricted to the loci that were heterozygotls and or not
amplified in BCIFIgeneration background screening.
Table 4.28 Details of BCZFIfore ground selected plants chosen for back ground
screening
Back cross population Plants selected Targeted QTLs
BClFl {[288(288) x RIL 189(312) x 18 5LGA
28B(288)]} x 28B(288) 4 LG E
4LGG
I LGJ
3 I,G E+G
1 LG A+J
RC2Fl [KR 192(304) x IS 18551(267) x KII 12 2 LG A
192(304)] x KK 192(304) 3LGG
7 LG A+G
BC2Fl [20B (186) x K1L 252(318) y 04 4LGE
20B(186)] x 20B(186)
UC2I2l [20B (179) x R1L 153(248) x 24 3 LGA
20B(17Y)] x 208(179) 4 1.O E
. G
3 1G
3 LC1 .I
3 LG A+G
5 LO A+J
3 LG A+G+J
BC2FI [KR 192(300) x RIL 252(319) x IiR I0 4LGA
l92(300)] x KR l92(300) 2LGG
3 1.G J
1 LG A+J
DISCUSSION
 CHAPTER V
DISCUSSION
Shoot fly is a major insect pest of cultivated sorghun~.Adoption of chemical control
method is not economically feasible for most of the sorghum-growing farmers.
Therefore host plant resistance is itself excellent pest controlling method, and when
integrated with other methods of insect control offers a sound approach to deal with
insect pest. This approach holds great potential for sorghum. which is known to be
poor mans crop. Shoot fly resistance involves number of component traits. which are
quantitative in nature and influenced by G x E interaction. There fore the phenotypic
selection for this trait is difficult Marker-assisted selection could increase efficiency
of breeding of such traits. Efforts are being made in this study to carry out
experinients on genetic diversity analysis, Q'I'L mapping and marker-assisted
selection for shoot fly resistance in sorghum. The discussions on results obtained are
presented below objective wise.
5.1 Application of SSR markers in diversity analysis of sorghum insect resistance
germplasm accessions
Genetic resources have evolved as a product of domestication. intensification.
diversification and improvement through conscious and unconscious selection by
countless generations of farmers. 'fhese landraces and improved cultivars provide the
basic and strategic raw materials for crop improvement the world over for present and
future generations (Mangala Rai. 2002).
Sorghum has an immense range of genctic resources, with much of the genetic
variability available in the African regions where domestication tirst occurred, and in
the Asian regions of early introduction. In Africa, the genetic variability occurs as
cultivated species, wild crop progenitors and wild species (de Wet and Harlan. 1971;
Gebrekidan. 1982). Landraces and wild relatives of cultivated sorghum from the
centers of diversity have been rich sources of resistance to new pathogens, insect pests
and other stresses such as high tetllperature and drought, as well as sources of traits to
improve food and fodder quality. animal feed and industrial products. Ilowever. this
natural genetic diversity is under threat through natural selection, the destruction of
habitats, by the spread of agricultural practice, and adoption of improved varieties.
To prevent the extinction of landraces and wild relatives of cultivated sorghum
the collection and conservation of sorghum germplasm was accelerated about four
decades ago. Since then, germplasnl collection and conservation has become an
integral component of several crop improvement programs at both national and
international levels (Rosenow and Dalhberg. 2000).
Analysis of the extent and distribution of genetic variation in a crop are
essential in understanding the evolutionary relationship between accessions and to
sample genetic resources in a more systematic fashion for breeding and conservation
purposes. Traditionally, taxonomists classified genetic resources in sorghum based
mainly on morphological markers (Harlan and de Wet. 1972; Murty er nl., 1967: ).
Harlan and de Wet (1972) used a small number of easily recogtiizable traits, including
grain shape, glumes, and panicle shape, to partition variability in cultivated sorghums
into five races and 10 intermediate (hybrid combinations of the major races) forms.
'fhe morphological traits used by Harlan and de Wet (1972) are conditioned by a
relatively small number of genes (Doggett. 1988). However, several complex
quantitative traits, which are related to habitat adoption and particular end use of the
crop, exhibit enornious variability among sorghum accessions within each race (de
Wet er al., 1976). Thus, classifying germplasm accessions based solely on a few
discrete morphological characteristics tilay not provide an accurate indication of the
genetic divergence among the cultivated genotlpes of sorghum. Since both natural
and human selection have contributed to genetic differentiation in sorghum (Murty rt
01.. 1967), landraces of the same race grown in different habitats may have greater
genetic dissimilarity than those of different races from the same habitat.
Biochen~icaland molecular markers are now widely used as tools to assess the
validity of taxonomic classification in cop plants. Allozyrne markers have been used
extensively to evaluate the extent and pattern of genetic %ariationin sorghum (Aldrich
et trl., 1992: Morden er 01.. 1989. 1990). Allozymes did not clearly separate the
various races of cultivate and wild sorghum. Instead. these markers showed some
degree of differentiation related to geographic area of origin. Because allozymes only
measure variation at a very limited number of sites, these results may not reflect
overall patterns of genetic variation throughout the genome (Aldrich er al., 1992).
Restriction fragments length polymorphisnl (RFLP) and RAPD markers can
overcome the limitation of allozymes because they have the potential to identify a
large number of polymorphism with good coverage of the entire genome (Melchinger,
1993). These techniques have been used to characterize genetic diversity and
phylogenetic relationship in sorghum (Aldrich and Doebley, 1992; Cui e/ al., 1995;
Deu e t a/., 1994; Tao e l a/., 1993). However, these studies provided conflicting results
concerning the degree of differentiation among cultivated races o f sorghum. These
studies also assessed genetic diversity either in a relatively small number o f
accessions (Aldrich and Doebley, 1992; Cui er a/., 1995; Tao er a/., 1993) or in non-
random accessions selected on the basis o f a prior information from other studies
(Due e t al., 1994). Thus, extensive random sampling from the world collection o f
sorghum gemplasm resources may allow a less biased assessment o f the genetic
diversity o f the crop. Recently microsatellite or SSR (Simple Sequence Repeat) loci,
which correspond to tandomly repeated DNA with a very short repreat unit have been
introduced as powerful genetic markers in plants (Morgante and Olivieri 1993; Powell
er a/., 1996a). Comparative studies in crop plants have shown that microsatellite
markers are more variable than most other molecular markers (Powell e l al., 1996a ;
Taramino and Tingey, 1996; Pejic er a/., 1998) and provide a powerful methodology
for discriminating between genotypes (Yang el a/., 1994; Russel et a/., 1997;
Bredmeijer et a/., 1998). SSRs are a highly useful class o f such PCR-based genetic
markers. Although costly to develop relative to some other classes o f genetic marker,
once developed their analysis is both easy and inexpensive. They are co-dominant,
occur in high frequency, and can display a high level o f polymorphism even among
closely related accessions. Their high information content and other favorable
characteristics made then excellent genetic markers for many types o f investigation
including marker-assisted selection and finger printng o f gel nplasm collections
(Brown e l al., 1996).
The analysis o f genetic diversity and relatedness between individuals within a
species or between different species or populations, is a central task for many
disciplines o f biological science. Genetic diversity and phylogenetic studies were
initially conducted using qualitative and quantitative traits, which are mostly
morphological. Using various statistical methods such as analysis o f variance
(ANOVA), covariance, and diversity measures such as Mahalnobis D~ statistic,
metroglyph analysis and Principal Components Analysis.These analyses are mostly
based on quantitative traits that are highly influenced by env~ronmentaleffects and
require tedious statistical procedures. Molecular markers are being widely used in
various areas o f plant breeding as important tools for evaluating genetic diversity and
determining cultivar identity (molecular fingerprinting). Establishlnent o f a molecular
marker and phenotypic assessment database o f crop germplasm will help breeders to
trace down the origins and degrees of relatedness of many landraces and cultivars.
considering the potential of molecular niarkers crop breeders can extend their hall&
to use these to supplement other tools currently exploited in their crop breeding
activities.
In this study, we tried to assess the genetic diversity of a set of 91 elite
sorghum germplasm accessions using SSR markers. The set includes 12 shoot fly
resistant accessions, 15 stem borer resistant accessions, 9 accessions iesistant to both
shoot fly and stem borer, 17 midge resistant accessions and 38 agronomically elite
recurrent parents that were used at ICKISA'I' to initiate large-scale marker-assisted
backcross program for the stay-green component of terminal drought tolerance frotn
donors 835 and E36-1, from which QTLs for this trait have been previously mapped
(St~budhier al., 2000; Haussmatln L./ rrl.. 2002).
Genomic D N A isolation was done by the C'TAB method. DNA of the 91
sorghum accessions were then genotyped usins 21 SSR primer pairs that detected loci
distributed over 9 of the 10 linkage groups in the sorghutii nuclear genome. The
NTSYS statistical software package was used for cluster analysis. Jaccard's similarity
coefficient between each pair of accessions was used to construct a dendrogran~using
the unweighted paired group method with arithmetic a\,erages (UPGMA).
5.1.1 Poly acrylamide gel electrophoresis (PAGE)
PCR products from 20 SSR primer pairs atid tetnplate DNA samples frotn 91
sorgllu~naccessions werc separated electrophoreticall~ using six percent denaturing
polyacrylamide gels. The allclic composition of' each genotype was determined b~
scoring silver-stained gels of the separated a~iiplifiedproducts. Twenty out of 21 SSK
primer pairs used provided amplificatio~iproducts. while 11 out of these 20 revealed
high levels of polymorphism (> 0 . 5 ) among the 91 sorghum accessions. A total of 69
alleles were detected by silver staining. An average 3.45 fragments \%ereamplified per
SSR locus among the 91 sorgliutn accessions studied.
The polymorphic information content (PlC) value range observed for these
SSR loci was 0.13 to 0.83. The highest level of polymorphism was found with the
primer pair for SSR locus Sgap84 (0.83). followed by those for Xtxp15 (0.82) and
Xtxp320 (0.77). The lowest polynlorphism was found with the primer pair for SSR
locus &'cup62 (0.13) (Table 4.1). These results in agreement with Smith el a[. (2000)
and Kamala el (2005) using the SSR molecular marker system in differelit
sorghum germplasm accessions.
5.1.2 PAGE dendrogram
Jaccard's coefficient of similarity for pairs of the 91 sorghum accessions studied
ranged from 0.28 to 1.00. The dendrograni for the genetic similarity between
accessions based on PAGE-generated SSR genotypic data showed clustering for
geographic origin, sorghum races, raw germplasm versus elite breeding lines and
specific traits such as insect resistance. The accessions studied were broadly grouped
into clusters representing 4 of the 5 sorghum races (L)rrrrrr.(budaturn (including elite
niaterial derived from Zero Zera landraces), Bicolor, and Gl~inra).
These results agree
with those observed previously by Tao et al. (1993) and Oliveria et al. (1996) using
other molecular marker systems. The 91 sorghum accessions studied diverged into 20
clusters at the 50% level of similarity (Fig 4.2). Among these the largest was cluster 4
(18 accessions) was followed by cluster 12 (13 accessions) and cluster 13 (12
accessions). However, some of the clusters (cluster 5. 7, 9, 10, 14 and 20)
accommodated only a single accession each and clusters 1. 3, and 6 accommodated
only 2 accessions each.
Cluster 4 included 18 genotypes. most of which exhibit resistance to sorghum
shoot fly and spotted stem borer. All sorghum genotypes in this cluster originated
from tlie Durra race. This group contains genotype IS 1855 I , which has been used as
tlie resistant parent in developnlent of two ICRISAT shoot fly resistance mapping
populations. Maiti et al. (1984) reported most of tlie shoot fly resista~itaccessions
belong to Dtrrrct group and some others to taxonomic groups such as Glii17ru.
('trr~drrrlrntand Bicolor. Premkisliore and Jotiwani ( 1 979). Shar~naet al. (1983), Prem
kishore (1984 reported most stem borer resistance sources belonged to the Dlnm
group among sorghums of Indian origin, followed by ('trrrcfattrr~i.('otispictc~~rn
(a
(a subgroup within the Kqffir race).
subgroup within the Gtrinrtr race). C'trffrorr~rn
Roxburghii (a subgroup within the Grrinrci race) and i k r ~ a . ~ r r(amsubgroup within the
Bicolor race).
These stem borer resistance sources are tall, late in maturity, susceptible to
lodging, photoperiod sensitive, low yielding and possess moderate to high degrees of
resistance. IS 21 121, IS 2265, IS 2312, IS 2195 and IS 2123 from cluster 4 have been
identified as sources of resistance against both sorghunl shoot fly and spotted stem
borer. These results are in agreement with reports by Jotliwani and Devies (197% and
Prem kishore and Jotwani (1982.), Most of the genotype pairs in this cluster have
operational bootstrap values greater than 50%. which provides confidence about their
clustering.
Cluster 8 contains 4 genotypes representing an intermediate population
developed from D ~ i r r ax Cloudarum crosses. These are all elite breeding lines and
known for their combination shoot fly resistance with better agronomic performance.
The genotypes from this group could be used as sources for the development of a new
mapping population for grain yield and shoot fly resistance.
Cluster 12 was comprised largely of ICRISAT-bred improved breeding lines
having sorghum midge resistance in agrolioniically superior C'atrdc~tumbackground
with Zera Zera landrace parentage and excellent grain quality. According to Rosscto
et al. (1975): Sharma et a!. (1992) and (199.3a). 'I'AM 2566. AF 28. DJ 6514 are stable
sources of resistance to sorghum midge. Jothwani and Davies (1979) and Sliarrna and
Davies(l981) reported that most of identified midge resistance sources belong to the
('~ludatum 1 Negriccms-%raZera, C'aridott~ni/Kqffir (ileguri) / Dt~rro-~Vegricc~t~~ri
C'tr~rdrrtunz-Bico(orgroups of sorghum.
Cluster 11 and 13 consisted of largely recurrent parents for tlie stay-green
backcrossing program and hence were conipriscd largely of improved C'trridottim race
materials including Zeru Zerrr landrace derivatives. According to Ilash et al. (2003)
marker-assisted selection for QTLs co~itrollingthe stay-green trait (a component of
terminal drought tolerance) in sorghum is in progress at ICKISAT Patancheru. Six
QTLs of relatively large effect from donor parent 8 3 5 , \\hich have independently
~ilappedby two or more groups of earlier workers. are targeted in this program with
agronomically elite genetically diverse sorgliuni Larietics R16. ICSV I I I . IRAT 204,
and ISIAP Dorado as recurrent parents.
Cluster 1 had only two genotypes, 296R and HC 260. Elite hybrid maintainer
line 2968 is susceptible to insect pests but is a potent combiner for high grain yield. It
has been used extensively in the khurif season hybrid breeding prograln in India. This
elite line has been used as the susceptible parent in a shoot fly rcsista~ice QTL
mapping population developed at ICRISAT.
Cluster 6 contains genotypes 8Tx263 and Supanburi I I. Both genotypes are
susceptible to sorghum shoot fly. BTx263 was used as the susceptible parent for the
first ICRISAT sorghum RIL population developed for QTL r n a ~ p i ~ iof
g shoot fly
resistance. Cluster 16 had three genotypes: LS 1 and LS 2 originated from the
Peoples' Republic of China. and the third genotype, i.e. Malisor 84-7, was developed
from Guinea race crosses to Cuudutum material in ICRISAT's breeding program in
Mali. All three are potential recurrent parents for the stay-green marker-assisted
backcrossing program. Single genotype clusters 19 and 20 appear to represent the
grassy Bicolor race of sorghum.
Molecular genetic diversity analysis was carried out on 91 sorghum accessions
differing in agronomic eliteness and the level of resistance in several insect pests
using the allelic information from 20 SSR loci as revealed by silver-staining of
PAGE-separated PCR products. This analysis revealed that the accessions studied are
genetically quite diverged, with sorghum accessions showing midge, shoot fly, and
stem borer resistance clustering in different groups. In addition, clusters of
agronomically superior recurrent parents have been identified that are genetically
quite divergent from each of these insect resistant clusters. flowever, some of the
midge, shoot fly and stem borer resistant accessions cluster separately indicating that
these accessions might contain new allelic variants that should be exploited in applied
breeding programs. This infornlation will be useful to identify parents for use in
marker-assisted backcross introgression of insect resistance QTLs from the currently
available mapping populations (in some cases from more agrono~nically elite
pedigree-derived insect resistant breeding lines) as well as for identifying additional
parental line pairs to use in developing new mapping populations to detect additional
insect resistance QTLs.
5.1.3 Capillary electrophoresis (ABI)
The genotypes studied using separation of PCR products on PAGE were also assessed
for their polymorphism using automated capillan electrophoresis (ABI 3100/ABI
3700 sequencing machine). A total of 1 18 alleles generated by 20 SSR primers were
detected. On average 5.1 fragments were amplified per SSR locus. Thirteen out of 20
(65 %) of SSK primer pairs detected high levels of polymorphism with PIC calues
>0.5. The PIC values observed ranged from 0.21 to 0.81. The highest level of
polymorphism was found with primer pairs for SSR locus Xtxp320 (0.81) and the
lowest polymorphism was found with primer pairs for SSR locus Xi.up60 (0.21)
(Table 4.1). Thus, the most of the polymorpl~icgroups of sorghum SSR markers did
not show substantial changes across the two PCR product separation and visualization
systems.
5.1.4 ABI dendrogram
The Jaccard's coefficient of similarity ranged from 0.21 to 1.OO among pairs of the 91
sorghum accessions studied. Marker alleles detected in the ABI-generated data sheets
gouped the 91 sorghum genotypes into 28 clusters at the 50% level of similarity.
When compared to the dendrograrn from the PAGE-generated data sheet. the number
of clusters detected with ABI-generated data sheet was comparatively higher (Fig
4.3). This may be due to the greater sensitivity of the automated sequencer. which
allows it to detect SSR alleles differing by smaller numbers of repeat units so that it
can effectively indicate higher levels of polymorphism.
The largest cluster was cluster 17, which co~isistedof 14 genotypes. Many of
the genotypes of this cluster show midge tly resistance and are agronomically elite
lines selected as a potentially recurrent parents for the ICRISAT marker-assisted
breeding program for the stay-green component of terminal drought tolerance. All
accessions in this cluster are agronomically elite ('u~durra~l-type
breeding lines and
released varieties.
'Twelve and seven genotypes were grouped in 3'* and 4"' clusters, respectively.
These originated from the Durru race and possess moderate to high lebels of
resistance to sorghum shoot fly and spotted stem borer. Actually the two above-
mentioned clusters based on the ABI-generated data sheet (3 and 4) formed a single
cluster in PAGE-generated data sheet. The use of a single representative from ABI-
generated cluster 3 and another from cluster 4 as resistant parents in two new mapping
populations targeting shoot fly andior spotted stern borer resistance would seem to be
a reasonable starting poitit.
Cluster 14 contains six improved genotypes, most of them with sorghum shoot
fly resistance and some of them with sorghum liiidge resistance. Cluster 15, which
could be designated as a cluster of agrononlically superior midge resistant breeding
lines. included five genotypes. Nearly all genotypes in clusters 14 and 15 were
developed at ICRISAT-Patancheru from crosses designed to introgressed insect
resistance into elite Zeru Zeru landrace background materials having a superior
agronomic characteristics and excellent grain quality. Single selected genotypes from
each of these two clusters could be used for the development of two new mapping
populations targeting QTLs for grain yield, grain quality and insect resistance.
Compared to the clustering pattern obtained from the PAGE-generated data set
many more genotypes fom~edsingle-genotype clusters at 50% similarity when the
ABI-generated data set was used. Among these single-genotype clusters, many of
them originated from different countries; i.e.. Suphanburil 1 came from Thailand.
Godamhuman originated from Sudan. and IS 18581 and IS 23637 are Nigerian
breeding lines. By and large most of the clusters that appeared from the PAGE-
generated SSR marker data set were separated further and their positions relative to
other clusters changed moderately in the dendrogram based upon the ABI-generated
SSR marker data set. This is expected as the ABI should give a more accurate picture
than PAGE because of its superior ability to detect the smaller polymorphisms
between the genotypes. For example, except for a very large cluster of related
breeding products that are insect resistant. all clusters detected based on the PAGE-
generated marker data set were separated into distinct sub-groups by the ABI-
generated SSR marker data set. If we look at around the 40% level of similarity, both
the PAGE- and ABI-generated data sets detect 12 clusters, but the positions of the
genotypes within the clusters were slightly lnoditied by the superior sensitivity of
I'CR product separation on the ABl machines. At the Yanie time if we look the
clustering pattern around the 70% level of similarity, both of the systems classified
the accessions into a larger number of clusters which indicates that the 91 genotypes
studied were well diverged in their genetic makeup.
5.1.5 Implication for sorghum breeding
'1-lie i~~forniationprovided by this study about the diversityisimilarity of the
gennplasm from different sources of region should prove extremely useful. The
results obtained can find using in heterosis breeding and in selecting parental lines for
specific breeding goals related to combining insect resistance with high grain yield
and mitigation of drought stress. Dendrogra~nsgenerated from both the ['AGE- and
AB1-derived molecular marker allele data sets provide useful i~lfornlationregarding
the relatedness of materials of similar and distinct geographical origin. and of
genotypes with varying level of agrotlomic eliteness. particularly with sources of
insect resistance that might be exploited in conventional or marker-assisted plant
breeding programs.
1. Development of new mapping population(s) for shoot fly resistance by crossing
296B. BTx623 and/or ICSV 88032 as the susceptible female parent and an improved
shoot fly resistant male parent (i.e. ICSV 705 and/or ICSV 708.
2. For spotted stem borer, three highly susceptible genotypes are available BTx623.
ICSV 745 and ICSV 88032. New mapping population(s) for spotted stem borer
resistance can be developed by crossing one of them as a female parent with
genetically distinct stem borer resistant parents such as IS 2367. IS 4756, PD 15881-3
and IS 18577, using one of the latter group as the male parent.
3. For development of a mapping population for both shoot fly and stem borer
resistance. three highly susceptible genotypes are available (i.e. BTx623, 2968 and
ICSV 88032), and any one of them can used as a female parent in a cross to a shoot
fly and stem borer resistant parent taken from any of the other clusters (i.e. ICSV 700
andlor ICSV 714) and as male parent.
4. For development of new mapping populations for resistance to sorghum midge fly,
susceptible female parent selected from among 296B, BTx623, PB 15881-3 and ICSV
714 can be crossed with a genetically distinct midge resistant male parent available
tiom other clusters such as AF 28, TAM 2566. ICSV 88032 andlor DJ 6514.
5. Almost all genotypes belonging to AH1 clusters 17. 18 and 19 are elite recurrent
parents in marker-assisted backcrossiiig programmes initiated at ICRISAT, in which
genes for the stay-green component of drought tolerance are being introgressed rronl
trait donors 8 3 5 and E 36-1. As genotypes within any one of these three clusters are
similar. it should be cost effective to reduce the number of recurrent parents that are
actually advanced in each of these clusters.
5.2 Phenotyping of RILs derived from cross 2968 x IS 18551 for components of
resistance to sorghum shoot fly
Shoot fly is major insect pest of sorghum. Shoot fly resistance is a complex trait.
involving a number of component traits. each of which are quantitative in nature and
influenced by G X E interaction. Therefore, direct phenotypic selection for this trait
will be difficult. Despite efforts made over the last two decades by utilizing the
existing cultivated sources of resistance. the level of resistance achieved so far in elite
backgrounds is limited. kiarker-assisted selection could increase efficiency of
breeding for such traits. As an initial step in this prograln, genornic regions associated
with resistance and its components are to be detected.
The present study, involving reconlbinant inbred lines (RILs) obtained from
cross 2968 (susceptible) X IS 18551 (resistant), was undertaken to detect and
estimate the effects of quantitative trait loci (QTL) for sorghum defense mechanisms
for shoot fly oviposition and deadheart incidence; and to identify simple sequence
repeat (SSR) markers linked to deadheart incidence QTL(s) for possible introgression
of these QTL(s) from more resistant donor genotypes into susceptible, agronomically
elite breeding lines.
5.2.1 Phenotypic and genotypic variation
Characterization of phenotypic and genotypic variation of complex traits like shoot
fly resistance is a pre-requisite to application of molecular genetic knowledge to
broaden our understanding of their genetic control.. Shoot fly resistance traits are
quantitative in nature. Genetic expectations of means and variances were obtained for
these resistance traits using 259 RlLs evaluated along with their parents. The genetic
variability was assessed under two levels of shoot fly infestation in this set of RlLs
derived from a cross between resistant and susceptible inbred lines. Estimates of
genetic variance components thus obtained have been used to compare tlie
heritabilities of different resistance components.
The pooled analysis of variance for different components o f shoot fly
resistance (Table 4.4) revealed highly signiticant genotypic (G). environmental (E).
and G X E interaction effects. The analysis not only depicts the variability that
existed in the two screening environnients, but also reflects the presence of genetic
variability among the tested genotypes for shoot fly resistance and its component
characters. Highly significant differences detected among the R1l.s and phenotypic
differences recorded bet\veen the parents fbr various resistance traits suggested that
sufficient variability exists in the experimental material for the purposes of this study.
Based on the v a ~ i n grange of phenotypic values for deadhearts (%) I1 in
susceptible parent 296B in tlie two screening environments. these environments were
categorized as having high shoot tly pressure (late khtrrifl and optimum shoot fly
pressure (rubi). The phenotypic values for deadhearts (%) I1 in the susceptible parent
were 96% and 77% in the late khrrrif'and rahi screening environments, respectively.
Borikar et 01. (1982b) observed that selection of shoot fly tolerant genotypes was
effective under optimum shoot fly pressure with 67% to 75% seedling mortality on
susceptible control entries, while Rana et ul. (1975) suggested that selection for shoot
fly resistance was effective under conditions when mortalities ranged from 6.7% to
67%.
According to Rao et al. (1974), the level of seedling mortality in a field crop
due to shoot fly deadhearts is a function of the intensity of insect infestation. plant
growth rate and inherent genotypic differences. It appeared that the extent of
deadhearts observed was primarily related to the level of shoot fly pressure. As the
rate and level of shoot fly population build-up varies with season and location,
sorghum genotypes also exhibit variable degrees of shoot fly damage in different
environments. Hence the screening of experiniental material during the present study
was done in two different screening environments. 'This provides opportunity to study
the genetics of adaptation for shoot fly resistance by genotypes. In previous genetic
study of shoot fly resistance in three different levels of shoot fly infestation (Borikar
and Chopde. 19801, it was indicated that variation within and between genotypic
groups became more apparent under high levels of shoot fly infestation. Considering
the level of shoot fly damage in terms of deadhearts (%) 11, the rubi screening
environment (E2) then should be considered as more favorable for selection of shoot
fly resistant and/or tolerant genotypes.
In general, the two parents (296B and IS 18551) differed significantly for all
observed important shoot fly resistance traits (Table 4.2). Lander and Rotstein (1989)
indicated that the ability to map QTLs underlying a quantitative trait depends on the
magnitude of phenotypic difference existing between parents of the mapping
population. The differences observed between the parents involved in this study
satisfy this requirement.
Parental performance and R l l s mean perforluance for various shoot fly
resistancc colnponents (Table 4.2 and Table 4.3) are discussed in following
paragraphs.
5.2.1.1 Glossiness
Parental performance and RIL niean performance for glossiness was consistent across
both screening environments. indicating consistent and reliable evaluation of this trait.
Jayanthi et crl. (1999) also reported that glossiness expression was stable across
seasons. The wide range of phenotypic values and high genotqpic variance (Table 4.4)
for glossiness indicates that selection for this trait will be effective. Expression of
differences in glossiness between the parents was greater in kharifthan in rcrhi. These
results corroborate those of Agrawal and Abraham (1985) and Jayanthi ct 01. (1999).
5.2.1.2 Seedling vigor
Significant differences in parental means was observed for this character in the rabi
screening environment. Wide variations was observed for this character among the
RIL means in the same season. The mean RIL values for character indicate that most
of RILs had moderate vigour and were less susceptible to shoote fly.
5.2.1.3 Oviposition incidence ( O h )
The phenotypic values of parents for oviposition (%) varied significantly across the
two screening environments indicating environmental intlue~lceon shoot fly egg
laying. Parental mean values and RIL ranges for oviposition (%) 11 indicated greater
shoot fly pressure in the khurfseason screening environment than in the rahi season
screening. Variability observed for oviposition depends on the level of shoot fly
pressure prevailing in screening environments and on the breeding material screened.
'The efficiency of the ovipositional non-preference mechanism of shoot fly resistance
is not stable and it is ineffective at high levcls of shoot fly pressure (Singh and
Jotwani, 1980a; Borikar rr a/.. 1982a: Sharnia e/ ul., 1097). The range of the
phenotypic values for oviposition (%) in tliis RIL population clearly indicate that the
shoot fly population pressure was optimum in rcrhi screening environment, which was
expressed in terms of wide ranges of pltenotypic value for this trait. Borikar et ol.
(1982b) reported higher variability when the material was tested under o p t i ~ ~ i ushoot
m
fly population level.
5.2.1.4 Deadhearts Incidence ('YO)
The most direct measure of shoot tly damage is that recorded in terms of deadhearts
incidence (96). The phenotypic value of parents for deadhcarts incidence (%) varied
significantly across two screening etivironments. Parental mean values and RII, mean
ranges for tliis traits indicated higher shoot tly pressure in the khurif'season screening
environment than in the rtrhi season screen. Variation in phenotypic values of RILs
for deadhearts (%) was lower in khtrrif season than in rtrhi season. Therefore, the
variability observed for this trait in the rahi season screening is likely to be of greater
importance for selecting resistant a~ldiortolerant genotypes. Borikar et al. (1982b)
reported that estimates of variability were higher for seedling mortality when the
material was screened under optimum shoot fly population levels. The results
obtained for deadhearts (%) in this study confir~tlstltat the screenillg of the RIL
population in rahi season was done under optimum shoot fly population.
5.2.1.5 Seedling height (cm)
Significant differences in parental mean and RIL ranges for seedling height I observed
in the rabi screening environnients indicated that the behavior of this character is
under genetic control. Sharma et (11. (1977) pointed out the predominance of fixable
genes in F2 population for seedling height; RILs are regarded as an immortalized F2
with all the genetic variation fixed (Burr er al., 1988).
5.2.1.6 Trichome density
The parental mean values exhibited a consistent trend for tricho~nedensity (both
upper and lower surface of leaf blade) across both screening environments. IIowever,
RIL population ranges indicated that trichome density was greater in rub; season than
in kharif season. This indicates that the character is under genetic control with some
environmental influences in its expression in different seasons. Maiti and Gibson
(1983) also indicated that expression of trichomes was comparatively lower in khrrrif
than in rabi season. In both seasons trichome density on upper leaf blade surface is
greater than lower leaf blade surface. Similar observations were recorded by Gibson
and Maiti (19831, Borikar and Chundurwar (1989). and Sajjanar (2002).
5.2.1.7 Time to 50% flowering, plant height and grain yield
Parental mean values for these characters revealed that 296B was later in tlowering,
shorter in height, and lower yielding than IS 18551 in both screening environments.
Susceptible parental genotype 296B was heavily infested by shoot Ily, producing
large ~ i i ~ m b e of
r s tillers. The time to 50% flowering recorded for 2960 takes into
account time taken for formation of these tillers also. As the tillers are later in
flowering than main culms. floweri~ig296B was observed to be later than it would
have been had the trial been protected fro111damage by shoor fly.
The range of values for KIL means indicated that selection for flowering could
better be done in the kharif'season due to a wider range of flowering time in this
environment; while for plant height wide variation in phenotypic values was observed
in both screening environments. Wide variation in phenotypic values for grain yield
was also observed in both screening environments.
5.2.1.8 Overall recovery score and aphid damage score
Significant differences in parental means were observed for both characters in the
khrrrif'screening environment; however, no significant difference in parental means
for either character were observed in the ruhi screening environment. The value of
N L s means indicated that larger numbers of RlLs having high recovery score and
low aphid damage were observed in the rabi screening environment than in the kharif
season screen. Wide variation in phenotypic value for these characters were observed
in both screening enviorments.
5.2.1.9 Pigmentation (scale), midge incidence score and agronomic score
Significant differences in parental means were observed for these three characters in
the rabi screening environment. Wide variations were observed for these characters
among the RIL means in the same season. The mean RIL values for these characters
indicate that most of RILs had non-tan foliage pigmentation. were susceptible to
midge fly, and had moderate agronomic adaptation.
In general, parental and RIL mean values revealed wide variation in
he no typic values for shoot fly resistance and its component traits in both screening
environments. Resistance in terms of oviposition non-preference is due to the
component traits that prevent egg laying. This is due to a combination of characters
expressing in favorable directions. It was previously reported that lear color. texture,
width (Raina, 1981) and hairiness (Bapat and Mote, 1982b) were important factors in
selection of the oviposition substrate by female shoot flies. The wide variation
observed in this RIL population for the shoot fly resistance component traits like
glossiness: seedling vigor and trichome densit) indicates that these traits can be used
as siniple criteria for selection of rcsistarit getlotypes.
5.3 Analysis of variance
The analysis of variance for different shoot fly resistance comporlents and related
traits observed in this study revealed that genotypic variances were significant for all
the observed traits in both of the individual screening environments as well as in the
across-season analysis. The genotypic variances for shoot fly resistance component
traits like glossiness, oviposition (%) I and 11. and deadliearts (%) I1 was greater than
the corresponding variances for G X E interaction. This indicated that these resistance
component traits are mainly under genetic control, but that there are significant effects
of environment in expression of these traits. Cilossi~iesswas mainly under genetic
co~itrolin agreement with earlier reports (Tarumoto. 1980: Agrawal and House, 1982;
Jayanthi el al., 1999); while for trichorne densities (both upper and lower leaf
surfaces) the genotype variance was greater than (i X E variance, indicating these
traits are mainly under genetic control but that there are significant effects of
environment in their expression. Jayanthi et crl. (1999) too observed the season effect
on expression of trichome density.
5.4 Inheritance of components of shoot fly resistance
The continuous distribution of RILs for various shoot fly resistance co~nponenttraits
observed in this study revealed that most of the traits studied were polygenic.
According to Menendez and Hall (1995). the absence of discrete segregating classes
for a trait suggests that its inheritance should be determined either by a large number
of genes with small effects or a few major genes with substantial environmental
effects. The observations made in the present study are supported by previous workers
findings that resistance to Atherigonu .Toccata is quantitatively inherited (Agrawal and
Abraham. 1985) and polygenically controlled (Goud e t a [ . , 1983; Halalli eta!.. 1983).
Sharma er al. (1977) and Borikar and Chopde ( 1 980) observed continuous variation in
different generations and indicated that shoot fly resistance is due to gradual
accumulation of resistance alleles at many genes.
The genetic analysis of colnponents of shoot fly resistance is discussed below.
5.4.1 Broad-sense heritability
Heritability is a useful quantitative statement of the relative iniportance of heredit).
and environment in determining the expression of the character (Allard. 1960). The
estimates of heritability help the plant breeder in selection of genotypes from diverse
genetic populations. Effective selection can be achieved when additive effects are
substantial and environmental effects are small. so that heritability estiniatcs are Iligli.
5.4.1.1 Glossiness
Consistently high heritability estimates observed for glossi~iessin two individual
screening environments and moderate estimates across these same two test
c~ivironlnentsindicate that contrihutio~isto phenotypic variance due to environmental
kctors and G X E interaction are less than genotypic factors as evidcnt from the
values of genotypic and G X E variances (Table 4.4). The G X E variance component
is. however, significant indicating the complex nature of glossiness. The Q'TL
analyses results revealed a significant epistatic interaction (additive x additive) for
this trait with presence of niultiple loci controlling the rxpressio~i of this trait.
Glossiness has previously been reported to he controlled by a single recessive gene
(Tarumoto. 1980: Agrawal and House. 1982: Jayanthi el cil.. 1999). while Agrawal
and Abraham ( I 985) indicated that the seedling glossiness intensity is quantitative in
nature and controlled by both additive and non-additive genes. The current study also
revealed the complex inheritance of seedling glossiness score.
5.4.1.2 Oviposition incidence ( O h )
Oviposition incidence recorded low to moderate but co~isistentoperational heritability
estimates in the two individual screening enviornments and also at different stages of
observation .However operational heritability estimates for this trait across seasons
for both observation stages were low indicating a pro~ninent role of screening
environment andlor genotype x environment interaction in expression of the trait.
This observation corroborates those by Halalli er al. (1983). Borikar and Chopde
(1982), and Borikar et a[. (1982b), who indicated that estimates of heritability for
oviposition were high when the material was tested under optimum shoot fly
population levels. This also confirms the utility of screening test material under
optimum insect population levels so that selection for ovipositional non-preference
will be effective.
5.4.1.3 Deadhearts incidence (%)
Operational heritability estimates observed for this trait varied from low to moderate
and were consistent at both observation stages. However the heritability estiniatcs
obtained from the across season analysis were low. These results are in agreement
with the results obtained by EIalalli e! 01. (1982). According to Bluni (1969a), seedling
mortality is dependent on the intensity of insect infestation and hence ally data on
sorghum reaction to shoot fly must be interpreted with reference to the shoot tly
population level. Borikar and Chopde (1982) also observed that genetics of deadhearts
(%) is most influenced by shoot fly population level. The consistent estimates for
heritability observed in individual screening environments in this study. suggests high
to optimum insect population levels. which is also revealed by lower heritability
values in the across-season analysis. .l'his observation corroborates that by Borikar el

