2 Line Hybrid Rice Breeding

Two-Line Hybrid
Rice Breeding
Manual
S.S. Virmani
Z.X. Sun
T.M. Mou
A. Jauhar Ali
C.X. Mao
Two-Line Hybrid
Rice Breeding
Manual
S.S. Virmani
Z.X. Sun
T.M. Mou
A. Jauhar Ali
C.X. Mao
INTERNATIONAL RICE RESEARCH INSTITUTE

i
The International Rice Research Institute (IRRI) was established in 1960 by

the Ford and Rockefeller Foundations with the help and approval of the Government of the Philippines. Today IRRI is one of 16 nonprofit international
research centers supported by the Consultative Group on International Agricultural Research (CGIAR www.cgiar.org).
IRRI receives support from several CGIAR members, including the World
Bank, European Union, Asian Development Bank, International Fund for
Agricultural Development, Rockefeller Foundation, and agencies of the
following governments: Australia, Belgium, Canada, Peoples Republic of
China, Denmark, France, Germany, India, Islamic Republic of Iran, Japan,
Republic of Korea, The Netherlands, Norway, Philippines, Portugal, Sweden,
Switzerland, Thailand, United Kingdom, United States, and Vietnam.
The responsibility for this publication rests with the International Rice
Research Institute.
Copyright International Rice Research Institute 2003
Mailing address: DAPO Box 7777, Metro Manila, Philippines
Phone: +63 (2) 580-5600, 845-0563, 844-3351 to 53
Fax: +63 (2) 580-5699, 891-1292, 845-0606
Email: irri@cgiar.org
Home page: www.irri.org
Riceweb: www.riceweb.org
Riceworld: www.riceworld.org
Courier address: Suite 1009, Pacific Bank Building
6776 Ayala Avenue, Makati City, Philippines
Tel. (63-2) 891-1236, 891-1174, 891-1258, 891-1303
Suggested citation:
Virmani SS, Sun ZX, Mou TM, Jauhar Ali A, Mao CX. 2003. Two-line
hybrid rice breeding manual. Los Baos (Philippines): International Rice
Research Institute. 88 p.
Editing: Bill Hardy

Cover design, print production coordinator,
and page makeup and composition: George R. Reyes
Figures and illustrations: Emmanuel Panisales and George R. Reyes
Cover photos: IRRI photos
ISBN 971-22-0185-6
ii
Contents
FOREWORD
CHAPTER 1
Hybrid rice and heterosis breeding
Types of heterosis
How is heterosis measured?
Genetic basis of heterosis
Molecular basis of heterosis
Approaches for using heterosis
Methods of using heterosis
Hybrid rice
1
1
1
2
2
2
3
3
CHAPTER 2
Male sterility systems in rice
Cytoplasmic genetic male sterility
Hybrid seed production using the
CMS system
Environment-sensitive genic
male sterility
Chemically induced male sterility
5
5
5
7
12
CHAPTER 3
Comparative organization of twoand three-line hybrid breeding
programs
Similarities in three-line and two-line
breeding nurseries
Differences in three-line and two-line
breeding nurseries
15
15
16
CHAPTER 4
Inheritance of EGMS
Procedure for carrying out inheritance
studies on the EGMS trait
Inheritance of TGMS
Inheritance of PGMS and PTGMS
19
19
19
19
CHAPTER 5
Breeding procedures for developing
EGMS lines
Screening of existing varieties for EGMS
Induced mutagenesis
Hybridization method
23
23
23
24
CHAPTER 6
Characterizing EGMS lines under field
and controlled conditions
Characterization of EGMS lines under
field conditions
Characterization of EGMS lines under
controlled conditions
Evaluation of EGMS lines
31
31
32
34
CHAPTER 7
Developing pollen parents for two-line
hybrids
Characteristic features of an elite pollen
parent
Breeding methods for identifying pollen
parents for two-line hybrids
New strategies for developing pollen
parent lines
37
37
37
38
CHAPTER 8
Combining ability nursery
Definitions
Type of lines to be evaluated
Procedure using the line tester design
Composition of the combining
ability nursery
Field layout
Statistical analysis
41
41
41
41
41
41
41
iii
Analysis of variance
Interpretation of results
Using the results
42
44
44
CHAPTER 9
Evaluating two-line hybrids
Observation yield trial (OYT)
Preliminary yield trials (PYT)
Advanced yield trials (AYT)
Multilocation yield trials (MLT)
45
45
47
50
50
CHAPTER 10
Two-line hybrid rice seed production
Multiplication of EGMS lines
Differences between EGMS
and CMS line multiplication
Similarity of CMS and EGMS
line multiplication
High-yielding techniques for PGMS line
multiplication (Chinese experience)
High-yielding techniques for TGMS line
Three-line hybrid rice seed production
Two-line hybrid rice seed
production on a large scale
iv
CHAPTER 11
Two-line rice hybrids: maintenance
of genetic seed purity standards
1. Using nucleus seeds of EGMS
for seed production
2. Using anther culture for
purifying EGMS lines
3. Transferring a recessive
marker gene into EGMS lines
4. Insertion of a dominant marker
gene into the pollen parent
65
65
66
66
66
53
53
54
CHAPTER 12
Future outlook for two-line
rice hybrids
69
55
BIBLIOGRAPHY
AUTHORS
GLOSSARY
APPENDIX I: IDENTIFYING MALE STERILITY
APPENDIX II: PROTOCOL FOR ANTHER
71
76
77
85
87
55
55
CULTURE
56
62
APPENDIX III: DATA TO BE RECORDED

FOR HYBRID RICE EXPERIMENTS
88
Foreword
Hybrid rice technology has contributed significantly toward food security, environmental protection, and employment opportunities in China
for the past 25 years. Since the mid-1990s, this
technology has also been developed and introduced to farmers in India, Vietnam, the Philippines, Bangladesh, and the United States, either
independently or in close collaboration with IRRI.
Several other countries, such as Egypt, Indonesia,
Myanmar, Sri Lanka, Thailand, and the Republic
of Korea, are now developing this technology in
collaboration with IRRI.
The availability of adequately trained human
resources is an essential prerequisite for developing and using hybrid rice technology. Hybrid rice
breeding uses several concepts, skills, and procedures that are strikingly different from those used
for inbred rice breeding. Two male sterility systems (the cytoplasmic genic male sterility and environment-sensitive genic male sterility system)
have been used extensively to develop commercial rice hybrids in China and elsewhere. During
the past 25 years, IRRI and China have offered
several short-term training courses jointly and independently to develop the human resources in
countries interested in developing this technology. In 1997, IRRI also published a Hybrid Rice
Breeding Manual to serve the needs of those
training courses. This manual described concepts
and procedures to breed rice hybrids primarily
using cytoplasmic genic male sterility and the fertility restoration system. Since then, considerable
progress has been made in breeding rice hybrids
using environment-sensitive genic male sterility.

In August 2000, China and IRRI held a collaborative international training course on hybrid
rice breeding at the China National Rice Research
Institute in Hangzhou, which was funded by the
IRRI-ADB project on Development and Use of
Hybrid Rice in Asia. Several Chinese and IRRI
scientists participated in this course as resource
persons and provided the training materials focusing on two-line hybrid breeding using the environment-sensitive genic male sterility system.
Based on these materials, and the experience of
hybrid rice scientists from China and IRRI, this
training manual on the two-line hybrid breeding
method has been prepared, which expands upon
the hybrid rice breeding manual published earlier
by IRRI. The authors have described the concepts
and procedures stepwise and in a systematic manner so that trainees can learn them easily. This
should make an excellent manual for future hybrid rice breeding training courses organized at
IRRI, in China, and in other countries.
I compliment the authors for preparing this
training manual and thank Bill Hardy for editing
it. The assistance of the Asian Development Bank,
which provided financial support under RETA
6005 on Sustaining Food Security in Asia Through
the Development of Hybrid Rice Technology, is
gratefully acknowledged.
RONALD P. CANTRELL
Director General
CHAPTER 1
Hybrid rice and heterosis

breeding
Heterosis is a phenomenon in which F1 hybrids

derived from diverse parents show superiority over
their parents in vigor, yield, panicle size, number
of spikelets per panicle, number of productive
tillers, etc.
Heterosis is expressed in the first generation only.
Heterosis varies according to the level of
parental diversity and or presence of heterotic gene blocks in parental lines; indica
japonica crosses show maximum heterosis vis--vis any other combination between
other subspecies. The crosses showing heterosis in descending order are indica
japonica > indica javanica > japonica
javanica > indica indica > japonica
japonica > javanica javanica.
Heterosis can be positive or negative. Both
positive and negative heterosis can be useful depending on the trait, for example, positive heterosis for yield and negative heterosis for growth duration.
Farmers tend to use a lower seed rate for
hybrids than for conventional varieties because of their better seed quality and higher
seed cost. However, it is necessary to purchase fresh seeds every season to raise a commercial crop.
Types of heterosis
Heterosis is expressed in three ways, depending
on the reference used to compare the performance
of a hybrid:
Mid-parent heterosis is the increase or decrease in the performance of a hybrid in comparison with the mid-parental value.
Heterobeltiosis is the increase or decrease
in the performance of a hybrid in comparison with the better parent of the cross combination.
Standard heterosis is the increase or decrease
in the performance of a hybrid in comparison with the standard check variety of the
region.
From the practical viewpoint, standard heterosis is the most important because we aim to
develop hybrids that are better than the existing
high-yielding varieties grown commercially by
farmers.
How is heterosis measured?

Measurement of heterosis is quite simple. It is generally expressed as the percent increase or decrease
in the performance of a hybrid in comparison with
the reference variety or a parameter.
F1 mid-parent
Mid-parent
100
=
heterosis (%)
Mid-parent
Heterobeltiosis (%) =
F1 better parent
Better parent
100
F1 check variety
Standard
100
=
heterosis (%)
Check variety
Genetic basis of heterosis

Two major hypotheses have been proposed to
explain the genetic basis of heterosis: the dominance hypothesis (Davenport 1908) and overdominance hypothesis (East 1908, 1936).
Dominance hypothesis
This states that heterosis is due to the accumulation of favorable dominant genes in a
hybrid derived from the two parents (Fig.
1).
This was demonstrated in a pea hybrid
whose parents had different dominant genes
for node number and internodal length. The
hybrid was much taller than either parent.
The increased height was due to the accumulation of four dominant genes in the hybrid.
Overdominance hypothesis
This states that the heterozygote (Aa) is more
vigorous and productive than either homozygote (AA or aa). This has been proven
in traits controlled by a single or a few genes.
The heterozygote performs a given function, over a range of environments, more
efficiently than either homozygote (East
1936).
Studies on the genetic basis of heterosis for
polygenic traits in various crops have shown that
heterosis is the result of partial to complete dominance, overdominance, and epistasis, and it may
be a combination of all these (Comstock and
Robinson 1952). Evidence of real overdominance
for quantitative traits is hard to find. However,
apparent overdominance caused by nonallelic
interaction and linkage disequilibrium is a common contributor to heterosis (Jinks 1983).
Heterosis may also be due to the specific positive effects of the cytoplasm of the maternal parent on the nuclear component of the paternal parent. Differential heterosis observed between the
same pollen parent and cytoplasmic male sterile
(CMS) lines from different cytosterility sources is
an example of this kind of heterosis.
Molecular basis of heterosis

Several molecular studies support the overdominance hypothesis (Stuber et al 1992, Yu et al 1997,
2
P1
AAbbCCdd
(2 dominant genes)
P2
aaBBccDD
(2 dominant genes)
F1
AaBbCcDd
(4 dominant genes)
Fig. 1. Illustration of dominance hypothesis to explain
genetic basis for heterosis.
Li et al 2000) except for a few that support the

dominance hypothesis (Xiao et al 1995). Yu et al
(1997) reported overdominance at several maineffect quantitative trait loci (QTLs) and a stronger
additive epistasis affecting grain yield and its components in F3 progenies from the most widely grown
hybrid in China, Shan You 63. Zhang et al (2001)
demonstrated the involvement of large numbers
of two-loci interactions or epistasis as the genetic
basis of quantitative traits and heterosis. Furthermore, Li et al (2000) concluded that most QTLs
associated with inbreeding depression and heterosis in rice appeared to be involved in epistasis.
And almost 90% of the QTLs contributing to heterosis appeared to be overdominant. Zhang et al
(2001) assessed the relationship between gene
expression and heterosis by assaying the patterns
of differential gene expression in hybrids relative
to their parents in a diallel cross. The analysis revealed that differentially expressed fragments occurring in only one parent of the cross were positively correlated with heterosis and fragments detected in F1s but not in the respective parents were
negatively correlated with heterosis. A total of 384
fragments recovered from gels were hybridized
with mRNAs from seedling and flag-leaf tissues
and thereby Zhang et al (2000) detected an overall elevated level of gene expression in the hybrid
compared with the parents. Several fragments
showed a higher expression in the highly heterotic
hybrid than in the other hybrids.
Nonetheless, a lack of a clear understanding
of the genetic or molecular basis of heterosis has
not prevented plant breeders from exploiting this
phenomenon to raise crop yields.
Approaches for using heterosis

Currently, hybrid rice technology mainly uses
intrasubspecific heterosis, that is, indica indica
and japonica japonica. The high-yielding
intrasubspecific hybrids yield nearly 15% to 20%

more than the best inbred varieties grown under
similar conditions. It has been quite difficult to
create or widen the genetic difference among parents belonging to the intrasubspecific hybrids and
they have almost reached their yield ceiling.
Indica and japonica (both tropical and temperate) are the two main subspecies of Oryza sativa in Asia. Among them, the indica and temperate japonica subspecies are the most apparently
different in their morphological and agronomic
traits apart from the genetic distance between them.
Therefore, it is now well understood why the indica japonica hybrids show maximum heterosis. But intersubspecific heterosis is limited because of high spikelet sterility and long growth
duration. With the discovery of wide compatibility (WC) genes, it has been possible to exploit
intersubspecific hybrids with normal seed setting
and suitable growth duration. Efforts are under
way in IRRI, China, and India to develop
intersubspecific hybrids. Recently, Chinese scientists have developed super high-yielding rice
hybrids from crosses involving indica/indicajaponica derivative parents.
Most of the interspecific crosses in cultivated
species pertain to only O. sativa and O. glaberrima,
which are heterotic but not so useful in terms of
yield and plant stature. Most interspecific hybrids,
resulting from wide hybridization, result in genetic variability and bring in desirable genes for
resistance to several biotic and abiotic stresses,
for example, O. sativa O. longistaminata, O. sativa O. rufipogon, and O. sativa O. perennis.
In rice, the interspecific F1 hybrids cannot be used
commercially.
Methods of using heterosis

The three-line method is based on cytoplasmic
genic male sterility and the fertility restoration
system and involves three linesthe CMS line
(A), maintainer line (B), and restorer line (R)for
the commercial production of rice hybrids. The
seed of the male sterile line is multiplied by crossing A and B lines in an isolation plot. Hybrid seed
is produced by crossing the A line with an R line
in isolation in another plot. Seed production techniques are now developed to produce up to 3 t
ha1 (mean 1.2 t ha1) of hybrid seed in the tropics
and up to 6 t ha1 (mean 2.7 t ha1) in subtropical
and temperate regions of China.
In the two-line method, the two lines are involved in a cross for hybrid rice seed production.
One is a male sterile line in which male sterility is
genetically controlled by recessive genes, the expression of which is influenced by environment
(temperature, photoperiod, or both) and the other
is any inbred variety with a dominant gene for
that locus. The male sterile lines in which sterility
expression is controlled by temperature are known
as thermosensitive genic male sterile (TGMS) lines
and those in which expression is controlled by
daylength are called photoperiod-sensitive genic
male sterile (PGMS) lines.
Another two-line approach for hybrid rice
seed production is by spraying chemical hybridizing agents (CHAs)ethrel, ethyl 4 fluoro
oxanilate, or sodium methyl arsenatethat selectively sterilize the male reproductive organs of
any one parent and planting the other line (not
sprayed) close to the pollinator rows. China is the
only country that used CHAs such as sodium methyl arsenate and zinc methyl arsenate on a commercial scale. Because of the inefficient seed production related to nonsynchronous tillering and
flowering as well as health hazards associated with
the use of arsenic compounds, CHA use in China
was discontinued.
To use three-line and two-line rice hybrids,
farmers have to buy fresh seed every season. This
seed is produced by a proficient seed production
agency in the public or private sector.
The one-line method involves the use of apomixis to develop F1 hybrids. This represents true
breeding so that farmers can use the harvest from
the hybrids as seed for the next crop as with any
inbred rice variety. Attempts to discover apomixis
have not succeeded so far; however, research is still
under way at IRRI, in China, and in some other
countries using genetic engineering techniques.
Hybrid rice
What is hybrid rice?
Hybrid rice is the commercial rice crop grown from
F1 seeds of a cross between two genetically dissimilar parents.
Good rice hybrids have the potential of
yielding 1520% more than the best inbred
variety grown under similar conditions.
To exploit the benefits of hybrid rice, farmers have to buy fresh seeds every cropping
season.
3
Why hybrid rice?

The need for hybrid rice has been felt because
Yield levels of semidwarf varieties of the
Green Revolution era have plateaued.
The demand for rice is increasing rapidly
with the increase in population, especially
in less developed countries.
More and more rice has to be produced on
less land and with less inputs.
Hybrid rice varieties have already shown a
1520% higher yield potential than inbred
rice varieties under farmers field conditions
in several countries.
Hybrids have also shown an ability to perform better under adverse conditions of
drought and salinity.
How is hybrid rice developed?