ul. (1982b), who reported that estimates of heritability were moderate to high for
seedling mortality when the material was tested undcr optimunl shoot fly population
levels
S.4.1.4 Trichome density
Consistently high heritability estimates (h' > 0.95) were recorded in both screening
environments for trichome density on both upper and lower leaf surfaces, while
across-season analyses revealed lower heritabilit) estimates for trichome densities of
both leaf surfaces (h' = 0.51 for upper and h" = 0.49 for lower leaf surfaces). This
indicates the role of environtnental factors and G X E factors in expression of this
trait. Shanna er al. (1977), Gibson and Maiti (1983). and Tarumoto (1980) studied the
nature of gene action for non-preference and each found that the presence of
trichomes was governed by a single recessive gene: however, the inheritance of
trichome density appeared to be more complex in the current study. The results
obtained in this study reveals that there is a strong seasonal effect on the expression of
trichome density. These results are in agreement with those of Jayanthi e! al. (1999)
and Sajjanar (2002).
5.4.1.5 Time to 50% flowering and plant height
Operational heritability estimates were high and consistent for time to 50% flowering
(days) and plant height (cm) in both screening environments. However, in the across-
season analyses a moderately lower estimate of heritability was recorded for time to
50% flowering and a high estimate of heritability was recorded for plant height. This
indicates these characters are under genetic control and that there is a moderate level
of interaction of genotype with the environment in expression of these traits. This
observation agreed with the observation of higher genotypic variances than G x E
interaction variances for these two agronomic traits.
5.4.1.6 Overall recovery score and aphid incidence score
Moderate operational heritabilities were recorded for these traits in individual
screening environments and low to moderate operational heritability estimates were
obtained in the across-season analysis for these traits. This suggests that shoot fly
population pressure, aphid density andlor other seasonal effects contribute to
expression of these traits.
5.4.1.7 Grain yield (g\ plot)
Moderate to low operational heritability was recorded for grain yield under conditions
of moderate to severe shoot fly pressure in individual screening environments. and a
low operational heritability estimate was obtained in the across-environment analysis.
IIowever, the genotypic variancs was higher than G x E variances, indicating grain
yield is indeed under genetic control although there is substantial interaction with the
environment in expression of this trait.
5.5 Transgressive segregation
In the absence of epistasis and in the presence of linkage equilibrium, the tnean of
RILs will be the midparental value (average of two parents) (Jinks and Pooni, 1981).
Epistasis leads to asymmetry in the distribution of derived inbreds relative to the
initial inbred parental means. In other words deviation of the mean of the population-
derived inbreds from the midparental value (either positive or negative) indicates the
presence of epistasis. In the present study, an attempt has been made to elucidate the
genetic constitution of parental inbreds 2968 and IS 18551, and the nature of gene
action involved in controlling shoot tly resistance components in the RlLs based on
means (of the parents and their derived RILs) and the appearance of transgressive
segregants.
 In general, for traits with RIL means less than midparental values. the
~roportionof RlLs outside the low scoring parent was greater than that outside the
higher scoring parent, and vice versa. The expectation of equal frequencies of inbreds
lying outside the parental limits of P1 and P2 was not observed in any case. This
shows that for each trait observed there were epistatic interactions influencing trait
expression. Further the occurrence of transgressive segregants indicates that the two
~arentallines of the RIL population both carried desirable and undesirable alleles at
various proportions of loci governing the various traits observed.
5.5.1 Glossiness intensity
Continuous distribution, from high intensity of glossiness (i.e. score I ) to non-
glossiness (i.e. score 5). with an apparent valley in the frequency distribution graph
between scores 3.5 to 4.0 (Fig 4.4.4) indicated the involvenient of major genes
controlling the glossiness trait. A previous study reported that the glossiness character
is controlled by single recessive gene (Tarumoto, 1980). A major gene will have
ma,jor effects that will be larger than those arising from non-heritable agencies; its
effects will be precipitated in phenotypic expression of a trait. The presence of such
major genes is also supported by the consistency of high heritability estimates
observed for glossiness in the individual screening environments (Table 4.6).
The observation that no transgressive segregant RlLs were recorded with
plienotypic scores falling outside the high intensity (low score) of glossiness of the
resistant parent indicates that the alleles for this trait are predominantly in coupling
phase. The positive deviation of the RIL population mean for glossiness from the
midparental value indicates the presence of epistasis for this trait. Agrawal and
Abraham (1985) also indicated that the seedling glossiness intensity is quantitative in
nature, controlled by both additive and non-additive genes. The high mean value of
the RILs, approaching that of high scoring parent (2960) indicates that the frequency
of RILs with high scores (indicative of low intensity of glossiness) was greater than
that of highly glossy (Fig 4.4A) individuals (indicated by low glossy scores). his
could be explained by the fact that as the number of e~istaticallyinteracting gelles
controlling the trait increases the probability of obtaining individuals homozygous for
favorable alleles for high glossiness intensity (low score) at all the concerned loci will
be reduced.
5.5.2 Oviposition incidence (9'0)
It was observed in the current study that, for oviposition 11. the RIL mean did not
differ significantly from that of resistant parent IS 18551. The RIL population mean
deviated from midparental values towards that of resistant parent for this trait.
Favourable transgressive segregants were observed for this trait at both observation
stages. However, the proportion of RILs out-side the parental mean of the low scoring
parent IS 18551 was comparatively higher for oviposition I1 than oviposition 1. This
suggested that the trait is mainly under the control oradditive as well as
additive X additive genetic effects. Borikar and Chopde (1981b) also indicated the
predominance of additive gene action for oviposition incidence.
5.5.3 Deadhearts incidence (%)
The continuous distribution of the RIL population for deadhearts incidence indicated
tlie qualitative nature of this trait. The transgressive segregants were observed for this
trait at both observation stages. The proportion of IllLs lying outside the low scoring
resistant parent IS 18551 was comparatively higher for deadhearts I1 than for
deadhearts I. The KIL population mean for deadhearts incidence were observed near
to be close to the midparental value indicating predomi~ialiceof additive as well as
additive X additive gene action for this trait. Borikar and Chopde (1981b) also
indicated the predominance of additive gene action for deadhearts incidence.
5.5.4 Trichome density (upper leaf blade surface)
The continuous distribution of tlie RIL population for trichome density on the upper
surface of seedling leaf blades indicated the quantitative nature of this trait. 'The
positive deviation of the RIL population mean from the midparental value indicates
the presence of epistasis as Jinks and Pooni (1976) have reported that any deviation of
the population mean from the midparental value reveals presence of epistasis. Sharma
et al. (1977) and Tarumoto (1980) reported that presence of trichomes was recessive
in nature. but the inheritance of trichome density was complex. The appearance of
transgressive segregation in this RIL population might due to complementation of
favorable and unfavorable alleles received from both parents. Due to
complementation of positive and negative alleles in the FI and subsequent
recombination events, individual RILs with higher than the ~arentalproportion of
favorable alleles have been observed as transgressive segregants having tri~home
density higher than that of the higher scoring parent. The presence of additive gene
action supplemented by additive X additive gene interaction in these RlLs might be
main reason for occurrence for these transgressive segregants with trichome density
higher than the high scoring parents.
5.5.5. Trichome density (lower leaf blade surface)
It was observed that the mean value of RIL population was on par with the
midparental value for trichome density on the lower surface of seedling leaf blades.
The distribution was observed to be skewed towards trichomelessness. This deviation
of the RIL population mean fro111 the midparental value points out the presence of
epistasis for this trait. The occurrence of transgressive segregarlt RlLs rvilh value
higher than the higher scoring parent indicates that favorable alleles were contributed
by both parents. for the presence of RlLs with larger number of favorable alleles than
high scoring parent IS 18551. This in turn could be attributed to additive and
additive X additive gene action.
5.5.6 Plant height (em)
The high mean plant height of RlLs ludy be due to transgressive segregants with
values lying outside the taller parent IS 18551. The low freque~icy of such
transgressive segregants indicates that the most of the alleles for greater plant height
might be associated in coupling phase. However. some favorable parental alleles
might be in repulsion phase resulting in limited opportunity for occurrence of
transgressive segregants.
5.5.7 Overall recovery score
I t was observed that the mean value of the RIL population was at par with the
rnidparental value for this trait. A 10% proportion of transgressive segregants was
observed with a few RILs having values lying outside the high recovery score parent
IS 18551. The proportion of transgressive RlLs was greater for individuals with
values lying outside the low recovery score parent 296B. The appearance of high
numbers of transgressive segregatit RlLs for low overall recovery score might be due
to the complementation of more unfavorable alleles received from both parents. It
indicates that presence of addithe gene action supplemented by additive x additive
gene interaction.
5.5.8 Aphid damage score
The positive deviation of the RIL population niean from the midparental value
indicates the presence of epistasis. Transgressive segregants were observed outside
the limits of both the high and low scoring parents. However the proportion of RILs
lying outside the low scoring (more resistant) parent IS 18551was comparatively
higher. The high frequency of such transgressive segregants indicated that a low
number of parental alleles might be in coupling phase. This in turn indicated the
possible role of additive additive gene interaction in control of this trait.
5.5.9 Grain yield
The two parents of this RIL population differ significantly for grain yield under
conditions of moderate to severe shoot fly pressure. The presence of epistasic gene
action for grain yield was revealed by negative deviation of the RIL population from
the midparental value. The appearance of transgressive segregalits might be due to the
accumulatioli of unfavorable alleles for lower grain yield in the RlLs during the
process of inbreeding and also some amount ofepistasic gene action
5.6 QTL mapping
5.6.1 Glossiness
For glossiness, three QTLs were detcctcd in the khor~f'screetiingenvirontnent (El).
whereas two QTLs were detected in the rtrhi screening environments (E?) and two
QTLs were detected in the across-environment analyqis (Table 4.7 and 4.8). One QTL
on LG J explaining 37% of observed phenotypic variance in late khurif. 21% in rabi,
and 34% across these two screening environments, and would be considered a major
QTL for this trait. The favorable allele for this QTL originates from the IS 18551
parent. The identification of a QTL explaining a high proportion o f the phenotypic
variance indicates a strong association between genotype and phenotype. According
to Terwilliger (2001). if the test locus getlotype-phenotype relationsliip is strong, the
power of QTL identification is solely a function of the strength of linkage
relationships.
The identification of a major QTL for glossiness explaining a high proportion
of observed phenotypic variance in different screening environments confirms the
high heritability (Table 4.6) of this trait and lou influence on it by environment and
G X E interaction (Table 4.4). This trait has been reported to be controlled by a
single recessive gene (Tammoto. 1980), but the current results indicate its inheritance
is more complex in the (296B X IS18551)-derived RIL population. The low
frequency of highly glossy RlLs does not indicate that the trait is controlled by a
single recessive gene (in which case a I:I segregation for homozygous glossy and
non-glossy RILs would have been expected), but instead that it is controlled by
epistatic interactions involving several loci. Aganval and Abraham (1985) indicated
that the intensity of seedling glossiness is quantitative in nature, controlled by both
additive and non-additive genes. This observation can be explained by the presence of
multiple QTLs controlling this trait and a significant additive X additive interactiotl
component in their final simultaneous effect. This study indicates that glossiness is
controlled a major QTL on LG J. accounting for 34% of the phenotypic variation. and
one minor QTL on LG G accounting for 8% of the phenotypic variation in the across-
seasons analysis. Over all, the two QTLs mapped in across-season analysis of these
two screening environnients explained 31% of genetic variation after correcting for
non-significant QTL x envirolirnelit interactions. This suggests the presence of
unmapped genetic effects in areas on the genetic map that are presently poorly
covered by SSR markers.
Two environment-specific glossiness QTLs were detected in the khurif'
screening environment, and mapped on L O E and I.(; H, respectively. One minor
QTL for glossiness detected on LG G was co-localired with important oviposition,
deadhearts, trichome density (upper and lower surfaces of seedling leaf blades) and
grain yield QTLs, indicating that glossiness is indeed an important component in the
interaction between shoot fly and sorghum and should be targeted for marker-assisted
selection for shoot fly resistance. When compared with the previous study to map
putative QTLs for shoot fly resistance by Folkertsma el NI. (2005. unpublished) in a
(BTx623 X IS 18551)-derived RIL population, the present study also located the
same major QTL originating from IS 1855 1 for glossiness on LG J. This confirms that
this could be a region with a candidate gene for shoot fly resistance. In both these
studies, resistant parent IS 1855 1 contributed the additive genetic effects for increased
glossiness. Glossiness has been described as one of the major factors determining
sorghum resistance against shoot fly (Omori el ul., 1983). The positive correlation
between glossiness score and deadhearts incidence; and between glossiness score and
oviposition signifies the importance of low glossiness score ( i . e . , high intensity of
glossiness) in imparting resistance of sorghum to shoot tly and should be targeted for
marker-assisted selection of shoot fly resistance (Folkertsma e t 01.. 2005;
unpublished).
5.6.2 Seedling vigor I and seedling height I
For seedling vigor I and seedling height 1, he no typic observations were not recorded
in the late kharif screening environment (El). Three seedling vigor I and two seedling
height 1 QTLs, respectively, were detected in the rubi screening environment. The
three seedling vigor I QTLs together explained 24% of the observed d he no typic
variance and the QTL mapped on LG B appears to be the most important QTL as it
alone explained 11% of the observed phenotypic variance in seedling vigor I. The two
QTLs for seedling height I explained together 14% of observed phenotypic variance,
For these two characters one common QTL has been mapped on LG B in marker
interval Xtxp01-Xtxp348. Favorable additive genetic effects at this QTL are
contributed by resistant parent IS 18551.The relatively low correlation between vigor
and shoot height on one hand. and deadhearts incidence I and I1 on the other hand,
and the absence of the seedling vigor and shoot height Q'H. co-localizing with any
deadhearts QTL indicates that seedling vigor and seedling height (at least as assessed
in this study) are of limited relevance to improvement of shoot fly resistance in
sorghum, which is in agreement with tlie findings of Folkerstma e1 al. (2005
unpublished).
5.6.3 Oviposition incidence (%)
Oviposition (%) expressed as the percentage plants of a genotype with shoot fly eggs,
is highly correlated with deadhearts, as expected. This indicates that there is a direct
relationship between the percentage of the plants with eggs and tlie percentage of the
plants showing deadhearts. Two and four QTLs for oviposition I and 11, respectively,
were detected across environments. with significant QTL x environnient interaction
observed for the ovipositio~i1 VfI, detected on 1.G I: and the oviposition 11 QTLs
detected of LG F and LG G. The two Q'rLs for oviposition I together explained 6% of
genetic variation for this trait, with the two-QTL liiodel having a peak LOD value of
3.9. The four oviposition I1 QT1.s together explained 15% of the genetic variation for
this trait, with a major QTL on LG G. In the rabi screening envirotl~uenta common
QTL for oviposition I and I1 was mapped on LG F. The ovipositioll I QTLs mapped
on LG C and LG F are co-localized with oviposition I1 QTLs in the across-
environment analysis. In addition, two more oviposition I1 QTLs were detected in the
across-environment analysis, which ~iiappedon LG G and LG J. The QTL for
oviposition detected on LG G in the across-environment analysis co-localized with a
deadhearts QTL, a major QTL for trichome density (upper and lower surfaces of
seedling leaf blades) and a minor QTL for grain yield under conditions of optimum to
high shoot fly pressure. In addition, the QTL for oviposition detected on 1.G J in the
across-environment analysis co-localized with the major QTL for glossiness intensity.
The presence of an environment-sensitive QTL for oviposition I1 was detected on LG
E (kharifscreening environment El). Variation in phenotypic values for oviposition 1
and 11 across the two screening environments might account for the significant QTL
X environment interaction observed for this pair of traits. ~ ~ parelltal
t h lines
contributed favorable alleles for oviposition. According to Ramie e,
(1998) and
Agrama et a/. (2002,) correlated traits often have QTLs mapping to the
clvomosomal locations. Two different types of correlation between the traits have
been observed in the current study. Correlation between different evaluation times for
a specific trait and correlation between different traits for a specific evaluation timet
or across different evaluation times. I-Iigh correlation between different evaluation
times for the same trait indicates that the expression of the trait at different evaluation
times is under control of similar genes and therefore Q'TLs detected for different
evaluation times of this should be co-localized.
5.6.4 Deadhearts incidence (%)
No QTL was detected for deadhearts 1; however, two QTLs, one each on LG F and
LG G. were mapped for dcadhearts I1 across the two screening environments.
Signiticant QTL x environmental interactions were observed for both of the detected
QTLs for deadhearts 11. The two QTLs together explained 11% of the genetic
variation for this trait with a possible major QTL on LG F itself explaining 12% of the
observed phenotypic variation. This QTL on LG F was detected in tlie across-
environment analysis as well as in both of the individual screening environments.
This QTL on I,G F for deadhearts 11 co-localized with an oviposition QTL and a QTL
for trichome density of the lower surface of the leaf blade in the across-environment
analyses. Another important QTL for deadhearts TI mapped on LG ti in the across-
environment analysis and in the rahi screening environment, and this co-localized
with a major QTL for trichome density (upper and lower leaf blade surf:dccs) and a
minor QTL for overall recovery score and grain yield. For oviposition and deadhearts
incidence several QTLs were mapped but only a few Q'I'1,s were co-localized in the
two individual screening environments and few mapped for both traits in the across-
season analysis. More over, no QTLs was found to be a major contributor to
phenotypic variation for either of these traits. When compared with the previous
studies where QTLs for oviposition and deadhearts incidence were co-localized on
LG J in the (BTx623 X IS 18551)-derived RIL population (Folkerstma er al.. 2005
unpublished), it can be clearly observed that the genomic regions harboring these
QTLs were not mapped with SSR markers in the RIL population used in the current
study so that it has not yet been possible to detect these QTLs in both of the IS 18551-
derived RIL populations.
5.6.5 Trichome density
One QTL on LG G accounting for 29% (in El). 15% (in E2) and 30% (across
environments) of observed phenotypic variation of trichome density on the upper
surface of the leaf blade was detected (Tables 4.7 and 4.8). Moreover. this QTL on
LG G for upper leaf trichonie density co-localized with a niajor QTL for trichome
density on the lower leaf surface explaining nearly 27% of the phenotypic variation in
the latter trait across the two screening environnients, pointing to similarities in
genetic control of tricliomes densities on either side of the sorghum seedling leaf
blade. Observations from this study indicated that the QTL on LG G is involved in the
control of trichome density on both sides of the seedling leaf blade. The importance of
tricliomes in connectio~i with shoot fly resistance has beell reported by several
research groups (Blum. 1968; Maiti and Hidinger. 1979; Maiti er al., 1980; Taneja
and Lcuschner 1985). Gibson and Maiti (1983) atid Maiti and Gibson (1983)
indicated that presence of abaxial trichome control by a single recessive gene.
However, IIalalli er a/. ( 1 982) reported that trichome density is under the control of
both additive and non-additive genetic effects. The QI'I. mapped o n LG G for both
upper and lower trichome density in this study corresponds to the QTL mapped in the
BTx623 X IS 18551 population (Folkerstama et al., 2005 unpublished). This niajor
QTL mapped on LG G for both upper and louer trichome densities is co-localized
with a minor QTL for glossiness, as well as Q'T1.s for oviposition, deadhearts. and
grain yield across the two testing environments. The one minor QTL for trichome
density detected on LG F co-localized u,ith a minor QTL for deadhearts and
oviposition across the two testing environments. A QTL was detected across
environments on LG F for tricliome density of the lower surface of the seedling leaf
blade; however, the same QTI. was not detected in either individual screelling
environment. Similarly. one minor QTL for trichome density of the lower leaf surface
was detected on LG C in the rabi screening environment, but the same QTL was not
detected in the across-environment analysis. The high frequency of transgressive
segregant RIL individuals with high trichome densities clearly indicated that this
character is not under control of a single recessive gene. Also, the portion of
phenotypic variation explained by the putative QTLs detected is low and there is a
strong possibility that other QTLs are influencing the trait and can be detected if
genomic region not Yet covered by SSR markers can be added to the linkage map for
this RIL population. Even though the role of higher trichome density in reducing
shoot fly oviposition and deadhearts incidence had been previously documented by
several researchers (Blum, 1968; Maiti and Bidinger. 1979; Maiti rt ai. 1980; Taneja
and Leuschner. 1985; Karanjkar et a[.. 1992). the observations of the current study
strengthen the case for using this trait as a selection criterion in sorghum shoot fly
resistance breeding.
5.6.6 Plant height (cm)
One QTL for plant height was detected in each of the kharf and rnbi screening
environments, and in the across-environment analysis. Tlie detected Q'TL was mapped
to a common position on L.G I, and accounted for 17.% (in khuril).lS% (in ruhi) and
16% (across these two environments) of the observed phenotypic variation for plant
height (Tables 4.7 and 4.8). This QTL. for plant height that niapped on LG I co-
localized with a QTI. for midge resistance in the rnhi screening environment. with the
allele for greater height associated with higher midge damage score. 'l'he
chromosonlal region to which this QTL ]napped could he near to the Dw2 dwarfing
gene locus. The additive genetic effects for increased plant height were co~itributedby
alleles from taller parent IS 18551 for both screening environments.
5.6.7 Time to 5UU/o flowering
Two QTL each for the khurif ( E l ) screening environment and across-environme~It
analyses, and one QTL in the r ~ t h (E2)
i screening environnlent were detected for time
to 50% flowering. One common QTL mapped on LG A were detected in El. E2 and
the across-environment analyses. In addition, one QTI. mapped on LG E was detected
in the across-environment analysis, but the same QTI. was not detected in either of
the individual screening environments. One QTL for this trait mapped on 1,G I in the
khc~rif screening environment, but the same QTL was not detected in the across-
environments analysis. The two QTLs detected in the across-environment analysis
together explained 16% of the genetic variation. Tlie major QTL for this trait niapped
on LG A, which alone explained 18% of the observed phenotypic variation, had a
peak LOD value of 9.15. Both QTLs exhibited non-significant QTL X environmental
interaction. Favorable additive genetic effects for this trait were contributed by IS
18551 alleles for the QTL on LG A, and by those from 296B for the QTLs on LG E
and LG I. The QTL mapped on LG A for time to 50% flowering might be mapped
near the ma3 locus, which contributes to early flowering. The QTL mapped on LG A
for this trait co-localized with QTLs of seedling vigor I, pigmentation and agronomic
score in the cooler rahi screening environment and with a QTL for grain yield that
was detected in the across- enviroment analysis.
5.6.8 Overall recovery score
Two QTLs were detected for this trait in across environment analysis and four Q'SLs
were detected in the late kharifscreening etivironment. No QTL was detected in the
rahi environment. One common QTL. which mapped on LG E was detected in the
klzr~rifandacross-environment analyses. One additional QTL mapped on LG J was
detected only in the across-environnient analysis. Three other Q'rl,s were detected in
the khurif screening environment, one each mapped on LG B. LG C and LG G. These
three Q T I S were not detected in the across-enviro~inie~itanalysis. 'The two QTLs
from the across-environment analysis together explained 9% of genetic variation, with
the Q'TL mapped on LG E explaining 6% of observed phenotypic variation. 'This latter
QTL co-localized with QTLs for glossiness intensity, oviposition 11. and grain yield in
the kharif screening environment, and with a QT1, for aphid damage score in across-
season analysis. In addition. the QTL mapped on LG J for overall recoverj score also
co-localized with a QTL for aphid damage score in across-environment analysis.
In botli cases the alleles associated with lower overall recovery score were also
associated with lower aphid damage score. One QTL mapped on LG G detectcd for
this trait in the kharif screening environmenl co-localized with QTLs for both
oviposition 11 and deadhearts 11, and the major QTL of trichome density (upper and
lower leaf surface). Favorable additive genetic effects Tor better overall recovery were
contributed by alleles from shoot fly resistant parent IS 1855 1.
5.6.9 Aphid damage score
Two QTLs were detected for this trait in the across-season analysis. These Iwo QI'Ls
were mapped on LG E and LG J. One common QTL mapped on LG J was detected in
botli individual screening environments and across these environments. llowever, the
QTL mapped on LG E in across-season analysis, could not be detected in the rahi
screening environment. The two detected QTLs across seasons together explained
19% of genetic variation for this trait while the major QTL mapped on LG J for
explained 20% of observed phenotypic variation and had a LOD peak value of 9.3.
The QTL mapped on LG E exhibited significant QTL x environment interaction. The
favourable additive genetic effects for both of these QTLs were contributed by IS
18551 alleles. The QTL mapped on LG E for aphid damage score co-localized with
glossiness intensity, oviposition 11, overall recovery score and grain yield QTLs in the
kharifscreening environment.
5.6.10 Grain yield
Tliree QTLS were detected for this trait under conditions of moderate to severe shoot
fly pressure in the across-season analysis. These niapped each on LG A. LG G and
LG I, and together explained about 12% of genetic variation for this trait in the RIL
population. All of these three QTLs exhibited non-significant Q x E interaction. The
favorable additive genetic effects were all contributed by alleles from resistant parent
IS 18551. The QTLs mapped on LG G and LG 1 in the across-environment analysis
were also detected in the kharif'screening environment. However. tlie grain yield QTL
mapped on LG A in the across-environnient analysis was not detected in either
individual screening environment. In addition, a QTL for grain yield mapped on LG E
in tlie kharif screening environment was not detected in the rahi and across-
environment analyses. Similarly, a QTL mapped on LC3 C in the rrrhi screening
environment was not detected in either the kharifor across-environment analyses. The
QTL mapped on LG A for grain yield in the across-environment co-localized with
time to 50% flowering, with the allele for early flowering associated with higher grain
yield. The QTL mapped on LG G for grain yield co-localized with QTLs for
glossiness intensity, oviposition II, deadhearts 11 and a major QTL for trichonie
density on upper and lower surfaces of seedling leaf blades. The QTL mapped on LG
I for grain yield co-localized with a seedling height QTL detected in the rubi
screening environment. 'The QTL mapped on LG E for grain yield in the late-kharif
environment co-localized with QTLs for glossiness intensity and oviposition I1 in the
late kharif season and QTLs for overall recovery score and aphid damage score in the
across-seasons analysis. The QTL mapped on LG C for grain yield in the rabi
screening environment co-localized with a QfL. for midge damage score in that same
rub; screening environment.
5.6.11 Pigmentation score
Pigmentation score was not recorded in the 2002 late-khary screening environment.
Two QTLs were detected for this trait in the rubi screening environment, one each
mapped on LG A and LG I and explaining 7% and 12% observed phenotypic
variance, respectively. The additive genetic effects for darker foliage color were
contributed by alleles from IS 18551, as expected given that 296B has lighter tan
foliage color. The QTL mapped on LG A for this trait co-localized with QTLs for
seedling vigor I, time to 50% flowering and agronomic score, in the vicinity of the Y
locus for yellow seed color. The QTL detected on LG I co-localized with a QTL for
time to 50% flowering in the longer-daylength khcrr~fscreening environment, which
maps in the vicinity of the Rsl locus governing seedling color.
5.6.12 Midge damage score
Two QTLs were detected in the rnbi screening environment. One each mapped on LG
C and LG I, explaining 6% and 16% of observed phenotypic variance. respectively.
The favourable additive genetic effects for low midge score were contributed by
2968. The QTL mapped on LG C for midge damage score co-localized with a grain
yield QTL in this environment, while the Q ' T L mapped on LG I for this trait co-
localized with a QTL for plant height that was detected in both the khtrrij and rahi
screening enviro~inientsas well as in the across-cnvironment analysis.
5.6.13 Agronomic score
One QTL was detected in rohi screening environment explaining 6% of the obscrved
phenotypic variance in agrononiic score. The additive effects for desirable agro~lo~nic
score were contributed by the 296H allele. This QTL for agronomic score, which
mapped on LG A, co-localized with QTLs for seedling vigor, time to 50% flowering.
and pigmentation score in this rrrhi screening environment.
5.7 Correlation between phenotypic traits
Correlated traits often have QTLs niapping to the same chromosomal locations
(Veldboom et ol., 1994; Xiao et al., 1996). According to IIcrnamalini c/ trl. (2000) co-
segregation may be due to tight linkage, pleiotropy or a causal relationship between
the traits. Two different types of correlations between traits have been observed in the
current study: correlations between different evaluation times for a specific trait and
correlations between different traits for a specific evaluation time or across different
times. High correlations between different evaluation times for the saliie trait indicate
that expression of the trait at different evaluation times of observation is under control
of similar genes and therefore QTLs detected for different evaluation times of the
same trait should co-localize (Folkertsma rr trl., 2005; unpublished). Correlation
between upper and lower leaf surface trichome densities was high in the current study.
The high correlation (0.82) between these two traits suggests similarity in genes
involved in the expression of the density of trichomes on both sides of seedling leaf
blades. This suggestion is further supported by a common major QTL on LG G. CO-
localization of QTLs of different traits will have implications for marker-assisted
breeding approaches to transfer traits from a donor genotype to a more elite recipient
genotypes. ~ o b u s phenotyping,
t genotyping and QTL mapping will help to identify
~ o s s i b l eunfavorable pleiotropic effects before embarking on a marker-assisted
backcross project.
5.8 RILs with potentially useful trait combinations
The across-season means of the RILs obtained from phenotypic screening for shoot
fly resistance component traits (Table 5.1) revealed that there exist some Rl1.s that are
either statistically superior to or on par with the resistant parent IS I8551 for low
deadhearts incidence. These RlLs were scored for plant pigmentation and for oberall
agronomic adaptation. The visual scores for plant piglnentation and overall agronomic
adaptation revealed that only a few individual RlLs comparable or superior to IS
18551 for deadhearts incidence coupled this with better agronomic adaptation.
Among them non-tan RILs 97 and 174 exhibited better adaptation for rahi season
while for kltorif adaptation tan RlLs 222 and 223 were found better. All RlLs with
better performance for deadliearts incidence than IS 18551 were scored (based on
their marker genotypes) for the presence or the absence of the putative QTLs (Table
5.1) mapped for important shoot fly resistance component traits. 'These QTLs were
said to be present (score 1) when the flanking markers to the targeted Q'TLs were
homozygous for the donor parent allele. When one or more flanking markers were
heterozygous for the IS 18551 allele, otherwise the targeted QTL was said to be
absent (score 2). Non-amplified and off-type alleles were scored 3. All RlLs with
better performance for deadheart incidence than the resistant parent were found to
harbor one or more important putative Q'I'Ls (plate5.1) for shoot fly resistant
component traits. RILs 130, , 208 and, 222 each hod all of the QTLs mapped for the
most direct measure of shoot fly resistance, i.e., the shoot fly deadhearts incidence
trait, which was reflected in tenns of their exhibition of deadhearts incidence at par to
lower than IS 18551 across the two screening environments. RILs 130 and 179 were
non-tan, rcrhi-adapted lines intermediate to their parents for plant height. These two
RlLs can be used for the robi season AiB-line improvement program, Intermediate
tan-foliaged RlLs 208, 222 and 223 have khauif adaptation and harbor most of the
putative QTLs imparting resistance to shoot fly. These lines can be used for
development of AIB-line pairs for shoot fly resistance and yield and also for varietal
development programs targeting the kkurifseason.(plate).
 The glossy trait. which is characteristic of most of the kharif and rabi season
landrace sorghum grown in India (Blum. 1972; Rao er a[., 1978). is reported to be
associated with shoot fly resistance (Bapat et al., 1975; Taneja and Leuschner, 1985;
Omeri el al., 1983). Interestingly, six RILs (47, 51, 82, 97, 130. 174) combining
agronomically desirable plant type with dark pigmented (non-tan) foliage, and
intermediate rabi adaptation were found to harboring (plate5.1) the major QTL for
high glossiness intensity. These RILs may be used as donors for transferring the
glossiness trait to rabi-adapted breeding lines. Similarly, intermediate tan foliaged,
agronomically desirable and khtrrif+tdaptated RlLs 46, 208 and 222 can be used as
donors for transfer the glossiness trait to kharif'breeding lines. As glossiness intensity
was observed to have a high broad-sense heritability, it can easily be manipulated
through conventional breeding practices and fixed in breeding lines.
5.9 Implication for breeding approaches
Following iniplications for breeding approaches can be drawn from the present study.
1. Shoot fly resistant component traits like glossiness, seedling vigor, trichonie
density, oviposition % and deadliearts % were observed to have moderate to
high heritability. These traits can be used as reliable parameters for large-scale
screening of germplasni and breeding material aimed at improving shoot fly
resistance.
2. Expression of glossiness is little affected by season and it is observed to be a
reliable coniponent parameter of shoot fly resistance. Glossiness can be used
as a selection criterion for shoot fly resistance screening and fixed in breeding
material f o l l o ~ i n gconventional breeding procedures. RILs 47. 51, 82, 97.
130, and 174 with intermediate height and pigmented foliage, and KILs 46.
208, and 222 with non-pigmented (tan) foliage, intermediate height, all harbor
the major glossiness QTL and can be used as donors for transferring high
glossiness intensity in to more agronomically elite tan and non-tan plant
genotypes.
3. Transfer of shoot fly resistance from donor IS 18551 probably requires more
cycles of selections. In the present study shoot fly resistant RlLs that are
agronomically more desirable than original donor parent IS 18551 were
identified with the help of molecular analysis. It is suggested to now utilize
these RILs as donor parents in crosses with the most agronomically elite
Screening of plants with several different pathogens and their pathotypes, or pests and
their biotypes, simultaneously or even sequentially is difficult if not in~possibleor
impractical. Availability of tightly linked genetic markers for resistance genes will
help in identifying plants carrying these genes simultaneously without subjecting
them to the pathogen or insect attack in early generations. The breeder would require
only a small amount of DNA from each of the individual plants to be tested, and this
can be obtained without destroying the plant. Using the known set of DNA primer
pairs for PCR, the products of the reaction would have to be run out on PAGE gels
and the genotypes of the individual plants assessed to predict resistance or
susceptibility by the presence or absence of the resistant parent's marker band on the
gel. Only the nlaterials in the advanced generations would be required to be tested in
disease and insect nurseries. Thus, with MAS. it is now possible for the breeder to
conduct many rounds of selection in a year without depending on the natural
occurrence of the pest or pathogen, and theoretically without the pest or pathogen as
well. However, the presence of different races or biotypes complicates the
development and application of lnolecular marker-assisted selection. Markers
developed for one pathotypes or biotype may not have application in other locations
in which different pathotypes or biotypes occur unless resistance to these is controlled
by the same gene.
The selection of sorghun~genotypes for resistance to shoot fly by utilizing one
or a few resistance parameters is inefficient because several components are involved
in resistance and one or more genes govern each of these resistance cotnponents.
Further expression of many of these components is influenced by environmental
variation, hence shoot fly resistance is a quantitative trait and shows a large amount of
genotype X environmental interaction.
Marker-assisted selection has a considerable potential to improve the
efficiency of the selection for quantitative traits (Hash and Bramel-Cox, 2000) such as
shoot fly resistance, for which expression is sensitive to the testing environment. As
the components to resistance to shoot fly are mostly quantitative in nature it is
potentially important to identify quantitative trait loci (QTLs) for these from the
viewpoint of applied genetics and plant breeding. The ultimate goal of such QTL
analysis is to develop tools that are useful for marker-assisted selection. In a practical
breeding program the aim is to increase the level of resistance in agrotiomically elite
backgrounds. If this is successful, these marker-assisted breeding tools can also be
helpful in pyramiding genes for hoot fly resistance.
Conventional quantitative genetic studies on shoot fly resistance with different
sorghum genetic materials have been reported by many workers. Recently QTL
analysis for shoot fly resistance component traits has been done using a set of
sorghum recombinant inbred lines (a RIL population) derived from cross BTx623 X

IS 18551. Sajjanar (2002) and Folkertsma et 01. (2005 unpublished) have reported on
252 RILs of a (BTx623 x IS 18551)-derived mapping population that were screened
for shoot fly resistance component traits in three environmetits. The same set of RlLs
was genotyped using 109 SSR marker loci and QTL analysis performed with the aim
of identifying the genomic region associated with shoot fly resistance. Composite
interval mapping QTL analysis using PLAB QTL version 1.1 revealed the presence of
28 QTLs detected at least two of the three screening environments. Closely linked
markers were identified for four QTLs for deadhearts incidence. In the present study
efforts are being made to transfer these four deadhearts incidence QTLs by marker-
aided selection from one or more donors to three elite recurrent parent lines (28B.
2 0 0 and KR 192) developed at the Sorghum Research Station Parbhani.
Using marker-assisted selection, we able to introgress genomic regions for
shoot fly resistance from donor parents (i.e., IS 18551 and RlLs 189, 153 and 252
derived from BTx623 X IS 18551) to recurrent parents (28B. 20B and KR 192) over
two generations. Markers linked to shoot fly resistant QTL regions to be transferred
from donor to recurrent parents were used for foreground selection. where as
polymorphic markers evenly distributed over genomic regions to be retained from the
recurrent parents were used for background selection. Based on the genotype data,
individuals heterozygous (BCIF1and BCzFl generations) for markers spanning shoot
fly resistance QTLs were selected during the first step of selection (foreground
selection). Among these initially selected individuals. those with background
genotypes having minimal presence of donor alleles unlinked to shoot fly resistance
QTLs were selected during the second step of selection (background selection).
5.10.1 Criteria for selecting the individuals
Markers, especially foreground markers, were taken into consideration for selection of
the individual segregants to be advanced. The individuals scored as 'A', 'H' or 'B' for
markers used in foreground selection, and 'A' for most background markers, were
selected for generation advance. Individuals scored 'H' at a particular marker locus
are expected to produce progeny segregating 1:2: 1 for homozygosity for the recurrent
parent allele (scored 'A'), heterozygosity (scored 'H') and homozygosity for the
donor parent allele (scored 'B') if they are advanced by selfing, or segregating 1: I for
homozygosity for the recurrent parent allele (scored 'A') and heterozygosity (scored
'H') if they are advanced by backcrossing to the recurrent parent. Presence of 'A'
genotypes for background markers and 'H' genotypes for foreground markers
flanking a particular shoot fly resistant QTL ensures the recovery of the recurrent
parent genome (288, 2 0 8 or KR 1921, while advancing introgression of the genomic
region contributing to shoot fly resistance traits. Individuals meeting these criteria
were selected and advanced to the next generation by selfing ahd backcrossing.
Individuals scored 'A' (for all foreground and background) markers were found to be
very similar to their recurrent parents (in fact they should be identical to the recurrent
parent except for small introgressions) and could be selected as control entries for use
in field trials to assess the efficiency of MAS for the shoot fly resistance component
traits.
For selected individuals. the markers scored 'H' and those that did not amplify
during the BCIFI background screening were screened again during the next
generation. The markers scored 'A' (homozygous for the allele of the recurrent
parent) were not tested further in advanced generations because recovery of tlie
recurrent parent genotype at these loci has been completed and their genetic
constitution is not expected to change further assuming a negligible rate of mutation
and no outcrossing to non-recurrent parent genotypes. Once tlie recurrent parent
genome has been recovered for all the background markers, a generation of selfing
and selection for homozygous donor parent marker alleles at loci flanking specific
target shoot fly resistance QTLs will be conducted and the selected genotypes then
multiplied by selfing prior to being tested niultilocationally to evaluate them
phenotypically for shoot fly resistance component traits and other agronomic traits.
After testing, if the progeny with the introgressed shoot fly resistance QTLs are found
to be significantly superior compared with the recurrent parent, they can be released
as improved versions of the recurrent parental lines for use in breeding agronomically
elite hybrids with improved shoot fly resistance. The improved potential for shoot fly
resistance of these new elite parental lines will be due to introgression of shoot fly
resistance QTLs by marker-assisted backcrossing.
 Twelve SSR marker loci linked to the four targeted shoot fly resistance QTLs
were used for genotyping recurrent and donor parents. The details of the parental
polymorphism detected at these loci were presented in the Results chapter of this
thesis (Table 4.9). After detecting the polymorphism between the recurrent and donor
parents, homozygous parental-type plants uzere selected at seedling stage for
subsequent plant-to-plant crosses, which were effected by manual emasculation
followed by controlled pollination. Finally, we succeeded to effect five FI hybrids
involving true-to-type parental plants. Out of these fibe hybrids, two were based on
20B, two were based on KR 192, and one was based on 28B
5.10.2 F I and BClFl generations for recurrent parent 28B
i Five plants putatively produced from cross 280 (288) X RII, 189 (3 12) were
getlotyped at seedling stage using (Table 4.10a) four SSK loci linked with
targeted shoot fly resistance QTLs. Two heterozygous Fl hybrid plants were
selected as a females and crossed with pollen from selfed progeny of the
original recurrent parent 2 8 8 (288) to produce BClFl seed.
r Thirty B C I F l plants of the [28R (288) x KIL 189 (312)l x 28B (288)
backcross populatio~~
were screened at seedling stage at loci detected by 11
SSR primer pairs that targeted four shoot fly QTLs (Table 4.10h). Fourteen
heterozygous plants (SP nos. 612, 613, 617. 619, 710. 71 1, 719, 81 1, 812, 814,
815. 816. 817 and 818) were selected at seedling stage and crossed with the
pollen from the selfed progeny of recurrent parent 2 8 8 (288) to generate
BC2F, populatio~ls.According to Tanksely er cri. (1989), computer stimulation
using the tomato as a model showed that by selecting the best plant out of a
total of 30 per generation. the whole recurrent genome could bc covered in
two generations
5.10.3 F I a n d B C I F l generations for recurrent parent 206
'r Ten plants putatively produced from the plant-by-plant cross 20B (186) x
RIL 252 (318) were genotyped at seedling stage with four SSR loci linked
with targeted shoot fly resistance QTLs. Two heterozygous F I hybrid plants
were identified and backcrossed with pollen from the selfed progeny of the
original recurrent parent 20B (186) to generate BCIFI populations (Table
4.12a).
); Thirty BCIFl plants of the [20B (186) X RIL 252 (3 18)] X 20B (186)
backcross populations were genotyped at seedling stage with 1I SSR loci
 associated with shoot fly resistance traits (Table 4.12b). Twelve plants (SP
nos. 649, 650, 651, 655, 656, 657, 754, 757, 848, 849, 850 and 853)
heterozygous for one or more targeted QTL introgressions were selected and
used as females in backcrosses with selfed progeny of the original recurrent
parent 2 0 8 (1 86) to advance to the BCIFl generation.
k Ten plants putatively produced from the plant-by-plant cross 20B (179) X

RIL 153 (248) were genotyped at seedling stage with four SSR loci linked
with targeted shoot fly resistance QT1.s. Eight heterozygous F, hybrid plants
were selected (Table 4.13a) and used for backcrossing as females with pollen
from selfed progeny of the original recurrent parent 2 0 8 (179) to advance to
the BClFl generation.
k Thirty plants of these backcross populations 1208 (1 79) X RIL 153 (248)l X

20B (179) were screened at seedling stage with I I SSR marker loci associated
with shoot fly resistance traits (Table 4.13b). Nineteen BCIFI plants
heterozygous for one or more targeted QTL(s) introgressions were identified
(SP nos. 668. 669, 670. 672, 673, 674. 767, 771, 772, 773, 775. 776. 867, 868,
870, 871. 873, 874 and 876) and backcrossed selfed progeny of the original
recurrent parent 20B (179) to advance to the B C ~ Fgeneration.
I
5.10.4 F l a n d BClFl generations for recurrent parent KR 192
i Ten plants putatively produced from cross KR 192 (304) X IS 18.551 (267)
were genotyped using four SSK loci linked to targeted shoot fly resistance
QTLs (Table 4.11a). All ten plants were identified as heterozygous and
backcrossed with pollen from selfed progeny of the original recurrent parent
KR 192 (304) to generate B C I F Iseed.
k Thirty BCIFl plants of the [KR 192 (304) x IS 1855 1 (267)l X KR 192 (304)
backcross populations were genotyped at seedling stage with 11 SSR loci
linked to the four targeted shoot fly resistance QTLs (Table 4.1 Ib). Fifteen
heterozygous plants having one or niore targeted QTL introgressions were
selected (SP nos. 629, 630. 633, 636, 729, 731. 732, 736. 737, 830, 832, 833.
834, 835 and 836) and used as females for backcrossing sJith pollen from
selfed progeny of the original recurrent parent KR 192 (304) to generate
BC2Fl seed.
P Nine plants putatively produced from cross KR 192 (300) x RIL 252 (319)
were screened with four SSR loci at the seedling stage (Table 4.14a). Eight
 heterozygous plants were identified and crossed as females with pollen from
selfed progeny of the original recurrent parent KR 192 (300) to generate
BClFl seed.
P Genotyping at seedling stage of 30 BClFl [KR 192 (300) X RIL 252 (319)l X