Hybrid rice is developed by exploiting the phenomenon of heterosis. Rice, being a strictly selfpollinated crop, requires the use of a male sterility
system to develop commercial rice hybrids. Male
sterility (genetic or nongenetic) makes the pollen
unviable so that rice spikelets are incapable of
setting seeds through selfing. A male sterile line
is used as a female parent and grown side by side
with a pollen parent in an isolated plot to produce
a bulk quantity of hybrid seed because of cross
pollination with the adjoining fertile pollen parent. The seed set on male sterile plants is the hybrid seed that is used to grow the commercial hybrid crop.
CHAPTER 2
Male sterility systems

in rice
Male sterility can be defined as a condition in

which the pollen grain is unviable or cannot germinate and fertilize normally to set seeds.
The following genetic and nongenetic male
sterility systems are known for developing rice
hybrids (Fig. 2):
Environment-sensitive genic male sterility

Male sterility is controlled by the interaction of a
genetic factor S present in the cytoplasm and
nuclear gene(s). It is now known that the male
sterility factor S is located in the mitochondrial
DNA. A line is male sterile when the male sterility-controlling factor S in the cytoplasm and recessive alleles (rf ) of fertility-restoring genes are
present in the nucleus. The maintainer line (B line)
is iso-cytoplasmic to the CMS line since it is similar to it for nuclear genes but differs in cytoplasmic factor (N), which makes it self-fertile, but it
has the capacity to maintain the sterility of the A
line when crossed with it. A restorer or R line possesses dominant fertility-restoring genes (Rf) and
it is dissimilar to or diverse from the A line. Crossing a restorer line as a pollen parent with a CMS
(A) line as a female parent restores the fertility in
the derived F1 hybrid.
The restorer gene in the dominant homozygous (RfRf) or heterozygous (Rfrf) state can
restore the fertility in the F1 hybrid despite
the presence of sterility factors in the cytoplasm derived from the A line. The CMS
system is illustrated in Figure 3.
Hybrid seed production using

the CMS system
Hybrid seed production involves two steps: multiplication of the CMS line and production of
hybrid seeds.
Multiplication of the CMS line
This requires crossing of the CMS line with its
maintainer line by outcrossing by hand (for a small
quantity of seed) or in the field under isolation by
space or time (to produce a bulk quantity of seed).
For successful production of the CMS line, it is
grown in six or eight rows interspersed by two
rows of a maintainer line in an alternating manner.
Because there is a small difference between
the growth duration of A and B lines, their sowing
dates are adjusted to achieve good synchronization of their flowering. Several other techniques
(such as flag-leaf clipping, GA3 application, and
supplementary pollination by rope pulling or the
bamboo pole method) are used to improve the
outcrossing rate and seed yield of the CMS line.
Production of hybrid seeds
This involves the use of CMS lines with a selected
restorer line (R line) by growing them in a specific
female:male ratio in the field under isolation by
space or time. The CMS line is usually grown in
eight or ten rows interspersed with two rows of
restorer lines in an alternating manner. The sowing dates of A and R lines are staggered to achieve
synchronization of their flowering. To increase
the outcrossing rate and hybrid seed yield, the
techniques described in the previous section are
used.
5
Hybrid
Hybrid
NonTGMS
Normal
temp.
Low-temp.
seed
multiplication
TGMS
(normal temp.)
Non-TGMS
TGMS
TGMS
rTGMS
(low temp.)
5
NonrTGMS
Normal
temp.
Hybrid
High-temp.
seed
multiplication
Non-rTGMS
Reverse (r)
TGMS
rTGMS
Environment-sensitive genic
male sterility
Fig. 2. Classification of male sterility systems in rice.
Environmental
conditions for hybrid
seed production
Cytoplasmic genetic
male sterility
Genetic
Male sterility
NonPTGMS
NonPTGMS
Non-PTGMS
Hybrid
PTGMS
(normal
temp. and long
daylength)
Low-temp. and
shorter photoperiod seed
multiplication
PTGMS
PTGMS
NonPGMS
Normal
daylength
Non-PGMS
Hybrid
PGMS
(long daylength)
Shorter
photoperiod
PGMS
PGMS
rPGMS
(short daylength)
Longer
photoperiod
rPGMS
Hybrid
NonrPGMS
Normal
daylength
Pollinator
male parent
Hybrid
Female parent
(sprayed with
CHA)
Non-rPGMS
RPGMS
Induced by
chemical
hybridizing agents
(CHA)
Nongenetic
B line
A line
R line
rf
rf
rf
N/s
rfrf
rfrf
Rfrf
RfRf
N/s
B line
(fertile)
A line
(sterile)
Hybrid
(fertile)
R line
(fertile)
Fig. 3. Schematic description of the cytoplasmic genIc

male sterility system. N = cytoplasmic factor, S = male
sterility factor.
Environment-sensitive genic male sterility

This male sterility system is controlled by nuclear
gene expression, which is influenced by environmental factors such as temperature, daylength, or
both. This male sterility system was first observed
in pepper by Martin and Crawford in 1951 and
subsequently in different crops (Table 1). However, this system has been exploited commercially
only in rice because of the pioneering work of
Chinese scientists (Tables 2 and 3).
Advantages of the EGMS system
There is no need for a maintainer line for
seed multiplication, thus making seed production simpler and more cost-effective.
Table 1. Some reports on environment-sensitive genic male sterility systems in crop plants.
Crop
Environmental factor
Reference
Pepper
Temperature
Temperature
Temperature
Temperature
Temperature
Temperature
Temperature
Temperature
Temperature
Temperature
Daylength
Temperature
Daylength and temperature
Daylength
Temperature
Daylength
Temperature
Light intensity
Temperature
Temperature
Temperature
Temperature
Temperature
Daylength
Temperature
Copper, boron, and molybdenum
deficiency in soil
Boron deficiency in soil
Copper deficiency in soil
Martin and Crawford (1951)

Peterson (1958)
Daskaloff (1972)
Rundfeldt (1960)
Duvick (1966)
He et al (1992, c.f. Yuan 1997)
Rick and Boynton (1967)
Abdallah and Verkerk (1968)
Steven and Rudich (1978)
Sawhney (1983)
Fisher (1972)
Jan (1974)
He et al (1992, c.f. Yuan 1997)
Tan et al (1992, c.f. Yuan 1997)
Batch and Morgan (1974)
Sharma and Reinbergs (1976)
Ahokas and Hocket (1977)
Berthelem and Le Guen (1975)
Duc (1980)
Rudich and Peles (1976)
Shi (1981, 1985)
Zhou et al (1988), Sun et al (1989)
Maruyama et al (1991)
Virmani and Voc (1991)
Brar (1982)
Murty (1995)
Wei et al (1994)
Xi et al (1997)
Agarwala et al (1979, 1980)
Cabbage
Maize
Tomato
Wheat
Barley
Vicia faba
Cucurbits
Rice
Sesame
Sorghum
Soybean
Brassica napus
Wheat
Maize, barley, oats,

and sunflower
Rerkasem and Jamjod (1997)

Dell (1981)
Table 2. Origin and fertility-sterility transformation behavior of photoperiod- and temperture-sensitive male sterile
sources in rice.
Source
Varietal
group
Photoperiod-sensitive
male sterile (h)
Nongken 58S
MSr 54A (B)
CIS 28-10S
26 Zhai Zao
EGMS
M 201
Japonica
Japonica
Indica
Indica
Japonica
Japonica
Temperature-sensitive
male sterile (oC)
5460 S
IR32364
IR68945
Indica
Indica
Indica
IR68949
H 89-1
Annong 1S
R 59TS
Xianquang
26 Zhi Zao S
N5088 S
SM 5
SM 3
JP 2
SA 2
F 61
JP 8-1A-12
JP 24A
JP38
Dianxin /A
Hennong S
IV A
J207S
CDLa (h)/CSP/
CFP (oC)
Origin
Spontaneous mutation, China

Induced (R), China
Induced (C), USA
Inducec (R), USA
Induced (R), China

Induced (R), IRRI
Introgression from Norin PL 12,
Japan
Indica
Introgression from Norin PL 12,
Japan
Japonica Induced (R), Japan
Indica
Indica
Induced (R), China
Indica
Breeding population, China
Indica
Induced (R), China
Indica
Introgression from Nongken 58 S,
China
Indica
Spontaneous, India
Indica
Spontaneous, India
Indica
Spontaneous, India
Indica
Induced mutation (C) India
Indica
Induced mutation (C) India
Indica
Breeding population, India
Indica
CMS, India
Indica
Spontaneous mutation, India
Japonica CMS, China
Indica
Cross breeding, China
Indica
Cross breeding, China
Indica
Reference
14.0013.45
14.0013.00
14.0012.00
14.0012.00
14.0013.00
14.0012.00
Shi and Deng (1986)

Lu and Wang (1988)
Huang and Zhang (1991)
Shen et al (1994)
Rutger and Schaeffer (1989)
Oard and Hu (1995)
28.026.0
32.024.0
30.024.0
Yang et al (1990)
Virmani and Voc (1991)
Virmani (1992)
30.024.0
Virmani (1992)
31.028.0
30.227.0
30.024.0
23.025.0
30.022.0
Maruyama et al (1991)
Tan et al (1990)
Yang and Wang (1990)
Cheng et al (1995)
Shen et al (1993)
Zhang et al (1994b,c)
32.322.0
32.022.0
33.923.0
31.720.0
30.922.0
30.920.0
33.823.0
24.030.5
23.020.0
30.029.0
24.028.0
31.0>31.0
Ali et al (1995)
Ali et al (1995)
Ali et al (1995)
Ali et al (1995)
Ali et al (1995)
Ali et al (1995)
Ali (1993)
Ali (1993)
Lu et al (1994)
Lu et al (1994)
Zhang et al (1991)
Jai et al (2001)
a
CDL = critical daylength, CSP = critical sterility point, CFP = critical fertility point, R = irradiation, C = chemical mutagen. Several introgressed forms
from Nongken 58S and Annong 1S developed by Yang (1997) and Mou et al (1998) not included here.
Any fertile line can be used as a pollen par-
Since there is no need for restorer genes in
ent (PP); therefore, the frequency of heterotic

hybrids is higher among two-line hybrids
than among three-line hybrids, thereby increasing hybrid breeding efficiency.
Negative effects of sterility-inducing cytoplasm are not encountered.
The EGMS trait is governed by major genes,
thus enabling their easy transfer to any genetic background and thus increasing diversity among the female (EGMS) parents,
which helps in reducing potential genetic
vulnerability among the hybrids.
the male parents of two-line hybrids, this

system is ideal for developing indica/
japonica hybrids because most japonica
lines do not possess restorer genes.
Disadvantages of the EGMS system

Since the sterility trait is conditioned by
environmental factors, any sudden change
such as temperature fluctuation because of
a thunderstorm, typhoon, etc., will influence
the sterility of EGMS lines.
Table 3. Two-line rice hybrids released up to 2001 in China.

Hybrid
Pedigree
Ejingza No.1
Huajingza No. 1
Huajingza No. 2
70 you 9
70 you 99
70 you 04
Peiliangyou Teqing
Peiliangyou 288
Peiliangyou Yuhong
Liangyou Peiju
Liangyou 923
Liangyou 681
Xiangliangyou 68
8 Liangyou 100
Tainliangyou 402
An Liangyou 25
Peiza Shangqing
Jinliangyou 36
Peiza Shuangqi
Liangyou 2163
Liangyou 2186
Fu Liangyou 63
Pei Liangyou 275
Pei Liangyou 99
Peiza Maosan
Peiza Maoxuan
South China/first and
second Peiza 67
Yunguang No. 8
Type
Year of release
Cultivation region/crop
N5088S/R187
7001S/1514
N5088S/65396
7001S/Wanhui 9
7001S/99
7001S/Xiushui 04
Pei-ai 64S/Teqing
Pei-ai 64S/R822
Pei-ai 64S/Yuhong No. 1
Pei-ai 64S/9311
W9593S/Shengyou No. 2
Shuguang 612S/881
Xiang 125S/D 68
Annong 810S/D100
TianfengS/R 402
1356 S/Zao 25
Pei-ai 64S/Shanqing
HS-3/946
Pei-ai 64S/Shuangqizhan
SE21S/Minghui63
SE21S/Minghui86
FJS-1/Minghui63
Pei-ai 64S/275
Pei-ai 64S/Gui99
Pei-ai 64S/Maosan
Pei-ai 64S/Maoxuan
Pei-ai 64S/G67
Japonica
Japonica
Japonica
Japonica
Japonica
Japonica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
Indica
1995
1995
2001
1994
1997
1994
1994
1996
1997
1999
2001
1999
1998
1998
1998
1998
1997
2000
1998
2000
2000
2000
1999
1998
2000
2000
2000
Yangtze Valley/second
Yangtze/single or second
Yangtze/first
Yangtze/first
Yangtze/first
Yangtze/first
South China/first and second
N5088S/Yunhui 11
Japonica
2000
West China/single
The multiplication of EGMS lines and hybrid seed production are restricted by space
and season. This means that an EGMS line
is used in a given region and season.
Characteristic features of ideal EGMS lines
The proportion of male sterile plants in a
population of more than 1,000 plants during the critical sterility period should be
100%.
The pollen sterility of each male sterile plant
should be more than 99.5%.
EGMS lines should have clearly defined sterility-fertility alteration regimes.
The male sterile phase should last for more
than 4 consecutive weeks.
Seed setting during the fertile phase should
be more than 30%.
The critical temperature or photoperiod for
inducing sterility should be as low as possible for more stability of the EGMS lines,
for example, <27 C (maximum) or <13 h

photoperiod.
In addition, these lines should have a good
plant type possessing high yield, acceptable grain
quality, pest and disease resistance, and adaptability to the target environment.
Classification of the EGMS system
Depending on the environmental factor(s) influencing expression of the sterility-inducing gene(s),
EGMS is classified in the following categories
(Figs. 47):
1. TGMS: temperature-sensitive genic male
sterility
2. rTGMS: reverse temperature-sensitive
genic male sterility
3. PGMS: photoperiod-sensitive genic male
sterility
4. rPGMS: reverse photoperiod-sensitive
genic male sterility
5. PTGMS: photothermosensitive genic male
sterility
9
Seed setting (%)

70
60
Seed setting (%)

25
30.1 C
24.1 C
23.1 C
20
50
15
40
30
10
20
10
0
30.1 C
24.1 C
23.1 C
15.0 h
14.0 h
12.5 h
15.0 h
14.0 h
12.5 h
Fig. 4. Fertility of PGMS line N9044S in phytotron

conditions.
Fig. 7. Low fertility of PTGMS line Miai 64S in phytotron

conditions.
Seed setting (%)

35
TGMS lines are sensitive to the temperature

for the expression of male sterility or fertility. For
example, most TGMS lines remain male sterile at
high temperature (day temperature >30 C/night
>24 C) and they revert back to partial fertility at
a lower temperature (day <24 C/>16 C night),
for example, 5460S, IR68945, H89-1, and SA2.
Reverse TGMS (rTGMS) lines are sensitive
to low temperature (<24 C day/>16 C night) for
the expression of male sterility, whereas, at a higher
temperature (>30 C day/24 C night), they become male fertile, which is just the reverse of the
TGMS system, for example, JP 38, Dianxin 1A,
and IVA.
PGMS lines are sensitive to the duration of
daylength for the expression of sterility or fertility. For example, most PGMS lines remain male
sterile under long-day (>13.75 h) conditions and
revert back to fertility under short-day (<13 h)
conditions, for example, N9044S and N5088S.
PGMS lines that express sterility under short
daylength and fertility under long daylength are
known as reverse PGMS (rPGMS). This category
is yet to be found.
PTGMS lines are sensitive to both photoperiod and temperature. Temperature is the key factor since PTGMS lines become completely male
sterile or fertile beyond a particular temperature
range, that is, >30 C or <24 C, without any influence of photoperiod. But, within this temperature
range (2432 C), photoperiod influences the
PTGMS lines, that is, longer photoperiod hours
will enhance male sterility at lower temperatures
30
30.1 C
24.1 C
23.1 C
25
20
15
10
5
0
15.0 h
14.0 h
12.5 h
Fig. 5. Fertility of TGMS line W9046S in phytotron

conditions.
Seed setting (%)

25
20
30.1 C
24.1 C
23.1 C
15
10
5
0
15.0 h
14.0 h
12.5 h
Fig. 6. High fertility of PTGMS line Xinguang S in

phytotron conditions.
10
vis--vis a shorter photoperiod (i.e., 14 h at 30 C

will make the PTGMS line sterile in comparison
with 13 h at 30 C), for example, Nongken 58S,
Xinguang S, and Miai 64S. Figure 7 characterizes
the model for PTGMS lines.
Male sterility expression in EGMS lines is
governed by a single nuclear recessive gene or
pair of nuclear recessive genes that are sensitive
to environmental conditions such as photoperiod,
temperature, or a combination of both.
Under natural conditions, there is a constant
interaction of photoperiod and temperature and it
is therefore difficult to separate the effects of photoperiod and temperature on fertility. Using statistical methods, you can separate the effects of
photoperiod (P), temperature (T), and P T on the
fertility of EGMS lines. EGMS lines can also be
classified by evaluating them in a combination of
photoperiod and temperature treatments in a twoway factorial analysis. Because of the limitation
of available phytotrons, the best combinations of
P and T should be selected.
While setting up P and T treatments, you need
to consider (1) the characteristics of fertility responses to P and T in the EGMS line and (2) the
ecological conditions in the target area in which
the EGMS system will be deployed. In China, several EGMS lines were evaluated in a combination
of three P and three T treatments (see Table 4).
Data processing
Pollen and spikelet fertility of the test entries must
be analyzed on time. Pollen samples must be collected at the time of heading by taking five apical
spikelets from the panicle on the primary tiller
from each plant and immersing them in a prepared
fixative solution (alcohol:acetic acid, 3:1). Percent fertile pollen is observed under a microscope
using the IKI staining procedure (Appendix 1). At
Table 4. Photoperiod and temperature conditions in

the phytotrons for evaluation of EGMs lines in China
from 1993 onward.a
Low temp.
(oC)
Time
05000800
08001100
11001500
15001800
18000100
01000500
Daily average
22.0
25.0
27.0
25.0
22.0
19.0
23.1
Medium temp.
(oC)
23.0
26.0
28.0
26.0
23.0
20.0
24.1
High temp.
(oC)
29.0
32.0
34.0
32.0
29.0
26.0
30.1
Illumination time: 12.5 h (06001830), 14.0 h (05301930), 15.0 h

(05002000).
the same time, panicles on primary and secondary

tillers that head synchronously with the main stem
(45 days) are bagged. At maturity, spikelet fertility percent is calculated by the number of filled
spikelets divided by total spikelets per panicle,
multiplied by 100.
Both pollen and spikelet fertility data from
the test EGMS lines need to be transformed in the
form sin1 x before analyzing them further.
Two-factor variance analysis can be used to
measure P, T, and P T effects (Table 5).
EGMS lines can be classified on completion
of the variance analysis as summarized in Table 6.
Different types of EGMS are characterized in
the next section.
Characteristics of different EGMS types
PGMS. Fertility change in PGMS lines is characterized by significant P and P T interaction effects (P<0.05) and nonsignificant T effects. PGMS
lines are completely or highly sterile under the
combinations of long daylength with high or low
temperature and highly male fertile under the com-
Table 5. Results of two-factor variance analysis.