KR 192 (300) individuals with 11 SSR loci linked to targeted shoot fly
resistance QTLs (Table4 .14b) identified 9 lieterozygous plants (SP nos. 688.
692, 887, 889. 890, 892. 893, 894 and 895), which were selected and crossed
with pollen from selfed progeny of the original recurrent parent KR 192 (300)
to generate BC2Fl seed.
One hundred and fifty plants froni five RCIFI populations (described in the above
paragraphs) were screened at seedling stage with 1 1 SSR loci linked with the four
targeted shoot fly resistance QTLs. On the basis of molecular data, 69 plants having
appropriate allelic constitution (heterozygous for one, two or more targeted QTL
introgressions) were identified ('fables 4.10b-4.14b) and crossed as females with
pollen from selfed progeny of their respective original recurrent parents to advance to
the BCzFl generation. 'The details of these 69 selected plants heterozygous for various
targeted QTLs and the associated characters for the targeted genomic regions are
presented in Table 5.2.
 162
Table 5.2 List of selected QTL-introgressed heterozygote plants and associated
characters of five backcross (BCIFI)populations produced in this study
No. of Targeted QTL Shoot fly resistance traits associated
heterozygote linkage groups
plants selected
20 LG A Oviposition I. Deadliearts 1
5 LG E Oviposition I. Deadhearts I
10 LG G Trichome density (upper and lower leaf blade
surfaces). Glossiness, Seedling vigor 11.
Oviposition I and 11, Deadhearts 1 and I1
LG A+E Oviposition I, Deadhearts I
LG A+J Glossiness, Seedling vigor 11. Ovipo$ition I and
11. Deadhearts I atid I1
1 3 A+G I'richome density (upper and lower leaf blade
surfaces). Glossiness. Seedling vigor 11,
Oviposition I and 11, Deadhearts I and 11
LG E+J Glossiness, Seedling vigor 11. Oviposition I and
11. Deadhearts 1 and ll
LO E+G 1richome density (upper and lowcr leaf blade
surfaces), tilossincss, Seedling vigor 11,
O\iposition I and 11. Deadhearts 1 and 11
LG G+J Trichome density (upper and lower leaf blade
surfaces). Glossiness, Seedling vigor 11.
Oviposition 11. Deadhearts 11
LG E+G+J 'Trichome density (upper and lower leaf blade
surfaces), Glossiness, Seedling vigor 11.
Ovipositio~i11. Deadhearts I1
LG A+E+J Glossiness, Seedling vigor 11, Oviposition 1 and
11, Deadhearts 1 and I1
LG A+G+J Tricho~nedensity (upper and lower leaf blade
surfaces), Glossiness, Seedling vigor 11.
Oviposition 11, Deadhearts I1
LG A+E+G Trichome density (upper and lower leaf blade
surfaces). Glossiness. Seedling vigor 11
Oviposition I and 11, Deadhearts I and I1
5.10.5 Efftciency of marker-assisted selection
5.10.5.1 Background genotyping of BCIFIindividuals selected on the basis of
foreground selection
A total of 61 B C I F Iplants selected through foreground screening from five backcross
populations were genotyped with a set of polymorphic SSR markers (Table 4.21)
covering the entire genome except the genomic regions harboring the four targeted
shoot fly resistance QTLs.
The main objective of background selection &as to confirm and hasten
recovery of the recurrent parent genome in these genomic regions distant from the
four targeted jntrogressions. Tsve)ve plants were selected from thesr five backcross
populations, each of which carry homozygous recurrent parent alleles (.A' genotypes)
at most of the background SSR loci and have a few heterozygous ('H' genotypes)
background SSR loci. Those individuals Iiomozygous for any donor parent allele ('B'
genotype) were rejected as they c o ~ ~ only
l d have resulted lion1 failure of backcrossing
(i.e.. selfing) in the previous generation. 'She 12 selected individuals fro111five BCIFl
populations each had been crossed with their respective recurrent parents to advance
this marker-assisted QTL introgression program to the B C ~ F Igeneration. Details
regarding selected individuals including targeted QTLs, number of SSR loci
genotyped as 'A' allele, 'H' allele, andlor 'not amplified SSR loci' in the BCIFl are
presented in Table 5.3.
The result of the RCIFl generation background screening ('Sable 5.3) revealed
that in the BClF, generation for recurrent parent 28B on average 21 SSR marker loci
were tested and in the BC2FIof recurrent parent 28B the number of these background
markers requiring genotyping reduced to 3-13 per population. Further, in the BCIFI
generation for recurrent parent KR 192 (304) on average 28 SSR marker loci
/population were tested for background screening, and in the B C ~ F generation
I for
this recurrent parent the number of these background markers requiring genoty~ing
reduced to 5-14 per population. Moreover, in the BCIF: generation for recurrent
parent KR 192 (300). an average 30 of marker locilpopulation were tested for
background screening and in the BCZFl generation for KR 192 (300) the number of
these background markers requiring genotyping reduced to 12 per population. Finally,
in the BCIFl generation for recurrent parents 20B ( I 86) and 20B (1 79), an average 18
and 23 SSR marker pairs1 population, respectively, were tested during background
genotyping, and in the BCIFl generation for 20B (186) and 20B (179) the number of
these background markers still requiring genotyping reduced to 11 and 5,
respectively. The decreasing numbers of background markers still requiring
genotyping in the B C ~ F generation
I populations reflects the increasing percentage of
recurrent parent alleles fixed in each advancing backcross generation. Hospital er al.
(1 997), based on the stimulation studies, recommended an optimal distance between
two adjacent markers of about 5-10 cM. We used niuch larger intervals between both
foreground marker pairs and between adjacent background marker loci. However, we
still observed that the frequency of recurrent genotypes among the selected progeny
increased as the selection intensity for recurrent genotype increased, as reported by
Knapp (1998). Practically speaking. the nuniber of markers that must be used
decreases in each successive hackcross generation (Table 5.3) because once the
recurrent parent allele has been fixed at any given non-targeted locus, it is not
necessary to continue screening at that locus in subsequent generations as the locus
will remain homozygous for the remainder of the backcross and selting generations
(Moris et crl., 2003). Marker-assisted selection has the potential to greatly reduce the
cost and time for selecting desirable genotypes with traits of interest (Moris c/ (11..
2003). Marker-assisted selection is more efficient and cost cl'fective than conventional
selection for traits with a low heritability and high phenotypic trait effect (Ilospital el
(11.. 1997). Using MAS the current study was able to advance through four generations
within two years. When conventional breeding strategies are applied. the
advancement of the four backcrossing generations with phenotypic selection for shoot
Ily resistance traits will take at lcast four years. Conventional breeding schemes
feature low cost per unit timc during the research stage but require a longer time to
conlplete. where as marker-assisted breeding features high cost during the research
stage, but takes less time to conlplete. Kelcase stage and adaptation stages of
conventional and marker-assisted breeding are assumed to be identical in terms of
cost as well as duration. From an economic point ol' view the advantages of MAS thus
derives from the fact that the release and adaptation stages move forward in time, so
that the benefits from crop varietal improvement reach farniers and consumers earlier
(increasing the rate of return on the research investment). This suggests that while
MAS needs more initial investment, it is worthwhile in at least certain cases because
via the accelerating rate of varietal release, MAS generates large additional economic
benefits (Moris el al., 2003).
 The fact that MAS technology is so challenging should not be a reason for
discouragement, but instead, a wake-up call for more ingenuity, better planning and
execution of marker-assisted breeding programs. MAS for quantitative traits is in an
important transition phase, and the field is on the verge of producing convincing
results. Technology development, including automation, allele-specific diagnostics
and DNA chips, will soon make marker-assisted selection approaches based on large-
scale screening much more powerful and effective (Young, 1999).
5.10.6 Genotyping of the five B C ~ Fpopulations
I with SSK markers for
foreground selection
Screening 224 BClFl plants from five populations was completed at the seedling
stage with 10 SSR marker loci linked to targeted QTLs associated with shoot fly
resistance traits in four linkage groups (Table 4.22-4.26). Around 100 heterozygous
plants have one or more targeted QTL introgressions were selected and crossed as a
female parents with the selfed progeny of their respective recurrent parents to
generate BC3FI populations. Details of the numbers of' plants introgressed with each
combinatio~iof targeted Q'1'Ls and associated shoot fly resistance traits across these
five populations are presented in Table 5.4.
Table 5.4 Details of foreground selection of five BClFl populations with targeted
shoot fly resistance QTL and character association
No. of introgressed plants Targeted QTL Shoot fly resistance trait
selected linkage group associations
37 LG A Oviposition I and Deadhearts I
16 LG E Oviposition I and Deadhearts 1
13 LG G 'frichome density (upper and
lower leaf blade surfaces).
Seedling vigor 11. Deadhearts I
and 11, Oviposition I and 11,
and Glossiness
Glossiness, Seedling vigor 11,
Oviposition I and 11,
1)eadhearts 1 and 11
Tricho~nedensity (upper and
lower leaf blade surfaces).
Seedling vigor 11, Deadhearts I
and 11, Oviposition 1 and 11.
and Glossiness
Glossiness, Seedling vigor 11.
Oviposition 1 and II.
Deadhearts l and 11
Trichome density (upper and
lower leaf blade surl8ces).
Seedling vigor 11, Deadhearts I
and 11. Oviposition I and 11,
and Glossiness
Glossiness, Trichome density
(upper and lower leaf blade
surfaces). Oviposition I and 11.
Deadhearts l and ll
5.10.7 Genotyping of foreground-selected BCIFI plants for background selection
Out of 100 BCzFl plants selected on the basis of foreground screening only 68 plants
from five backcross population will be genotyped with a set of polymorphic loci
covering the entire genome except regions associated with targeted shoot fly
resistance QTLs. This background screening will be restricted to the SSR loci that
were heterozygous or not amplified in the BClF, generation background screening
(Table 4.28). Details of the 68 foreground-selected B C ~ Fplants
I from five backcross
populations chosen for background screening are presented in Table 5.5.
Table 5.5 Details of foreground-selected BCIFI plants chosen for background
genotyping from five BCZFIpopulations with targeted QTL and character
associations
No. of plants chosen for Targeted QTL Shoot fly resistance trait associations
background screening linkage groups
14 1.G A Oviposition I and Deadhearts I
12 LG E Oviposition I and Deadhearts I
12 LG G Tricholue density (upper and lower),
Glossiness. Seedling vigor ll.
Oviposition I and 11. Deadhearts I and I1
Glossiness. Seedling vigor 11.
Oviposition I and 11. Deadhearts I and I1
Trichome density (upper and lower),
Glossiness, Seedling vigor 11.
Oviposition I and 11. Deadhearts I and 11
Glossiness, Seedling vigor 11.
Oviposition 1 and 11. Deadhearts I and I1
Trichome density (upper and lower),
Glossiness, Seedling vigor 11,
Oviposition I and 11, Deadhearts I and 11
Trichome density (upper and lower),
Glossiness, Seedling vigor 11.
Oviposition I and 11, Deadhearts I and I1
5.10.8 Recommendations for future studies
Based on the Present study, the following two routes for future pyramiding of shoot
fly resistance QTLs in elite agronomic backgrounds are suggested.
1 The first avenue consists of identifying the best BC,FI plants, i.e. those showing the
highest amount of favorable QTL introgression for shoot fly resistance traits and
fixing the favorable recurrent parent alleles in BC3FI generation background
screening. Such selected B C ~ Fplants
I can then be crossed with other BC3FI plants of
different descent but having a common recurrent parent in order to pyramid as many
as shoot fly resistance QTLs as possible (each contributing to different traits) within
the same genome (selective pyramiding).
2 The second avenue consists of seedling stage genotyping of the BC3FI population
for each recurrent parent with linked SSR marker pairs targeting each of the four
shoot fly resistance QTLs for foreground selection. Foreground--selected individuals
will be background genotyped using a set of polymorphic SSR markers unlinked to
targeted QTL regions. Those individuals appearing to fully recover of recurrent parent
marker genotype across these background marker loci and phenotypically most nearly
identical to their recurrent parent will selfed to fix the introgressed QTL alleles. The
resulting segregating progenies will be used for development of near-isogenic lines
(NILS). Such plant material should prove useful to study the effects of any given
single QTL on the phenotypic value of a plant harboring it. In case the introgressed
QTL proves to contribute significantly to the i~nprovementof any given trait the line
can then be used directly in hybrid breeding. Further, such QTL-NILS could also be
used as donor or recurrent parents in a short series of crosses for QTL pyramiding
3 Phenotyping of the selected individuals for the shoot fly resistance components
traits is needed. lntrogression of any trait is confirmed phenotypically after several
generation of genotyping. In this context, the selected BC3F2genotypes (in 28B, KR
192 and 20B recurrent parent backgrounds) will be evaluated for shoot fly resistance
component traits during late kharif2006 and rabi 2006107 seasons. Fine mapping for
these shoot fly QTLs will be a practical possibility only once the Presence of the
different shoot fly resistance QTLs is phenotypically confirmed. Such c~nfirmed
shoot fly resistance QTL introgression lines can then be used to to generate ESTs for
better understanding of this complex trait.
SUMMARY AND CONCLUSIONS
 CHAPTER VI
SUMMARY AND CONCLUSIONS
The present investigation entitled. "Genetic diversity analysis, QTL mapping and
marker-assisted selectiocl for shoot fly resistance in sorgllum [Sorghl~rrrhicolor (L.)
Moench]" was aimed, 1) to assess genetic dhersity by SSR markers in a set of insect
resistant lines. 2) to understand tlie genetics of tlie shoot fly resistance and locnte the
chromosomal regions harboring the quantitative trait loci (QTLs) for shoot fly
resistance and related component traits. and 3 ) introgression of shoot fly resistance
co~nponents traits in agronomically superior genotypes using ~nolecular marker-
assisted selection. For all the sti~dicsinvolving the use ol' SSIi ~narkers,DNA was
extracted with a modified CLAD metllod. 111 the first ubjective. genetic diversity
analysis by SSK markers in a sct of sorghulu lines. 91 sorghu~naccessions were
genotyped using 20 SSIi primel- pairs that detected loci distributed over 9 of the 10
linkage groi~psi l l the sorgliuni ~iuclenrgenome. PCR amplilication of a targeted
sequence was conducted in an Applied Biosysteln GecleALIP I'CK systeni 0700
tliermocycler using a touch d o ~ v nPCK pl.otocril. Ampliiied products were separated
by electrophoresis on 6% non-denaturing polyncr)lamide gels. and by capillar).
electsophoresis using automated DNA scqucnccrs. PCR products separated by PAGE
\\ere visualized by silver staining. Illtensely amplified specilic bands representing the
corresponding allclic products oS thc SSR 111arliel.sin each accession \bere scored to
produce tlie PAGE data set. 'l'hc N'I'SYS statistical solinare package was used for
cluster atialysis of the SSR niarker allele data sets f ~ o mthe PAGE and capillary
electrophoresis I'CR product sepalation ~netliods. Jaccard's siciiilarity coefficient
for each of these
bet\veen each pair of accessions was used to construct de~idrogra~ns
two data sets usi~igthe un-weighted paired group method with arithmetic averages
(UPGMA).
The seco~idob,jective included plienotypitlg and mapping of QTL(s) for shoot
fly resistance and its component traits in a (2960 x IS 18551)-derived RIL mapping
population. The 259 RILs along with their two parental lines were evaluated in late
khtrrif 2002 and early rohi 2004-2005 screening cnviron~~lents
with standard shoot fly
resistance screening procedure at ICRISAT. Patancheru. Observations for shoot fly
resistacice and cotiiponent traits were recorded during phenotypic screening in each
environment. The analysis of variance for phenotypic data sets was performed using
the residual maximum likelihood algorithm (ReML), which provides the best linear
unbiased predictors (BLUPs) of performance of the genotypes. The BLUPs of 213
uniform RILs, along with their genotypic data from I I I marker loci, were used for
QTL analysis. QTL analysis was performed using the composite interval mapping
(CIM) method. The required computations were performed using PLABQTL version
1.1, which uses a regression approach.
The third objective included introgression of shoot fly resistance components
traits into agronomically superior genetic backgrounds using marker-assisted
selection, in which two potential maintainer lines (20B and 288) and one potential
restorer line (KR 192) used as a recurrent parents, and RILs 153, 189 and 252 derived
from cross (BTx623 x IS 1855 I), and IS 1855 1 (resistant parent) were used as donor
parents. Three shoot fly resistance QTLs had been previously mapped in each of these
selected RILs, and the resistant parent carries all the four QTLs imparting shoot fly
resistance. Eleven primer pairs for SSR loci flanking to the four target shoot fly
resistance QTLs from 4 linkage group were used for genotyping the parental
population at seedling stage, polymorphic SSR loci were identified, true-to-type
parental plants were selected and crosses (plant-to-plant) between recurrent and donor
parents were effected. Five hybrids were developed and thest. five hybrids were
advanced for backcrossing. Now these five populations are in B C ~ F Istage. The
genotyping procedure described for the first objective above was also followed in this
objective.
The research results and conclusion^ for each of these objectives are briefly
summarized below.
I. Application of SSR markers in diversity analysis of sorghum insect resistant
germplasm accessions
1. In this study we tried to assess the genetic diversity of a set of 91 elite
sorghum germplasm accessions using SSR markers. The set include 12 shoot
fly and 15 stem borer resistant accessions, 9 accessions resistant to both shoot
fly and stem borer, 17 midge resistant accessions, and 38 agronomically elite
recurrent parents that were in use at ICRISAT to initiate a large-scale marker-
assisted backcross program for the stay-green components of terminal drought
tolerance from donor parents 835 and E 36-1.
2. In case of PAGE electrophoresis sep~rationof the PCR products for diversity
analysis, 20 out of 21 primer pairs used provided amplification products, while
 11 out of these 20 revealed liigli level of polymorphism. A total 69 alleles
were detected by silver staining. for an average of 3.5 fragme~ltsamplified per
SSR locus across the 91 sorglit~maccessions studied. In case of capillaq
electrophoresis (ABI) separation of these PCR products, a total of 118 alleles
were detected. for an average of'5.1 fragments amplified per SSR locus. Based
on capillary electrophoretic separalio~iof their PCR products. 13 out of 20
(65%) SSK primer pairs were able to detect a high lcvel ofpolymorphism.
3. Jaccard's coefficient of similarity for pairs of the 91 sorghum accessions
studied ranged form 0.28-1.00. The dendrogrncus for the si~ililaritybetween
accessions based on PAGI+- and ABI-generated SSR genotype data showed
clustering for geographical origins. races and specific traits such as insect
resistance.
4. 111 case of the PAGE dendrogram. the 91 sorplium accessiolis diverged into 20
clusters at tlie 50% level ( ~ sinlilari(y.
f Among these. the largcst cluster was
clirster 4 ( 1 8 accessions). which Mas iollouetl by cluster 12 (13 accessions)
and cluster 13 (12 accessions). I l ~ \ v e \ ~ esomc
r. of the clusters (tj g. clusters 5.
0111). a si~igleaccession each. and
7. 9. 10. 14. 19 and 20) acco~n~ilodated
clusters 1 . 3 and 6 accotnniodated only 2 accessions each. 111case of the ARI
de~idrogram.the 91 sorghum genot3pes grouped into 28 clusters at the 50%
level o f si~nilarity. Whet1 compared to the dendrogram fro111 the PAGE-
generated d a t i ~sets. tlie ~ ~ i ~ m of
h e clusters
r detected with 12F31-generated data
was moderately Iiighe~..Tliis is likely due to tlie greatcl se~lsitivity of tlie
automated sequencer. \vliicli alloxv it to detect SSR alleles differing by sc~ialler
t~icmbersof repeated units than was p o s i b l e \bit11 the silver-stained PAGE
gels. so tliat it coulde effectively detect a higher level of polymorphism.
5 . The information pro\,ided by this study about the diversity/si~niIar~ty
of the
ger~nplasmfrom dilf'erent so~trcesof origin or region s h o ~ ~ prove
ld extremely
useft~l for heterosis breeding and in selecting parental lilies for specific
breeding goals related to combining insect resistance with high grain yield and
mitigation of drouglit stress.
6 . Tllis diversity analysis for 01 sorghum accessions revealed that the genotypes
studied are genetically quite diverged with sorghum lines showing midge.
shoot fly and stem borer resistance clustering in different groups. In addition. a
cluster of agronomically superior recurrent parents was identified that is
 getietically quite divergent from each of tliese insect resistance clusters.
, However, sonie of the accessions with resistance to midge. shoot fly and stem
borer clustered separately. indicating that these lines might contain new allelic
variants that can be exploited in breeding prograni. This information will be
useful for identifying parents for marker-assisted bacltcrossing programmes to
i~itrogress insect resistance QTLs from the currently available mapping
populations. Further, newly identified agronomically elite and genetically
diverse insect resistant breeding lines can be used for developilig new
mapping populations to detect additional insect resistance QTLs.
11. Plienotyping n set of R l L s anrl identification of Q T L s f o r shoot fly
resistnr~cecolnponents of the RIL population derived from cross 296B x

IS 18551.
I. In general. the mapping populntion parents (296B and IS 18551) differed
phenotypically for all observed parameters of slioot fly resistance.
2. Parental and RlLs tiiean values re\,ealed wide \ariation in phenotypic values
for slioot fly r e s i s t a ~ ~ cand
e its component traits ill both o f the screening
environments. Wide variation was observed in the KIL ],op~~lationfor slioot
fly resistance component traits like glossiness intensity. trichome density
(upper and lower surfaces of seetlling leaf hltitles). seedling vigor. oviposition
incidence (Oh). and deadliearts incidence ("A). These traits can be used as
simple criteria f'or selection of resistant genotypes.
3. 'The genotypic variances for shoot fly resistance traits were significant ill both
of tlie itidividual screening en\ironments as \\,ell a s in the across-season
analysis.
4. Glossiness intensity. tricliome density (both upper and lower surfaces of
seedling leaf blades). o\,i],osition incidence (%). dearlhearts incidence (%). and
seedling vigor recorded consistent heritability (broad sense) estimates in
individual screening en\~ironmelits.hut low to moderate heritability estimates
in the across-season analyses indicating that these traits are under genetic
control but that there is a suhstantial role of genotype (G) x environtnent (E)
interaction in expression of these traits.
5. RIL population means differed significantly from those of both tlie parents for
important slioot fly resistance component traits like glossiness. oviposition
incidence (%). deadliearts incidence (%) and tricliome density (upper and
 lower surfaces of seedling leaf blades). Transgressive segregants with
phenotypic values out side tlie parental linlits were observed for glossiness,
oviposition incidence. deadhearts incidence nnd trichome density.
6. QTL analysis revealed presence of putative QT1.s for all i~nportantshoot fly
resistance and resistance component traits like glossiness, oviposition
incidence (%). deadhearts incidence (%), and trichome density. The portioli of'
observed plienotypic variance explained by different putative Q'I'Ls varied
fro111 6 to 34%. Glossiness intensity was largely controllecl hy a major QTL on
1.G J. accounting 34% of observed phenotypic variation. and one minor QTI.
on LG G. accounting for 8% of observed phenotypic variation in thc across-
season analysis. After adjusting fc~rQTL x environn~ental interaction. thesc
two QTLs explained 31% ol' genetic variation in glossiness intensity in this
KIL population. Resistant paretit IS 18551 contributed the additive genetic
effects for increased glossitiess at both ol'these QTLs.
7 . For oviposition 11 and deadliearts 11 incidence ('$6). two common QTLs (one
on LC; F and one on 1-Ci G ) \+ere mapped in tlie across-season analysis. Both
0.1-Ls together explained 17% and 19%) plien<~typicvariation in oviposition I 1
and deadhearts 11, respectivelq. in the across-season analysis. Significant QTI.
\\ere obscr\cd for both of these QTLs for
e ~ ~ v i r ~ n ~ i l ei~iteractio~ls
~ital
oviposition I1 and deadhearts 11 resistance traits. I-he Q 1.1. d napped on LG G
for deadhearts ailtl oviposition co-locali~etl\\it11 a niajor QTL for trichome
delisit!, (upper and loher surfaces of seedling leaf blades) and a minor Q1'1. of
glossiness intensity. The QTL mapped on 1.G F for deadliearts and oviposilion
co-localized with a minor QTL for tricliolne tlciisity of tile lower leaf blade
surface.
8. For trichome density. one QTL \+as clelectetl on LG G accounting for 30% of
observed pllenotypic variance in tlie across-season analysis for trichonle
density on the upper leaf blade surface. This QTL for trichome density on tlie
upper leaf surface co-localized \vith a QTL for trichome density on the lo\ver
leaf surface that explains nearl! 27% ol'ohservcd plienotypic variance across
the two screenillg environments. 'l'his indicates similarity in genetic control of
trichome density on either side of sorghum seedling leaf blades.
9. The major QTL for glossiness intensity and minor QTL for oviposition (LG J)
and major QTL for trichome density and minol. QTLs for glos~illess.
 deadliearts and oviposition (LG G) detected in this study have previously been
mapped at the same location in another sorghum RIL population derived fro111
cross RTx623 IS 18551. This confirms that these chromoso~nalregions
might be llarboring candidate genes contributing to shoot fly resistance of IS
18551.
10. T w o aphid resistance QTLs were detected in the across-season analysis. These
mapped on LG E and I.Ci .I and togcther explain 26% of observed phenotypic
variation in aphid score. The Q'I-L mapped on LC .I was a major one.
accounting for 20% of observed plicnotypic variance and having non-
signiiicant Q x E interaction. The favorable additive genetic effects for both
aphid resistance Q'SLs \+ere contributed by IS 18551 alleles.
1 1. Lltilization of RlLs n o . 47. 51. 82. 97. 130 and I74 (!,oh;season adaptation)
and KlLs no. 46. 2U8. 222. and 223 ( k l ~ l ~ r ~ ~ f s e aadaptation)
son in sorghum
improvement profranis aimed at improving shoot fly resistance of elite
cultivars and hybrid parental lines is likely to be more fruitful than direct use
of an agronotnically poor source like IS 18551
111. lntrogression of s l ~ o o tfly resist:~ncecomponent traits in agronornically
superior genotypes using tnolccular marker-assisted selectiur~s
I. I.'lanliing SSK marker loci closely linked to four QTLs for shuot Ily resistance
coniponents in tlie (B'I'x623 x IS 1x55 I )-based K11. population and resistant
p r e n t IS 18551 mere identified in an earlie]. studies. In the present study
eflbrts are being made to transfer these fain OTLs by marker-assisted
selection fro111 donor parents (IS 18551. and lZlL nos. 1XL1. 153, and 252
derived from BTx623 x IS 1855 1 ) to three elite recurrent parental lines (28B.
20B and KR 192) developed at SRS. hlAU. Parbha111.Mahalzshtra. India
2. Twelve SSR marker loci linked to the targeted shoot fly resistance Q'fLs from
four linkage groups (ILG A. LG E. 1.6 Ci and LG J ) were used for genotyping
the three recurrent anrl four donor parents. After detecting SSK marker
polymorphism between the recurrent and donor parents at seedling stage,
homozygous parental-type plants were selected and subsequent plant-to-plant
crosses were effected. Finally we succeeded in developing five F I hybrids: i . e .
1. 2 8 0 (228) x RIL. 189 (312). 2. KR 192 (304) x IS 18551 (267). 3. 20B
( I 86) x RIL 252 (3 18). 4. 20B ( 1 79) x RIL 153 (248). 5. KIZ 192 (300) x RIL
252 (319).
3. Five to ten putative F I plants produced from each of hybrid cotnbinations
were genotyped at the seedling stage using four SSR loci linked with targeted
QTLs. Heterozygous F I hybrid plants in each of these combinations were
selected and used as a fe~nalesin crosses with pollen froni selfed progeny of
their respective recurrent parent to produce B C I F l generation seeds.
4. Around IS0 plants liom f i ~ BClFl
e populations were screened at the seedling
stage with 11 SSR loci linked with the four targeted slioot fly resistance QTLs
(foreground selection). O n tlie basis of this genotypic data. 69 lieterozygous
plants having one, two or more targeted QTL introgressic~n(s)were identitied
and crossed a s feliiales with pollen from selfed progeny of their respective
recurrent parent to produce BC2F1 seed.
5. In case of background genotyping of these BC1FI individuals a total of 61
BClFl platits selected through foreground screening from five backcross
populations were genotyped with a set u f polymorphic SSK markers covering
the entire genome except tlie geno~nicregions llarboring the four targeted
shoot fly resistance QTLs.
'f\celve plants \\,ere selected from across tliese tive hackcrossing populations.
whichcarry licimozygous recurrent parent alleles at most of the backgrou~id
SSR marker loci ant1 have only a few Iieteroz!~gous loci to he used for Suture
background screening. The baclicrossed seed of tlie 12 plants were advanced
to raise the 13Cr121penelxtion.
6. Arouliil 224 BC2FI pla~lts kiim live populatio~is Icerc gcnotyped at the
seedling stage with 10 SSK ~narkcr loci lililicd to four targeted QTLs
associated \\it11 sliout 11y resistance traits. Around 100 I~etero'ygous plants
have one. two or more QTI.. introgression(s) \\ere selected and crossed as a
female parents with tlie selfed progeny of theie respective recc~rl.entparents to
generate B C ~ Fseed.
I
7. Out of 100 BC2FI plants sclected in roregroulid screening only 68 plants from
five backcross populatio~iswill he genotyped (background selection) with a set
of polymorphic SSR loci covering tlie entire genome except tlie region of
targeted QTIA This background screening will be restricted to the loci that
were either lieterozygous or not amplified ill the B C I F I generation background
screening. On tlie basis of background and foreground screening data.
individuals having essentially fill1 genome of recurrent parent recovered along
 with one or more of tlie four targeted QTLs will be selected. Cross seed of
such individuals will be advanced to raise tlie BC3FI generation for
advancement to tlie BC,F2/BC4FI generations.
8. 111 the B C ~ F Igeneration sizable liuliibers of plants in each backcross
populatio~iwill be genotyped at tlie seedling stage with 10 SSR markers linked
with the four targeted QTLs (foreground selection). 011the basis of foregroulid
molecular data heterozygous individuals will be identified that carry one or
more of the targeted QTLs. Such individuals will be selfed to fix tlie favorable
allele(s). and segregating progeny \+ill be used for development of near-
isogenic lines (NILS) and pyramiding of particular resistance QTL
combinations in the genetic backgro~~lids
of each of the three recurrent parents.
9. Phenotyping of shoot fly resistance co~iipo~ient
traits for the resulting selected
Iiomozygous ilidividuals for each near-isogenic line pair is suggested. I11 this
context tlie selected genotypes in five BCjF? populations will be evaluated for
shoot fly resistance component traits in a slioot fly screening nurserq. in late
klior.[f 2006, atid rtrhi 2006-2007. Fine mapping for these slioot fly QSLs is
possible once the presence of different slioot tly resistatice Ql'Ls is
phenotypically confinlied. ESTs can be generated sltbsequently from such
confirmed QTL introgression lilies in order to i~iiproveour ~~nderstanding
of
this complex and challengilig trait. for the economic bc~iefitof poverty-striken
poor sorghum growing farnlers in tlie Semi-Arid Tropics (SAT).
LITERATURE CITED
 LITERATURE CITED

Agrama, H A . G E W~dle,J C Reese, L R Campbell and M R Tulnstra 2002 Genetlc

mapplng of QTLs assoc~atedw ~ t hgreen bug reslstance and tolerance In
Sorghum hrtolor Theor Appl Genet 104 1373-1378

Agrabal, B L and C V Abraham 1985 Breeding sorghuni for resfstance to shoot fly
[Athengonu I O C C U ~ U ]and midge [Conlirrrnru \ o r g h ~ ~ o l aProceedlngs
] of
the Internatlonal Sorghum Entomology Workshop, 15-21 lul) 1984, 'Texas
A & M Un~vers~ty.
College Stat~on.TX. USA. pp 171-183

Agrawal, B L and L R House 1982 Hreedlng for pert resirtance 111 sorghum In
Sorghum In the F~ght~es.
Proceedlngs of the Internatlonal Symporlum on
Sorghum, 2-7 November 1981 Intemat~onalCrops Research Institute for
the Seml-And Troplcs. Andhra Pradesh. Indla. pp 435-446

.
.4hnert. D M Lee. D F Austln. C L I ~ I ~\hiI I-. Woodman. H J Openjhan. J A C
Sni~th.K Porter and N G IIalk) 1996 Genetic diverr~t?among elite
sorghum ~nbreedl~nesdss15tedwlth DNA Crop SLI 36 1385-1192

AICSIP 1984 Progress reporl, Section IV (Entomology) All lndla C oordlnated Sorghum
lmprovenlent Project. Ind~an Counc~l of Agrlcultural Resedrch, IARI
Reglonal Statfon. tlyderdbad

AlCClP 2003 Progress Report-Cntomology, All lndla Coord~nated Sorghum

Improvement project (AICSIP). lndlan Counclal of Agrlcultural Kemrch,
Kat~onal Research Center for Sorghum. Ralendra Ndgar. Hyderabad-
500030. Ind~a

.
Akkaya, M S A A Bhagwat and P B Cregan 1992 Length polymorph~smsof clmple
sequence repeat DNA In soybean Genet~cs.132 1 131-1445

Alan~,S N and M B Cohen 1998 Detect~onand analys~sof QTLs for resistance to the
brown plant hopper, hi~luparvu/a lugenc, In a doubled-haplold rlce
population Theor Appl Genet 97 1370-1379
Aldrich, P.R. and J. Doebley. 1992. Restriction fragment variation in the nuclear and
chloroplast genomes of cultivated and wild Sorghum bicolor. Theor. Appl.
Genet. 85: 293-302.

Aldrich. P.R., J. Doebley, K.F Schertz and A. Stec. 1992. Patterns of allozyme variation
in cultivated and wild Sorghum hicolor. Theor. Appl. Genet. 85: 451-460.

Allard, R.W. 1960. Principles of plant breeding. John Wiley and Sons, Inc.. Ne% York

Balakotaiah, K., B.S. Rana, D.P. Tripathi and N.G.P. Rao. 1975. Genetic analysis of some
exotic x Indian crosses in sorghum. IX. Inheritance of resistance to shoot
fly. Indian J. Genet. 35: 344-349.

Balikai. R.A.. C.1). Singh and D.I. Jirali. 1908. Screening of sorghum genotypes against
the shoot fly, Atherigonu snccrriu Rond. Kamataka J. Agric. Sci. 1l(1):
208-21 1.

Bapat. D.R. and U.N. Mote. 198%. Upgrading the resistance lcvels of derivatives from
Indian x Indian crosses of sorghum against shoot fly. J. Maharashtra
Agric. Univ. 7: 171-173.

Bapat. D.R. and U.N.Mote. 1982h. Sources of shoot fly resistance in sorghum. J.
Maharashtra Agric. Univ. 7: 238-240.

Bapat. D.R.. K.A. Napeem and H.M. Markande. 1975. Correlation studies of HCN
content and shoot fly attack in sorghum. Sorghum Newslett. 18: 50.

*Barbosa-Neto .I.F.. M.E. Sorrels and G. Cisar. 1996. Prediction of heterosis in wheat
using coefficient of parentage and RFLP-based estimates of genetic
relationship. Genome. 39: 1142-1 149.

*Basten, C.J.. W.S. Weir and Z.B. Zeng. 1998. QTL Cartographer: a suite of programs
for mapping quantitative trait loci. Plant and Animal Genome VI. San
Diego, CA. January 1998, pp. 76.

Beckmann. J.S. and M. Soller. 1983. Restriction fragment length polymorphism in

genetic improvement: methodologies. mapping and costs. Theor, Appl
Genet. 67: 35-43.
*Bennetzen, J.L.. V. Subrammian. J. Xu. S.S. Salimath. S. Subramanian, D.
Bhattramakk~and G.E. Hart. 2001. A framework genetic map of sorghum
containing RFLP, SSR and morphological markers. In: DNA-based
markers in plants (Eds. Phillips. R.L. and 1.K. Vasil), 2nd Edn, Chapter 20.
Kluwer Academic Publishers. Boston, pp: 347-356.

Berhan. M., S.H. Hulbert, L.G. Butler and J.L. Bennetzen. 1993. Structure aid evolution
of the Genomes of Sorghum hicoior and Zeu mccys. Theor. Appl. Genet.
86: 598-604.

Bernacchi D. T. Beck-Bunn. D. Emmaty. Y. Eslied. S.lnani.. J. I,opez, V. Petiard. H.

Saysma, J. IJhlig, D. Zamir, and S. Tanksley. 1998 Advanced backcross
QTL analysis of tomato. 11. Reviewuation of near-isogenic lines carrying
single donor introgression for desirable wild QTL-alleles derived from
I,ycopersicon hir.\u/fr~,rand L. ~~im~~incllifoli~rnt.
'lheor. Appl. Genet. 97:
170-180.

Bhattacharjee. R.. P.J Braniel. C.T. Hash. M.A. Kolesnikova and I.S. Khairwal. 2002.
Assessment of genetic diversity within and betwen pearl millet landraces.
Theor. 4ppl. Genet. 105: 666-673.

Bhattramakki, D.,J. Dong, A.K. Chhabra and G.E. I-iart. 2000. An integrated SSR and
RFLP linltage map of Sorgh~imhicoior (I..) Mocncli. Genome. 43: 988-
1002.

Binclli, G.. L.M.E. Gianfranceschi. G. Pe. C. 'l'araniino. J. Husso. J . Stenhouse and E.

Ottaviano. 1992. Siniilarity of maize and sorghum genomes as revealed by
maize RFLP probes. Theor. Appl. Genet. 84: 10-16.

Biradar, S.G. and S.T. Borikar. 198.5. Genetic analysis of shoot fly resistance in relation
to growth stages in Sorghum. Z. Pflanzenziichtg. 95: 173-178.

Biradar. S.G., S.T. Borikar and R.D. Chundurwar. 1086. Trichonie density in some
progeny of sorghum. Sorghum Newslett. 29: 107.

*Blum. A. 1965. Screening of sorghum varieties for resistance to Athrrigoncr ~irriuand

preliminary investigation into the nature of this resistance. Volcani. Inst.
Agrie. Prelim. Rep. No. 503.
Blum, A. 1967. Varietal resistance-of sorghum to the sorghum shoot fly [Alherigona
varia soccala]. Crop Sci. 7: 461-462.

Rlum. A. 1968. Anatomical phenomena in seedlings of sorghum barieties resistance to

sorghum shoot fly (Atherigonu vuriu soccata). Crop Sci. 8: 388-391.

Blum, A. 1969a. Factors associated with tiller survival in sorghum varieties resistance to
sorghum shoot fly (Atherigona vrrria socccrtu).Crop Sci. 9: 508-5 10.

Blum, A. 1969b. O\aiposition preference by the sorghum shoot f l j (Alherigontr rariu

soccara) in progenies of susceptible x resistant crosses. Crop Sci. 9: 695-
696.

Blum, A. 1972. Sorghum breeding for shoot fly resistance in Israel. In: Control of
Sorghum Shoot Fly. (Eds, Sotwani. M.G. and W.R. Young). Oxford and
IBH Publishing Co.. New 1)elIii. pp. 180-191.

Bohn. M.. M.M. Khairallah. 1).C i o n ~ n l ede

~ Leond. D..4. Iloisington. H.F. Cltz, J.A.
Deutsch and A.E. Melchinger. 1006. Q'I'L, mapping in tropical maize I.
Genomic regions affecting leaf feeding resistance to sugarcane borer and
other traits. Crop Sci. 36: 1352-1361.

Bohn, M., M.M. Khairallah, C. Jiang, D. Gonzalez de Leon. D.A. Hoisington. H.F. Utz.
J.4. Deutsch. I1.C. Jewcl. J.A. Mihnl and A.E. Melchinger. 1997. Q'TL
mapping in tropical maize: 11. Comparision of genomic regions for
resistance to Llicitrctrc spp. Crop Sci. 37: 1892-1902.

*Bohn. M., S. Groh. M.M. Khairallah. D.A. Hoisington. 11. Futz and A.E. Melchinger.
2001. Re-evaluation of the prospects of marker assisted selection for
improving insect resistance against Diatruea ,spp. In tropical maize by
cross validation and independent validation. Theor. Appl. Genet. 103:
1059-1067.

Boivin, K., M. Deu, J.F. Rami, G. Trouche and P. Hamon. 1999. Towards a saturated
sorghum map using RFLP and AFL,P markers. Theor. Appl. Genet. 98:
320-328.
*Boom, K.S., R.A. Frederiksen and C.W. Magill. 1998. DNA-based markers for a
recessive gene conferring anthracnose resistance in sorghum. Crop Sci. 38:
1708-1 709.

*Boom. K.S., R.A. Frederiksen and C.W. Magill. 1999. A molecular marker that
segregates with sorghum leaf blight resistance in one cross is maternally
inherited in another. Mol (ien. Genet. 261 : 3 17-322.

Borad, P.K. and V.P. Mittal. 1983. Assessment of losses caused hq pest complex on
sorghum hybrids CSH-5. Pagcs 271-288 in Proceedings of the National
Seminar on Crop Losses Due to Insect Pests, 7-9 January 1983,
Hyderabad. A.P.. India (Krishnamurthy Rao. B.II. and Murty. K.S.R.K.,
eds.). Indian Journal of Entomology Vol. 11 (Special Issue). Hyderabad.
Andhra Pradesh. India: Entonlological Society of India.

Borikar. S.T. and P.K. Chopde. 1980. Inheritance of shoot fly resistance under threc
levels of infestation. Mqdica. 25: 175-183.

Borikar. S.T. and I'.K. Chopde. 1981a. Inheritance of shoot fly rcsistance in sorghum. J.
Maharashtra agric. Univ. 6: 47-18.

Borikar. S. t'. and P.R. Chopde. 1081b. Shoot fly resistance in sorghum. Indian .I. Genct.
41: 191-199.

Borikar. S.T. and P.R. Chopde. 1982. Genetics of rcsistance to sorghum shoot fly. Z.
Pflanzenzuchtg. 88: 230-224,

Borikar. S.T. and R.D. Chundunvar. 1989. Genetic analysis for trichome density in
Sorghum hicolor (I,.) Moench. J. Maharashtra Agric. llniv. 14: 304-305.

Borikar. S.T.. R.D. Chunduruar and P.R. Chopde. 1982a. Note on genetic variability for
traits related with primary and recovery resistance to shoot fly in sorghum.
Indian J. agric. Sci. 52: 867-869.

Borikar, S.T., R.D. Chundurmrar and P.R. Chopde. 1982b. Cienctic variability for shoot fly
resistance in sorghun~.J. Maharashtra Agric. Univ. 7: 223-225.

Bothe, N.N. and R.N. Pokharkar. 1985. Role of silica content in sorghum for rcaction to
shoot fly. J. Maharashtra Agric. Univ. 10: 338-339.
Bovin, K., M. Deu, J.F. Rami, G. Trouche and P. Hamon. 1999. Towards a saturated
sorghum map using RFLP and AFLP markers. Theor. Appl. Genet. 98:
320-328.
Bowers, J.E.. C. Abbey. S. Anderson. C. Chang. X. Draye, A.H. Hoppe. R. Jessup. C.
Lemke, J. Lennington, Z. Li. Y. Lin, S.Liu. L. Luo. B.S. Marler, R. Ming,
S.E. Mitchell. D. Qiang, K. Reischmann. S.R. Sc1iulr.e. D.N. Skinner. Y.
Wang. S. Kresovich, K.F. Schertz and A.H. I'aterson. 2003. A high-
density genetic recombination map of sequence-tagged sites for Sorghum,
as a framework for comparative structural and evolutionary genomics of
tropical grains and gasses. Genetics. 165: 367-386.
*Bowers, J.E.. K.F. Schertz. C. Abbey, S. Anderson. C. Chang, I..M. Chittenden, X.
Draye. A.H. Hoppe, R. Jessup. .I.Ixnnington. Y. Lin. S. Liu. K. Ming, D.
Qiang, K. Reischmann, N. Skinner, Y. Wang and A.Ii. Patcrson. 2000. A
high-density 2,399-locus genetic map of sorgliu~ii.1:inal abstracts guide.
Plant and Animal Genome VIII Conference. Jan 9-12, 2000. Sali Diego.
Califon~ia,http:/luwu.i1itl-pag.org/pagiX!abstracts/pa&712.htn1l.
Bredenieijer Gmm. P. Arens, D. Wruters. I). Visser and B. Vosman. 1998. 'lhe use of
semi-automated flurescent microsatellite analysis for tomato cultivar
identification. Theor. Appl. Genct. 97: 584-590.
Brown. S.M.. M.S. Hopkins. S.E. Mitchell. M.L. Senior. 'f.Y. \hlang. R.R. Duncan. F.
Gonzalez Candelas and S. Kreso~ich.1996. Multiple methods for the
identification of simple sequence repeats (SSRs) in sorgl~um(Sorghum
hicolor (L.) Moench). Theor. Appl. Genet. 93: 190-198.

Bubeck, D.M.. M.M. Goodman. \'.I). Beavis and D. Grant. 1993. Quantitative trait loci
controlling resistance to gray leaf spot in maize. Crop Sci. 33: 838-847.

Bunyamin. T.. E. 'Thomas. M. Chales and K. Perrpauls. 2003. Marker-assisted selection

for complex trait in common bean using Q'l'l.bascd index. Euphytica. 130:
423-432.

Burr, B., F.A. Burr, K.II. Thomson, M.C. Albertson and C.W. Stubbcr. 1988. Gene
mapping with recombinant inbreds in maizc. Genetics. 118: 518-526.

*Bushara, A. 1972. The world literature on the central shoot fly. Ahlerigontr \nricrvcrr
soccata Rond. with some preliminary contribution to the study of its
biology. B. Sc. Dissertation University of Khartoum. Sudan.

*Callow, J.A. 1997. In Biotechnology and Plant Genitics Resources Conservation and use
(eds. Callow. J. A. and Newbury. H.J.).
Cardinal, A.J., M. Lee, N. Sharpova, L. Wendy, Woodman Clikeman and M.J. Long.
2001. Genetic mapping and analysis of quantitative trait loci for
resistance to stalk tunneling by the European corn borer in maize. Crop
Sci. 41: 835-845.

Cardinal, A.J. and M. Lee. 2005. Genetic relationships between resistance to stalk
tunneling by the European corn borer and cell wall components in
maize populations. Theor. Appl. Genet. 11 1 : 1-7.

Casa, A.M., S.E. Mitchell, M.T. Hambin, H. Sun, J.E. Bowers, A.H. Paterson, C.F.
Aguadro and S. Kerswich. 2005. Diversity and selection in sorghum:
simultaneous analysis using simple sequence repeat. Theor. Appl.
Genet. 11 1: 23-30.

Charmet G, N. Robert, M.R. Perretant, G. Gay, P. Sourdilla, C. Groos, S. Bernard and

M. Bernard. 1999. Marker-assisted recurrent selection for cumulating
additive and interactive QTLs in recombinant inbred lines. Theor.
Appl. Genet. 99: 1 143-1 148.

Chens, S., X.H. Lin, C.G. Xu and Q. Zhang 2000. lmrpovement of bacterial blight
resistance of 'minghui 63' an elite restorer line of hybrid rice by
molecular marke-assisted selection. Crop Sci. 40: 239-244.

Chens, S., C.G. Xu, X.H. Lin and Q. Zhang. 2001. lm~roving bacterial blight
resistance of '6078' an elite restorer line of hybrid rice by molecular
marker-assisted selection. Plant Breed. 120: 133-137.

'Childs, K.L., F.R. Miller, M.M. Cordonnier Pratt, L.H. Pratt, P.W. Morgan and J.E.
Mullet. 1997. The sorghum photoperiod sensitivity gene Mu3 encodes
a phytochrome BI. Plant Physiol. 113: 61 1-619.

Chittenden, L.M., K.F. Schertz, Y.R. Lin, R.A. Wing and A.H. Paterson. 1994. A
detailed RFLP map of Sorghum bicolor X S. propinquurn, suitable for
high-density mapping, suggests ancestral duplication of Sorghum
chromosomes or chromosomal segments. Theor. Appl. Genet.
87: 925-933.

Cho, Y.G., M.C. Eummy, S.R. Couch and Y.A. Chae. 1994. The semi dwarf gene
sd-l of rice (Ovza sari:^,^ L.) 11. Molecular mapping and marker-
assisted selection. Theor. Appl. Genet. 89: 54-59.
*Chowdari, K.V., A.P. Daviewala V.S. Gupta. P.K. Ranjekar and O.P. Govila. 1998.
Genotype identification and assessment of genetic relationship in
pearlmillet using micro satellites and RAPIDS. Theor. Appl. Genet. 97:
154-162.

Chunduwar, R.D. and S.T. Borikar. 1983. Stahility analysis for shoot fly resistance in
sorghum. Paper presented at the National Seminar on Breeding to Pests
and Diseases. May 25-27 1983. School of Genetics. TNAU. C'oimhatore,
641 003, India, pp.13-15.

Chunduwar. R.D.. R.R. Karanjkar and S.T. Borikar. 1992. Stabilib of shootfly resistance
in sorghum. J. Maharashtra Agric. Univ. 17: 380-383.

*Clegg. M.T. 1991. Molecular evaluation of plant genetic resources. Stadler Genet.
Symp. Ser. 20: 67-86.

*Coulibaly. S. 2002. Marker-assisted selection for stay-green in ,Yor,qhum hico/or L.

Moench. Ph.D. Dissertation submitted to Texas Tech Universily. Lubbock.
TX.

Crasta, O.R.. W.W. Xu. D.T. Rosenow, J.E. Mullet and H.T. Nguyen. 1999. Mapping of
post-flowering drought rcsistance traits in sorghum: association between
QTLs influencing premature senescence and maturity. Mol. (;en. Genet.
262: 579-588.

Cregan, I'.B.. J. Mudge, E.W. Fickus. L.F. Marck, D. 1)anesh. R. Ilenny. R.C.
Shoemaker. R.E. Mathe~vs.'T. Jarvik and N.D. Young. 1999. Targeted
isolation of simple sequence repeat markers through the use of bacterial
artificial chromosomes. Theor. Appl. Genet. 98: 919-928.

Cui. Y.X., G.W. Xu. C.W. Magill and K.F. Scherz. 1995. RFLP- based assay of sorghum
bicolor (L.) Moench. Genetic diversity. Theor. Appl. Genet. 90: 787-796.

Dabholkar. A.R.. G.S. Lal. R.C. Mishra and N.B. Hrache. 1989. Combining abil~ty
analysis of resistance of sorghum to shoot fly. Theor. Appl. Genet. 49:
325-330.
Dahlberg. J.A. 2000. Classification and characterization of sorghum In: Sorghum. Origin,
History, Technology and Production (Smith C.W. and Frederikasen R. E
(Eds)) Willey series in Crop Sci. John Wiley and Sons. Inc. New York 99-
130.

Dalvi. C.S., V.P. Dalaya and V.G. Khanvilkar. 1990. Screening some sorghum varieties
for resistance to shoot fly Aihcr~gonusocctriu (Kondani). Indian J. Ent. 52:
279-289.

*Dallas. J.F. 1988. Detection of DNA "fingerprints" of cultivates rice by hybridization

with a hunlan minisatellite DNA probe. I'roc. Nat. Acad. Sci. USA. 85:
6831-6835

*David, H. and S.E. Moorthy. 1988. Physical resistance mechanisms in insect plant
interaction. In: Dqanamics of insect-plant interaction: recent advances and
future vends (Ananthkrishna T. N.. and Karllan A. cds) Oxford IBH
I'ublshing Co. New Oelhi India. pp 45-70.

Dean. R.E.,J.A. Dahlberg. M.S. FIopkins, S.E. Mitchell and S.Kreso\ich. 1990. Genetic
redundancy and diversity among "Orange' accessions in the U.S. national
sorghum collection as assessed with simple sequence repeat ISSR)
markers. Crop Sci. 39: 1215-1 22 1.

*Delobel. A.G.L. 1982. Effect of sorghum density on oviposition and survival of the
sorghum shoot fly. Entomol. Exp. Applic. 31: 170-174.

*de Wet. J.M.J. and Harlan, J.R. 1971. Thc origin and domestication of S o r g k ~ mhicolor.
~
Econ. Bot. 25: 128-135.

*de Wet. J.M.J.. Harlan J.R. and E.G. Price. 1976. Variability inSor~humhicolor. In:
Origins of African plant domestication. Ilarlan, J.R.. J.M.J. de .Wet and
A.B.C. Stemler (eds). The Mountain press. The Hague. 453-463,

Deu. M.P.J. Hamon. P. Chantereau. A.U. Dufour. A. D'Hont and C. Lanaud. 1995.
Mitochondria1 DNA diversity in wild and cultivated sorghum. Genome.
38: 635-645.
Deu, M., D. Gonzalez de Leon, J.C. Glasmann, I. Degreniont. J. Chantereau. C. Lanaud,
and P. Hamon. 1994. RFLP diversity in cultivated sorghum in relation to
racial differentiation. Theor. Appl. Genet. 88: 838-844.

Devicente, M.C. and S.D. Tanksley. 1993. QTL analysis of transgressive segregation in
an interspecific cross. Genetics. 134: 585-596.

Deshpande S.P. 2005. QTL analysis for shoot fl) resistance in sorghum. Ph. D. Thesis
submitted to Marathwada Agricultural Ilniversitp, I'arbliani (Maharashtra.
India).

*Dhams, R.G. 1943. Insect resistance in sorghum and cotton. Amcr. Soc. Agron J. 35:
704-71 5.

Dhauan. P.K.. S.P. Singh. A.N. Verma and 1). R . Arya. 1993. Antibiosis mechanism of
resistance to shoot fly. Alherigonu soccarrr (Rondani) in sorghum. Crop
Res. 6: 306-310.

Dhillion. M.K. 2004. Effccts of cytoplasmic male-sterility on expression of resistance to

sorghum shoot fly. Ph. D. Thesis submitted to CCS Haryana Agricultural
University, Hissar India.

fliers. B.W.. P. Keim. W.R. Fehr and R.C. Shoemaker. 1992. RFLP analysis of soybean
seed protein and oil contcnt. Theor. Appl. Genet. 83: 608-612.

Dje. Y., M. Heuertz, C. Lefebvre and X.Vekmans. 2000. Assessment of genetic diversity
uithin and among germplasm accessions in cultivated sorghum using
micro satellite markers. Theor. Appl. Genet. 100: 918-925.

D,je. Y..D. Forcioli. M. Ater, C. LeTkbvre and X.Vekemans. 1999. Assessing population
genetic structure of sorghum landraces from North-western Morocco using
allozyme and microsatellite markers. Theor. Appl. Genet. 99: 157-163.

*Doggett, H. 1972. Breeding for resistance to sorghum shoot fly in Uganda. In: Control
of Sorghum Shoot Fly. (Eds. Jotwani. M.G. and W.R. Young). Oxford and
IBH Publishing Co., New Delhi. pp. 1925-201.

Doggett. H. 1988. Sorghum. Second edition. Longnian. London. 688 pp.

Doggett, H.,K.J. Starks and S.A. Eberhart. 1970. Breeding for resistance to the sorghum
shoot fly. Crop Sci. 10: 528-53 I.

Dreher, K., M. Khairallah. J. Ribaut and M. Morris. 2003. Money matters (I): Cost of
field and laboratory procedures associated \+ith conventional and marker-
assisted maize breeding at CIMMYT'. Mol. Breed. 11: 221-234.

Dudley, J.W. 1993. Molecular markers in plant improvenlent: Manipulation of genes

affecting Quantitative traits. Crop. Sci. 33: 660-668.

*Duffey, S.S. 1986. Plant glandular trichomes: their partial role in defence against
insects: In: Insects and the plant surface (Juniper, B. E.. and Southwood. T.
R. E.eds) Edward Arnolds Publisher Ltd. London U K . Pp 15 1-172.

Dufour. P.. M. Deu. L. Grivet. A. D'Hont. F. Paulel. A. Bouet. C. Lanaud. J.C.

Glaszmann and 1'. Hamon. 1997. Construction of a composite sorghum
map and comparison with sugarcane. a related complex polyploid. fheor.
Appl. Genet. 93: 409-41 8.

Eathington, S.K., J.W. Dudley and G.K. KeSener 11. 1007. Usefulness of marker QTL
associations in early generation selection. Crop Sci. 37: 1686-1693

*Edwadrs., M.D., and L. Johnson. 1994. RFI.Ps for rccurrent selection, In: Proc. Joint
plant Breeding. Synlp. Series of CSSA and ASHA Conallis OR. pp 33-40.
Am. Soc. Hort. Sci. Alexandria. VA.

Edwards. M.D.. C.W. Stubcr and J.F. Wendfel. 1987. Molecular-marker ljcilitated
investigations of quantitative-trait loci in maize. I. Numbers, genomic
distribution and types of gene action. Genetics, I 16: 113-125.

Elhadawi. A.A., A.M. Nour and R.S. I'athak. 1997. Combining ability of resistance to the
sorghum shoot tly. Aiherigona soccafo and stem borer C'hiio purrellus
grain sorghum, ,%~rghurnhicolor. Insect Sci. Applic. 17: 329-333.

Eshed. Y.. and D. Zamir. 1994. lntrogression fro111 Lycoper.\icoft pe~melliican improve
the soluble solids yield of tomato harvest. Theor. Appl. Genet. 88: 891-
897.
Everson, E.H. and C.W. Schaller. 1955. The genetics of yield differences associated with
awn bearing in barley hybrid (Lion X ~ t l a s " ) X Atlas. Agron. J. 47: 276-
280.

Falconer, D.S. 1989. Introduction to quantitative genetics jrdEdn. Ronald Press Co.. Net%
York, pp.187-193.

FAOSTAT. 2004. http://apps.Pdo.org/default.htm

FAS. 2005. Iittp:llfas.usda.uo\~/FAS Online. United States Dcpartnicnt of Agriculture.

Foreign Agricultural Service.

*Fatocun, C.A., D.1. Menancio llautea. D. Danesh and N.D. Young. 1902. Evidence for
orthologous seed weight genes in cowpea and niung bean based on RFLP
mapping. Genetics, 132: 841-846.

*Fischer. R.A. 1918. The correlations butwen relatives on the supposition of Mendellian
inheritance. Transactions and Proceedings. Hotanlcal Society of
Edinburgh. 52399-4.33.

Fisch. M.. M. Bohn and A.E. Melchingcr. 1900. Comparison of selection strategies for
marker-assisted back crossing of a gene. Crop Sci. 39: 1205-1301.

Folkertsnia. R.T.. G.M. Sajjanar, S. Chandra. D.V.S. Reddy and C.T. Hash. 2005.
Mapping quantitalive trait loci associated with sorghum shoot tly
(..lihcrigonu socculu Rondai) resistance in sorghum [Sor~lrtrmhicolor (L.)
Moench]. ICRISA'I' unpublished.

*Franks, C.D. 2003. The efficacy of marker aasisted selection for grain mold resistance in
sorghum. Ph.D. Dissertation, Texas A & M Ilniv., College Station. 'TX.

'Frederiksen. R.A. and F. Miller. 1972. Proposal for release and increase: A'I's622.
BTx622, ATx623, B'fx623. A'rx624, B'rx624. TAES From 96-72: Seed
Release Committee of the 'Texas Agricultural Experinlent Station, College
Station.

Frisch M. M. Bohn and A.E. Melchinger.1999 Comparison Selection Strategies for

Marker-Assisted Backcrossing of a Gene. Crop Sci. 39: 1295-1301.

Frisch, M., M. Bohn and A.E. Melchinger. 1999. MinimumSampleSize and Optinial
Positioning of Flanking Markers in Marker-Assisted backcrossing for
Transfer of a Target Gene. Crop Sci. 39: 967-975
'Gale, M.D. and J.R. Witcombe. 1992. DNA markers and marker-mediated
applications in plant breeding with particular reference to pearl millet
breeding. In: Biotechnology and crop improvement in Asia. (Ed. Moss, J.P.),
International Crops Research Institute for the Semi-Arid Tropics, Patancheru,
Andhra Pradesh, India, pp. 323-332.

Galtan, P.T. 1889. Natural inheritance, Macmillan. London

Gebrekidan, B. 1982. Utilization of Germplasm in Sorghum Improvement, In: House,

L.R., L.K. Mughogho and J.M. Peacock (eds). Proceedings of the
International Symposium on Sorgltum, 2-7 November, 198 1 ICRISAT,
Patancheru, A.P India, I , 335-345, 1982.

Ghareyazie, B., N. Huang, G. Second, J. Bennett and G.S. Khush. 1995. Classification
of rice germplasm I. Analysis using ALP and PCR-based RFLP. Theor. Appl.
Genet. 91 : 21 8-227.

Ghebru, B., R.J. Schmidt and J.L. Bennetzen. 2002. Genetic diversity of eritrean
sorghum land races assessed with simple sequence repeats (SSR) markers.
Theor. Appl. Genet. 105: 229-236.

Gibson, P.T. and R.K. Maiti. 1983. Trichomes in segregating generations of sorghum
matings. I. Inheritance of presence and density. Crop Sci. 23: 73-75.

Gimelfarb, A. and R. Lande. 1994a. Simulation of marker-assisted selection in hybrid

populations. Genet. Res. 63: 39-47.

Gimelfarb, A. and R. Lande. 1994b. Simulation of marker-assisted selection for non-

additive traits. Genet. Res. 64: 127-136.

Gimelfarb, A. and R. Lande. 1995. Marker-assisted selection and marker-QTL

association in hybrid population. Theor. Appl. Genet. 91: 522-528.

Godshalk, E.B., M. Lee and K.R. Lamkey. 1990. Relationship of restriction fragment
length polymorphism to single-cross hybrid performance of Maize. Theor.
Appl. Genet. 80: 423-432.

Goud, J.V., K.H', Anahosur and K.A. Kulkarni. 1983. Breeding for multiple resistance
in sorghum. Proceedings of National Seminar on Breeding Crop Plants for
Resistance to Pests and Diseases, 25-27 May 1983. Tamilnadu Agricultural
University, Coimbatore. pp. 3.
Gowda, P.S.. G.W. Xu. R.A. Frederiksen and C.W. Magill. 1995. DNA markers for
downy mildew resistance genes in sorghum. Genome. 38: 823-826.

Graham, G.I.. D. Wolff and C.W. Stuber.1997 Characterization of a yield quantitative

locus on chromosome 5 of maize by fine mapping. Crop Sci. 37: 1601-
1610

Grenier, C., M. Deu. S. Kresovich. I'.J. Bramel-Cox and P. Hamon. 2000. Assessment of
genetic diversity in three subsets constituted from the ICRISAT sorghum
collection using randoni versus non-random sampling procedures using
molecular marker. Theor. Appl. Genet. 101 : 197-202.

Groh. S., D. Gonzaler. de L.eon. M.M. Khairallali, D.C. Jiang, D Bergvinson. M. 13ohn.
D.A. Hoisinbeon and A.E. Melchinper. 1998. QTL mapping in tropical
maize: 111. Genomic regions for resistance to Ilzc~lrerrspp. and associated
traits in two RIL populations. Crop Sci, 38: 1062-1072.

Gupta. M..Y.S. Chgi. J.R. Scverson and J.L. Owen. 1994. Amplification of DNA
markers from evolutionarily diverse genomes using single primers of
simple-sequence repcats. Theor. Appl. Genet. 89: 998-1006,

Ilalalli. M.S.. B.T.S. Gowda. K.A. Kulkami and J V. Goud. 1082. Inheritance of
resistance to shoot fly (Arhcrigona soccu~crRond.) in sorghum I.Sor,yhtlm
hrcolor (L.) Moench]. SABRA0 J. 14: 165-170.

Halalli. MS.. B.T.S. Gowda. K.A. Kulkarni and J.V. Goud. 1983. [:valuation of advanced
generations progenies for resistance to shoot fl) in sorghum. Indian J.
Genet. 43: 291-293.

Haldane. J.B.S. 1919. The combination of linkage values and the calculation of distance
between the loci of linked factors. J. Genet. 8: 299-309

I-laley, C.S, and S.A. Knott. 1992. A simple regression method for mapping quantitative
trait loci in the line crosses using flanking markers. Ilereditl. 69: 3 15-324.

*Han, F., I. Romagosa. S.E. Ullrich. B.1,. Jones. P.M. Hayes and D.M. Wesenberg. 1097.
Molecular marker assisted selection for the malt quality traits in barley.
Mol. Breed. 3: 427-437

Harlan, J.R. and J.M.J. Dewet. 1972. A simplified classification of cultivated sorghum.
Crop Sci. 12: 172-176.
'Harlan, J.R. 1989. Plant Gentic Resources : A conversation I ipreative (eds Yetman,
C.W. Kafton, D.and Wilkas, G.) pp. 1 1 1-129
*Hamood, R.R., Y.R. Granados, S. Jamornman and R.G. Granados. 1972. Breeding
for the sorghum shoot fly resistance in Thailand. In: Control of
Sorghum Shoot Fly. (Ed. Jotwani, M.G. and W.R. Young), Oxford and
IBH, New Delhi, pp. 208-217.
Hash, C.T. and P. Bramel-Cox. 2000. Survey of molecular marker applications. In
Application of Molecular Markers in Plant Breeding. (Ed. Haussmann,
B.I.G., H.H. Geiger, D.E.Hess, C.T. Hash and P. Bramel-Cox).
Training manual for seminar held at IITA, Ibadan, Nigeria, 16-17
August 1999. International Crops Research Institute for the Semi-Arid
Tropics, Patancheru, India, pp. 3-12.
Hash, C.T., R.T. Folkertsma, P. Ramu, B.V.S. Reddy, V. Mahalakshmi, H.C. Sharma,
H.F.W. Ranunde, E.R. Weltzien, B.I.G. Haussmann, M.E. Ferguson
and J.H. Crouch. 2003 Marker assisted breeding across lCRlSAT for
terminal drought tolerence and resistance to Shoot fly and S~rigain
sorghum. Poster presentation in 'In the wake of the double helix: From
the Green Revolution to Gene Revolution'. International symposium at
Bologna, ITALY, May 27-3 1
Haussmann, B.I.G., D.E.Hess, N. Seetharama, H.G. Welz and H.H. Geiger. 2002.
Construction of a combined sorghum linkage map from two
recombinant inbred populations using AFLP, SSR, RFLP and RAPD
markers and comparison with other sorghum maps. Theor. Appl.
Genet. 105: 629-637.
Haussmann, B.I.G., V. Mahalakshmi, B.V.S. Reddy, N. Seetharama, C.T. Hash and
H.H. Geiger. 2003. QTL mapping of stay-green in two sorghum
recombinant inbred populations. Theor. Appl. Genet. 106: 133-142.