Source of
variance
Degrees of
freedom
Sum of
squares
Mean of
squares
Calculated F
value
Tabular F0.05
value
Replication
r1
Treatment
pt 1
P
p1
T
t1
PT
(p 1)(t 1)
Error
(r 1)(pt 1)
Total
rpt 1
11
Table 6. Classification of EGMS lines based on statistical

significance of effects of various treatments.a
Photoperiod Temperature
(P)
(T)
*
*
ns
ns
ns
a
ns
ns
*
*
ns
PT
*
ns
*
ns
*
Classification
category
PGMS
TGMS
PTGMS
* = significant at the 5% level and ns = nonsignificant at the 5% level.
bination of short daylength with low temperature

in the phytotron. Nongken 58S and most japonica
EGMS lines derived from it belong to this group.
An example of a typical PGMS line is N9044S.
Figure 4 shows its behavior in response to P and T
treatments.
PGMS lines are stably male sterile in the summer and high in fertility recovery in autumn under natural conditions in temperate regions. In
Central China, most PGMS lines showed stable
sterility, making them suitable for hybrid seed
production in the summer and easy to be self-multiplied in autumn. But low temperature under long
daylength or high temperature under short
daylength induces partial fertility in such lines.
TGMS. Most indica EGMS lines studied belong to this group. These lines are characterized
by nonsignificant P effects and significant T effects. TGMS lines are completely or highly sterile
under high temperature and highly fertile under
low temperature irrespective of photoperiod in the
growth chamber. They are stable in male sterility
in the summer and high in fertility recovery in
autumn in the northern hemisphere under natural
conditions in the tropical rice-growing regions but
unstable in temperate regions. Typical examples
of TGMS lines are W9046S (in China), Norin PL
12 (in Japan), and IR32364TGMS (at IRRI). Figure 5 shows the behavior of W9046S (an indica
TGMS line derived from Nongken 58S) in response
to P and T treatments.
TGMS lines have limited utility in temperate
conditions because of their unstable sterility in
the summer caused by the occurrence of sudden
low temperatures. However, such lines can be easily multiplied with high yields in autumn because
of the occurrence of low temperature. In tropical
regions, where abnormally low temperature seldom occurs, TGMS lines can be used to produce
hybrid seeds. These lines can be multiplied at
12
high-altitude areas and/or in a season when the

requisite temperatures occur under natural conditions for a prolonged period. Practically speaking, TGMS lines can be used in the plains during
the summer for hybrid seed production and these
can be multiplied in a mountainous area or in a
season when the temperatures are low. Chinese
scientists have also found that irrigating TGMS
lines with deep cool water (with temperature >17
24 C) at the critical stage also helps to induce the
fertility required for their seed multiplication.
P-TGMS. PTGMS lines show nonsignificant
effects for both photoperiod and temperature but
significant effects for P T. Fertility changes in
PTGMS lines are closely related to the combination of photoperiod and temperature. These lines
are fertile only under the combination of short
daylength and low temperature and are completely
or highly sterile under all other combinations in
the phytotron. About half of the EGMS lines studied belong to this group and most of them had
originally been classified as PGMS. Xinguang S
and Miai 64S are typical examples. Their behavior in response to P and T treatments is given in
Figures 6 and 7, respectively.
PTGMS lines have stable sterility in the summer in large areas of China and can be used for
hybrid seed production. However, PTGMS lines
are difficult to multiply because of low fertility
induction, which limits their wide use. Like TGMS
lines, their multiplication is affected by abnormally high temperature in autumn.
Validation under natural conditions. To use
different EGMS lines in the two-line hybrid rice
breeding program safely, carry out a validation
test under natural conditions to determine the best
dates for hybrid seed production and EGMS seed
multiplication.

Since the early 1970s, attempts have been made
to identify and use potential chemical hybridizing agents (CHAs) for hybrid rice seed production. Various chemicals tried so far include ethylene-releasing compounds, highly carcinogenic
arsenic compounds, and growth hormones (Table
7). China is probably the only country where
gametocides were used in commerical hybrid seed
production, but their use has been reduced because they were found to be unsafe for human
health. Rice hybrids developed by using CHAs
Table 7. List of chemical hybridizing agents and their efficacy on the rice plant.
Chemical
Male sterility induction
Ethrel
Ethrel + isourea
Partial
Partial
Partial
Partial
High
Partial
Partial
High
Arsenates
Sodium methyl arsenate (MG2)
Zinc methyl arsenate (MG1)
Monosodium methane arsenate
Complete
Complete
Complete
Chen et al (1986), Ali (1990, 1993)

Anonymous (1978)
Wang and Que (1981)
Oxanilates
Ethyl 4 flouro oxanilate
Ethyl 4 metho oxyoxanilate
Ethyl 4 bromo oxanilate
Ethyl 4 chloro oxanilate
Complete
Complete
High
High
Ali
Ali
Ali
Ali
Complete
Prevents anther dehiscence
High to complete
Perez et al (1973)
Long et al (1973)
Zhangzing and Chunnong (1980)
Complete
Complete
High
Complete
Complete
Complete
Partial to high
High
Wang et al (1981)
Takeoka et al (1990)
Chen (1985)
Luo et al (1988)
Song et al (1990)
Wang et al (1991a,b,c)
Atsumi et al (1992)
Astumi et al (1992)
Other chemicals (with and without codes)

RH 531
DPX 3778
3-(-p-chlorophenyl) 6-methoxy-s-triazine
2-4 (1H,3H) dione tri-ethanolamine
Sodium sulfate
HRG 626
Ammonia sulfonic acid
HAC 123 + N 312
MHC
CRMS
Kasugamycin
AOA
to
to
to
to
high
high
high
high
to high
to high
have been tested along with 3-line bred hybrids

and were reported to give consistently comparable
and often higher yields. Over the years, seed yields
have increased from 0.4 t ha1 with 4060% seed
purity to 1.5 t ha1 with 8090% seed purity. CHAs
must be able to selectively induce total male sterility. The effectiveness of CHAs is highly stagespecific (i.e., these should be applied at the stamen and pistil primordia formation stage or stage
IV) and genotype-specific (i.e., the gametocidal
effect varies from variety to variety). In India,
oxanilates, when sprayed at stage IV (meiotic
stage) of rice development, were found to be effective and variety Pusa 150 was sterilized more
effectively by the gametocidal spray than other
varieties, thus indicating genotype specificity (Ali
1993). Figure 8 illustrates an example of two-line
hybrid rice seed production using a CHA.
Reference
Perez et al (1973)
Cheng and Huang (1978, 1980)
Parmar et al (1979)
Chan and Cheah (1981)
Wang and Que (1981)
Kaul (1988)
Song et al (1990)
Kitaoka et al (1991)
(1990,
(1990,
(1990,
(1990,
1993)
1993)
1993)
1993)
Properties of an ideal CHA

An ideal CHA should have the following properties:
1. Wide-spectrum action to induce sterility in
successively emerging panicles.
2. Selective and total sterilization of stamens
without affecting ovular fertility.
3. Be less phytotoxic, noncarcinogenic, and
without residual toxicity that could harm
human beings and animals.
4. Be easy to apply and economical.
Advantages of the two-line approach
via CHA
1. A wide range of varieties can be used for
making superior hybrid combinations.
2. The method of seed production is simpler
than that of three-line breeding as it does
13
X1
F1
Y1, X2
Y2, X3
F1
Y3,... Xn 5 Yn
F1
F1
Select heterotic combinations (e.g., X2 5 Y2)
Sprayed
with CHA
Sprayed
with CHA
Sprayed
with CHA
XXOOOOOOOOXXOOOOOOOOXXOOOOOOOOXX
Y2
X2
Y2
X2
Y2
X2
Y2
(pollen
parent)
(seed
parent)
Harvest 2-line hybrid from seed parent
Fig. 8. Schematic description of the use of a CHA to

develop two-line hybrids.
14
not require development of three lines (A,

B, and R).
3. If the CHA is not effective because of
nonsynchronization of flowering or continuous rain during the critical stage, heavy
crop losses can still be averted because the
yield of the sprayed unaffected female would
still be good enough.
4. Partial CMS lines and EGMS lines can be
made completely sterile with a spray of an
ideal CHA.
5. The narrow genetic base for cytoplasmic
genic male sterility, inherent in three-line
rice hybrids, ceases to be a problem in CHAderived hybrids.
Disadvantages of CHAs
1. Production of impure hybrid seeds if the
CHA is not effective because of unfavorable weather conditions or nonsynchronized
tillering and growth.
2. Health hazards of some CHAs (such as zinc
methyl arsenate or sodium methyl arsenate).
3. High cost of the chemicals.
CHAPTER 3
Comparative organization
of two- and three-line hybrid
breeding programs
Organization of a hybrid breeding program is strikingly different from the conventional inbred breeding program in rice. Three-line and two-line hybrid rice breeding have some similarities as well
as differences in their organization. These are described in this chapter.
Similarities in three-line and two-line

breeding nurseries
The breeding nurseries that are identical in both
breeding programs are as follows.
Source nursery
This nursery contains elite lines that have the potential to become parents of commercial hybrids.
Include the best available CMS, TGMS, and PGMS
lines in this nursery. To raise the source nursery,
grow 20 plants (with a single seedling per hill) per
line in rows. Use these as single plants for testcrosses with the best available CMS or EGMS
lines. To synchronize flowering and to make as
many testcrosses between elite lines as possible
and to make sterile lines, plant the sterile lines on
three dates with a 10-day interval.
Testcross nursery
This nursery contains the testcrosses made in the
source nursery along with the single-plant progenies of the male parents used to make the testcrosses. After growing every ten pairs of testcross
and the corresponding male parent progenies,
grow single rows of inbred and hybrid checks. In a
three-line hybrid breeding program, screen the
testcross progenies for pollen sterility/fertility,
spikelet fertility, and other agronomic traits to
identify potential maintainers and restorers and
heterotic hybrids. In the two-line hybrid breeding

program, screen the testcross progenies for pollen
fertility, spikelet fertility, and other agronomic
traits to identify the potential elite male parents
and heterotic hybrids. Plant the F1 and their male
parents side by side in three rows each with a single
plant per hill.
This breeding nursery contains a set of crosses
derived from promising CMS and restorer lines
(in the three-line breeding program) and promising EGMS and elite pollen parent lines (in the
two-line breeding program) made in a line tester
design to evaluate the parental lines for their combining ability, that is, their ability to produce superior progenies when crossed with several male
parents.
Yield evaluation nurseries
The procedure adopted for heterosis evaluation
for two- and three-line breeding is similar. It contains the following five different trials:
1. Observational yield trial (OYT). This contains hybrids derived from commercially
usable CMS lines (showing stable pollen
sterility, high general combining ability,
and good outcrossing potential) and promising restorer lines identified in the testcross
nursery. The trial also contains inbred check
varieties of different growth duration.
Groups of test hybrids are compared with a
set of inbred checks and hybrids. Thus, the
test hybrids are unreplicated, whereas the
check varieties and hybrids are replicated
across different groups. The OYT uses an
augmented design in which test hybrids are
15
arranged in groups. Each plot is about 6 m2,

in which 125175 single plants per hill are
included. Data on agronomic traits, yield,
and disease and insect pest resistance are
recorded.
2. Preliminary yield trial (PYT). The promising hybrids that yield 1520% higher than
the check varieties in the testcross nursery,
combining ability nursery, and OYT are forwarded for further evaluation in the PYT.
Hybrids are grouped according to their maturity duration. This trial uses a randomized
complete block design with 34 replications, with an individual plot size of 7 m2.
Data on agronomic traits, yield, disease and
insect resistance, and grain quality (of heterotic hybrids only) are recorded. The plot
design and data recorded are similar to those
of the OYT.
3. Advanced yield trial (AYT). The promising hybrid entries from the PYT are included
in the AYT. The plot design and data recorded are similar to those of the PYT, but
the plot size is 10 m2.
4. Multilocation trials (MLT)/national hybrid trial (NHT). The outstanding hybrid
entries in the AYT are nominated into the
MLT/NHT. The experimental design and
data collected remain the same as in the AYT,
except that the plot size is increased to 15
20 m2.
5. On-farm testing (OFT). One to two hybrids
performing consistently better than the regional check and local checks over 23
years are recommended to the OFT. The plot
area is about 0.1 ha, with or without replication. The best local check variety (hybrid)
is used as the control.
Hybrids that pass all the above trials and have
resistance to major insect pests and diseases and
acceptable grain quality are registered and/or released.
16
Differences in three-line and two-line

breeding nurseries
1. Male sterile maintenance and evaluation
nurseries
In three-line breeding, CMS lines are maintained
and evaluated in a CMS line maintenance and
evaluation nursery by growing A and B lines side
by side. Their pollen sterility/fertility is monitored
and lines are evaluated on a single-plant basis for
phenotypic acceptability (on a 19 scale, where 1
= excellent and 9 = poor), days to 50% flowering,
outcrossing ability, and other desired traits. The
CMS lines are maintained by hand-crossing of A
and B lines (on a single-plant basis) to maintain
their purity. Hand-crossed seeds of the commercially usable CMS lines are produced in larger
quantities (5001,000 seeds) so that they can be
used for nucleus seed production.
In the two-line breeding program, EGMS lines
are maintained by selfing by growing them in appropriate daylength and temperature conditions
that induce fertility. These EGMS lines are evaluated for sterility separately by growing them under suitable daylength and/or temperature conditions in the phytotron and/or field. Simultaneously, they are also evaluated for phenotypic
acceptability, flowering, outcrossing rate, and
other desired traits in comparison with suitable
checks (such as popular commercial varieties,
popular CMS lines, and the best available EGMS
lines), which should be grown side by side.
2. EGMS breeding nurseries
These are described in the next chapter. A comparative flow chart of two-line and three-line hybrid rice breeding nurseries appears in Figure 9.
17
F1 seed for commercial

cultivation
Release for cultivation
On-farm testing of hybrids
Hybrid seed production for on-farm trials
Hybrid seed production for on-farm trials

Foundation seed
production of pollen
parent by seed
production agencies
Seed production
agencies
Breeder seed
of A and B
lines
Seed increase of
elite A and B lines
CMS line
maintenance
and evaluation
New A and B
lines
Backcross nursery
to develop A lines
Elite CMS lines
For CMS
F1 seed for commercial

cultivation
Seed production for on-farm trials
AYTs replicated
multilocation
Seed production for AYT
PYTs replicated
Seed production
agencies
Seed increase of
elite lines
Restorer evaluation
nursery
Lines from restorer and

maintainer breeding nursery
Elite lines from

different sources
Seed production of elite hybrids for PYT
OYTs
Evaluation of A/R hybrids
Seed production
for OYT
Combining ability
nursery
Testcross nursery:
to identify B & R lines
Source nursery: to
evaluate and testcross
Fig. 9. Operational flow chart of hybrid rice breeding using EGMS and CMS lines. OYT = observation yield trial, PYT = preliminary yield trial, AYT = advanced
yield trial, MLT = multilocational trial.
Foundation seed
production of EGMS
lines by seed
production agencies
Evaluation of hybrids in AYT/MLT
Hybrid seed production for AYT
Evaluation of hybrids in OYT/PYT
F1 seed production for OYT/PYT
Breeder seed production of

pollen parent
Nucleus seed multiplication of

pollen parents
Nucleus seed multiplication of

EGMS lines
Breeder seed production

of EGMS
Pollen parent source nursery
EGMS evaluation nursery
For EGMS
CHAPTER 4
Inheritance of EGMS
Procedure for carrying out inheritance

studies on the EGMS trait
The following stepwise procedure explains how
inheritance studies on the EGMS trait are conducted:
Purify the EGMS lines and fertile parents
through bagging.
Prepare all generations for genetic analysis
and obtain sufficient seeds. The generations
involve P1, P2, F1, BC1F1, BC2F1, and F2.
Plant different generations with 1020 P1,
P2, and F1 plants, 60120 BC1 plants, and
more than 200 F2 plants.
Observe the pollen and spikelet sterility of
each plant by putting the plants under appropriate sterile environmental conditions
during the sensitive stage.
Classify the sterile:fertile plants accordingly.
Use the 2 test to analyze the segregation
pattern and determine the genes that control EGMS.
Inheritance of TGMS
Genetic studies in Japan, at IRRI, and in India
(Maruyama et al 1991, Borkakati and Virmani
1993, Ali 1996) indicated that the TGMS trait in
Norin PL 12, IR32364S, and SA 2 was controlled
by a single recessive gene. Allelic relationship
studies indicated that the TGMS genes in Norin
PL 12 and IR32364S mutants were different.
TGMS line 5460s developed in China carries the
tms1 allele on chromosome 8, whereas IR32364S
carries the tms3 allele on chromosome 6. The
TGMS allele in Norin PL 12 is designated as tms2
(Fig. 10). On the basis of this information, other

TGMS sources from India have also been studied
for allelic relationships and Reddy et al (2000)
found a new nonallelic TGMS trait designated as
tms4 in SA 2, a sodium azide-induced TGMS mutant. Other details are given in Table 8.
The transfer of the Norin PL 12 gene (tms2)
into IRRI cultivars such as IR68945S, IR68949S,
and IR68294S revealed varied sterility-fertility
altering conditions, indicating a change in the
expression of the TGMS trait in a different genetic background.
Inheritance of PGMS and PTGMS

Single-locus genetic model
In an initial study, many conventional japonica
and indica varieties were reciprocally crossed to
Nongken 58S (NK58S) and all the F1s were fertile.
In F2 populations of reciprocal crosses between
NK58S and three conventional japonica rice lines,
a 3:1 ratio of fertile:sterile plants was observed in
each population grown under long-day conditions.
The sterility was therefore considered to be controlled by a single recessive nuclear gene. Yang et
al (1992a) designated the PGMS gene as Ps. It was
also noted that minor genes might also be responsible for the sterility, as few sterile F2 plants were
100% sterile. Many other studies showed that fertility segregation in crosses between NK58S and
NK58 was controlled by a single nuclear gene.
Two-loci genetic model
Many studies showed that PTGMS in NK58S-derived P(T)GMS lines was conditioned by recessive alleles at two loci.
The two-loci genetic control of PTGMS was
19
20
Chromosome 9
OPA12
TS200
EAA/MCAG
tms4
RM257
EACT/MCAG
RZ649
RG136
RZ66
RG1
RG978
TGMS 1.2
tms1
RG333
RZ562
RG20
tms1 (Wang et al 1995)
9.3
7.2
7.6
9.0
12.7
12.6
3.2
29.7
cM
Chromosome 8
Fig. 10. Location of TGMS genes on the linkage map in rice.
tms4 (Reddy et al 2000)
23.1
1.8
5.3
6.2
17.3
cM
R646(D23951)
R1440(D24156)
tms2
R643A(23948)
R1788(D24362)
tms2 (Yamaguchi et al 1997)
0.4
1.7
1.0
cM
Chromosome 7
OPB19-750
OPAA7-550
OPAC3640
tms3(t)
OPF18-2600
RZ2
RZ398
RZ516
tms3(t) (Subudhi et al 1997)
10.6
10.0
7.7
2.4
16.7
5.7
10.6
cM
IR32364TGMS/IR38
Chromosome 6
Table 8. Different alleles for the TGMS trait among

available TGMS sources.
Gene symbol
TGMS source
tms1
tms2
tms3
JP 1 and 5460S
IR68945 (Norin PL 12)
JP 8-8-1-S, 1C 10,
IR32364
SA2
Annong S-1
J207 S
tms4
tms5
rtms1 (reverse TGMS)
well demonstrated in the work of Mei et al (1990).

The F1, F2, and BC1F1 populations were derived by
crossing NK58, NK58S(R), 1149, and NK58S with
each other. NK58S(R) is a fertility revertant mutant of NK58S, whereas l149 is a fertile line developed from the cross of NK58S and japonica variety Nonghu 26. A 2 test indicated that sterility in
NK58S was conditioned by recessive alleles at
two loci. The fertility-restoring alleles in
NK58S(R) and l149 are allelic to each other, but
they are nonallelic to that in NK58.
Designating one gene as msph and the other as
ph
rf , the genotypes of NK58S, NK58, and NK58S
(R8), a fertility revertant mutant of NK58S, were
the following:
Nongken 58S: msph msph rfph rfph
Nongken 58: Msph Msph rfph rfph
Nongken 58S(R): msph msph Rfph Rfph
Complete dominance is displayed at the Msph
locus and plants possessing Msph would be fertile.
When partial dominance is displayed at the Rfph
locus, plants possessing Rfph alone would be partially fertile.
Diverse genetic model
Results obtained from crosses between NK58S and
japonica varieties other than NK58 were quite
complex. Although the fertility segregation in the
majority of crosses was attributed to a single dominant gene or two dominant genes, it was also evident that multiple genes were involved in the genetic control of PTGMS.
The two genes segregated independently and
the interaction between them varied greatly depending on the genetic background. Different segregation patterns such as 9 fertile:6 partially fertile:1 sterile, 9 fertile:3 partially fertile:3 partially
sterile:l sterile, and 9 fertile:3 partially fertile:4
sterile were observed (Sheng 1992). A bimodal
distribution was generally observed in the fertil-
ity segregation but, on close scrutiny, the segregation was found to be continuously distributed
in each population. Mei et al (1990) showed that
the fertility segregation ratio varied greatly depending on the criteria for the classification of
sterile and fertile plants. Although Mei et al (1990)
suggested that PTGMS was quantitatively inherited, other reports suggested that the complexity
of PTGMS was conditioned by a major gene and
multiple minor genes, partial dominance of sterile alleles, and dosage effects. In addition, it was
suggested that both major and minor genes might
be responsible for the genotypic difference in the
low-sterility-inducing temperature. The concept
that the genetic control of PTGMS is now governed by a set of multiple genes having major and
minor effects is well accepted.
Allelic relationship studies
The PTGMS genes in NK58S-derived japonica
lines were allelic to the NK58S gene. In some of
the NK58S-derived indica lines, genes that were
nonallelic to NK58S were also found. Among
PTGMS lines having independent origins, the sterility in both Hengnong S-l and 5460S was controlled by two recessive nuclear genes, and that in
Annong S-l was controlled by a recessive nuclear
gene. The allelism test indicated that the genes in
NK58S, Annong S-l, 5460S, and Hengnong S-l
were nonallelic to each other. It is now understood
that five genes were involved in the genetic control of NK58S, NK58S-derived P(T)GMS lines
W6154S and W7415S, Annong S-1, and 5460S
(Table 9).
Genetic mapping of PTGMS
By employing marker-gene stocks, Zhang et al
(1990) found that the major gene for PTGMS msph
in NK58S was linked to d-1 on chromosome 5
with a recombination fraction of 28.4% and it segregated independently from the other 33 marker
genes located on the remaining 11 chromosomes
of rice. This result was confirmed by the detection
Table 9. Genotypes proposed for PTGMS in five lines.
Line
Genotype
Nongken 58S
W6154S
W7415S
Annong S-1
5460S
S1 S1 S 5 S 5
S 2 S2 S 5 S 5
S1S1 S4S4
S2S2
S 3 S3 S 5 S 5
21
of closer linkage (20.3 cM) between msph and another marker gene, gh-l on chromosome 5. Isozyme
markers Cat-1 on chromosome 6 and Adh-l on
chromosome 11 were also each linked to a gene
conditioning PTGMS in NK58S.
The PTGMS genes were mapped precisely
using RFLP markers by a new mapping approach
using the bulked extremes and the recessive class
for increasing mapping efficiency (Zhang et al
1994). In a cross involving NK58S-derived
PTGMS indica line 32001S as the PTGMS parent,
two PTGMS loci, pms1 and pms2, were located on
chromosomes 7 and 3, respectively. The effect of
pms1 was 23 times larger than that of pms2 and
dominance was almost complete at both loci. In
two crosses involving NK58S as the PTGMS parent, two PTGMS loci were located in both populations. One locus was pms1 on chromosome 7 and
the other was pms3 on chromosome 12 (Fig. 11).
Using the F2 population of NK58S/NK58, the lo-
cus relevant to the fertility difference between

NK58 and NK58S was confirmed to be pms3 on
chromosome 12.
Taking all the mapping studies into account,
genes conditioning PTGMS have been assigned
to six of the 12 rice chromosomes: chromosomes
3, 5, 6 , 7, 11, and 12 (Zhang et al 1990, Zhang Q
et al 1994, Wang et al 1997, Mei et al 1999). Only
a small number of crosses were used in the mapping studies and the PTGMS parents involved
only NK58S and an NK58S-derived PTGMS line.
Moreover, no reports to date have determined
minor genes for PTGMS. We could expect that
many genes are responsible for PTGMS. There
might be two groups of genes: some genes condition photoperiod sensitivity, whereas others control male sterility. Studies to distinguish between
genes for photoperiod sensitivity and for male sterility would be as important as studies to identify
gene locations.
Rice chromosome 7
cM
RG146B
5.4
RG650
7.4
RG678
8.5
1.6
5.6
3.5
Rice chromosome 3
WG719, RG30
CDO533
RG477
pms1
pms2
7.0
1.4
3.2
pms2
Marker
RG543
CDO344
1.2
RG511, RZ272
10.6
cM
2.3
15.0
RM348
cM
Rice chromosome 12
CDO459
R2708
pms3
C751
0.4
13.6
0.4
18.5
RG191 (RG266)
RG450 (RG117)
RG335
RG128
pms1
3.7
RZ261
2.5
C2
G2140
3.1
RG9
pms3
Fig. 11. Location of PGMS genes on the linkage map in rice (Zhang et al 1994, Mei et al 1999).
22
CHAPTER 5
Breeding procedures
for developing EGMS lines
EGMS lines can be developed by any of the following methods:

Screening of existing varieties
Induced mutagenesis
Hybridization followed by pedigree selection, anther culture, backcrossing, and
marker-aided selection (MAS)
Screening of existing varieties for EGMS

Existing varieties can be screened to detect spontaneous EGMS mutants. A large number of rice
germplasm materials should be screened over seasons involving various high-temperature or
daylength regimes to identify any spontaneous
mutants or lines showing differential pollen and
spikelet sterility during different temperature or
daylength regimes occurring during the panicle
initiation stage onward. Rice germplasm consisting of photoperiod-sensitive varieties or varieties
adapted to high altitudes provides a higher probability of success when using this method.
During this screening,
1. Select plants in which earlier panicles are
partly fertile and later ones are almost sterile or vice versa according to the environmental conditions. These plants are easily
identified by the combination of partly filled
bending panicles and sterile erect panicles
in the same plant.
2. Study pollen sterility of younger panicles
and determine whether sterility is higher
than 99% (see Appendix 1 for the detailed
procedure).
3. Multiply the suspected plants by separating the tillers and ratooning them.
4. Evaluate plants for their fertility behavior

under different temperature (TGMS) and
photoperiod (PGMS) using the growth chamber or phytotron or under field conditions
(as mentioned in Chapter 2).
Induced mutagenesis
EGMS lines, particularly TGMS lines, can also be
developed by the mutation breeding method described below.
1. Select the best available high-yielding rice
cultivars that are photoperiod-sensitive,
cold-tolerant, and adapted to high altitudes
for inducing EGMS mutants. These lines
stand a better chance for inducing such
mutations.
2. Select any of the physical (gamma rays, fast
neutrons) or chemical (sodium azide, ethyl
methane sulfonate, EMS, methyl methane
sulfonate, MMS, N-methyl-N-nitrosourea,
MNU, etc.) mutagens.
3. Treat the seed material with an appropriate
dose of mutagen and grow it as the M1 generation, for example, gamma rays25 Kr;
sodium azide0.002 M, pH 3 for 6 h;
EMS0.1%, pH 7.0 for 6 h; MNU1.5 mM
for 1 h.
4. Select 1 or 2 panicles from single plants in
the M1 generation.
5. Grow the M2 progenies under appropriate
temperature (such as >30/24 C) and
daylength (such as >13.45 h) conditions and
select plants showing complete sterility or
differential fertility of panicles within the
same plant. Multiply selected plants by
separating their stubbles and evaluate them
23
under appropriate fertility induction in the

growth chamber (as shown in Tables 10 and
11) or under natural field conditions. Those
that revert to fertility are suspected EGMS
plants. These may be TGMS, reverse TGMS,
PTGMS, PGMS, or reverse PGMS depending on their behavior.
Hybridization method
This is an important and simple method to develop new EGMS lines in diverse genetic backgrounds by using already identified and characterized EGMS lines. Under this method, the known
EGMS source is crossed as a female parent to any

of the popular high-yielding rice varieties (HYV)
possessing good combining ability, disease and
pest resistance, and acceptable grain quality. This
method is generally used in a situation in which
the available EGMS line is unadapted or unacceptable for one reason or another. The crosses
made above can be handled in any of the following procedures:
Pedigree selection
a) Pedigree selection procedure using method I
1. Grow F1 plants from the crosses made to
obtain F2 seeds.
Table 10. Simultaneous screening of suspected EGMS lines for identification of type of EGMS.a
Expected reaction of suspected EGMS on pollen
sterility/fertility
Category of growth chamber settings

High temperature (32 oC) and long photoperiod (14 h)
High temperature (32 oC) and short photoperiod (12 h)
Low temperature (24 oC) and long photoperiod (14 h)
Low temperature (24 oC) and short photoperiod (12 h)
Completely sterile (PGMS/TGMS/PTGMS)

Fertile (reject/reverse PGMS/reverse TGMS)
Completely sterile (TGMS/PTGMS/reverse PGMS)
Fertile (PGMS/reverse TGMS)
Completely sterile (PGMS/reverse TGMS)
Fertile (TGMS/PTGMS/reverse PGMS)
Completely sterile (reverse PGMS/reverse TGMS
accordingly)
Fertile (TGMS/PGMS/PTGMS)
The same set of lines must be placed in all four chambers simultaneously.
Table 11. Design of experiment to identify and differentiate PTGMS from TGMS lines.a
Photoperiod (daylength duration)
Temperature (oC)
12 h
>32
30
28
26
24
Sterile (TGMS and PTGMS)

(1)
Fertile (partial) TGMS
(4)
Fertile (partial) (TGMS)
More fertile than the value
in box 4 (PTGMS)
(7)
More fertile than the value
in box 7 (PTGMS)
(10)
Fertile (TGMS and PTGMS)
(13)
13 h
(2)
Less fertile than the value
in box 4 (PTGMS)
(5)
in box 7 (PTGMS)
(8)
in box 10 (PTGMS)
(11)
(14)
14 h
(3)
in boxes 4 and 5 (PTGMS)
(6)
Less fertile than the value in
boxes 7 and 8 (PTGMS)
(9)
Less fertile than the value in
box 11 (PTGMS)
(12)
(15)
Lines that show influence of temperature and photoperiod are PTGMS, whereas the lines that show no influence of photoperiod are TGMS. Pollen
fertility % is a better measure than spikelet fertility provided anthers from the top five spikelets at panicle emergence are used for making a squash
preparation with 1% IKI stain and the pollen fertility/sterility of the emerged panicle is related to the given treatment in the phytotron.
24
2. Grow F2 populations of the crosses made

under appropriate temperature (for TGMS)
or daylength (for PGMS) conditions to identify sterile single plants that combine most
of the useful traits of both parents. Grow the
popular HYV and the donor EGMS line beside the F2 population to select comparable
plants. Raise stubbles of the selected sterile
single plants under fertility induction conditions to produce F3 seeds. Seed setting on
the ratooned plants identified earlier as male
sterile will confirm the EGMS trait.
3. Grow F3 progenies of the selected EGMS
plants under appropriate fertility induction
temperature or daylength conditions to select desirable plants and harvest F4 seeds.
Concurrently evaluate the F3 progenies in
sterility-inducing temperature or daylength
conditions to select completely sterile F3
progenies to advance the generation.
4. Handle F4 and F5 progenies in the same manner as F3 progenies.
5. To confirm stability for sterility of an EGMS
line in any given generation, trials can be
conducted under varying temperature and
daylength regimes in various places.
6. Characterize identified TGMS or PGMS
lines for specific defined environmental
conditions at the sensitive stage both in the
growth chamber and under field conditions.
7. Test the specific TGMS and PGMS lines
across various locations and seasons to identify suitable locations for hybrid seed production and EGMS self seed multiplication.
b) Pedigree selection procedure using method II
This method is deployed when we do not have a
low temperature/shorter photoperiod conducive
to fertility near where the research is conducted or
the specific environmental condition is not available during the same season of selection.
1. Grow the F2 population under a sterility-inducing temperature regime. Raise the standard controls as mentioned earlier besides
the EGMS and the other elite parent used
for hybridization to develop superior EGMS
lines. Select desired fertile plants in the segregating population.
2. Grow the F3 to F5 generations under sterility-inducing temperature and select 810
desirable fertile plants from the progeny
rows segregating for sterility. The reason for
3.
4.
5.
6.
selecting so many fertile plants in the segregating population is to ensure the probability of selecting at least one heterozygous
fertile plant that would segregate for sterility in the next generation.
Grow F5 and F6 populations under sterilityinducing temperature or longer photoperiod.
Select the most desirable male sterile plants
and ratoon them.
Transfer the ratooned male sterile plants to
a phytotron or glasshouse with a day/night
temperature of 25/19 C or shorter daylength
(12 h) to induce fertility.
Select those plants that revert to fertility
under low-temperature or shorter photoperiod conditions and collect their seeds. These
are suspected TGMS/PGMS plants.
Grow progenies of the suspected TGMS/
PGMS plants under sterility-inducing temperature/photoperiod conditions in the field
and select those plants as TGMS/PGMS that
give completely male sterile progenies.
Backcrossing after hybridization

Backcrossing is the most suitable method when
we need to transfer oligogenes in a recessive condition to an already established variety as the recurrent parent.
The procedure involves the following steps:
Select a stable and suitable TGMS/PGMS donor
with well-defined critical sterility/fertility points
and cross it with an established variety to which
the TGMS/PGMS trait has to be transferred.
1. Grow the F1 generation.
2. Grow the F2 generation to select suspected
EGMS plants; ratoon these plants to confirm their TGMS/PGMS nature.
3. Backcross the EGMS plants with the recurrent parent.
4. Repeat steps 2 and 3.
5. After every two backcrosses, one generation
of selfing is required to verify that the recessive EGMS trait is being carried forward as
depicted in Figure 12. After six generations
of backcross, self the BC6F1 to select sterile
plants under high temperature (>30 C day/
24 C night)/long photoperiod (14 h
daylength). Ratoon plants may be sent to
the growth chamber or suitable natural conditions conducive to fertility reversion.
6. The TGMS development procedure is shown
in Figure 13 but the same is applicable to
25
EGMS
donor
Elite variety/line
(recurrent parent)
F1
F2 plants with
confirmed
EGMS trait
Recurrent
parent
BC1F1
BC1F2 plants
with confirmed
EGMS trait
Recurrent
parent
BC2F1
Recurrent
parent
BC3F1
BC3F2
with confirmed
EGMS trait
Recurrent
parent
BC4F1
BC6F1
BC6F2
with confirmed
EGMS trait
Screen EGMS for

low CSP (27/21 C)
Induce fertility
(25/19 C)
Fig. 12. Backcross procedure for developing TGMS/PGMS lines.
26
PGMS since the screening procedure varies

only under the BCnF2 (n = 1, 3, 5) generation. To develop PGMS through the backcross procedure, long-day conditions (>14
h) will be required for screening in the
BC1F 2, BC3F2, and BC5F2 generations to
identify the recessive sterile genotypes.
Anther culture or pollen culture
following hybridization
Anther culture or pollen culture accelerates the
breeding process and increases selection efficiency
in the breeding of EGMS lines over conventional

techniques. To use this method, rice breeders
should work closely with a tissue culture expert.
1. Make specific crosses to combine stable
TGMS or PGMS lines with promising local
cultivars. Grow F1 plants of these crosses.
2. Produce dihaploid lines through anther culture using anthers or pollen grains from the
F1 plants, adopting standard available protocols for rice (Fig. 13). (See also Appendix
II.)
Tiller collection, anther culture, and

plantlet transplanting from F1 generation
(by tissue culture specialist)
F1 plant anthers plated