Haussmann, B.I.G., D.E. Hess, (3.0. Omanya, R.T. FOIL-rtsma, B.V.S. Reddy,
M. Kayentao, H.G.Weiz and H.H. Geiger. 2004. Genomic regions
influencing resistance to the parasitic wed srriga hemonlhica in two
recombinant inbred populations of sorghum. Theor. Appl. Genet. 109:
1005-1016.

'Hearne, C.M., S. Ghosh and J.A. Todd. 1992. Microsatellites for linkage analysis of
genetic traits. Trends Genet. 8: 288-294.
Hemamalini, G.S., H.E. Shashidhar and S. Hittalamani. 2000. Molecular marker
assisted tagging o f morphological and physiclogical traits under two
contrasting moisture regimes at peak vegetative stage in rice (Olyza
saliva L.). Euphytica, 1 12: 69-78.

Hittalmani, S., A. Parco, T.V. Mew, R.S. Zeigler and N. Haung. 2000. Fine mapping
and D N A marker -assisted pyramiding o f the three major genes for
blast resistance in rice. Theor. Appl. Genet. 100: 1121-1 128.

Hittalmani, S., M.R. Foolad, T. Mew, R.L. Rodriguez and N. Huang. 1995.
Development of PCR-based marker to identify rice blast resistance
gene PI-2(e) in a segregating population. Theor. Appl. Genet. 91 : 9-14.

Hospital, F., L. Moreau, F. Lacoudre, A. Charcosset and A. Gallais. 1997. More on

the efficiency o f marker-assisted selection. Theor. Appl. Genet. 95:
1181-1189.

Hospital, F., C. Chevalet and P. Mulsant. 1992. Using markers in gene introgression
breeding programs. Genetics. 132: 1 199-1220.

Hospital, F. and A . Charcosset 1997. Marker Assisted Introgression o f quantitative

trait loci. Genetics. 147: 1469-1485.

'Huang, Z., G. He, L. Shu, X. L i and Q. Zhang. 2001. Identification and mapping two
brown planthopper resistance genes in rice. Theor. Appl. Genet. 102:
929-934.

Huang, N., E.R. Angeles, J. Domingo, G . Magpantay, S. Singh, G. Zhang, N .

Kumarvadivel, J. Bennett and G.S. Khush. 1997. Pyramiding o f
bacterial blight resistance genes in rice: marker-assisted selection using
RFLP and PCR. Theor. Appl. Genet. 95: 3 13-320.

*Hulbert, S.H., T.E. Richter, J.D. Axtell and J.L. Bennetzen. 1990. Genetic mapping
and characterization o f sorghum and related crops by means o f maize
D N A probes. Proc. Nat. Acad. Sci. USA, 87: 4. 51-4255.

International Crops Research Institute for the Semi-Arid Tropics (ICRISAT). 1991.
ICRISAT Annual Report 1990, Patancheru, A.P., India.

International Crops Research Institute for the Semi-Arid Tropics (ICRISAT). 1992.
ICRISAT Annual Report 1991, Patancheru, A.P., India.
International Crops Research Institute for the Semi-Arid Tropics (ICRISAT). 1996. The
world sorghum and millet economies: Facts. trends and outlook. Rome,
Italy: Food and Agricultural Organization of the United Nations (FAO),
and Patancheru. A.P. 502 324. India: lnternational Crops Research
Institute for the Semi-arid 'I ropics (ICRISAT). 72 pp.

International Crops Research Institute for the Semi-Arid Tropics (ICRISAT). 1009.
Progress report. Entomologj. Genetic resources and enhancement
program. 98: 1998-99. International Crops Research I~lstitutcfor the Semi-
Arid Tropics Patancehru. 502 324. AP.. India.

Jadhav, S.S., U.N.Mote and D.R. Bapat.1986. Biophysical plant characters contributing
to shoot fly resistance. Sorghum Newslett. 29: 70.

Jadhav, S.S.. I7.N. Mote and D.R. Bapat. 1988. I'erfonnance of sLceet stalked sorghum
genotypes to shoot fly reaction. 1. Maharashtra Agric. Univ. 13: 232-234.

Jain. K.K. and M.P. Bhatnagar. 1962. Studies on varietal resistance to jowar shoot fly.
Indian 1. Genet. 22: 224-229.

Jampatong. C., M.1). McMullcn. R. [lean Barry. L.L. 1)arrah. P.T. Byrne and H. Kross.
2002. Quantitative trait loci for first- and second- generation European
corn borer resistance dcrivcd from the niaiz,e inbred Mo47. Crop Sci. 42:
584-593.

Jansen, R.C. and 1'. Stam. 1994. High-resolution quantitative trails into nlultiple loci via
interbal mapping. Genetics. 136: 1447-1455.

Jayanthi, P.D.K.. B.V.S. Reddy. T.H. Ciour and D.D.R. Reddy. 1999. Genetics of glossy
and trichome characters in sorghum hybrids of cytoplasmic male sterilc
lines. J. Maharashtra Agric. Univ. 24: 251-256.

Jinks. J.L. and H.S. Pooni. 1976. Predicting the properties of the recombinant inbred lines
derived by single seed descent. Heredity. 36: 253-266.

Jinks. J.L. and H.S. Pooni. 1981. Comparative results of selection in the early and late
stages of an inbreeding programme. Heredity, 39: 345-355.
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 glass house screening detect the same QTL. Crop Sci. 42:
1316-1323.

Jordan. D.R., Y. Tao, 1.D. Godwin, R.G. Henzell, M. Cooper and C.L. Mcinture. 2003. A
prediction of hybrid performance in grain sorghum using RFLP markers.
Theor. Appl. Genet. 106: 559-567.

*Jotwani, M.G. 1978. Investigations on Insect Pests of Sorghum and Millcts with special
Reference to Host Plant Resistance. Final Technical Rcport (1972-1975).
Res. Bull. Div. Entomol. lndian Agricultural Kesearch Institute, Ncw
Delhi. India, 114 pp.

Jotwani. M.G. and J.C. Davies. 1980. lnsecl resistance studies in sorghum at I~iternational
Institutes and National Programmes with special reference to India. In:
Proceedings of the Short Course in Host Plant Resistance ( t d . Harris M.
K.). Texas A & M llniversity. College Station. 'Texas, pp. 214-236.

Sotwani. M.G.. K.K. Marwaha. K.M. Srivastava and W.R. Young. 1970. Seasonal
incidence of shoot fly (Alheri~ono.socci~~o
Kond.) in jowar hyhrids at
Delhi. lndian 1. Ent. 32: 7-15.

*Jotwani, M.G., G.C Shamia. B.G. Srivastava and K.K. Maruaha. 1071. Ovipositional
response of shoot fly At/?erigonu curicr .soccrr/ci Kondani on some
promising resistant lines of sorghum. In: Investigations on insect pests of
sorghum and millets. (Ed. Pradhan. S.). Final technical report. 1965-1970.
lndian Agricultural Research Institute. New Delhi, pp. 1 1 9-12?,

Jotwani. M.G. and K.P. Srivastava. 1970. Studies on sorghum lines resistant against shoot
fly Alherigona vuricr socculc~Rond. lndian J . Ent. 32: 7-15.

Jotwani, M.G., W.R. Young and G.I.. Teetes 1980. Elements of integrated control of
sorghum pests. FA0 I'lant Production and Protection Paper. FAO. Kame,
Italy, 159 pp.
*Jotwani, M.G. and J.C. Devies. 1979. Insect resistance studies on sorghum at
International Institutes and National programs with special reference to
India. pp 224-236. In: Biology and Breeding for Resistance to Arthopodas
and Pathogen in Agricultural Plants. (Harris, M. K. Eds).. College station.
Texas, USA. Texas. A and M University.
Kamala. V., P.J. Barmal-Cox, S. Sivaramkrishnan. S. Chandra. Seeta Kannan. S.
Harikrishna and D. Manohar Rao. 2005. Genetic and phenotypic dibersity
in downy mildew resistant sorghuni gemiplasm (In Press).

Kamatar, M.Y. and P.M. Salimath. 2003. Morphological traits of sorghum associated
with resistance to shoot fly Aiherigonu .socctr/u Rondani. Indian J. Plant
Prot. 3 1: 73-77.

Karanjkar. R.R., R.D. Chundurwar and S.T. Borikar. 1942. Correlations and path analysis
of shoot f l y resistance in sorghunl. J. Maharashtra Agric. Univ. 17. 789-
391.

*Katsar. C.S., A.H. Paterson. (i.I>. Tectcs and G.C. Peterson. 2002. Molecular analysis of
sorghum resistance to the greenbug (I lon~optera:Aphididae). J. Econ. Fnt.
95: 448-457.

Kearscy. M.J. and A.G.L. Farquhar. 1098. QTL analysis in plants: where are we no\v?
IHcredity, 80: 137-142.

Keim. P.. B.W. Diers. T.C. Olson and R.C. Shoemaker. 1990. RFLP mapping in soybean:
Association between marker loci and variation in quantitative traits.
Genetics. 126: 735-742.

*Kendall. M.G. and A. Stuart. 1961. The Advance Theory of Statistics. Vol. 11.. 3rd Ed"..
Charles Griffin 8: Co.. London.

Khurana. A.D. and A.N. Verma. 1982. Amino acid conrents in sorghum plants. resistant I
susceptible to stem borer and shoot fly. Indian J. Ent. 44: 184-188.

Khurana. A.D. and A.N. Verma. 1983. Some biochemical plant characters in relation to
susceptibility of sorghunl to stem borer and shoot fly. Indian J. Ent. 45: 29-
37.
Khurana, A.D. and A.N. Verma. 1985. Some physical plant characters in relation to
stem borer and shoot fly resistance in sorghum. Indian J. Ent. 47: 14-
19.
Khush, G.S. and D.S. Brar. 1991. Genetics o f resistance to insects in crop plants. Adv.
Agronomy. 45: 223-274.
Kim, J.S., P.E. Klein, R.R. Klein, H.J. Price, J.E. Mullet and D.M. Stelly 2005.
Chromosome identification and nomenclature o f Sorghum brcolor.
Genetics. 169: 1169-1 173.

Kishore, P., B.S. Rana and K.N. Agrawal. 1985. Selection o f shoot fly resistance in
high yielding varieties of sorghum. J. Entomol. Res. 9: 188-194.
Kishore, P. 1986. Breeding for resistance to shoot fly and stem-borer in sorghum.
Insect Sci. Applic. 7: 117-121.
Kishore, P. 2001. Current status o f host plant resistance in sorghum in India. J. Ent.
Res. 25: 1-20.
Klein, R.R., R. Rodriguez Herrera, J.A. Schlueter, P.E. Klein, Z.H. Yu and W.L.
Rooney. 2001a. Identification o f genomic regions that affect grain-
mould incidence and other traits o f agronomic importance in sorghum.
Theor. Appl. Genet. 102: 307-319.

Klein, R.R., P.E. Klein, A.K. Chhabra, J. Dong, S. Pammi, K.L. Childs, J.E. Mullete,
W.L. Rooney and K.F. Schertz. 200Ib. Molecular mapping o f the R f l
gene for pollen fertility restoration in sorghum. Proceedings o f the 21''
Biennial Grain Sorghum Research and Utilization Conference, pp. 16.

Knapp, S.J. 1994, mapping quantitative trait loci. In: DNA based markers in plants
(R. L. Philips and Vasil I K Eds.) Kluwer Academic: Dordrecht, The
Netherlands. Pp. 58-96

Knapp, S.J. 1998. Marker-assisted selection as strategy for increasing the probability
o f selecting superior genotypes. Crop Sci. 38: 1 164-1174

*Knapp, S.J. 1989. Quasi-Mendellian analyses of quantitative traits using molecular

linkage maps. In: Science for Plant Breeding, (Eds. Phillips, R.L. and
1.K. Vasil), Kluwer Academic Publishers, Boston, Netherlands, pp.
347-355.

*Kong, L., J. Dong and G. Hart. 1997. Linkage mapping o f simple sequence repeat
loci in Sorghum bicolor (L.) Moench. In: Plant and Animal Genome V .
The International Conference on the Status o f Plant and Animal
Genome Research, 12-16 January 1997, San Diego, California, USA,
pp. 100.
Kong, L., J. Dong and G. E. Hart. 2000. Characteristics, linkage-map positions and
allelic differentiation of Sorghum bicolor (L.) Moench DNA simple
sequence repeats (SSRs). Theor. Appl. Genet. 101 : 438-448.
Kosambi, D.D. 1944. The estimation of map distances from recombination values.
Ann. Eugen. 12: 172-175.
Krakowsky, M.D., M.J. Brinkman, W.L. Woodman Clikeman and M. Lee. 2002.
Genetic components of resistance to stalk tun1 *ling by European corn
borer in maize. Crop Sci. 42: 1309-1315.
Krakowsky, M.D., M. Lee, W.L. Woodman Clikeman, M.J. Long and N. Sharpova.
2004. QTL mapping of resistance to stalk tunneling by European corn
borer in RlLs of maize population B73 x De81 1. Crop Sci. 44: 274-
282.

Krishnananda, M., S. Jayaraj and T.R. Subramaniam. 1970. Resistance in sorghum to

stem fly, Atherogona varia soccata R. Madras Agric. J., 57: 674-679.
Kulakrni, N., K. Hussain Sahib and N.K. Murty. 1978. Combining ability for shoot
fly resistance in sorghum. Indian J. Genet. 38: 193-198.
Kundu, G.G., P. Kishore and M.G. Jotwani. 1977. Field evaluation of some sorghum
selections for resistance to shoot fly and stem borer. Entomon, 2: 153-
155.
Kundu, G.G. and J.K. Sharma. 1975. Field screening of some local germplasms of
sorghum from Rajasthan for resistance against shoot fly, Alherrgona
variu soccuta Rondani. Sorghum Newslett. 18: 58-59.

Lakshminarayana, K. and D.V. Subba Rao. 1978. Screening of sorghum varieties for
resistance to shoot fly. Sorghum Newslett. 18: 22-23.

Lander, E.S. and D. Botstein. 1989. Mapping Mendellian 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 package for
constructing primary genetic linkage maps of experimental and natural
populations. Genomics, I: 174-18 I .

Lande R. and R. Thompson. 1990 Efficiency of marker-assisted selection in the

improvement of quantitative traits. Genetics, 124:743-756.
Lander, E.S. and D. Botstein. 1989. Mapping Mendelian factors underlying quantitative
traits using RFLP linkage maps. Genetics, 121: 185-199

*Langeridge. P.. E. Lagudah, T. Holton. R. Appels. P. Sharps and K. Chalmers. 2001.

Trends in genetic and genome analysis in wheat- A reviev,. Australian J.
Agric. Res. 52: 1043-1077.

Lanza. L.L.B., C.L. de Souza Jr.. L.M.M. Ottoboni. M.L.C. Vieira. A.P. de Souza. 1997.
Genetic distance of inbred lines and prediction of maize single-cross
performance using RAPD markers. Theor. Appl. Genet. 94: 1023-1030.

Lawson. D.M.. C.F. Lunde and M.A. Mutschler. 1997 Marker assisted transfer of
acylsugar-mediated pcst resistance fiom thc wild tomato. Lj~copcr.sicon
prnnnellii. to the cultivated tomato. Ly~.ol)er.sicone.scu/rn/lrrn. Mol. Breed.
3: 307-317

*Lee. M. 1998. Genome. prqjects and gene pools, ncw germplasm for plant breeding.
Proc. Natl. Acad. Sci. USA. 95(5): 2001 -2004.

Levin. D.A. 1973. The role of trichomes in plant defense. Quart. Rev. Bio. 48: 3-15.

Lijavetzky. D.. M C. Martinei.. F. C'arrai and H.E. flopp. 2000. QTL analysis and
mapping of pre-harvest sprouting resistance in sorghum. Euphytica. 112:
125-135.

Lin. Y.. K.F. Schertz and A.H. Paterson. 1095. Comparative analysis of QTLs affecting
plant height and maturity across the I'oaceae, in refercnce to an intcr-
specific sorghum population. Genetics. 141 : 391 -4 1 1.

*Lindstroni, E.W. 1926. Hereditary correlation of size and color characters in tomatoes.
Research Bulletin. lo~vaAgricultural Experiment Station. 93: 99-128.

*Lindstrorn, E.W. 1931. Genet~ctests for linkage between row number genes and certain
qualitative genes in maize. Research Bulletin, Iowa State College of
Agriculture. 142: 250-288.

Mace. E.S, H.K. Buhariwalla and J.C. Crouch. 2003. A high-throughput DNA extraction
protocol for tropical molecular breeding programs. PMBR. 21 : 459a-459h.
Maiti, R.K. and F.R. Bidinger. 1979. A simple approach to identification of shoot fly
resistance in sorghum. Indian .I.
PI. Protec. 7: 135-140.

Maiti, R.K.. F.R. Bidinger. K.V.S. Reddy. P. Gibson and S.C. Davies. 1980. Nature and
occurrence of trichomes in sorghum lines with resistniice to the sorghum
shoot fly. Joint Progress report of sorghum physiologylsorghum
entoniology, vol. 3. IC'RISAT, Patancheru.

Maiti, R.K. and I'.T. Gibson. 1983. Tricholnes in segregating generations of sorghum
matings. 11. Association with shoot fly resistance. Crop Sci. 23: 76-79.

Maiti, R.K., K.E. Prasada Rao, P.S. Raju and L.R. House. 1084. The glossy trait in
sorghum: its characteristics and significance in crop improvement. Field
Crops Res. 9: 279-289.

*Manley, K.P. and R H. Cudmore. 1998. Map Maiiagcr XP. Plant and Animal (ienoiiie
VI. San Dlego. California. Januarj 1908, pp. 75.

Mangala Rai. 2002 Genetic resources and inlellectual property rights in agriculture
perspective. Indian J. Pulscs Res. IS: 1-18.

Martin. H.. J. Nienhuis. G. King and A. Schaefer. 1980. Restriction fragment length
polyniorpliisms associated with insect resistance in tomato. Science. 243:
1725-1728.

Martin, J.M.. L.E. Talbert. S.P. Lanning, K.K. Blake. 1995. Ilybrid performance in wheat
using coefficient ofparcntal diversity. Crop Sci. 35: 104-108.

Mather. K. 1938. The Measurement of Linkage in Heredity. Methuen. London

Mather, K. 1949. Biometrical Genetics. Methuen. London.

*Mehta. P.J. 2002. Anthracnose resistance genes in sorghum: characterization and

molecular markers. Ph.D. Dissertation. Texas A & M University. College
Station. TX.

*Melchinger. A.E. 1993. Use o f RFLP markers for analysis of genetic relationship anlong
breeding materials and prediction of hybrid performance. In: D. R. Buxton
et al. (ed.) International crop science I. CSSA. Madison. Wl. Pp: 621-628.
Melchinger, A.E., J. Boppenmaier, B.S. Dhillon, W.G. Pollmer, R.G. Herrmann.
1992. Genetic diversity for RFLPs in European Maize inbreds, 11.
Relation to performance o f hybrids within versus between heterotic
groups for forage traits. Thero. Appl. Genet. 84: 672-68 1.
Menendez, C.M. and A. Hall. 1995. Heritability o f carbon isotope discrimination and
correlations with earliness in cowpea. Crop Sci. 35: 673-678.

Menkir, A.P., P. Goldsbrough and G. Ejeta. 1997. RAPD based assessment o f genetic
diversity in cultivated races o f sorghum. Crop Sci. 37: 564-569

Menz, M.A., R.R. Klein, J.E. Mullet, J.A. Obert, N.C. Unruh and P.E. Klein. 2002. A
high-density genetic linkage map o f Sorghum bicolor (L.) Moench
based on 2926 AFLP@, RFLP and SSR markers. Plant Mol. Biol. 48:
483-499.

Michelmore, R. 1995. Molecular approaches to manipulation o f disease resistance

genes. Ann. Rev. Phytopa:!~. 33: 393-427.

Ming, R., S.C. Liu, J. Dasilva, W. Wilson, D. Braga, A. Vandeynze, T.F. Wenslaff,
K.K. Wu, P.H. Moore, W. Burnquist, M.E. Sorrells, J.E. lrvine and
A.H. Paterson. 1998. Detailed alignment o f Saccharaum and Sorghum
Chromosomes, comparative organization o f closely related diploid and
polyploid genomes. Genetics, 150: 1663-1682.

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 o f quantitative traits. Mol. Breed. 3: 87-
103

Moharrarnipour, H., K.S. Tsumuki and H. Yoshida. 1997. Mapping resistance to

cereal aphids in barley. Theor. Appl. Genet. 94: 592-596.

Monica, A,, R. Menz, K . Robert, C.U. Natalie, C.R. William, E.K. Patricia and J.E.
Mullet. 2004. Genetic diversity o f public inbreds o f sorghum
determined by mapped AFLP and SSR markers. Crop Sci. 44: 1236-
1244.

Morden, W.C., J. Doebley and K.F. Schertz. 1989. Allozyme variation in old world
races ofSorghum bicolor (Poaceae). Am. J. Bot. 76: 245-255.

Morden, W.C., J. Doebley and K.F. Schertz. 1990. Allozyme variation among the
spontaneous species o f sorghum section Sorghum. Theor. Appl. Genet.
80: 296-304.

Moreau L, S. Lamarie, A. Charcosset and A. Gallais. 2000 Economic Efficiency o f

One Cycle o f Marker-Assisted Selection. Crop Sci. 40: 329-337.
Morgante, M. and Olivieri. A.M. 1993. PCR-amplified microsatellites as markers in plant
genetics. Plant J. 3: 175-182.

Morris, M.. K. Dreher. J.M. Ribaut and M. Khairallah. 2003 Money matters (11) costs of
maize inbred line conversion schemes at CIMMYT using conventional and
marker-assisted selection. Mol. Breed. l I: 235-247

*Morton. N.E. 1955. Sequential tests for the detection of linkage. Am. J . tluman Genet.
7: 277-31 8.

Moser, H. and M. Lee. 1994. RFPL, variation of genealogical distance, multivariate

distance, heterotic and genetic variation in oats. Theor. Appl. Genet. 87:
947-956.

Mote. U.N. 1984. Sorghum spec~esresistant to shoot fly. Indian J. Ent. 46. 241-243

Mote, 1J.N. and D.R. Bapat. 1983. Evaluation of some sorghum genotypes for multiple
resistance to shoot fly and stem borer. .I.Maharashtra Agric. Ilniv. 8: 151-
153.

Mote. U.N.. J.R. Kada~nand D.R. Bapat. 1983a. Screening of new male-sterile lines for
shoot fly reaction. Sorghum Newslett. 26: 29.

Mote. 1J.N.. J.R. Kadam and D.R. Bapat. 1983b. Evaluation of high yielding sorghum
derivatives for resistance to shoot t1). Sorghum Ncwslert. 26: 30.

Mote, U.N., J.K. Kadam and D.R. Bapat. 1986. Antibiosis mechanism of resistance to
sorghum shoot fly. J. Maharashtra Agric. Ilniv. 11: 43-46.

Mote. U.N.. J.R. Kadam and D.R. Bapat. 1985. Kecovcry resistance to shoot fly in some
sorghum !lybrids. J. Maharashtra Agric. Ilniv. 10: 100-193.

Mote. U.N.. S.M. Shirole and D.R. Bapat. 1981. Screening of local khar~fbarieties of
sorghum for resistance to shoot fly. J Maharashtra Agric. Univ. 6: 165-
166

*Mowafi, K. 1967. Preliminary studies on the Durra shoot fl), A~hcrrgonu indicu
infucasta (Emdna). B. Sc. Dissertation. University of Khartoum. Sudan.

Murty, B.R. and V. Amnachalan. 1967. Factor analysis and genetic diversity in the genus
sorghum. Indian J. genet. 27: 123-135.
Nagaraj, N., J.C. Resse, M.R. Tuinstra. C.M. Smith, P.S. Amand, M.B. Kirkham.
K.D. Kofold, L.R. Campbell and G.E. Wilde. 2005. Molecular
mapping o f sorghum genes expressing tolerance to damage by
greenbug (Homoptera: Aphidide). J. Econ. Entomol. 98(2): 595-602.
Naik, S., K.S. Gill, V.S. Prakasrao, V.S. Gupta, S.A. Tamhankar, S. Pujar, B.S. Gill
and P.K. Ranjekar. 1998. Identification o f a STS marker linked to the
Aegilops spelroides derived leaf rust resistance gene LR28 in wheat.
Thero. Appl. Genet. 97: 535-540.

Nair, S., A. Kumar, M.N. Srivastava and M. Mohan. 1996. PCR-based DNA markers
linked to a gall midge resistance gene. GM 40, has potential for
marker-aided selection in rice. Theor. Appl. Genet. 92: 660-665.
Nair, S., J.S. Bentur, V. Prasadarao and M . Mohan. 1995a. DNA markers tightly
linked to a gall midge resistance gene (GM2) -re potentially useful for
marker aided selection in rice breeding. Theor. Appl. Genet. 91: 68-73.

Nair, S., V. Prasadarao, J. Bennett and M . Mohan. 1995b. Detection o f a highly

heterozygous locus in recombinant inbred lines o f rice and its possible
involvement in heterosis. Theor. Appl. Genet. 91 : 978-986.
Narayana, N.D. 1975. Characters contributing to sorghum shoot fly resistance.
Sorghum Newslett. 18: 2 1.

'Nelson, J.C. 1997. Qgene - Software for marker-based genomic analysis and
breeding. Mol. Breed. 3: 239-245.

Nimbalkar, V.S. and D.R. Bapat. 1987. Genetic analysis o f shoot fly resistance under
high level o f shoot fly infestation in sorghum. J. Maharashtra Agric.
Univ. 12: 331-334.

Nimbalkar, V.S. and D.R. Bapat. 1992. Inheritance o f shoot fly resistance in sorghum.
J. Maharashtra Agric. Univ. 17: 93-96.

NRCS, 1998. Vision 2020. NRCS Perspective Plan, National Research Center for
Sorghum, Rajendranagar, Hyderabad - 500 030, India, pp. 17.

*Oh, B.J., R.A. Frederiksen and C.W. Magill. 1994. Identification o f molecular
markers linked to head smut resistance gene (SHS) in sorghum by
RFLP and RAPD analysis. Phytopath. 94: 830-833.

*Oh, B.J., R.A. Frederiksen and C.W. Magill. 1996. Identification o f RFLP markers
linked to a gene for downy mildew resistance in sorghum. Can. J. Bot.
74: 31 5-3 17.
Oliveira, A.C. de. T. Richter and J.L Bennetzen. 1996. Regional and racial specificities

in sorghum germplasm assessed with DNA markers. Genome. 39: 579-87.

Omori, T., B.L. Agraual and L.R. House. 1983. Componental analysis of the factors
influencing shoot fly resistance in sorghum [Sorgh~rm hic,o/or (I..)
Moench] at he rig on^^ soccarr~].JARQ. 17: 2 15-218.

Omori, T., B.L. Agrawal and L.R. House. 1988. Genetic diversity for resistance to shoot
fly, Aiherigonci socculu Kond. in sorghum, ,Cor,ghutn hicolor (I..) Moench
and its relationship with heterosis. Insect Sci. Appl. 9: 483-488.

*Painter. R.H. 1951. Insect resistancc in plants. Macniillan. New York. LISA.

*Pakniyat. H.. W. Powell. E. Bair. L.L. Handley. D. Robinson. C.M. Scriongeous. E.

Nevo. C.A. Hackert. P.D.S. Caligari and B.P. Foster. 1997. AFLP
variation in wild barlcy (Ilordrum .s/~onitmetml)with reference to salt
tolerance and associated cco-geography. Genome. 40: 332-341.

Pammi, S.. K. Schcrtz. G. Xu. G. IIart and J.E. Mullet. 1994. Random-amplified-
polymorphic DNA markers in sorghum. Theor. Appl. Genet. 89: 80-88.

Patel, G.M. and T.R. Sukhani. 1990. Bioph!sical plant characters associated with shoot
11) resistancc. Indian J. Ent. 52: 14-17.

Patel R.H., M . K Kukadia. K.B. Desai and S.N. Patel. 1085. Inheritance of resistance to
sorghum shoot fly in grain sorghum. Sorghun1 Newslett. 28: 58.

Patel. R.H.. K.B. Desai. P.K. Borad. M.U. Kukadia and K.R.V. Raja. 1084. Combining
ability fbr shoot fly resistance in sorghum. Sorghunl Newslctt. 27: 87-88.

Patel, R.H., M.U. Kukadia. K.B. Desai and S.N. Patel. 1985. Inheritance of resistance to
sorghum shoot fly in grain sorghum. Sorghum Newslctt. 28: 58.

Paterson. H.D. and R. Thonlpson.l971. Recovery of inter-block inrormation when block

sizes are unequal. Bionietrika. 58: 545-554.

Paterson. A.H., E.S. Lander. J.D. Hewitt. S. Peterson, S.E. 1-incoln and S.D. Tanksley.
1988. Resolution of quantitative factors into Mendellian factors by using
complete linkage map of restriction fragment length polymorphisms.
Nature, 335: 721 -726.
Paterson. A.H.. K.F. Schenz, Y.R. Lin, S.C. Liu and Y.L. Chang. 1995. The weediness of
wild plants: molecular analysis of genes influencing dispersal and
persistence of johnsongrass. Sorghum hulpepen.se (L.) Pers. Proc. Natl.
Acad. Sci. USA, 92: 6127-6131.