H1 pollen
plants
Callus
Plantlets
Evaluation by
tissue culture specialist
and plant breeders
Regeneration
H2 evaluation onward by plant breeders
Fig. 13. Procedure for producing and evaluating dihaploid EGMS lines.
27
3. Seeds obtained from the dihaploid plants

should be sown and raised under sterilityconducive conditions for the selection of
male sterile lines. Monitor the pollen sterility of anther/pollen culturederived plants.
4. Ratoon sterile dihaploid plants and place
them under low temperature and shorter photoperiod to induce fertility and get selfed
seeds.
5. Screen for TGMS and PGMS lines based on
the criteria for ideal EGMS lines.
6. TGMS lines such as 6442S, 1286S, HS-1,
and HS-5 and one PTGMS line, Lu Guang
2S, have been developed through anther
culture and extensively tested in Jiangxi,
Sichuan, and Fujian provinces of China (Fig.
14).
EGMS gene transfer and pyramiding

through marker-aided selection (MAS)
after hybridization
Several closely linked markers have been identified with TGMS and PGMS genes to use the MAS
approach. These markers are listed in Table 12
and can be used for selection of the TGMS/PGMS
trait in segregating F2 and backcross breeding
populations without actually screening them under specific field or phytotron conditions. The
MAS approach can enhance the speed and efficiency for selecting EGMS plants without exposing them to conditions for expression of the EGMS
gene.
MAS also allows pyramiding of different
sources of TGMS or PGMS/PTGMS alleles into a
common genetic background. It would be inter-
Genetic resource plot/

source nursery
Materials for crossing
Seed propagation
Anther culture
Tube plantlets
H2 generation
H1 generation
Yield trial
Multilocational
evaluation
Regional test
Variety registration
Characterization
P(T)GMS
lines
Twoline
breeding
Crossing
Hybrid seed
production
Variety release
Fig. 14. Procedure for indica two-line hybrid rice breeding through anther culture (modified
from Zhu et al 1999).
28
Table 12. Closely linked flanking molecular markers for EGMS genes.
Trait
Gene
Chromosome
TGMS
tms1
tms2
tms3
tms4
rtms1
pms 1
pms 2
pms 3
8
7
6
9
10
7
3
12
PGMS/PTGMS
esting to study the manifestation of pyramided

EGMS genes under different environmental conditions. Pyramiding different sources of alleles
may improve the EGMS lines for their critical sterility points (CSP) and their critical fertility points
(CFP). A low CSPlow CFP line will be useful in
tropical conditions. Four different TGMS genes
tms1, tms2, tms3, and tms4exist, while there are
three reported PGMS genespms1, pms2, and pms3.
Efforts are under way to pyramid different TGMS
genes in China, at IRRI, and in India (Fig. 15).
Closest flanking
markers
RZ562RG978
R643AR1440 (D24156)
OPAC3 640 OPAA7 550
RM257TS200
RM239RG257
RG477RG511, RZ272
RG348RG191 (RG266)
R2708RZ261/C751
tms1 tms2
tms1/2
tms3 tms4
tms1/2/3/4
tms3/4
Select with the help of

close-flanking molecular
markers
Test these pyramided lines
under different ecological
situations
Fig. 15. Procedure for TGMS gene pyramiding.
29
CHAPTER 6
Characterizing EGMS lines under

field and controlled conditions
Photoperiod varies in a gradual and periodical

manner year-round based on the latitude and solar terms. For a given place, the photoperiod on
one day over different years is the same and, on a
given date, the photoperiod in different locations
at the same latitude is also the same. Likewise, the
daylength at different latitudes in the northern
hemisphere is the same on the spring (21 March)
and autumn equinox (23 September) because the
sun radiates vertically on the equator on those
days (Table 13). The longest photoperiod is observed on the summer solstice, when the sun radiates vertically on the Tropic of Cancer, and the
shortest photoperiod is on the winter solstice, when
the sun radiates on the Tropic of Capricorn. Data
in Table 13 also indicate that the higher the latitude, the longer the photoperiod on 22 June (summer solstice) and the shorter the photoperiod on
22 December (winter solstice). Likewise, in the
southern hemisphere, similar conditions exist except that the winter and summer solstice changes.
The winter solstice occurs in the southern hemisphere on 22 June when the daylength is shortest,
whereas the summer solstice occurs on 22 December when the daylength is longest. The higher the
latitude, the longer the observed photoperiod.
Under the winter solstice on 22 June in the southern hemisphere, the higher the latitude, the lower
the photoperiod or daylength (Table 13).
The temperature at any given location is influenced by factors such as solar radiation, latitude, altitude, local topography, and atmospheric
and marine phenomena such as El Nio and La
Nia. Many locations in the tropics still have less
influence from the abovementioned environmental factors on temperature and therefore those locations are quite stable in temperature and can be
easily identified based on 25 years of meteorological data.
Characterization of EGMS lines

under field conditions
Detailed meteorological data (such as minimum
and maximum temperature, daylength, sunshine
hours, humidity, etc.) are essential to characterize
EGMS lines at a given location. It is better if data
of the past 1015 years are available. To characterize given EGMS lines, the following procedure
can be used:
1. Identify 34 distinct periods of high and
low temperatures during the year. Likewise,
determine the longer and shorter daylength
durations during the year and over locations.
2. Seed/plant EGMS lines at 1525-day intervals in such a way that their heading coincides with the high temperature or longer
photoperiod.
3. Study pollen fertility of the EGMS lines
from the top five spikelets of primary
panicles under the microscope.
4. Relate the pollen sterility data to temperatures/daylengths prevailing during the period of 1525 days before heading (tracking technique). The temperature or
daylength that is just sufficient to make the
plant completely pollen sterile must be
noted among the several temperature or
daylength regimes to which plants were exposed during the period. Such a temperature point or daylength at which complete
pollen sterility is obtained is termed the
critical sterility point and the number of days
31
Table 13. The natural photoperiod change with dates and latitudes in the northern
hemisphere.
Latitude
(oN)
40
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
Spring equinox
21 March
12.09
12.08
12.06
12.06
12.06
12.06
12.06
12.06
12.06
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
Summer solstice Autumnal equinox

22 June
23 Sept.
(h.min)
15.01
14.04
13.58
13.52
13.48
13.44
13.38
13.34
13.30
13.26
13.22
13.18
13.14
13.10
13.06
13.02
12.58
12.54
12.52
12.48
12.44
12.40
12.36
12.32
12.30
12.26
12.22
12.18
12.16
before heading during which this behavior

is expressed is designated as the sensitive
stage.
5. Likewise, determine the critical fertility
point (i.e., the lowest temperature at which
maximum pollen fertility is achieved) by
using the tracking technique. Verify the
CSP and CFP information for each of the
EGMS lines under growth chamber or phytotron conditions after the field characterization studies have been done.
Characterization of EGMS lines

under controlled conditions
1. Controlled conditions can be created in the
growth chamber. A minimum of 4 to a maximum of 15 growth chambers are needed to
characterize EGMS lines.
32
12.12
12.12
12.06
12.06
12.06
12.60
12.06
12.06
12.06
12.06
12.06
12.06
12.06
12.06
12.06
12.06
12.06
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
12.04
Winter solstice
22 Dec.
9.20
10.13
10.14
10.20
10.24
10.28
10.32
10.36
10.40
10.44
10.48
10.52
10.56
11.00
11.04
11.08
11.12
11.16
11.20
11.22
11.26
11.30
11.34
11.38
11.40
11.44
11.48
11.52
11.54
2. The total number of growth chambers

should be equal to the photoperiod temperature treatments while keeping relative
humidity constant at 75%. The number of
seedlings per EGMS entry should be equal
to the number of growth chambers times the
number of plants in each treatment. For example, three temperature and three photoperiod treatments for a total of nine growth
chambers are required. At the rate of 10 seedlings per entry per growth chamber, a total
of 3 3 10 = 90 seedlings per entry will be
needed for evaluation. The number of seeds
to be sown for the experiment should be
three times more than the number of seedlings required per entry for the experiment.
This will make an allowance for poor germination, discarding of seedlings because
of nonuniform growth, and varied growth

duration of EGMS lines.
3. Sowing and nursery management. Use the
following procedures for sowing and nursery management.
Sow the seeds of EGMS lines with similar duration and a standard set of
controls in the nursery at the same time.
If there is a significant difference in
growth duration among the EGMS lines,
sow these in two stages (the first stage for
late-maturing material and the second
stage for early maturing material).
Manage the nursery using routine management to eliminate obvious mixed-rice
seedlings and weeds.
Select uniform and healthy rice seedlings
at the 5-leaf stage for transplanting t w o
plants per plastic pot with a single plant
per hill.
Use any convenient method to label the
pots. As an example, 5-digit labeling can
be used for each plant, that is, 30817
means number 3 photoperiod/temperature treatment, 08 is the number of the
EGMS line, and 17 is the number of the
individual plant.
Treat high-photosensitive japonica and
indica EGMS lines with a short-day pretreatment (10 h/28 C of photoduration
and 14 h/25 C of dark duration for 10
days) to speed up their development.
Adjust growth chambers prior to their use
for the experiment. These should be allowed to function for one or two days
using the specified temperature and relative humidity level. A placement plan
for each of the pots within a growth
chamber should be prepared and the
chambers should be sprayed with appropriate chemicals a few days before moving the plants to control pests and diseases.
Keep the progenies of suspected EGMS
plants in separate growth chambers with
varying day and night temperature regimes and separate daylength durations.
For easy adaptation of the EGMS lines,
temperature or daylength should be kept
as available in the natural situations exemplified as follows:
Situation 1 (maximum-minimum difference of 6 C)

Day temperature (C)
24
25
26
27 28 30 32
Night temperature (C) 18
19
20
21 22 24 26
Situation 2 (maximum-minimum difference of 8 C)

Day temperature (C)
24 25 26 27 28 29 30 31 32
Night temperature (C) 16 17 18 19 20 21 22 23 24
The lines identified to adapt to specific situations

can be deployed accordingly.
Place the plants in growth chambers at the
critical stage. The critical stage for photoperiod or temperature sensitivity is during
515 days after panicle initiation (PI). The
suspected EGMS plants grown in pots are
observed for PI. The plants that are used for
determining PI by physical opening must
not be used for the experiment.
Plants must be placed inside the growth
chamber at the PI stage.
Plants must be placed for 2 weeks or up to
heading in the growth chamber for treatment.
After the plants are treated in the growth
chamber, they need to be examined for pollen sterility from the primary tiller (see Appendix1).
Observations on pollen sterility must be recorded in relation to temperature and photoperiod.
Collect a pollen sample from the primary
tiller and bag the same panicle for spikelet
fertility percent (bagged). Likewise, the pollen and spikelet sterility of the first three
panicles from each plant need to be recorded
separately as panicles 1, 2, and 3.
The CSP and CFP vary from genotype to
genotype. It is therefore essential to know
the CSP and CFP of each EGMS line before
it is used for seed production (Table 2).
Sterile plants that remain completely sterile
in different temperature or photoperiod regimes are not considered as EGMS types
and are discarded.
The EGMS lines that have been strictly
evaluated in phytotrons for their fertility alteration are advanced for evaluation under
different ecological conditions to further
ascertain their suitability for hybrid seed
production and self seed multiplication.
33
Evaluation of EGMS lines

EGMS lines are also evaluated for their phenotypic acceptability, outcrossing rate, and combining ability as per the methods described for evaluating CMS lines. The operational flow chart describing the procedure for using the EGMS system is presented in Figure 9.
After a thorough examination of the EGMS
lines in the growth chamber for their fertility alteration behavior, they need to be evaluated across
several selected environments. The purpose of this
evaluation is primarily to identify suitable regions
or locations for EGMS self seed multiplication
and hybrid seed production. This evaluation also
provides information on their range of adaptability and their resistance to diseases and insect pests.
The key to successful multilocation evaluation
lies in the timely layout of the trials and consistent observations. It is therefore highly recommended that one researcher be in charge of the
multilocation evaluation.
Selection of locations based on environment
The location must represent the rice production region and must be in the vicinity of the
research institute/station, with specialists
equipped with minimum scientific instruments (e.g., microscope, refrigerator, hot dry
oven, daily weather-recording gadgets, seed
germinators, etc.) required for evaluating
EGMS lines.
Criteria used for identifying locations
a. Locations that have a different latitude
within the same longitude.
b. Locations with distinct differences in topography and temperature within the same
latitude.
c. Locations with different elevations within
the same region.
There should be proper representation of ecological conditions for better understanding of the
behavior of EGMS lines under different
agroclimatic conditions.
For the unified evaluation of EGMS lines, the
sowing and transplanting dates must be in accordance with the respective locations identified because of the different ecological conditions (Fig.
16). For all other purposes, a standard evaluation
34
procedure must be adopted for drawing meaningful conclusions, that is, planting design, pollen
and spikelet fertility, and their method of data collection should be consistent.
The identification of appropriate areas and
seasons for large-scale commercial production of
TGMS and hybrid seed needs special attention,
unlike for the PGMS lines, for which the daylength
is quite stable across a given region. On the basis
of analysis of about 1015 years of meteorological data, areas and ideal periods for seed production in different parts of the world can be identified.
Places located 500 and 700 m above sea
level are highly suitable during May to September for hybrid seed production and during November to March for TGMS self
seed production.
In the choice of place, in hills or coastal
plains or interior plains and plateau regions,
extra care needs to be taken to ensure that
the temperature range is <40 to > 16 C, beyond which physiological sterility occurs.
Four weeks of stable high temperature (>30/
>24 C day/night) are required for hybrid
rice seed production interspersing the sensitive stage of stage II (secondary branch
primordial) to stage IV (stamen and pistil
primordial stage) of the panicle development
stage of rice (Fig. 16).
Similarly, four weeks of low temperature
(<26/>16 C night) are required for TGMS
self seed multiplication, interspersing the
sensitive stage of stage II (secondary branch
primordial stage) to stage IV (stamen and
pistil primordial stage) of the panicle development stage of rice (Fig. 16).
The identified place must be suitable for rice
cultivation with space or time isolation.
Places with naturally chilled irrigation water of about >1724 C can be useful for
TGMS self seed multiplication under hightemperature fluctuations. The chilled irrigation water is effective from stage IV (stamen and pistil primordial stage) to stage V
(meiotic division).
All the data collected should be stored in a
computer and sent to the researcher in charge for
analysis and interpretation.
Ideal locations with four consecutive weeks with stable temperature or photoperiod
A. Hybrid seed production
Minimum of four continuous weeks with stable high temperature (>30/24 oC day/night) or long
photoperiod (>14 h)
1st week
1d
2nd week
7d
3rd week
14 d
4th week
21 d
28 d
25 24 23 22 21 20 19 18 17 16 15
Days before flowering
Adjust
sowing date accordingly so that 1525 days before flowering falls between the second and third
week of high temperature.
B. EGMS self seed multiplication

Minimum of four continuous weeks with stable low temperature (<24/>16 oC day/night) or short
photoperiod (<12 h).
1st week
1d
2nd week
7d
3rd week
14 d
4th week
21 d
28 d
25 24 23 22 21 20 19 18 17 16 15
Days before flowering
Adjust
sowing date accordingly so that 1525 days before flowering falls between second and third
week of low temperature.
Fig. 16. Graphic illustration for adjustment of sowing time for EGMS for hybrid seed and self seed production
in ideal locations.
35
CHAPTER 7
Developing pollen parents

for two-line hybrids
A pollen parent line is defined as a male parental

line that has the ability to restore the fertility in
the F1 of the EGMS line under a sterile phase regime that ensures complete male sterility. Most
conventional inbred lines are the source of pollen
parent lines or pollen parents. For hybrid rice seed
production using EGMS, any pollen parent line
can restore fertility in the F1, unlike CMS lines,
for which a set of restorer lines alone can restore
fertility in the F1. It was estimated that about 97%
of the japonica and indica inbred lines restored
the fertility of japonica PGMS (such as Nongken
58S) and indica TGMS (such as W6154S) lines.
Characteristic features of an
elite pollen parent
The following are the characteristic features of an
ideal elite pollen parent:
Strong fertility-restoring ability. When a
cross is made with the EGMS line, the hybrids have a normal seed setting percentage
(>80%) and are less affected by changes in
environmental conditions.
Good general combining ability. When
crossed with different EGMS lines, the F1
hybrids from many crosses perform well.
Good agronomic characters. The pollen
parent should be a high-yielding inbred line
with favorable traits for outcrossing, for example, good anther dehiscence, good anther protrusion, large anther size, high pollen load, etc.
Genetic distance. Considerable genetic distance from the EGMS lines will be the key
to enhanced heterosis.
Breeding methods for identifying pollen

parents for two-line hybrids
The following breeding methods can be used to
identify pollen parents for two-line hybrids:
screening available elite inbred lines by testcrossing with available EGMS lines, hybridization,
induced mutagenesis, and anther culture. Among
these methods, the testcross and hybridization
methods are the most common.
Screening available elite inbred lines by
testcrossing with available EGMS lines
This is the most effective method to screen for
suitable pollen parents from existing rice
germplasm. In China, most pollen parents for twoline hybrid rice breeding were screened from existing inbred lines. For example, the pollen parent
in hybrid Liangyou Peijiu (Peiai64S/9311) was
Yangdao b (9311), a medium-season indica variety developed by the Lixiahe Agricultural Institute, Jiangsu Province. The high-yielding indica
variety Teqing developed by the Rice Institute,
Guangdong Academy of Agricultural Sciences,
was used as the pollen parent for the TGMS line
Peiai64S and developed a hybrid, Liangyou Peite,
in Hunan Province.
Three fundamental steps are required in making testcrosses, as discussed below.
1. Preliminary testcross. Use existing elite
lines or varieties to cross to EGMS lines,
then conduct a preliminary evaluation of
the F1 hybrids based on their seed setting,
yield components, grain quality, and resistance to diseases and insect pests, etc. At
37
least 10 single plants are needed for a preliminary testcross.

2. Re-testcross. From the preliminary testcross
data, the varieties or lines that performed
well in terms of strong restoration and that
showed no segregation for agronomic characters can be advanced to the re-testcross to
confirm the results of the preliminary testcross. About 50100 single plants are
needed for the re-testcross. In the re-testcross,
yield will be considered as the key evaluating factor.
3. Identification. Only excellent pollen parents (varieties or lines) are selected for further use. After identification of suitable pollen parents, hybrid seeds will be produced
for field evaluation and multilocation trials.
Hybridization
This is an important and very popular method for
breeding pollen parents for two-line hybrids when
there is a lack of such parents among existing varieties or lines.
1. Single cross. Under this method, two varieties or lines with desired traits are crossed
and, in the F2, selection is exercised to identify single plants that possess the desired
traits. The standard selection pedigree procedure is followed to fix the lines. At the
F5F6 generation, a testcross with EGMS
lines is made. The best-performing pollen
parent showing high fertility restoration and
heterosis in testcross F1s is selected for the
production of hybrid seeds and subsequent
evaluation in the field.
2. Multiple crosses. This procedure uses two
or more varieties or lines to develop ideal
pollen parent lines. The procedure is almost
the same as the single-cross method except
that in the F2 generation the selections must
be made to combine the favorable traits from
all the parents used in the multiple cross.
3. Backcross. For a pollen parent line with several good traits except for one or two undesirable traits, the backcross method of breeding is the most suitable for removing and
replacing the undesirable traits with favorable traits from the donor parent. Select a
stable and suitable variety or line as the trait
donor and cross it with the pollen parent
line (recurrent parent) that needs to be im38
proved. Raise the F1 generation and backcross it with the recurrent parent and BC1F1
seeds obtained. The selection strategy may
vary according to the dominant or recessive
trait. For a dominant trait, it is simple to select the plants with the target trait in the
backcross generations. In the BC5F1 generation, self them once. After selfing, the selected plants are used to test-cross to screen
the strong F1 hybrids for evaluation. For the
recessive trait, care must be taken to ensure
that the trait is selected and confirmed by
selfing after every two generations of backcrossing.
New strategies for developing

pollen parent lines
The main difference in pollen parent and restorer
lines is that the former may not have special restorer genes, that is, Rf, like the latter, thereby
making them more convenient for using the heterosis between indica and japonica lines in twoline hybrid breeding.
To overcome hybrid sterility, longer growth
duration, taller plant types, and incomplete grain
filling, wide-compatibility genes can be used in
the development of pollen parent lines.
Exploiting intersubspecific indica japonica
crosses through the wide-compatibility
system
What is wide compatibility? It is well known that
indica and japonica crosses result in hybrid sterility despite their high heterosis for various agronomic traits. But, it was found that certain indica
and japonica hybrids showed normal spikelet fertility. One or both parents of these crosses must
possess a dominant wide-compatability gene (S5n)
and such lines are designated as wide-compatibility varieties (WCVs). When such WCVs are
crossed with indica or japonica, the hybrids show
normal spikelet fertility.
The concept of wide compatibility in rice was
first introduced by Ikehashi and Araki (1986) to
explain hybrid sterility at the subspecies level.
To study the interactions between the sporophyte
and gametophyte, Ikehashi and Araki (1986) conducted a survey for WCVs via a triple testcross.
They found that varieties such as Ketan Nangka,
Calotoc, and CPSLO-17 showed wide compatibility. The fertility of their hybrids was linked to the
cytochrome (C) and waxy (wx) gene. The WC gene

was located on chromosome 6 and three alleles of
the locus were identified: S-5i from indica rice, S5j from japonica rice, and S-5n from WC rice. The
sterility resulting from crosses between indica and
japonica rice is due to the interaction between the
alleles S-5i and S-5j. The S-5i/S-5j genotype produces semisterile panicles because of the partial
abortion of female gametes carrying the allele S5j. Such abortion does not occur in S-5n/S-5i and
S-5n/S-5j genotypes. The donor of S-5n is referred
to as a WCV (Table 14).
Developing pollen parents with wide compatibility. Pollen parents with wide compatibility
were developed by introducing the WC gene into
conventional indica and japonica rice.
Pollen parents are generally crossed with WC
lines and then selection is made in the segregating generation using marker genes linked with
the WC trait. Sometimes, the backcross method is
used when the traits of WCVs are not ideal.
In China, at the National Rice Research Institute (CNRRI), pollen parent R2070 with WC was
successfully developed by using the cross
Minghui 63 (Lunhui422 and javanica line
WL1312). R2070 can be used in both three-line
and two-line hybrid rice breeding. The three-line
hybrid II You 2070 (II-32A/R2070) and two-line
hybrid Guangya 2 (M2S/R2070) performed well
in the yield trial and in farmers fields and both
were registered in Zhejiang Province.
To evaluate indica/tropical japonica hybrids,
Make several crosses between TGMS lines
(WC) and tropical japonica varieties. If some
tropical japonicas have the WC gene, they
can be crossed to any indica TGMS line.
Evaluate the hybrids in the OYT and a series of trials to identify the most promising
hybrids with enhanced heterosis. The best
hybrids that are grown in the locality should
be used as standard checks. Emphasis should
be given to monitoring spikelet fertility
during the evaluation.
Developing tropical japonica pollen parents,
the same method described for transferring the
desirable tropical japonica pollen parent can be
followed. However, the recipient tropical japonica
should possess a WC gene.
When developing desirable indica pollen
parents possessing wide compatibility, the fre-
Table 14. Wide-compatibility varieties in different

varietal groups.
Indica
Japonica
BPI 76
Dular (aus)
N22 (aus)
NK 4
Norin PL9
02428
Tropical
japonica
Banten
Calotoc
CP-SLO
Ketan Nangka
Moroberekan
Palawan
Padi Bujang
Pendec
IR64446-7-3-2-2
IR65598-112-2
quency of elite indica lines is quite high. Hence,

these can be good male parents developing indica/tropical japonica hybrids, if only WC genes
are transferred into them. The procedure for this is
as follows:
Select indica lines with very good restoring
ability and WC donors with a marker gene
(a purple apiculus or Amp32).
Make crosses between the indica line and
WC donors.
Grow F1 and evaluate all crosses for spikelet
fertility. Choose the highly fertile cross for
further selection.
Grow F2 and select good recombinants looking like the restorer but with the apiculus
pigmentation or Amp32 allele.
Grow F3 and F4 and select the best plants in
the best families. Keep track of the original
plant type of the pollen parent and the Amp32
allele.
In the F5, test-cross the selected lines on a
single-plant basis with a tropical japonica
EGMS line having no WC gene.
Evaluate the testcross progenies for fertility. Select lines that show high fertility. They
are the pollen parents possessing WC genes.
Once the requisite pollen parents are developed, they can be test-crossed with the best available female parents and the experimental hybrids
can be evaluated as discussed in Chapter 9 on
Evaluating two-line hybrids.
39
CHAPTER 8
Assessing the combining ability of parental lines

is extremely useful in a hybrid breeding program,
especially when many prospective parental lines
are available and the most promising ones are to
be identified on the basis of their ability to give
superior hybrids. The line tester method
(Kempthorne 1957) is commonly used for this
purpose.
All the l lines should be crossed to each of

the t testers to produce l t experimental
hybrids.
Composition of the combining ability

nursery
All the l t hybrids along with the parents
(lines + testers).
Definitions
Suitable check varieties may also be included for working out standard heterosis.
Combining ability refers to the ability of a
genotype to transfer its desirable traits to its

progenies.
General combining ability (GCA) is the average performance of a parent in a series of
crosses.
Specific combining ability (SCA) is the deviation in the performance of a hybrid from
the performance predicted on the basis of
the general combining ability of its parents.
Field layout
Choose a fairly homogeneous plot for grow-
Type of lines to be evaluated
The most stable CMS and EGMS lines pos
sessing high phenotypic acceptability and

a fair to excellent outcrossing rate.
Effective restorers/pollen parents adapted to
the target area.
Procedure using the

line tester design
ing the combining ability nursery in a replicated trial using a randomized complete
block design (RCBD).
Use several replications to ensure a minimum of 12 degrees of freedom for error to
have statistically valid comparisons.
Plant a single seedling per hill with a spacing of 20 15 or 20 20 cm.
Plot size may depend on the amount of F1
seed available. However, a minimum of 50
plants per plot is essential. The larger the
plot size, the better it is for evaluation.
Avoid collecting data from border plants.
Each three-row plot of hybrids can be flanked
by a border row of a check variety.
Let us suppose we have l lines (elite pol-
If we have five lines (pollen parent lines)
len parents) and t testers (elite CMS and/

or EGMS lines).
and four testers (EGMS lines), the total number of crosses will be 1 t = 5 4 = 20.
41
Rep 1
5
16
12
3
10
14
25
9
21
7
20
22
15
11
1
23
C
6
Rep 2
18
2
17
4
26
24
13
28
8
27
29
19
24
19
5
29
12
27
C
14
28
7
1
25
4
23
16
3
18
15
Rep 3
21
13
9
11
22
8
6
26
17
10
20
2
25
19
21
26
29
9
17
10
29
6
13
18
27
1
3
23
5
15
14
20
16
8
24
11
7
12
2
C
22
4
120 hybrids, 2125 lines, and 2629 testers. C = check variety.

Fig. 17. Layout of combining ability trial.
Let these 20 crosses along with five lines

and four testers (29 entries) be tested in an
RCBD with three replications with one
check variety (Fig. 17).
with the table value of F for 28 and 56 degrees of

freedom at the 5% or 1% level of significance.
Treatment SS can be further partitioned into
SS from parents, SS from crosses, and SS from the
interaction of parents vs crosses.
Analysis of variance
Correction factor =
(CF)
(Grand total)
Treatment SS
Total no. of observations
Total sum of squares (TSS) = Yij2 CF

Y.J2
Replication SS (RSS) =
t
Yi2
Replication SS (RSS) =
SS from parents =
Replications
Treatments
Error
Total
SS
(r l) [2]
(t l) [28]
(r l)(t l) [56]
(rt l) [86]
MSS
RSS/2
TrMSS/28
ErMSS/56
df = degrees of freedom, SS = sum of squares,

MSS = mean sum of squares, F = F value.
To test the significance of the genotypic difference, compare the calculated F (TrMSS/ErMSS)
42
CF (parents
with 8 df)
SS from interaction = TrSS SS (crosses)

of parents vs crosses
SS (parents) (with 1
df)
Analysis of variance table

df
P2ii
TSS TrSS RSS

a
Source
CF (crosses with 19 df)
CF
r
Error SS (ErSS)
CF
r
Cij = observation for ijth cross
Pii = observation for ith parent
r = number of replications
C2ij
SS from
=
crosses
r
CF
C2ij + P2ii
ANOVA with parents and crosses

Source
df
Replications
Treatments
Crosses
Parents
Parents vs crosses
Error
2
28
19
8
1
56
SS
MSS
Test all sources of variation against error variance.
Line tester analysis
Note: MSS from lines and MSS from testers are to

be tested against MSS from lines testers. MSS
from lines testers is to be tested against MSS
from error.
Construct a two-way table.

Testers
Lines
Yij
Yi..
Sometimes line tester analysis is done by

using cross means (means of crosses over replications). In that case, MS from error that is used for
testing the significance of MS (lines tester) should
be divided by the number of replications before
testing.
Y.J.
Y..
Estimation of GCA effects:
1
1
2
3
4
5
Total
SS from lines =
Total
i) GCA effects of lines
Yi2..
CF (crosses)
rt
gi =
Y.J.2
1r
CF (crosses)
SS from lines testers = SS (crosses) SS (lines)

SS (testers)
df
Lines
Testers
Lines testers
Error
4
3
12
56
SS
MSS
df
SS
where Yi.. = total of ith line over testers, Y.. =

grand total, and ltr = number of lines, testers, and
replications.
Work out GCA effects for g1 to g5. See whether
gi = 0.
MSS
gt =
2
28
8
1
19
4
3
12
56
86
Y.j.
lr
Y..
ltr
iii) Estimation of SCA effects

F
Sij =
Replications
Treatments
Parents
Parents vs crosses
Crosses
Lines
Testers
Lines testers
Error
Total
ltr
where Y.j. = total of jth tester over lines, Y.. =

grand total, and ltr = number of lines, testers, and
replications.
Work out GCA effects for g5 to g9. See whether
gt = 0.
ANOVA for line tester analysis including

parents
Source
Y..
ii) GCA effects of testers
ANOVA for line tester analysis

Source
tr
where r = replications and t = testers.
SS from testers =
Yi..
Yij.
r
Y..
rt
Y.j.
rl
Y..
ltr
where Yij. = value of jth line with ith tester, Yi.. =

total of ith line over all testers, Y.j. = total of jth
tester over all lines, Y.. = grand total, and ltr =
number of lines, testers, and replications.
Work out SCA effects for all hybrids. See
whether i j Sij = 0.
43
Testing the significance of combining ability

effects
SE (standard error) (GCA for lines) =
Me
rt
SE (standard error) (GCA for testers) =
Me
rl
SE (standard error) (SCA effects)
Me
r
SE (standard error) (gi gj) line
2Me
rt
SE (standard error) (gi gj) testers
2Me
rl
SE (standard error) (Sij Skl)
2Me
r
Me is the error mean sum of squares.
Interpretation of results
The statistical significance of treatments
indicates that the entries have genotypic differences between them. If the treatment differences are significant, we can use further
partitioning.
44
Partitioning of treatment SS into SS from
crosses and parents helps to test the significance of these two components individually.
The parents with higher positive significant
GCA effects are considered to be good general combiners, whereas those with negative
GCA effects are considered to be poor general combiners.
The hybrids with significant SCA effects in
a positive direction are considered to be the
most promising ones.
Using the results

The EGMS lines with good GCA are chosen
for developing experimental hybrids for testing in observation yield trials.

The pollen parents with good GCA will be
used for crossing with EGMS lines to produce experimental hybrids for testing in
observation yield trials.
Hybrids with higher positive significant
SCA effects are chosen for evaluation in preliminary yield trials.
CHAPTER 9
Evaluating two-line
hybrids
Many hybrids developed through the two-line

approach using EGMS lines are heterotic. Therefore, an elaborate evaluation of the hybrids is required to identify the most promising ones in terms
of yield heterosis per se and their suitability for
high seed yields and self seed yields of the female
parent during its multiplication. Experimental
hybrids should be evaluated in a series of trials,
using an appropriate statistical design to ensure
unbiased comparisons. The choice of design depends on the number of entries and the quantity
of hybrid seed available. During the initial stages,
the number of entries to be tested is large and the
quantity of hybrid seed available is limited; therefore, the hybrids are tested in unreplicated trials.
However, in subsequent evaluation, hybrids need
to be tested in replicated trials with a larger plot
size. The performance of hybrids may be locationspecific. Therefore, it is necessary to conduct
multilocation trials to identify hybrids having
wide adaptability and those that are specifically
adapted to certain locations. Testing the performance of hybrids in farmers fields along with local check varieties of the region is necessary before these hybrids are released for commercial cultivation.
Experimental design and field layout

Since the number of experimental rice hybrids is large (100500) and the amount of
hybrid seed is limited, it is convenient to
conduct the OYT by using the augmented
design.
In this design, the whole experimental area
is divided into several blocks.
The check varieties are replicated in each
block, whereas the test entries are not replicated but are assigned to the remaining plots
randomly.
The yields of test entries are adjusted for
block differences based on the yield of check
varieties in each block.
The block size is determined as follows:
If c = number of check varieties, v = number
of test hybrids, and b = number of blocks, the number of test entries in a block (n) = v/b, the number
of plots/block (P) = c + n, and the total number of
plots (N) = b (c + n).
The total number of blocks should ensure at
least 12 df for error in ANOVA.
b>
Observation yield trial (OYT)

Composition
Re-testcrosses are made between the commercially
usable EGMS lines and effective restorers identified in the testcross nursery. Use three to four check
varieties representing different growth durations
(such as very early, early, medium, and late).
12
c1
+1
Let us take 40 hybrids and four check varieties.