*Paterson, H.D. 1997. Analysis of series of variety trials. In Stuti.sticc11rncthods fir plant

vuriey evuluution. Eds. Kcnipton. R. A. and P. N. Fox. Chapman and

Hall, London.

*Payne, R.W. (Ed.) 2002. 'The Guide to Genestat'B Release 6.1. Part 2: Statistics. VSN
lnternatiol~alI.td., Oxford. II.K.

Pejic. 1.. P.A. Marson, M. Morgante. V. Kozumplick. P. Castaglioni. G. l'aramino and M.

Motto. 1998. Comparative analysis of gcnctic similarity among maize
inbred lines detectcd by RFLPs. RAPlls. SSRs. atid AFl.I)s. 'Theor. Appl.
Genet. 97: 1248-1255.

Peng. Y.. K.F. Schcrtz. S. Cartinhour and G.L. tlart. 1999. Coniparati\c gcnomc mapping
of i%~rghumhicolor (L.) Moench using a RFLP map constructed in a
population of recombinant inbred lines. Plant Breed. 118: 225-235.

Pereira. M.G. and M.1,ee. 1995. Identification of genomic regions affecting plant height
in sorghum and maize. 'Theor. Appl. Genet. 90: 380-388.

Pereira. M.G., M. Lee. P.Bramel-Cox. W. Woodman. J. Docbley and K. Whitkus. 1994.

Construction of an RFLP niap in sorghum and cornparatibe mapping in
maize. Genome, 37: 236-243.

Pereria, M.G.. D. Ahnert. M. Lee and K. Killer. 1995. Genetic mapping of quantitative
trait loci for panicle characteristics and seed weight in sorghum. Rev.
Brasil Genet. 18: 249-257.

Peter, J.A.. T.G. Shanower and J . Romeis. 1995. The role of plant trichomes in insect
resistance: A selective review. Phytophaga. 7: 41-63.

Ponnaiya . B.W.X. 1951a. Studies on genus sorghum I. Field observations on sorghunl

resistance to insect pest, Atherigonu indicci. M. Madras IJniv. J.21: 96-1 17.
Powell, W., G. Machray and J. Provan. 1996a. Polymorphism repeated by simple
sequence repeats. Trends Plant Sci. 1: 215-222.

*Pradhan, S. 1971. Investigation on Insect Pests of Sorghuni and Millets. Final Technical
Report (1965-1970). Indian Agricultural Research Institute. Division of
Entomology. New Delhi. India. 157 pp.

Prasad, M.. R.K. Varshney. A.. Kumar, F1.S. Balyan. P.C. Sharnin. K.J. Edwards. H.
Singh. H.S. Dahliwal, J.K. Roy and P.K. Gupta. 1999. A micro satellite
niarker associated ~vitha QI'I. for grain protein content o n chromosome
arm 2DL of the bread whcat. Theor. Appl. Genet. 99: 341-345.

I'rem Kishore and M.G. Jotwani. 1982. Integrated pest nlanagenlent in sorghum. J.
Environ. Res. I11 (1): 107.

Prenl Kishorc. 1984. Development of stahle resistance cultivars for stem borer (('hi110
pariel1u.s) in sorghum. lnscct Sci. Appli. 5 ( 6 ) :475-179.

I'rem Kishore and M.G. Jotwani. 1979. IIh73-A promising selection showing recovery
resistance to stem borer (('hi/1ipor/cliut). Bull. Ent. 20: 147-149.

Pujar. S.. S.A. 'l'amhankar. V.S. Ran. V.S. Gupta. S. Naili and P.K. Ranjekar. 1909.
Arbitrarily primcd-PCR based diversity assessment reflects hierarchical
groupings of Indian tetraploid wheat genotypes. rheor. Appl.
Genet.99:868-876

Ragab. R.A., S. Dronavalli. M.A. Saghai Maroof and Y.G. Yu. 1994. Construction of a
sorghum RFLP linkage map using sorghum and m a i ~ eDNA probes.
Genome. 37: 590-594.

Kaina. A.K. 1981. Movement. feeding behavior and growth of larvae of sorghum shoot
fly Athcrigonrr soccilla. Insect Sci. Appl. 22: 77-8 1.

Raina, A.K. 1985. Mechanisms of resistance to shoot fly (Atherigontr boccuru) in

sorghum: a review. Proceedings of the International Sorghum Entonlology
Workshop. 15-21 July 1984. Texas A & M University. College Station.
TX, USA, pp. 131-136.

Raina. A.K.. H.Z. Thindwa. S.M. Othieno and L.W. Douglass. 1984. Resistance in
sorghum to shoot fly (Diptera: Muscidae) oviposition on selected cultivars.
J. Econ. Ent. 77: 648-651.
Raina, A.K., H.Z. Thindwa, S.M. Othieno and R.T. Cork Hill. 1981. Resistance in
sorghum shoot fly: larval development and adult longevity on selected
cultivars. Insect Sci. Appl. 2: 99-103.

Ramakrishna, W., M.D. L.agu, V.S. Gupta and P.K. Ranjekar. 1994. DNA linger printing
in rice using oligonuclcotide probes specific to simple repetitive DNA
sequences. Theor. Appl. Genet. 88: 402-406.

Rami, J.F., P. Dufour. G. Trouche. C. Mcstres. F. Davrieux. P. Blanchard and P. Ilamon.

1998. Quantitative trait loci for grain quality, productivity. morphological
and agronomical traits in sorghum (Sorghum hicoior L. Moench). Theor.
Appl. Genet. 97: 605-616.

Rana. B.S.. B.U. Singh and N.G.P. Rao. 1085. Breeding for shoot fly rcsistance and stem
borer resistance in sorghum. I'roceedings of the International Sorghum
Entomology Workshop, 15-21 .July 1084. Texas A & M Ilniversity.
College Station. TX. IISA, pp. 347-350.

Rana. H.S.. B.11. Singh. V.J.M. Rao. H.H. Keddy and N.G.P. Kao. 1984. Inheritance of
stem borer resistance in sorghum. Indian J. Genet. 44: 7-14.

Rana. B.S.. D.P. Tripathi, K. Ralakotaiah. R. Damodar and N.G.P. Rao. 1075. Genetic
analysis of some exotic X Indian crosses in sorghum XI. Sclcction for
shoot fly resistance. lndian J. Genet. 35: 350-354.

Kana, B.S., M.G. Jotuani and N.G.P. Rao. 1981. Inheritance of host plant resistance to
the sorghum shoot fly. Insect Sci. Appl. 2: 105-109

*Kangdang, Y., S. Jamorman and R.G. Granados. 1970. Entomological rcscarch at Farm
Suuan during 1970. rhailand Nation~l Corn and Sorghum Programme
Annual Report. 1970.

*Rao. N.G.P. 1972. Discussion. In: Control of Sorghum Shoot F1) (Ed. .lot\+'ani M. C;.
and Young W.R.). Oxford & IBH. New Delhi. pp. 222-224.
Rao. N.G.P.. B.S. Kana and M.G. Jotwani. 1978, llost plant resistance to ma.jor insect
pests of sorghum. In: Plant Breeding for Resistance to Insect Pests:
Considerations About the Use of Induced Mutations. IAEA-215. Vienna.
pp. 63-78.
Rao. N.G.P.. B.S. Rana K. Balakotaiah, D.P. Tripathi and M.F.S. Fa)ed. 1974. Genetic
analysis of some exotic x Indian crosses in sorghum. VIII. FI analysis of
ovipositional non-preference underlying resistance to sorghum shoot fly.
Indian J. Genet. 34: 122-127.

*Rao, S.B.P. and D.V.N. Rao. 1956. Studies on the sorghum shoot fly. Alherigonu indicu
Malloch (Anthoniyiidae. Diptera) at Srirguppa. Mysore Agric. J. 31: 158
174.

Reddy. K.V. and J.C. Davis. 1979. A new medium for mass rearing of sorghuni stem
borer. C'hiio parlei1rr.s (Swinhoe) Lepiodeptera: Pyralidac and its use in
resistance screening. Indian J. I'lant Prot. 6: 8-55.

Resseto, C.J., K.V. Banzatto, J..F..M. Lira and J.L. Ovcnnan. 1975. AF-2XA Sorgh~ml
hicolor variety resistant to sorghum midge ('onrtrrrniu sorghicola.
Sorgliuni Newslett. 18: 5.

Ribaut. J.M.. G.O. Edmeades. F.J. Bctran, C. Jiang and M.Banziger. 1000. Marker-
assisted sclcction for improving drought tolerance in tropical m a i ~ eIn:
. 0.
Ito, J.O. Toole, and R. Hardey (Eds) Cienetic l~ilprovenient of Rice for
Water-Limited In[. Kicc Rcscarcli Institute. Manila.
I'hillipines. Pp: 193-209.

Kibaut. J.M. and D. tioisington. 1998. Marker-assisted selection: New tools and
strategies. Trends Plant Sci. 3: 236-239.

Ribaut. J.M.. X. Hu. I). Hoisington and D. Cionzatez-l)e-Leon. 1997. Use of S'I'Ss and
SSlis as rapid and reliable pre selection tools in marker-assisted sclcction
backcross scheme. Plant Mol. Biol. Rep. 15: 156-164

*Roder. M.S., V. Korzun, K. Wendehake. J. Plaschke. M.H.Tixicr. P. 1,eroy and M.W.

Ganal. 1998. Microsatellite map of wheat. Genetics. 149: 2007-2023.

Rongwen, J., M.S. Akkaya, A.A. Bhagwat. 17. Lavi and P.B. Cregan. 1995. The use of
niicrosatellite DNA markers for soybean genotype identification Theor.
Appl. Genet. 90: 43-48.chek

Rooney. W.L. 2004. Sorghum improvement - integrating traditional and new technology
to produce improved genotypes. (Ed. Sparks. D.L.). Adv. Agron. 83:38-
111.
Rosenow, D.T. and J.A. Dalhberg. 2000. Collection, conversion and utilization o f
sorghum, In: Smith, E.W and Frederiksen, R.A. (eds.), Sorghum:
origin, hltory, ~echnologl~
ondproduction, John Wiley and Sons, Inc.
New York, 309-328.

Russell, J., J. Fuller, G. Young, B. Thomas, G. Taramino, M . Machary, R. Waugh and

W. Powell.l997.Discriminating between barley genotypes using
microsatellite markers. Genome, 40: 442-450.

Sadder, M.T. and G. Weber. 2002. Comparision between genetic and physical maps
in Zea mays L. o f molecular markers linked to resistance against
Diatraea spp. Theor. Appl. Genet. 104: 908-915.

*Saghai Maroof, M.A., R.M. Biyashev, G.P. Yang, Q. Zhang and R.W. Allard. 1994.
Extraordinarily polymorphic microsatellite DNA in barley: Species
diversity, chromosomal locations and population dynamics. Proc. Nat.
Acad. Sci. USA, 91: 5466-5470.

Sajjanar, G.M. 2002. Genetic analysis and molecular mapping o f components o f

resistance to shoot fly (Alherigona socca!~) in sorghum [Sorghum
bicolor (L.) Moench.]. Ph.D. Thesis, University o f Agricultural
Sciences, Dharwad.

Salunke, G.N., D.N. Gandhale. T.K. Murti and L.M. Naik. 1982. Field screening of
sorghum lines for resistance to shoot fly (Alherrgona soccata Rond.) J .
Maharashtra Agric. Univ. 7: 270.

*Sanchez Gomez, A. 2002. Identification o f quantitative trait loci for grain yield in a
recombinant inbred B-line population in sorghum. Ph.D. Dissertation,
Texas A & M University, College Station, TX.

Sanchez, A.C., D. S. Brar, N . Huang, Z. L i and G.S. Khush. 2000, sequence tagged
site marker-assisted selection for three bacterial blight resistance genes
i n rice. Crop Sci. 40: 792-797.

Sax, K. 1923. The association o f size differences with seed coat protein and
pigmentation in Phaseolus vulgaris. Genetics, 8: 552-560.
Sharma, G.C.. M.G. Jotwani, B.S. Rana and N.G.P. Rao. . ,77. Resistance to the
sorghum shoot fly, Alheripono soccata (Rondani) and its genetic
analysis. J. Ent. Res. 1: 1-12.
Sharma, H.C., Faujdar Singh and K.F. Nwanze (eds.). 1997. Plant resistance to insects
in sorghum. lnternational Crops Research lnstitute for the Semi-Arid
Tropics, Patancheru 502 324, Andhra Pradesh, India.

Sharma, H.C., S.L. Taneja and K. Leuschner. 1983. Screening sorghums for
resistanceto insects. Paper presented at A l l India Co-ordinated
Sorghum Improvement Project Workshop, Haryana Agricultural
University, Hissar, India. April, 19-22.
Sharma, H.C., S.L. Taneja, K. Leuschner and K.F. Nwanze. 1992. Techniques to
screen sorghum for resisatnce to insect pests. lnformation Bulletin No.
32, lnternational Crops Research Institute for the Semi-Arid Tropics
Patancheru, Andhra Pradesh 502 324, India.
Sharma, H.C., B.C. Agrawal, P. Vidyasagar, C.V. Abraham and K.F. Nwanze. 1993a.
Identification and utilization o f resistance to Sorghum midge
(Conrarinia sorghicola (Coqillett)) in India. Crop Prot. 12: 343-350.
Sharma, H.C. and J.C. Davies. 1981. A literature review on the souces and
mechanism of resisatnce to the sorghum midge (Srenodiplosis
sorghicola). Progress report 5.Sorghum Entomology, lnternational
Crops Research lnstitute for the Semi-Arid Tropics, Patancehru, A.P.,
India.
Sharma, H.C., S.L. Taneja, N. Kameswara Rao and K.E. Prasad Rao. 2003.
Evaluation o f sorghum germplasm for resistance to insect pests.
lnformation Bulletin no. 63. lnternational Crops Research lnstitute for
the Semi-Arid Tropics, Fatancheru 502 324, Andhra Pradesh, India,
pp. 17-27.

Shen, L., B. Courtois, K.L. Mc Nally, S. Ribin and 2. Li. 2000. Evaluation o f near
isogenic lines o f rice introgressed with QTLs for root depth through
marker-aided selection. Theor. Appl. Genet. 10,: 75-83.

Shinde, C.B., G.S. Bharaj and M . Sharma. 1985. Screening o f breeders material for
resistance to the sorghum shoot fly (Antherigona soccata Rondani).
Sorghum Newsletter. 28: 5 5 .

Singh, R.K. and B.D. Chaudahary. 1996. Biometrical methods in quantitative genetic
analysis. Kalyani Publishers, New Delhi.
Singh. B.U. and B.S. Rana. 1986. Resistance in sorghum to shoot fly, dfherigonu soccuru
Rondani. Insect Sci. Appl. 7: 577-587.

Singh. B.U. and H.C. Sharma. 2002. Natural enemies of sorghum shoot fly. Alherigona
soccula Rondani (Diptera: Muscidae). Biocon. Sci. Tech. 12: 307-323.

Singh, B.U.. B.S. Rana and N.G.P. Rao. 1981. Host plat resistance to mite (Oligonjchus
indicus Hiirst) and its relationship with shoot fly (Alhcrigona .roccuiu
Rond.) resistance in sorghum. J. 6nt. Kes. 5: 25-30,

Singh. R. and K.L. Narayana. 1978. Influence of different varieties of sorglium on the
biology of sorghum shoot fly. Indian J. Agric. Sci. 48: 8-12.

Singh, S.P. 1998. Association of sorghum seedling characters wit11 resistance to shoot fly.
Afherigonu .roccu[cr (Kondani). ISMN 38: 88-89,

Singh. S.P. and A.N. Verma. 1988. Inheritance of rcsistance to shoot fly. -llhcrigonrr
.~occatrr(Kondani) in forage sorghum. J, Insect Sci. I: 40-52.

Singh. S.P, and ,M.G. Jotwani. 1080a. Mechanism of resistance in sorghum to shoot fly. I.
Ovipositional non-preference. Indian 1. Ent. 42: 8-12,

Sinyh. S.P. and M.G. Jotwani. 198Oh. Mechanism ofresismnce in sorghun~to slloot fly.
11. Antibiosis, Indian .I. Ent. 42: 353-357.

Singh. S.P. and M.G. Jotwani. IOXOc. Mechanism of resistance in sorghum to shoot fly.
111. Biochemical basis of resistance. Indian J. Ent. 42: 5.51-556.

Singh. S.P. and M.G. Jotwani. 1980d. Mechanism of resistance in sorgllu~nto shoot fly. IV
Role of niorphological characters of seedlings. Ind..l. Ent. 42: 806-808.

Singh, S.P.. M.G. Jotwani. B.S. Kana and N.(i.P. Rao. 1978. Stability of host-plant
resistance to sorghum shoot fly, Arhcrigontf voccaru (Rondani). Indian J.
Ent. 40: 376-383.

*Singh. S.R., G. Vedamoorthy. V.V. 'rhobii. M.G. Jotwani. W.R. Young. J.S. Blan. K.P.
Srivastava. G.S. Sandhu and N. Krishnananda. 1068. Kesistance to stem
borer. Chilo zoncllus (Swinhoe) and stem fly. Alherigonrr 1.ctriu Rond. in
the world sorghum collection in India. Memoirs Ent. Soc. India No.7.
*Smith, C.M. 1989. Plant resistance to insects. Neb York, USA: John Wiley and Sons..
286 PP
Smith, J.S.L. and O.S. Smith. 1991. Restriction fragment length polymorphisms can
differentiate among C.S. maize hybrids. Crop Sci. 31: 893-899.

Smith, O.S.. J.S.C. Smith, S.L. Bowen. R.A. 'l'enborg and S.J. Wall. 1990. Similarities
among a group of elite maize inbreds as measured by pedigree. F1 grain
yield, grain yield heterosis and RFLPs. Theor. Appl. Genet. 80: 833-840.

Smith J.R., J.D. Carpten, M.J. Brownstein, S. Ghosh. V.L. Magnuson. D.A. Gilbert, 1.M.
Trcnt and F.S. Collins. 1995. Approach to genotyping errors caused by
non-templated nucleotide addition bq Tuq DNA polymerase. Genome Res.
5: 312-317.

Smith. J.S.C., S. Kresovich. M.S. Holpkins. S.E. Mitchell. R.E. Dean. W.L. Woodman.
M. Lce and K. Porter. 2000. Genetic diversity arnong elite sorghum inbred
lines assessed with simple sequence repeats. Crop Sci 40: 216-133

Soller. M. and J.S. Deckmann. 1983. Genet~cpolymorphism in varictal identification and

genetic improvement. Theor, Appl. (;met. 67: 25-33.

Soller, M.and 'I'. Hrody. 1976. On the poucr of experimental designs for the dctcction of
linkage hetween marker loci and quantitative loci in crosses between
inbred lines. 'fheor. Appl. Genet. 47: 35-39.

Soto. I'.E. 1974. Ovipositonal preference and antibiosis in relation to resistance to

sorghum shoot fly. J. Eco. Ent. 67: 165-167.

Soundararajan, R.P.. P. Kadirvel. K. Gunathilagaraj and M. Mahes\varan. 2004. Mapping

of quantitative trait loci associate with resistance to brown planthopper in
rice by means of doubled haploid population. Crop Sci. 44: 2214-2220.

*Stam, P. 1993. Construction of integrated genetic linkage map b) means of a new

computer package: Joinmap. Plant J. 3: 739-744.

Starks, K.J., S.A. Eberhart and H. Dogyet. 1970. Kecovery from shoot fly attack in
sorghum diallel. Crop Sci. 10: 51 9-522.
Stuber, C.W. and M.D. Edwards. 1986. Genotypic selection for improvement for
quantitative traits in corn using molecular marl .r loci In: Proc. Annual
Corn and Sorghum Res. Conference. Pp 63-70.

Stuber, C.W. 1994. Success in the use o f molecular marker for yield enhancement in
corn. Proc. 491h Annual Conference. And Sorghum industry Res. Conf.
American Seed Trade Association. 49: 232-238.

'Stuber, C.W., M.D. Edwards and J.F.Wendel. 1987. Molecular marker-facilitated

investigations o f quantitative trait loci in maize: 11. Factors influencing
yield and its component traits. Crop Sci. 27: 639-648.

Stuber, C.W., S.E. Lincoln, D.W. Wolff, T. Helentjaris and E.S. Lander. 1992.
Identification o f genetic factors contributing to heterotic in a hybrid
from two elite maize inbred lines using molecular markers.
Genetics, 132: 823-839.

Subudhi, P.K. and H.T. Nguyen. 2000. Linkage map alignment o f sorghum RFLP
maps using a RIL mapping population. Genome, 43: 240-249.

Subudhi, P.K., D.T. Rosenow and H.T. Nguyen. 2000. Quantitative trait loci for the
stay green trait in sorghum (Sorghum bicolor L. Moench): consistency
across genetic backgrounds and environments. Theor. Appl. Genet.
101: 733-741.

Sukhani, T.R. and M.G. Jotwani. 1980. Efficacy o f mixtures o f carbofuran treated and
untreated sorghum seed for the control o f shoot fly, Alherigona
soccara (Rodani). I. Ent. Res. 4: 186-189.

Taramino, G., R. Tarchini, S. Ferrario, M. Lee and M.E. Pe. 1997. Characterization
and mapping of simple sequence repeats (SSRs) in Sorghum bicolor.
'Theor. Appl. Genet. 95: 66-72.

Taneja , S.L. and K . Leuschner. 1985. Resistance screening and mechanism o f

resistance in sorghum to shoot fly. Proceeditfigs o f the International
Sorghum Entomology Workshop, 15-21 July 1984, Texas A & M
University, College Station, TX, USA, pp. 115-129.

Tanksley, S.D. 1983. Molecular markers in plant breeding. PMBR. 1: 3-8.

 214
Tanksley, S.D.and J.C. Nelson. 1996. Advanced backcross QTL analysis: a method
for the simultaneous discovery and transfer o f valuable QTLs from un-
adopted germplasm into elite breeding lines. Theor. Appl. Genet. 92:
191-203.

Tanksley, S.D. and C.M. Rick. 1980. lsozyme gene linkage map o f the tomato:
application in genetics and breeding. Theor. Appl. Genet. 57: 161-170.

Tanksley, S.D., N.D. Young, A.H. Paterson and M.W. Bonierbel. 1989. RFLP
mapping in plant breeding: new tools for an old science. Biotech. 7:
257-264.

Tanksley, S.D. and J. Hewitt. 1988. Use o f molecular markers in breeding for soluble
solids content in tomato- a re-examination. Theor. Appl. Genet. 92:
191-203.

Tanksley, S.D. and J.C. Nelson. 1996. Advanced back cross QTL analysis: a method
for the simultaneous discovery and transfer o f valuable QTLs from
unadapted germplasm into elite breeding lines. Theor. Appl. Genet. 92:
191-203.

Tanksley, S.D., H.M. Filho and C.M. Rick. 1982. Use o f naturally occurring enzyme
variation to detect and map genes controlling quantitative traits in an
interspecific backcross o f tomato. Heredity, 49: 11-25.

Tao,Y., J.M. Manners, M.M. Ludlow and R.G. Henzell. 1993. DNA polymorphism in
grain sorghum (Sorghum bicolor (L.) Moench). Theor. Appl. Genet.
86: 679-688.

Tao, Y.Z., A. Hardy, J. Drenth, R.G. Henzell, B.A. Franzmann, D.R. Jordan, D.G.
Butler and C.L. Mclntyre. 2003. Identifications o f tho different
mechanisms for sorghunl midge resistance through QTL mapping.
Theor. Appl. Genet. 100: 1225-1232.

Tao, Y.Z., D.R. Jordan, R.G. Henzell and C.L. Mclntyre. 1998a. Construction o f a
genetic map in sorghum RIL population usir, probes from different
sources and its comparison with other sorghum maps. Aust. J. Agric.
Res. 49: 729-736.
Tao, Y.Z., D.R. Jordan, R.G. Henzell and C.L. Mclntyre. 1998b. ldentification o f
genomic regions for rust resistance in sorghum. Euphytica, 103: 287-
292.
Tao, Y.Z., R.G. Henzell, R.G. Jordan, D.G. Butler, A.M. Kelly and C.L. Mclntyre.
2000. ldentification o f genomic regions associated with stay green in
sorghum by testing RlLs in multiple environments. Theor. Appl.
Genet. 100: 1 125-1232.
Taramino, G. and S. Tingey. 1996. Simple sequence repeats for germplasm analysis
and mapping in maize. Genome, 39: 277-287.
Taramin0.G.. R. Tarchini, S. Ferrario, M . Lee and M.E. Pe. 1997. Characterization
and mapping o f simple sequence repeats (SSRs) in Sorghum bicolor.
Theor. Appl. Genet. 95: 66-72.
.Tarumoto, I. 1980. Inheritance o f glossiness o f leaf blades in sorghum, Sorghum
bicolor (L.) Moench. Japanese J. Breed. 30: 237-240.
Tarumoto, 1. 1983. Multiple alleles on glossy locus in Sorghum b~color(L.) Moench.
Abstracts o f contributed papers o f the Fifteenth International Congress
o f Genetics, 12-21 December, 1983, New Delhi, 2: 643.
Tegelstrom, H. 1992. Detection o f mitochondria1 DNA fragements in molecular
genetic analysis o f populations: A pratical Approach (Ed.Hoelzel, AR),
IRL Press, Oxford. pp. 89-1 14.
Terwilliger, J.D.2001. On the resolution and feasibility o f genome scanning
operations. In: Genetics dissection o f complex traits. (Eds. Rao, D.C.
and M.A. Province), pp. 351-392.
Thoday, J.M. 1961. Location o f polygenes. Nature, 191: 368-370
Toojinda, T., E. Baird, A. Booth, L. Broers, P. Hayes, W. Powell, W. Thomas, H .
Vivar and G. Young. 1998 lntrogression o f quantitative loci (QTLs)
determining stripe rust rc-sistance in barley: an example o f marker-
assisted line development. Theor. Appl. Genet. 96: 123-13 1
Tuinstra, M.R., E.M. Grote, P.B. Goldsborough and G. Ejeta. 1996, ldentification of
quantitative trait loci associated with pre-flowering drought tolerance
in sorghum. Crop Sci. 36: 1337-1344.
Tuinstra, M.R., E.M. Grote, P.B. Goldsborough and G. Ejeta. 1997. Genetica analysis
o f post-flowering drought tolerance and components o f grain
development in Sorghum bicolor (L.) Moench. Mol. Breed. 3: 439-
448.
'Ulstrop, A.J. 1978. Agric. Handbook, pp. 199-200.
U ~ i t h a n G.C.
. and K.V.S. Reddy. 1985. Oviposition and infestation of. sorghum shoot
fly Aiherigonu soccatu Rondani, on certain sorghum cultivars in relation to
their relative resistance and susceptibility. Insect Sci. Appl. 6: 409-41 2.

*Usman, S. 1972. Efficacy of granular insecticides for sorghum slioot fly in India.
Pesticide, 8: 409-412.

Utz, H.F. and A.E. Melchinger. 2000. PIABQT'L: A computer program to map QTI.
(version 1.1). Institute fir I'flanzenziichtung. Saatgutforschung und
Populationsgenetik. llniversitaet Ilohcnhcim, 1)-70593 Stuttgart.
Germany.

*Utz. H.F. and A.E. Melchinger. 1994. Comparison of dirferent approaclies to interval
mapping o f quantitative trait loci. In: Biometrics in Plant Breeding:
Applications of Molecular Markers. (Ed. Van Ooiien J.W. and .I.Jansen).
Wageninfen, The Netherlands. pp. 195-204.

Utz. H.F. and A.E. Melchinger. 1906. I'L.ABOTL: A program for conlposite interval
mapping of QTL J. Agric. Genornics 2( 1 ).
htlp://www.ncgr.~~rgiagipapersO6/p'dperl96/indexplY6.l~tml

Van Bcrloo. R, and P. Stam 1908 Marker-assisted selection in autogalllous RII.

populations: a simulation study. Thcor. Appl. Genet. 06: 147-154

Veldboom. L.R.. M. Lee and W.1,. M'oodman. 1994. Molecular niarker-ihcilitnted studies
of morphological traits in an elite mai7e population. 11. Determination of
Q'fL.s for grain yield and yield components. 'Sheor. Appl. Genet. 89: 45 1 -
458.

*Venkateswaran. K. 2003. Diversity analysis and identification of sources of resistance to

donny mildew. slioot tly and stem borer in wild sorghum. Ph. D. 'l'hcsis
Dept. of Genetics Usmania IJniversity. Hyderabad.

*Vidyabhushanam. R.V. 1972. Breeding for shoot fly resistance in India. In: Control of
Sorghum Shoot Fly (Eds. Jotwani, M.G. and W.R.Young). Oxford and
IBH Publishing Co.. New Delhi-I. pp. 218-219.

Vierling, R.A.. 2. Xing, C.P. Joshi. M.L.. Gilbert and l1.T. Nguyen. 1904. Genetic
diversity among elite sorghum lines revealed by restriction. Theor. Appl.
Genet. 87: 816-820.
Virk, P.S., H.S. Newury, M.T. Jackson and R.V. Ford Llyod. 1995. The Identification of
duplicating accessions within rice germplasm collection using RAPD
analysis. Theor. Appl. Genet. 90: 1049-1055.

*Virk, P.S.. H.J. Newury, M.T. Jackson and B.V. Ford I,lyod. 1997. In Biotechnology
and Plant Genitics Resources Conservation and use (eds Callow. J. A,. and
Newbury. H.J.).

Vos. P., R. Ilogers. M. Bleeker. M. Reijans, 'I'. Van dc Lee. A. Hornes. M. Frijters. J. Pot.
J. Peleman, M. Kuiper and M. Zabeau. 1995. AFLP: a nea technique for
DNA fingerprinting. Nucl. Acids Res. 23: 4407-4414.