The number of blocks =
12
41
12
3
+1
+1 = 5
45
Number of check varieties = 4

Number of test hybrids = 40
Number of blocks = 5
Number of hybrids per block = 40/5 = 8
Number of plots per block = 8 + 4 = 12
Agronomic management for yield trials

The trial should be managed so that all entries
receive appropriate water, fertilizer, and measures
to control weeds, insects, and diseases. Yield trials should be harvested carefully.
Total number of plots = 60
Data recording
Collect data on the following parameters:
Vegetative vigor (on a 19 scale).
Days to 50% flowering.
Visual score for spikelet fertility (on a 19
scale).
Yield plot1 for conversion to yield ha 1.
Phenotypic acceptability score (on a 19
scale).
Layout for augmented design

The plot size should be at least 5 m2 for each
entry.
Plant a single seedling per hill with a spacing of 20 15 cm.
First assign the check varieties randomly in
each block.
Assign the test hybrids randomly to the remaining plots.
The field should be properly leveled.
Take up gap filling 710 days after transplanting to obtain a uniform plant population.
Care should be taken for a uniform distribution of fertilizers and plant protection
chemicals.
Uniform water control is a must for valid
comparisons.
A worked-out example of the OYT is described below.
Layout and yield figures (t ha1) for OYT (augmented design).
Blocks
1
17 (4.6)
C (3.8)
9 (5.6)
13 (5.3)
D (5.7)
29 (5.2)
6 (4.9)
A (4.5)
31 (3.2)
18 (2.6)
B (4.6)
21 (6.1)
23 (7.1)
A (4.0)
3 (5.6)
36 (7.0)
B (5.1)
7 (6.3)
38 (4.6)
15 (3.9)
C (4.1)
20 (5.3)
11 (5.0)
D (4.9)
B (4.0)
28 (5.0)
14 (3.6)
D (4.3)
24 (7.5)
30 (6.2)
A (4.5)
C (3.9)
39 (4.3)
1 (3.6)
32 (5.9)
19 (5.4)
12 (7.9)
37 (5.3)
A (3.9)
8 (3.4)
33 (5.2)
C (4.2)
5 (2.9)
40 (6.8)
26 (7.8)
D (3.9)
34 (3.8)
B (4.9)
2 (5.0)
25 (3.2)
B (3.2)
27 (5.4)
16 (6.0)
35 (2.6)
D (4.6)
22 (3.9)
A (3.8)
4 (4.7)
C (3.3)
10 (5.2)
46
Construct a two-way table of check yields (in t
ha1) and means.
Check
variety
A
B
C
D
Total
Mean
Blocks
Total Mean
1
4.5 4.0 4.5 3.9 3.8

4.6 5.1 4.0 4.9 3.2
3.8 4.1 3.9 4.2 3.3
5.7 4.9 4.3 3.9 4.6
18.6 18.1 16.7 16.9 14.9
4.65 4.52 4.17 4.22 3.72
20.7
21.8
19.3
23.4
85.2
4.14
4.36
3.86
4.68
4.26
Compute the block effect: rj = Bj M

where rj = block effect of jth block, Bj = mean of
all checks in jth block, and M = grand mean of the
checks.
Block effects of different blocks are
Block
rj
1
2
3
4
5
0.39
0.26
0.09
0.04
0.52
See whether rj = 0
Construct a table of unadjusted and adjusted

yields. The adjusted yield for each test entry is obtained by deducting the block effect from the unadjusted yield.
Difference between two check means

2MSE/b
2 0.258/5 = 0.32
Difference between adjusted yields of two

Adjusted (AD) and observed (O) yields (t ha1) of
test hybrids in the OYT.
hybrids in the same block

2MSE
Yield
Hybrid Block
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
3
5
2
5
4
1
2
4
1
5
2
4
1
3
2
5
1
1
3
2
2 0.258 = 0.72
Yield
Difference between adjusted yields of two
Hybrid Block
O
AD
3.6
5.0
5.6
4.7
2.9
4.9
6.3
3.4
5.6
5.2
5.0
7.9
5.3
3.6
3.9
6.0
4.6
2.6
5.4
5.3
3.69
5.52
4.74
5.22
2.94
4.51
6.04
3.44
5.21
5.72
4.74
7.94
4.91
3.69
3.64
6.52
4.21
2.21
5.49
5.04
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
1
5
2
3
5
4
5
3
1
3
1
3
4
4
5
2
4
2
3
4
AD
6.1
3.9
7.1
7.5
3.2
7.8
5.4
5.0
5.2
6.2
3.2
5.9
5.2
3.8
2.6
7.0
5.3
4.6
4.3
6.8
5.71
4.42
6.84
7.59
3.72
7.84
5.92
5.09
4.81
6.29
2.81
5.99
5.24
3.84
3.12
6.74
5.34
4.34
4.39
6.84
To work out the standard errors for comparing the means, an ANOVA table is prepared
by using the replicated data of check varieties.
hybrids in different blocks

2MSE (1 + 1/c) = 0.80
Difference between an adjusted yield of a

hybrid and a check mean
MSE (b + 1) (C + 1)/bc = 0.62
Use of results
The test entries are classified based on different maturity groups and their performance
is compared with that of the check variety
of the corresponding duration by using the
standard errors calculated for the purpose.
The hybrids that yield significantly higher
than the check varieties are identified and
promoted for the preliminary yield trial.
Preliminary yield trials (PYT)

Composition
Identify promising hybrids in observation yield
trials. Use hybrids showing apparent heterosis in
the testcross nursery and significant heterosis in
the combining ability nursery. Use check varieties of different growth duration (very early, early,
medium, and late).
ANOVA for check varieties. ns = nonsignificant.

Source
df
SS
MSS
Block
Checks
Error
Total
4
3
12
19
2.068
1.804
3.096
6.968
0.517
0.601
0.258
2.32 ns
The standard errors are worked out as follows for different comparisons:

The RCBD is ideal for conducting the preliminary yield trials. The steps involved are
as follows:
The number of blocks or replications should
be such that the error degree of freedom
should be at least 12.
The ideal plot size is about 10 m2.
If the fertility gradient is unidirectional, the
blocks should be perpendicular to the fertility gradient.
47
Hybrids should be grouped according to

their growth duration. Each group or subgroup should have 1520 hybrids and suitable checks.
An example
Let us use 16 hybrids and four check varieties to be tested in four replications.
Divide the field into four equal blocks.
Subdivide each block into 20 experimental
plots.
Assign the treatments to each plot randomly.
Each treatment should appear in every block.
Layout of the PYT (with RCBD). Numbers in parentheses represent yield (in t ha1).
Rep 1
Rep 2
Rep 3
Rep 4
3 (5.3)
D (4.3)
8 (6.5)
10 (7.5)
2 (6.0)
A (3.8)
7 (5.6)
4 (7.9)
12 (6.0)
6 (7.1)
14 (8.2)
1 (3.8)
B (5.9)
15 (7.9)
16 (5.8)
C (6.5)
13 (7.2)
5 (6.9)
11 (4.9)
9 (5.3)
4 (8.3)
11 (5.6)
A (4.2)
2 (5.8)
D (3.9)
6 (5.6)
3 (6.2)
13 (5.9)
B (6.5)
9 (5.9)
15 (8.2)
16 (4.9)
5 (7.6)
14 (7.6)
7 (4.6)
10 (7.6)
12 (5.9)
1 (4.6)
C (7.8)
8 (7.2)
9 (4.7)
5 (7.3)
7 (5.0)
C (6.9)
10 (6.8)
1 (4.7)
15 (6.9)
D (5.2)
8 (5.9)
14 (8.2)
A (4.6)
4 (8.0)
11 (5.9)
3 (5.8)
6 (5.3)
2 (7.3)
16 (5.9)
13 (6.3)
12 (4.9)
B (5.2)
C (6.0)
2 (7.1)
13 (6.0)
6 (5.9)
B (6.0)
11 (6.1)
4 (7.6)
3 (6.0)
12 (5.6)
7 (4.9)
15 (7.8)
D (4.6)
9 (5.0)
8 (6.7)
16 (6.1)
A (5.0)
1 (5.1)
14 (7.2)
10 (7.3)
5 (8.1)
Data recording
Observations are recorded on the following parameters:
Days to 50% flowering
Plant height
Spikelet fertility (%)
Grain yield (kg ha-1)
1,000-grain weight
Reactions to major diseases/insects
48
Group the data by treatments (entries) and
replications and calculate the treatment total (T), replication total (R), and grand total
(GT).
Table of means. Yield in t ha1.
Treat- Rep1 Rep2 Rep3 Rep4 Total Mean
ments
1
3.8 4.6
4.7
5.1
18.2
2
6.0 5.8
7.3
7.1
26.2
3
5.3 6.2
5.8
6.0
23.3
4
7.9 8.3
8.0
7.6
31.8
5
6.9 7.6
7.3
8.1
29.9
6
7.1 5.6
5.3
5.9
23.9
7
5.6 4.6
5.0
4.9
20.1
8
6.5 7.2
5.9
6.7
26.3
9
5.3 5.9
4.7
5.0
20.9
10
7.5 7.6
6.8
7.3
29.2
11
4.9 5.6
5.9
6.1
22.5
12
6.0 5.9
4.9
5.6
22.4
13
7.2 5.9
6.3
6.0
25.4
14
8.2 7.6
8.2
7.2
31.2
15
7.9 8.2
6.9
7.8
30.8
16
5.8 4.9
5.9
6.1
22.7
A
3.8 4.2
4.6
5.0
17.6
B
5.9 6.5
5.2
6.0
23.6
C
6.5 7.8
6.9
6.0
27.2
D
4.3 3.9
5.2
4.6
18.0
Total 122.4 123.9 120.8 124.1 491.2
4.55
6.55
5.82
7.95
7.47
5.97
5.02
6.57
5.22
7.30
5.62
5.60
6.35
7.80
7.70
5.67
4.40
5.90
6.80
4.50
6.14
Compute the correction factor and various

sums of squares as follows:
CF =
(GT) 2
N
241,277.4
80
= 3,015.96
Total SS = (3.8)2 + (6.0)2 + (4.6)2 CF

= 3,128.76 3,015.96
= 112.80
Replication SS =
(122.4)2 + + (124.1)2
t
CF
= 3,016.32 - 3,015.96
Analysis of variance (ANOVA) table.
= 0.36
Source
2
2
2
Treatment = (18.2) + (26.2) + (18.0) CF
SS
r
= 3,110.82 CF
= 94.86
Error SS = Total SS RSS TrSS
DF
SS
Computed Table F
Fa
5% 1%
MSS
Replication 3
0.36
Treatment 19 94.86
Error
57 17.58
Total
79 112.80
0.12
4.99
0.30
16.6**
a**
A highly significant F value indicates that the test entries differ significantly among themselves.
= 112.80 0.38 94.86
Compute the coefficient of variation (CV)

= 17.58
Compute the mean sum of squares by divid-
CV =
ing each sum of squares by its corresponding degree of freedom.
Replication MSS =
RSS
r1
0.36
=
= 0.12
3
Error MS
100
GM1
0.30
100
6.14
= 8.92
Treatment MS =
Error MS =
Tr.SS
t1
Er.SS
(r 1)(t 1)
94.86
19
17.58
57
Compute the critical difference (CD)
= 4.99
CD = t 0.05
2 EMS
= 0.30
= 3.44 0.387
Compute the F value for testing the treatment differences.
F value =
Treatment MS
Error MS
= 1.33
4.99
The hybrids with a difference of more than
0.30
the CD value from the check variety are considered significantly superior to the check
variety.
= 16.6
Compare the calculated F value with the
table F value.
Prepare the analysis of variance table by
including all the computed values.
Use of results
The performance of the hybrids is compared
with that of the check variety of corresponding duration or the highest-yielding check
variety.
GM = grand mean.
49
The hybrids that have a significantly higher
Mere statistical significance is not sufficient
yield than the check variety are identified

and promoted to the advanced yield trial. A
significant yield advantage of more than 1 t
ha1 would also be the ideal criterion for selecting the best hybrids after testing the significance.
to consider a hybrid as promising. Therefore, an advantage of about 1 t ha1 over the

check variety is specified, which would result in a real benefit to farmers.
Advanced yield trials (AYT)

Composition
Identifying promising hybrids in the preliminary
yield trials. Use three or four check varieties of
different duration (very early, early, medium, and
late).
The RCBD is ideal for conducting the AYT.
The number of entries in the AYT is much
lower than in the PYT. It is helpful to increase the plot size to 15 m2.
Entries should be divided into at least two
maturity groups (1very early and early;
2medium and late).
The field layout and agronomic management are similar to those for the PYT.
Data recording is essential.
The following observations are recorded for
the AYT:
Plant height
Panicles m2
Number of filled grains panicle1
Yield ha1
1,000-grain weight
Reactions to major pests and diseases
Remarks on special features
The method of statistical analysis is the same as
the one explained for the preliminary yield trials.
Use of results
The performance of the hybrids is compared
with that of the check variety of corresponding duration or the highest-yielding check
variety.
The hybrids that have significantly higher
yield (>1 t ha1) than the check variety are
promoted for multilocation trials.
50
Multilocation yield trials (MLT)

The major objective of multilocation yield trials
is to identify the hybrids that have a wider adaptability or those that are specifically adapted to a
particular location. This exercise is essential as
hybrids perform differently in different environments. This also provides an opportunity for breeders to see the performance of hybrids bred by them
in other locations, even though these hybrids may
fail to perform well in the location where they are
developed. The concept of multilocation yield
trials has really improved the efficiency of rice
breeders and this is more so in hybrid rice breeding.
Composition
Identify the promising hybrids in the AYT from
different centers, including those introduced from
abroad. Use three or four check varieties of different duration (very early, early, medium, and late).
If the trials are constituted based on duration, it
would suffice to include a check variety of corresponding duration in the trial.
The locations for MLT should be selected
carefully so that each location serves as a
distinct environment. The location selected
should be in the proposed target area for the
cultivation of hybrid rice.
An RCBD is most commonly used for conducting MLT.
It is necessary to have common guidelines
for agronomic management and data collection from different centers.
General guidelines for conducting MLT
Specify the design to be adaptedan RCBD
with four replications.
Specify the entries and the check varieties.
Besides the common check, each center can
choose a local check for comparison.
The trial should be conducted during the
same season at all locations.
Specify the seedling age at transplanting
2125 days old.

The spacing adopted should be uniform
20 20 or 20 15 cm.
Specify planting a single seedling per hill.
The fertilizer dose may depend on the native soil fertility and recommendations in
the local area.
Plant protection should be need-based.
The plot size should be uniform in all the
locations to the extent possible (at least 15
m2).
Agronomic management
Agronomic management should be uniform in all
locations so as to have valid comparisons, except
for some specific recommendations made for a
particular location.
The statistical analysis for G E interactions

and interpretation of results are covered in the
chapter on G E analysis.
Use of results
The hybrids with higher yield potential and
wider adaptability are identified based on
stability analysis. These are promoted for
on-farm testing in different areas, before their
release for commercial cultivation.
Those hybrids that are found to be suitable
for a particular location are promoted for
on-farm testing in that particular region
only.
Data recording
Data sheets are circulated to all the cooperators
for collecting data on important parameters, such
as
Plant height
Panicles m2
Number of filled grains panicle1
Yield plot1
Yield ha1
1,000-grain weight
Reactions to pests and diseases
Weather data of each location
51
CHAPTER 10
Two-line hybrid rice

seed production
The production of two-line rice hybrids involves

two major steps: (1) multiplication of EGMS lines
and (2) hybrid rice seed production.
Each of these steps requires specific environmental conditions for seed production. Therefore,
the locations/seasons for producing the two types
of seed have to be different. In Chapter 5, we suggested that suitable specific locations must be
identified for elite EGMS lines for their self seed
multiplication. Locations must also be identified
that are suitable for hybrid seed production using
the same EGMS lines.
Multiplication of EGMS lines

EGMS lines, if multiplied continuously for several generations without any selection, may segregate for CSP, thereby causing major problems
in maintaining purity of the hybrid seeds. Therefore, nucleus and breeder seed production must
be taken up on a continual basis.
Method I
Nucleus seed production of an EGMS
(TGMS or PGMS) line begins in the fertility-inducing environment. Seeding of
TGMS or PGMS lines is arranged in such a
way that the sensitive stage occurs when the
temperature or photoperiod is favorable for
a higher seed set.
At the time of flowering, about 100 typical
plants are selected from the population of
an EGMS (TGMS or PGMS) line and their
panicles are bagged. The selection process
should be completed within 1 week.
After the harvest, the selected plants are
scored for spikelet fertility (based on the
main panicle) and 50 plants with higher

spikelet fertility (above 30%) are selected.
Progenies of the selected plants are grown
in the sterility-inducing environment. About
30 seeds are taken from each of the selected
plants to grow single-row progenies and the
remaining seeds are stored carefully. The
balance of the seeds of the progenies that
are uniform and completely male sterile
must be marked and bulked to form the
nucleus seed (Fig. 18).
Nucleus seed of the EGMS line is used for
producing breeder seed under strict isolation. Breeder seed for the EGMS line is produced in the fertility-inducing environment.
The breeder seed produced under the direct
supervision of the plant breeder has high
genetic purity and is used for producing
foundation seed of parental lines, which in
turn will be used for producing hybrid seed.
Method II
Select a completely male sterile plant with
typical characteristics of the original EGMS
line under a sterility-inducing environment.
Ratoon the selected plant and clone it for as
many plants as you need. Multiply the ratooned stubbles under a fertility-inducing
environment. The nucleus seed will be harvested from the ratooned stubbles.
The nucleus seed is used for producing
breeder seed and the latter for producing
foundation seed.
Preserve the selected stubbles under favorable temperature conditions with good management as long as you need them. The new
nucleus seed will be produced continuously.
53
TGMS population
Step
1
Plants selected
Progeny rows of selected plants
Progenies selected
Breeder seed production
Select about 100 typical plants

from a population of TGMS lines
grown in a fertility-inducing
environment and bag all the
panicles. Select 50 highly
fertile plants among them
based on their higher spikelet
fertility determined after
harvest.
Step Grow single-row progenies of

2a 50 selected plants in a sterilityInducing environment and
mark those progenies that
are completely male sterile.
Step Bulk the balance of seed kept

2b earlier as a reserve of the
progenies selected in step 2a
to form the nucleus seed.
Step Multiply the nucleus seed from
3 step 2b to produce breeder
seed under strict isolation.
Step Multiply the breeder seed to
4 produce foundation seed
under strict isolation.
Step Repeat steps 1 to 3 in the
5 breeder seed production plots
to continue the process of
producing nucleus and breeder
seed.
Fig. 18. Procedure for nucleus and breeder seed production of TGMS lines
(Virmani et al 1997a).
With this system of multiplication, the