Wall, J.S.. D.A. I:e? and J.W. Paulis. 1984. Improved two-dimensional electrophoretic
separation of zein protcins: Application to study of zein inheritancc in corn
genotypes. Cereal Chem. 61 : 141-146.

*Wang. C.. H. Yasui. A. Yoshimura. 11. Zhai and J. Wan. 2004. Inheritance and Q'fL
mapping of antibiosis to green leafliopper in rice. Crop Sci. 44: 389-393.

Wang. Z.. J.L. Weber. G . Zhong and S.1). Tanksley. 1994. Survey of plant short tandem
DNA repeats. Thcor. Appl. Genet. 88: 1-6.

Waxelson. ti. 1033. L.inkage between quantitative and qualitative characters in barley.
Hereditas. 17: 323-341.

Westnian. A.L. and S. Kresovich. 1907. Use of ~nolecular marker techniques for
description of plant genetic variation. In: Biotechnology and Plant Genetic
Resources: Conservation and Use. (Eds. Callow. J.A.. J.A. Ford-Lioyd and
H.J. Newbury). CAB International. Oxford. IJK. pp. 9-48.

Whitkus. R.. J. Doebley and M. Lec. 1992. Comparative niappihg of sorghum and maize.
Genetics, 132: 11 19-1 330.

*Williams. J.G.K.. A.R. Kubelik, K.J. Livak, J.A. Rafalski and S.V. 'l'ingeq. 1990. DNA
polymorphisms amplified by arbitrary primers arc useful as genetic
markers. Nucl. Acids Res. IS: 6531-6535.
'Wiseman, B.R. and Morrison, W.P. 1981. Components for management o f field corn
and grain sorghum insects and mites in the United States. United States
Department o f Agriculture - Agricultural Research Service 9USDA-
ARS), A R M 5-18. Washington, D.C., U.S.A.: Government Printing
Press. 18 pp.

Wittaker, J.C., R.N. Curnoul, C.S. Haley and R. Thompson. 1995. Using marker-
maps in marker-assisted selection. Genet. Res. 66: 255-265.

Witcombe, J.R. and C.T. Hash. 2000 Resistance gene development strategies in cereal

hybrids using marker-assisted selection: Gene pyrimiding, three-way hybrids and

synthetic parent populations. Euphytica. 112: 175-186

Wright, S. 1934. An analysis o f variability in number o f digits in an inbred strain o f

guinea pigs. Genetics, 19: 506-536.

Wu, K.S. and S.D. Tanksley. 1993. Abundance polymorphism and genetic mapping
o f microsatellites in rice. Mol. Gen. Genet. 241: 225-235.

Xiao, J., J. Li., L. Yuan, S.R. McCouch and S.D. Tanksley. 1996a. Genetic diversity
and its relationship to hybrid performance and heterosis in rice as
revealed by PCR-based markers. Theor. Appl. Genet. 92: 637-643.

Xiao, J., S. Grandillo, S.N. Ahn, S.R. M c Couch, S.D. Tansksley, J. L i and L . Yuan.
1996b. Genes from wild rice improve yield. Nature, 384: 223-224.

Xiao, J., J. Li and S.D. Tanksley. 1996c. Identification o f QTLs affecting traits o f
agronomic importance in recombinant inbred population derived from
a sub-specific rice cross. Theor. Appl. Genet. 92: 230-244.

Xu, G.W., C.W. Magill, K.F. Schertz and G.E. Hart. 1994. A RFLP linkage map o f
Sorghum bicolor (L.) Moench. Theor. Appl. Genet. 89: 139-145.

Xu, W.W., P.K. Subudhi, O.R. Crasta, D.T. Rosenow, J.E. Mullet and H.T. Nguyen.
2000. Molecular mapping o f QTLs conferring stay green in sorghum.
Genome, 43: 461 -469.

Xu, X.F., H.W. Mei, L.J. Luo, X.N. Cheng and Z.K. Li. 2002. RFLP-facilitated
investigation o f the quantitative resistance o f rice to brown planthopper
(Nilapawara lugens). Theor. Appl. Genet. 104: 248-253.
*Yadav, R.S., C.T. Hash, F.R. Bidinger. G.P. Cavan and C.J. Howrth. 2002. Quantitative
trait loci associated with traits determining grain and stover yield in peral
millet under terniinal drought strcss conditions. Thero. Appl.Genet. 104:
67-83.

Yadav. R.S., C.T. Hash. F.R. Bidinger. K.M.Devos and C.J. fiowrth. 2004. Genomic
regions associated with grain yield and aspects of post-tlowering drought
tolerance in pearl millet across stress environnients and tester background.
Euphytica, I36(3): 265-277,

*Yamsaki, M.. A. Yoshiniura and I-I. Yasui. 2000. Mapping of quantitative trait loci of
Iirgens L..). Asta
ovicidal response to brown plant hopper (Niltr~~rrr~rr/cr
Agriculturae Universitatis Zliejiangensis. 18: 109-1 17.

*Yang, G.K.. M.A. Saghai Maroof. C.G. Xu, 0. %hang. and R.M. Lliyasliev. 1994.
Comparative analysis of niicrosatellite DKA polymorpliisni in laridraces
and cultivars of rice. Mol. Gen. Genet. 245: 187-194.

Young. N.D. 1996. QTL mapping and quantitative disease resistance in plants. Ann. Re\'.
I'lnnt I'ath, 34: 470-501.

Young. N.D. 1099, A cautiously optimistic vision for marker-assisted breeding. Mol.
Brccd. 5: 505-510.

*Young, N.D. 2001. Constructing a plant genetic linkage map with I>NA markers. In:
DN.4 Based Markers in Plants. (Eds. Phillips. R.L. and I.K. Vasil). Kluwer
Academic Publishers. The Netherlands, pp. 3 1-48.

Young. N.D. and S.D. 'Panksley. 1989. restriction fragment length polymorphism maps
and the concept of graphical genotypes. Theor. Appl. Genet. 77: 3 1-48.

*Young. W.R. 1972. Sources of resistance to the sorghum shoot fly. Athcrigonrr vuriu
Rond. In: Control of Sorghum Shoot Fly (Eds. Jotwani. M. T'. and Young
W. R.). Oxford & IBH. New Delhi. pp. 168-179.

Yu. K..S. Park and V. Poysa. 2000. Marker-assisted selection of conmion beans Tor
resistance to common bacterial blight. Efficacy and economics. Plant
Breed. 119: 411-415.
Yuan, L.P. 2002. Future out look on hybrid rice research and development. P3 in Abstract
of the Fourth International Symposium on Hybrid rice 14-17 May 2002,
Hanoi, Vietnam.

Zamir, D. and Y. Eshed. 1998. Case historj and germplasm introgression : tomato
genetics and Breeding using nearly isogenic introgression lines derived
from wild species: In: Paterson. A. 11. (Edb) Molecular dissection of
complex traits. CRC. Press Ltd.

Zeng, Z.B. 1994. Precision mapping of quantitative trait loci. Genetics, 136: 1457-1468.

Zeinel Abdin A.M. 1981. Review of sorghum shoot fly research in Sudan. Insect Sci.
Appl. 2: 55-58.

*Zhang, Q. and N. Huang. 1998. Mapping and molecular marker-based genetic analysis
for efficient hybrid rice breeding. P. 256. In: S. S. Virmani. E. A. Siddique.
and K. Murlidharan (Eds) Advances in Ilybrid Rice Technology. Proc.
Third Int. Symposium Hybrid Rice 14-16 Nov. 1996. Hyderabad. India.
Int. Rice Research Institute. Manila. Phillipines.

Zhang. W. and C. Smith. 1992. Computer simulation of marker-assisted selection

utilizing linkage disequlibrium. Theor. Appl. Genet. 83: 813-820.

Zhang. W. and C. Smith. 1993. Computer simulation of marker-assisted selection

utilizing linkage disequlibrium: The effects of several additional factors.
Theor. Appl. Genet. 86: 492-496.

Zhang. Q.. Y .J. Gao, S.H. Yang, R.A. Ragab. M.A. Saghai Maroof and Z.B LI. 1994 A
diallel analysis of heterosis in elite hybrid rice based on RFLPs and
microsatellite. Theor. Appl. Genet. 89: 185-192.

Zhang, Q., Y.J. Gao, M.A. Saghai Maroof, S.H. Yang and J.X. Li. 1995. Molecular
divergence and hybrid performance in rice. Mol. Breed. 1: 133-142.

* Originals not seen

APPENDICES
APPENDIX I: Particular characteristics ofsorghum accessions used i n this sorghum diversity study
Mapping
Stay- population Recurrent Parent forDonor
SI. No Genotype name Origin Shoot fly Stem borer Midge green parent Shoot fly resistance- parent

1 BTx623 TAMU, USA Susceptible Susceptible Susceptible Senescent Shoot tly

2 IS2122 Resistant
3 I S 2123 Resistant Resistant
4 IS2195 India Resistant Resistant
5 IS 2205 India Resistant
6 IS2265 Sudan Resistant Resistant
7 IS2312 Sudan Resistant Resistant
8 IS2367 Nigeria Resistant
9 IS 3962 India Resistant
10 IS4756 India Resistant
II IS 5469 India Resistant
I2 IS 5470 India Resistant
13 IS 5480 India Resistant
14 I S 5490 India Keaistdnt
15 IS5571 India Kesistant
16 IS5658 India Resistant
17 IS 17948 India Resiatar~t
18 IS 18573 Nigeria (breeding line) Resistant
19 IS 18577 Nigeria (breeding line) Resista~it
20 I S 18579 Nigeria (breeding line) Resirtar11
21 IS 18581 Nigeria (breeding Ilnr) Kesista~~t
22 IS22148 India Resistant
Stem borer and Stem borer
23 ICSV 700 ICRISAT Resistant Resi5tant Susceptible midge fly resistance
24 P B 12779-2 ICRISAT Resistant
Stem borer and Stem borer
25 PB 15881-3 ICRISAT Re~istant Smceptible midge fly resistance
26 IS 1034 Resistant
27 IS2269 Resistant
28 IS2291 Re~istant Resistant
29 IS 18366 Resistant
 - - - .
SI. N o Mapp~ng Recurrent Parent for
Sta\- population
. . Shoot fly resistance Donor
Genotype name Origin Shoot fl\ stem borer M i d g c green pare* -- parent
Shoot fly
Re\istanr Shoot fly Resistance
Resistant Resistant
Shoot fly
32 ICSB457 ICRISAT llesistant Shoot fly resistance

ICSV 705 ICRISAT Resistant

ICSV 707 ICRISAT Re~istant
ICSV 708 ICKISAT Resistant
ICSV 713 ICRISAT Resistant
Stem borer and Stem borer
Resistant Rezistant Succepihlo mid%" resistance
Stem borer and Stern horer
PB 15520 ICRISAT Resistant Resictant midge resistance
I S 18695 Resictanl
19 19476 Resistant
IS 22806 Resistant
ICSV 197 Rc~istanl
ICSV 388 Resistant
ICSV 39 1 Rcci~tant
ICSV 730 Resistant
ICSV 736 Resistanl
Stem horer and Midge
47 ICSV 745 ICRISAT Susceptible Resistant midge re5i\tance
48 ICSV 746 ICRISAT Rcrlstanl
49 ICSV 748 ICRISAT Resistant
50 ICSV 757 ICRISAT Rrsistant
51 ICSV88014 ICRISAT Res~stant
Stem borer and Midge
52 ICSV 88032 ICRISAT Susceptible Susceptible Resistant midge resistance
53 ICSV 88041 ICRISAT Rc>i$tant
54 ICSV 89051 ICRISAT Resistant
55 ICSV 89054 ICRISAT Resistant
56 ICSV90004 ICRISAT Rcslstant
 APPENDIX 11
P r e p a r a t i o n of S t o c k S o l u t i o n s

C T A B (Cetyl Trimethyl Ammonium Bromide) (2 %) buffer

CTAB 20 g
I M Tris 200 ml
5 MNaCl 280 ml
0.5 M EDTA 40 nil
Na2S03 2.5 g
Distilled water 460 ml
Add mercaptoethanol (0.1 %) fresh while using CTAR ( 2 %) solution
Rnase (10 mg/ ml)
Dissolve Rnase in water. place in a tube in a boiling water bath for 10 minutes. Allow
this to cool on a bench and store at -20"~.
Chloroform: isoamyl alcohol (24:l)
Chloroform 240 ml
lsoamyl alcohol 10 ml
Store in dark at room temperature. Mahe up and dispenses the solution in a fumed
cupboard.
Ethanol (70 %)
Absolute alcohol 70 ml
Distilled watcr 30 1n1
NaCI (5 M)
Dissolve 292.2 g NaCl in 750 ml watcr. Make up to 1 liter with water, filter and
autoclave
Phenol1 Chloroform
Mix equal volume of the buffered phenol and chloroform: isoamyl alcohol (24:l).
Store at 4 "c.
Sodium acetate (2.5 M, pH 5.2)
Dissolve 340.2 g sodium acetate in 500 ml water. Adjust pII to 5.2 with glacial acetic
acid and make volume up to 1 liter and autoclave.
Tris HCI (lM, pH 8.0)
Dissolve 121.1 g Tris in 800 ml of water. Adjust pH to 8.0 with conc. HCI make
volume up to 1 liter and autoclave.
EDTA (0.5 m, Ph 8.0)
Dissolve 186.1 g Na2 EDTA.ZII20 in 800 rill water. Adjust pH to 8.0 with Sodium
hydroxide pellets. Make up volume to 1 liter and autoclave.
TluEl buffer
1M Tris HCI pH 8.0 10 ml
1 M EDTA pH 8.0 1 ml
And make up to 1 liter with sterile distilled watcr.
T S O E buffer
I~
1M Tris I1CI pH 8.0 50 ml
0.5 M EDTA pl-l 8.0 20 ml
Make volume up to 1 litcr with sterile distilled water.
10X Tris-Borate Buffer (TBE) (per liter)
Tris buffer
Boric Acid
EDTA
108 g Tris base, 55 g Boric acid and 9.3 g EDTA. Add dciotiised 1120to I litcr. The
pI1 is 8.3 and requires no adjustment.
6X Gel loading buffer (0.25 '%, Hr~~rnophenol
hlue, 40 % sucrose)(l0 ml)
Sucrose 4g
Bromophenol Blue 7.5 nil
dH2C) up to 10 ml
Store at 4'c.
Ethidium bromide (I0 mglml)
Dissolve 100 nig ethidiutil bromide in 10 nil of distilled water; wrap tube in
aluminium foil and store at 4"c.
Caution: Ethidium hromide is extremely mutagenic.
Acrylamide / biacrylamide 29:l (w/w)
Acrylamide 29 g
Biacrylamide 1E
Water (deionised distilled) up to 100 ml
Store at 4'c for <= 1 month.
Acrylamidelbisacrylamide 29:l (v/v)
87 ml Acrylamide
3 ml Bisacrylamide
Add deionised distilled water to 300 ml. Solution can he stored up to 1 month at 4°C.
10 %(WN) Ammonium Per Sulphate
Ammonium per Sulphate I g
Water (deionised distilled) 10 ml
Make fresh stock every week and store at 4OC.
TEMED (N, N, N', N'-tetramethylethylenediamine)
Ready made, store between 10 and 30°C (check label flask).
Loading buffer for non-denaturing PAGE (5X)
50 mM EDTA ( I ml of 0.5 M EDTA. pH 8.0)
50 mM NaCl (I00 yl of 5 M NaCI)
50% (vlv) glycerol (5 nil)
Make up to 9 ml with sterilized deionised water. Add 10 mg fast orangc G dye and
adjust the volumc to 10 ml. If you are using bromophenol blue and cyanol then less is
required.
Binding silane
0.15 ml Bind silane
0.5 ml Acetic Acid
90.35 ml Ethanol
Mix the ingredients and store at 4'C.
100 base pairs ladder (50ngIml)
100 hp ladder (stock conc.1 pglpl) 50 pI
Blue (6X dye) 165 p1
TlOEl buffcr 785 PI
Repel silane
Ready made. store at PC.
Reagents used for the Silver staining for PAGE
0.1 O/o (wIv)
CTAB
2 gram CTAB in 2 liters of distilled deionised water
1 M NaOH
freshly prepared
0.3 % liquid ammonia
wear face mask when handling ammonia, should preferably be done in fume cupboard
Silver nitrate solution (freshly prepared)
2 gram silver nitrate
8 ml1M NaOH
6-8 ml 25% ammonia.
Dissolve the silver nitrate and NaOH into 2 liters of distilled deionised water. Titrate
with ammonia (on a shaker) until the solution becomes clear; add a further 1 nil of
ammonia solution.
Sodium Carbonatc solution
(fteshly prepared, mind that the Sodiu~iiCarbonate should not be older than 12
months)
30 g Sodium Carbonate
0.4 ml Formaldehyde
Dissolve the sodium carbonate in 2 liters of distilled dcionised water. Add 0.4 ml
fomialdehyde.
Glycerol solution
30 ml Glycerol into 2 liters distillcd deionised water.
Concentrated NaOH solution
40 gram into 1 liter of water.
 APPENDIX 111
Details on pedigree of 259 F,., R1L.a derived from cross 296B X IS18551
- - (cont...)
Sr.No. Plot No Pedigree 51. NO. Plot NO Peigree
1 43101 (2968 x IS 18551)-1-1-1 $8 43148 (2968 x IS 18551)-48-1-1
2 43102 (2968 x IS 18551)-2-1-1 19 43149 (2968 x IS 18551)-49-1-1
3 43103 (2968 x IS 18551)-3-1-1 50 43150 (2968 x IS 18551)-50-1-1
4 43104 (2968 x IS 18551)-4-1-1 51 43151 (2968 x IS 18551)-51-1-1
5 43105 (2968 x IS 18551)-5-1-1 52 43152 (2968 x IS 18551).52.1-1
6 43106 (2968 x IS 18551)-6-1-1 53 43153 (2968 x IS 18551)-53-1-1
7 43107 (2968 x IS 18551)-7-1-1 54 43154 (2968 x IS 18551)-54-1-1
8 43108 (2968 x IS 18551)-8-1-1 55 43155 (2968 x IS 185511-55-1-1
9 43109 (2968 x IS 18551)-9-1-1 56 43156 (2968 x IS 18551)-56-1-1
10 43110 (2968 x IS 18551)-10-1-1 57 43157 (2968 x IS 18551)-57-1-1
11 43111 (2968 x IS 18551)-11-1-1 58 43158 (2968 x IS 18551)-58-1-1
12 43112 (2968 x IS 18551)-12-1-1 59 43159 (2968 x IS 18551)-59-1-1
13 43113 (2968 x IS 18551)-13-1-1 60 43160 (2968 x IS 18551)-60-1-1
14 43114 (396B x IS 18551)-14-1-1 61 43161 (2968 x IS 18551)-61-1-1
15 43115 (2968 x IS 18551)-15-1-1 62 43162 (2968 x IS 18551)-62-1-1
16 43116 (2968 x IS 18551)-16-1-1 63 43163 (2968 x IS 18551)-63-1-1
17 43117 (2968 x IS 18551)-17-1-1 64 43164 (2968 x IS 18551)-64-1-1
18 43118 (2968 x IS 18551)-18-1-1 65 43165 (2968 x IS 18551)-65-1-1
19 43119 (2968 x IS 18551)-19-1-1 66 43166 (2968 x IS 185511-66-1-1
20 43120 (2968 x IS 18551)-20-1-1 67 43167 (2968 x IS 185511-67-1-1
21 43121 (2968 x IS 18551)-11-1-1 68 43168 (296BxIS 18551)-68-1-1
22 43122 (2968 x IS 18551)-22-1-1 69 43169 (2968 x IS 18551)-69-1-1
23 43123 (2968 x IS 18551)-23-1-1 70 43170 (2968 x IS 18551)-70-1-1
24 43124 (2968 x IS 185.71)-24-1-1 71 43171 (2968 x IS 18551)-71-1-1
25 43125 (2968 x IS 18551)-25-1-1 72 43172 (2968 x IS 18551)-72-1-1
26 43126 (2968 x IS 185.51)-26-1-1 73 43173 (2968 x IS 18551)-73-1-1
27 43117 (2968 x IS 18551)-27-1-1 74 43174 (2968 x IS 18551)-74-1-1
28 43128 (2968 x IS 18551)-28-1-1 75 43175 (2968 x IS 185511-75-1-3
29 43129 (2968 x IS 18551)-29-1-1 76 43176 (2968 x IS 185511-76-1-1
30 43130 (2968 x IS 18551)-30-1-1 77 43177 (2968 x IS 18551)-77-1-1
31 43131 (2968 x IS 18551)-31-1-1 78 43178 (2968 x IS 18551)-78-1-1
32 43132 (2968 x IS 18551)-32-1-1 79 43179 (2968 x IS 18551)-79-1-1
33 43133 (2968 x IS 18551) 33-1-1 80 43180 (2968 x IS 18551)-80-1-1
34 43134 (2968 x 18 18551)-34-3-1 81 43181 (2968 x IS 18551)-81-1-1
35 43135 (2968 x IS 18551)-35-1-1 82 43182 (2968 x IS 18551)-82-1-1
36 43136 (296B x IS 18551)-36-1-1 83 43183 (2968 x IS 18551)-83-1-1
37 43137 (2968 x IS 18551)-37-1-1 84 43184 (2968 x IS 18551)-84-1-1
38 43138 (2968 x IS 18551)-38-1-1 85 43185 (2968 x IS 18551)-85-1-1
39 43139 (2968 x IS 18551)-39-1-1 86 43186 (2968 x IS 18551)-86-1-1
40 43140 (296B x IS 18551)-40-1-1 87 43187 (2968 x IS 18551)-87-1-1
41 43141 (2968 x IS 18551)-41-1-1 88 43188 (2968 x IS 18551)-89-1-1
42 43142 (2968 x IS 18551)-42-1-1 89 43189 (2968 x IS 18551)-90-1-1
 (cont.
Pedigree
- -
.
Sr.No Plot NO
(296B x IS 18551)-96-1-1 144 43244
145 43245
I46 43246
147 43247
148 43248
149 43249
150 43250
151 43251
152 43252
153 43253
154 43254
155 43255
156 43256
157 43257
158 43258
159 43259
160 43260
161 4326 1
162 43262
163 43263
164 43264
165 43265
166 43266
167 43267
168 46268
169 43269
170 43270
171 43271
172 43272
173 43273
174 43274
175 43275
176 43276
177 43277
178 43278
179 43279
180 43280
181 4328 1
182 43282
183 43283
184 43284
185 43285
286
ix 287
188 43288
189 43289
190 43290
191 43291
192 43292
(cont...) (cont...)
ST.No. Plot No Peigree Sr.No. Plot No Peigree
193 43293 (296B x 1s 18551)-152-1-1 242 43342 (2968 x IS 18551)-176-2-1
194 43294 (296B X I S 18551)-152-2-1 243 43343 (296B x IS 18551)-177-1-1
195 43295 (296B x IS 18551)-153-1-1 244 43344 (296B x IS 18551)-177-2-1
196 43296 (296B x IS 18551)-153-2-1 245 43345 (296B x IS 18551)-178-1-1
197 43297 (296B x IS 18551)-154-1-1 246 43346 (296B x IS 18551)-178-2-1
198 43298 (296B x IS 18551)-154-2-1 248 43348 (296B x IS 18551)-179-2-1
199 43299 (296B x IS 38551)-155-3-1 249 43349 (296B x IS 18551)-180-1-1
200 43300 (296B x IS 18551)-1.55-2-1 250 43350 (296B x IS 18551)-180-2-3
201 43301 (296B X I S 18551)-156-1-1 251 43351 (296B x IS 18551)-181-3-1
202 43302 (296B x IS 18551)-156-2-1 252 43352 (296B x IS 18551)-181-2-1
203 43303 (296B x IS 18551)-157-1-1 253 43353 (296B x IS 18551)-182-1-1
204 43304 (296B x IS 18551)-157-2-1 254 43354 (296B x IS 18551)-182-2-1
205 43305 (296B x IS 18551)-158-1-1 255 43356 (296B x IS 18551)-184-2-1
206 43306 (296B x IS 18551)-158-2-1 256 43357 (296B X I S 18551)-185-1-1
207 43307 ( 2 9 6 B x I S 38551)-159-1-1 257 43358 (296B X I S 18551)-185-2-1
208 43308 (296B x IS 18551)-159-2-1 258 43359 (296B x IS 18551)-186-1-1
209 43309 (296B x IS 18551)-160-1-1 259 43360 (296B x IS 18551)-186-2-1
210 43310 (296B x IS 18551)-160-2-1 260 43361 296 6
211 43311 ( 2 9 6 B x i S 18551)-161-1-1 261 IS 18551 (Et>tornologysource)
212 43312 (296B x IS 18551)-161-2-1
213 43313 (296B X I S 18551)-162-1-1
214 43314 ( 2 9 6 B x I S 18551)-162-2-1
215 43315 (296B x IS 18551)-163-1-1
216 43316 (296B x IS 18551)-163-2-1
217 43317 ( 2 9 6 B x I S 18551)-164-1-3
218 43318 j206BxIS18551)-164-2-1
219 43319 ( 2 9 6 B x I S 18551)-165-1-1
220 43320 (296B x IS 18551)-165-2-1
221 43321 (296B x IS 18551)-166-1-1
222 43322 (296B x IS 18551)-166-2-1
223 43323 (296B x IS 18551)-167-1-1
224 43324 (296B x IS 18551)-167-2-1
225 43325 (296B x IS 18551)-168-1-1
226 43326 (296B x IS 18551)-168-2-1
227 43327 ( 2 9 6 B x I S 18551)-169-1-1
228 43328 (296B x IS 18551)-169-2-1
229 43329 (2Y6B x IS 18551)-170-1-1
230 43330 (296B x IS 18551)-170-2-1
231 43331 (296B x IS 18551)-171-1-1
232 43332 (296B x IS 18551)-171-2-1
233 43333 (296B x IS 18551)-172-1-1
234 43334 (296B x IS 18551)-172-2-1
235 43335 (296B x IS 18551)-173-1-1
236 43336 (296B x IS 18551)-173-2-1
237 43337 (296B x IS 18551)-174-1-1
238 43338 (296B x IS 18551)-174-2-1
239 43339 (296B x IS 18551)-175-1-1
240 43340 (296B x IS 185511-175-2-1
241 43341 (296B x IS 18551)-176-1-1
(cont...)
 ...
cont ... Xlll

RIL No. Glossiness

o
,re Egg(%) l Egg (76) 11 UH (%) I UH (7'0) I 1 Tri-u Tri-I Flower plht ORS AS GR-Yi

32 5 70.4 94.6 78 92 0.0 0.0 82 210.4 6 7 520

33 3 64.1 83.3 59 75 156.7 84.0 85 202.9 4 7 686
34 5 64.9 97.1 73 88 81.7 44.0 86 195.0 6 5 598
35 3 65.4 88.7 76 88 70.0 34.7 75 174.2 5 7 384
36 3 58.5 97.9 73 92 73.3 33.3 90 152.5 5 8 575
37 4 56.8 93.6 71 88 25.7 2.7 91 137.5 7 6 620
38 4 67.5 92.6 73 89 18.7 4.3 79 174.2 8 6 590
39 3 64.9 84.4 72 86 113.7 50.3 79 135.4 6 8 554
40 4 56.9 90.8 73 86 45.0 3.7 90 169.6 6 5 282
41 4 67.1 97.5 76 90 44 3 13.7 84 240.0 5 6 488
42 4 58.3 85.2 76 86 177.7 55.0 92 204.2 6 7 306
43 3 64.1 86.5 72 83 26.3 3.7 92 223.3 6 5 419
44 3 59.5 89.5 73 89 l 11.0 47.0 85 189.6 6 7 332
45 4 70.7 96.4 76 89 74.7 20.7 86 118.3 8 8 298
46 2 53.8 76.5 60 75 89.3 28.3 83 190.4 6 5 508
47 3 66 85.1 62 76 143.7 55.0 81 227.5 5 6 486
48 4 66.3 95.1 69 89 85..i 41.3 80 192.5 6 7 675
49 4 64.3 93.8 84 95 66.7 29 0 78 200.8 6 7 586
50 3 62.1 90.6 77 91 192.3 92.7 85 169.2 6 7 71 1
51 3 54 75.1 53 73 0.0 0.0 90 165.4 7 7 192
52 3 59.4 85.5 74 86 64.7 25.0 75 204.2 6 6 453
53 3 59.3 89 75 88 66.7 34.3 78 218.8 6 6 533
54 4 60.7 98.7 68 89 138 7 46.7 80 175.8 5 6 772
55 4 63.1 92.4 79 89 88.7 52.0 82 2 17.5 4 7 815
56 3 66.4 91.3 82 91 77.7 33.3 78 209.2 5 6 700
57 3 61.3 90.3 75 84 90.3 45.3 80 164.6 7 7 644
58 2 62.4 94.2 76 91 35.3 14.7 93 2 17.5 6 4 420
59 3 62.2 98.9 76 94 104.3 61.3 81 195.8 6 5 434
60 5 55.5 86.5 71 88 66.3 29.0 70 194.2 7 8 165
61 2 50 76.3 61 73 11.7 4.0 85 22 1.3 5 6 547
62 4 64.2 87.2 76 89 61.7 23.3 76 141.7 7 7 605
63 4 61.1 92.7 72 89 59.3 30.3 78 180.0 6 8 344
RIL No. Glossiness
Egg (%) l Egg (%) ll DH (%) I DH (%) II Tri-u Tri-l Floner plht ORS AS CR-Yi
256 4 60 7 89 2 76 87 80 0 38 0 88 205 0 5 7 459
g>o
w a$
5' z a,
2 g
E 3
- w 0
g
pro <
=:.$
eg Yi
8 5
3"
-% -&
O 9