EGMS line continues in the same generation and it will not segregate for its CSP. Its
general combining ability will also not
change.
In the tropics, since ratooning may not always be successful because of the high incidence
of diseases and insects, method II can be risky.
Foundation seed production. Fresh breeder
seed should be used by seed p r o d u c t i o n
agencies to produce foundation seed of
TGMS/PGMS lines in a strict isolation area
with suitable temperature and/or daylength
conditions.
54
Differences between EGMS

and CMS line multiplication
The fertility of the CMS line is controlled
by the CMS gene and a pair of recessive

nuclear fertility restorer genes. Expression
of male sterility of CMS lines is stable over
environments. The fertility of EGMS lines
is controlled by a recessive nuclear gene
alone, the expression of which is influenced
by environmental conditions.
The maintenance of the CMS line is through
its cross pollination with the maintainer line.
The yield of CMS seeds through this process largely depends on outcrossing and
other related traits of the CMS line. The

maintenance of the EGMS lines, in contrast,
is quite simple since it is through selfing
under fertility-inducing environments, especially during the sensitive panicle development phase (stage II to stage VI). The seed
yield of EGMS lines largely depends upon
the CFP and the favorable environment during fertility alteration.
The CMS line is purified through pair crossing (A B) in which the A and B lines are
selected based on their morphological characteristics. The purity of the EGMS line is
maintained by evaluating the morphological characters and the CFP under specific
locations.
The yield of CMS line multiplication is
about 2 t ha1 under suitable conditions and
it is unstable because of the changing
weather conditions during the flowering
period. Under favorable conditions, the
yield of EGMS line multiplication could
reach 46 t ha1. This yield is higher because it occurs due to self-pollination compared to seed yield in the CMS line, which
occurs due to cross-pollination.
Similarity of CMS and EGMS

line multiplication
CMS line multiplication (A B) requires
strict isolation to prevent contamination by

stray pollen. Under favorable environmental conditions, the EGMS line becomes fertile and seed setting can be as high as 75%;
however, some male sterile spikelets can still
receive pollen from another source. Therefore, strict isolation must also be provided
to multiply pure EGMS lines.
Strict roguing is necessary for CMS and
EGMS line multiplication.
High-yielding techniques for PGMS line

Autumn-season multiplication
Short daylength can induce PGMS lines to become fertile under proper temperature conditions.
In autumn, the daylength is getting shorter and a
high percentage of seed setting can be achieved if
the sensitive stage occurs under this condition.
However, seed quality becomes inferior, especially

when the temperature falls sharply in late autumn.
Winter-season multiplication
in lower-latitude areas
For PGMS self seed multiplication, short
daylength and low temperature in lower-latitude
areas in the winter season are the most suitable
environmental conditions. Seed quality is better
than in late autumn because the temperature gradually rises at the time of seed maturity.
High-yielding techniques for TGMS line

Spring-season multiplication
The fertility of TGMS lines is mainly influenced
by temperature. TGMS lines should be sown in
early spring so that the sensitive stage matches
the temperature for fertility induction based on
local meteorological data. The yield may not be
stable because of the occurrence of abnormally
high temperature, and the long growing period
may result in nonmatching of the sensitive stage
with the proper temperature regime.
Autumn-season multiplication
The temperature is low enough for fertility induction in the autumn season. The TGMS lines should
be sown at the proper time to make the sensitive
stage coincide with the low-temperature period in
order to obtain higher yields. Seed quality may be
influenced by low temperature in the late maturity period.
High-altitude multiplication
In high-altitude areas (with altitudes from 800 to
1,000 m), moderate temperature and longer
daylength conditions are suitable for the short
growing period for PGMS and TGMS line self seed
multiplication.
Chilled-water irrigation
Results showed that chilled-water (>17 C) irrigation can induce TGMS lines to produce fertile
pollen similar to ambient low temperatures. The
advantage of chilled-water irrigation is that it provides flexibility for adjusting the sensitive stage.
The TGMS self seed production is high and more
stable since the water temperature is relatively
stable and controllable.
55
Seed plot selection. A sufficient water supply

with suitable temperature is essential to the establishment of a seed plot for TGMS self seed multiplication. The water temperature should be higher
than 1718 C but lower than the CFP of the TGMS
line to be multiplied.
Season selection. The joint effect of lower
temperature and chilled-water irrigation is good
for increasing EGMS self seed yield.
Irrigation time. Chilled-water irrigation starts
from the young panicle differentiation of the stamen and pistil primordial (stage IV) and goes to
the meiotic division of the pollen mother cell
(stage VI).
Three-line hybrid rice seed production

The three-line hybrid rice seed production system
involving CMS lines is a relatively stable method
across normal rice-growing conditions, whereas
the two-line hybrid rice seed production system
involving EGMS lines has environmental limitations requiring strict adherence to the proper timing according to season and location. For successful hybrid rice seed production, the male sterility
trait of the female parent and effective pollen load
of the male parent with proper flowering synchronization are the key factors. For three-line seed
production, only the flowering synchronization
between the A and R lines is important for higher
hybrid seed yields, whereas, for two-line seed production, both the male sterility expression under
the sterile phase and the flowering synchrony between the EGMS and pollen parent influence seed
purity and bring about higher hybrid seed yields.
Two safe-period determinations
for hybrid seed production
In three-line hybrid rice seed production,
favorable climatic conditions for pollination are called the safe flowering period,
which is important for increasing the outcrossing rate. In two-line seed production,
the two safe-periods refers to the conducive
environmental conditions that support first
the induction of complete male sterility and
second the facilitation of proper pollen
movement from the pollen parent and fertilization of the EGMS parent.
Therefore, the first safe-period determines
seed purity, while the second safe-period determines hybrid seed yield.
56
The first safe-period, the sensitive stage of

fertility alteration, must be given priority if
there is a need for an adjustment according
to the prevailing environment at a given
location.
The desirable climatic conditions for pollination are as follows: (1) temperature of 2335 C
(minimum-maximum) in a day, (2) relative humidity around 7090%, and (3) no continuous rains
(that last more than 3 days) during the pollination
period.
Isolation of the hybrid
seed production plot
Terrain isolation. The selected seed plot is
either isolated by mountains or hills or by
other natural barriers.
Time isolation. The hybrid seed plot should
flower 2030 days earlier or later than any
other plot that could be a source of stray
pollen for the seed production plot.
Distance isolation. Keep a distance of 200
m between the seed plot and any other rice
pollen source.
Determining the seeding
interval for synchronization
Time method. The seeding interval is determined by the difference in growth duration
between the two parental lines. The one with
longer duration must be sown early according to the number of days of difference between the two parents in terms of days to
50% flowering.
Leaf number method. The total leaf number
of a variety is relatively stable at the same
site and in the same season in different years.
The rate of growth in terms of leaf number is
influenced by environmental temperature.
By observing the leaf number of the early
seeded parental line, you could determine
the seeding date of the later one because the
leaf number difference is rather stable between the two parents.
Heading date prediction
Remaining leaf number method. The
young panicle differentiation starts from
the reciprocal third-leaf emergence
(keeping the flag leaf as the first leaf and
the leaf before the flag leaf as the 2nd
leaf and the leaf before the 2nd leaf as
the 3rd leaf) and the time taken for com-
plete panicle differentiation is about 30

days from the start to heading. This
means that the parental line would head
after 30 days, when the reciprocal third
leaf emerges.
Stripping the young panicle. The young
panicle differentiation could be divided
morphologically into eight stages and
each stage takes about 34 days for developing (Table 15, Fig. 19).
Production of hybrid seeds

for preliminary yield trials
For preliminary yield trials, the hybrid seed requirement is usually small, but numerous hybrids
are produced simultaneously.
Planting techniques. To produce a small quantity of hybrid seeds for the OYT or PYT, the following are the planting designs: (1) chimney isolation, (2) modified chimney isolation, and (3) the
isolation-free method.
i. Chimney isolation procedure
The desired parental lines are sown on different dates to obtain synchronous flowering.
Twenty-five-day-old seedlings of EGMS
(female parent) and pollen parent lines are
planted in alternate rows of five plants each
at a spacing of 15 15 cm (Fig. 20).
Frames of 1 1 1 m are prepared with
either iron or aluminum angles.
Cubicles of 1 1 1 m are stitched with
muslin cloth, with a flap at the top.
The metal frame is placed around a 1-m2 area
where the EGMS (female parent) and pollen

parent lines are planted just before flowering.
The frame is covered with a muslin cloth
bag to prevent cross pollination.
During the flowering period, the pollen
plants are shaken to increase seed setting
on the EGMS line. This can be facilitated
by opening the flap.
Pollen parent plants are harvested first and
threshed separately. The EGMS line is harvested and threshed later to avoid possible
seed admixture.
ii. Modified chimney isolation procedure

The chimney isolation method has been modified
to overcome the problem of synchronization and
to simplify the supplementary pollination. The
basic layout is the same as that of the chimney
method, except for the following differences:
The EGMS and pollen parents are sown on
different days to achieve maximum flowering synchronization.
At the boot-leaf stage of the parental lines,
2-m-high barriers are erected to cover the
three sides of a 1-m2 plot, leaving a gap of
20 cm from the ground. The open side is
covered by the barrier of the opposite plot.
The space between the opposite plots is convenient for cultural operations, including
supplementary pollination.
Supplementary pollination is done by using sticks 34 times per day at peak anthesis during the flowering period of 710 days.
Table 15. Morphological stages in rice panicle development.

Morphological character
Morpho-physiological stages of rice

panicle development (equivalent)
Invisible
First bract primordium differentiation

(stage I) and primary branch
primordium (stage II)
Secondary branch primordium
differentiation (stage III)
Stamen and pistil primordium
differentiation (stage IV)
Mid meiotic stage (stage V)
Late meiotic stage (stage VI)
Pollen tetrad stage (stage IV)
Pollen maturation (stage VII)
Heading stage (stage VIII, stage IX)
Little white hairs

More large hairs
Can see the spikelet individually
The lemma and palea are visible
Spikelet has full size
Panicle has green color
Panicle is enclosed in leaf sheath
and is about to emerge
57
Stage no.
I
II
III
IV
V
VI
VII
VIII
XI
X
Development
stage
Panicle primordia
Primary branch primordium
Secondary branch primordium
Stamen and pistil primordia
Pollen mother cells
Meiotic division
Mature pollen
Ripe stage of pollen
Completed spikelets
Flowering
Approx. days
before heading
Approx. panicle
length (mm)
30
27
24
20
17
12
6
4
12
0.2
0.4
1.5
2
1025
80
190250
260
270
Fig. 19. Development stages of panicle formation to flowering.
58
1.0 m
Pollen parent
15 cm
1.0 m
20 cm
EGMS line
20 cm
15 cm
Fig. 20. Position of EGMS line and pollen parent line in the chimney isolation procedure.
iii. Isolation-free method

An isolation-free method developed at the International Rice Research Institute has been found
to be more practical and popular in tropical countries. This method is ideal for producing small
quantities of hybrid seed required for the OYT
and PYT.
Selected pollen parent lines are grown side
by side in 5 3-m plots. In each pollen parent line plot, four rows of pollen parent
plants are planted as border rows at 20 20cm spacing to provide isolation from adjoining plots. Four vacant spaces 40 cm in
width are left in the middle of the plot, which
are interspersed by single rows of pollen
parent line plants. About 68 EGMS plants
can be planted in these spaces at the time of
flowering (Fig. 21).
EGMS lines of experimental hybrids are staggered five times at 810-day intervals to ensure a continuous supply of EGMS plants at
the flowering stage to synchronize the flow-
ering of pollen parents in different seed production plots.

When primary tillers of EGMS and pollen
parent lines are in the boot-leaf stage, their
flag leaves are clipped except for the two
outermost border rows of pollen parent lines,
which act as a barrier for pollen from adjoining plots.
Three to five days after leaf clipping, the
EGMS lines are uprooted (preferably in the
morning, that is, 06000800) and are
planted in the vacant spaces of the plots.
To enhance outcrossing, supplementary
pollination is advocated at the peak anthesis period. Care should be taken to shake
only those pollen parent lines that are flanking the EGMS lines.
The pollen parent line is harvested first and
threshed separately, followed by the EGMS
line bearing the hybrid seeds.
By adopting this method, 35 g of hybrid
seed can be obtained from each EGMS/CMS
59
Unclipped R plants
Clipped R plants
20 cm
Transferred A line
20 cm
Restorer line no. 1
Restorer line no. 2
Restorer line no. 3
Restorer line no. 4
Fig. 21. Layout for isolation-free system for producing seeds of experimental two-line hybrids.
plant. A plot with 1540 EGMS plants can

yield 50200 g of hybrid seed, which will
be enough to conduct the OYT for two seasons (20 g per season) and replicated PYTs
also for two seasons (100 g per season).
Seed production for AYT and MLT
Strict isolation method. The hybrid seed
required for conducting the AYT should be
highly pure. About 12 kg of seed is required for this purpose. Therefore, the seed
has to be produced in a larger area (100
200-m2 plots) under strict isolation to en60
sure purity. The method is described below.
Isolation. A space isolation of 50 m is ideal
for hybrid seed production, which means

that within this range no other rice varieties
should be flowering except the pollen parent. If it is difficult to get space isolation, a
time isolation of more than 21 days would
serve the purpose. Distance isolation can be
reduced to 3040 m if the hybrid seed production plot is surrounded by an additional
1520 rows of pollen parents.
Seeding sequence. Parental lines of hybrid
combinations differ in their growth duration.
Therefore, they have to be seeded on different dates so that their flowering will be synchronous. A late parent is sown first and an
early parent is sown later, the difference being equal to the difference in their growth
duration. The EGMS line is seeded only
once so as to match the environmental conditions that favor complete male sterility.
The pollen parent is seeded three times with
3-day intervals, such that the difference between the second sowing of the pollen parent and that of the EGMS line is equal to the
seeding interval between the parental lines.
Row ratio and layout. The optimum row
ratio for hybrid seed production is 23
males:810 females. Pollen parent seedlings
are evenly mixed and planted in three rows,
at a spacing of 15 15 cm, leaving a space
for an EGMS line in between. The EGMS
seedlings are planted with a spacing of 30
15 cm. The spacing between the EGMS line
and the adjacent pollen parent line should
be 20 cm. Row direction should be perpendicular to the wind direction (Fig. 22).
Roguing. Roguing is an important operation in a hybrid seed production plot to ensure purity of hybrid seeds. Rogues can be
identified as those that are out of their row
and early in booting, and based on other
morphological characters. The off-types
observed during different growth stages are
to be removed. Before flowering, roguing is
essential, especially in experimental hybrid
seed production plots. Roguing at flowering is also extremely important as pollen
from off-type plants can cause irreparable
damage through cross pollination with male
sterile plants.
GA3 spray. Spraying of GA3 is recommended
to obtain good panicle exsertion. A dose of
4060 g ha1 by a knapsack sprayer or 15
20 g ha1 by a ULV sprayer is recommended
for desired results. The spray liquid required
is 500 L and 20 L for the knapsack and ULV
sprayer, respectively. GA3 should be sprayed
two times, the first when 1520% of the
tillers have started heading and the second
2 days after the first spraying or when 35
40% of the panicles of the seed parent have
emerged.
Supplementary pollination. At the time of
flowering, supplementary pollination is

done by shaking the pollen parents with either a rope or bamboo sticks. This operation
has to be done 34 times daily at peak anthesis for 610 days. The supplementary
pollination technique using bamboo poles
will substantially increase the experimental hybrid seed yield.
Harvesting and threshing. Extreme care
should be taken while harvesting and
threshing the hybrid rice plots. Harvest and
thresh the pollen parent first. Thoroughly
check and remove any panicles of the pollen parent separately. The seed should be
dried, processed, bagged, and properly labeled. The rogues must be removed before
the flowering of the parental lines and before harvest. The male rows or the pollen
parent lines must be harvested carefully and
threshed separately. Before threshing, the
threshing machine must be properly cleaned
to avoid seed admixture. The labels must be
carefully placed on each bag of hybrid seed
along with the parentage, date of harvest,
and field location.
Two-line hybrid rice seed

production on a large scale
Isolation from a pollen source is equally important for large-scale seed production, especially in
the initial stages, for a newly released hybrid. Total space isolation is the best way for large-plot
seed production. But, a minimum of 200 m distance from any other rice pollen source is essential for ensuring seed purity.
Sowing and transplanting
For proper flowering synchronization, the male
parental line should be raised on three different
sowing dates to match the estimated heading date
of the female line. Such a sowing plan will provide a continuous supply of pollen during the flowering of the female (EGMS) parent (Fig. 23).
At the time of transplanting, the top one-third
part of the leaves may be chopped off from each
uprooted nursery bundle of the pollen parent for
its easy identification. The pollen parent can be
transplanted first in two rows with a space of 45
61
Prevailing wind direction at time of flowering
1st seeding of pollen parent

2nd seeding of pollen parent
3rd seeding of pollen parent
Seed parent
Vacant row
Hill to hill spacing15 cm

Male to female ratio3:10
Alternate transplanting
of pollen parent
or
Random transplanting
of pollen parent
Fig. 22. Layout of breeder seed and hybrid seed production plots.
cm in between the two rows (Fig. 24). Later, the

EGMS lines can be transplanted in eight or ten
rows with 15-cm spacing in between them. A space
of 30 cm between the pollen parent and EGMS
line must also be maintained.
62
Before flowering, roguing is essential, especially in hybrid seed production plots. Rogues
can be identified as those plants that are out of
their row and early in booting, and based on other
morphological characters. The rogues must be re-
Pollen parent (3rd) -----------------------------------------Pollen parent (2nd) ------------------------------------------Pollen parent (1st) --------------------------------------------EGMS female parent flowering
Fig. 23. Proper flowering synchronization between pollen
and seed parent in a hybrid seed production plot.
moved before the flowering of the parental lines

and before harvest.
Supplementary pollination
techniques
Supplementary pollination techniques (i.e., pushing the plant over to facilitate pollen movement)
using rope or bamboo poles will substantially increase seed yield.
Harvesting
The male rows or the pollen parent lines must be
harvested carefully and threshed separately. Before threshing, the threshing machine must be properly cleaned to avoid seed admixture. The labels
must be carefully placed on each bag of hybrid
seed along with the parentage, date of harvest,
and field location for further seed processing and
packaging for distribution to farmers.
Problems and their solutions
in hybrid seed production
Male sterility of EGMS lines, particularly TGMS
and PTGMS lines, is highly influenced by tem-
perature rather than by photoperiod. Among the

problems, the most important is the adverse effect
of temperature fluctuations caused by sudden/unforeseen local weather changes.
During hybrid seed production at high-temperature locations, a sudden drop in temperature
(below the CSP) can be disastrous because of reversion to the fertile phase resulting in selfing in
the female parent or the TGMS line. Hence, care
must be taken to use TGMS lines of low CSP and
low CFP.
Equally important is the identification of locations with stable temperature based on several
years of meteorological data.
Further, the high-temperature regime can be
prolonged for a minimum of 4 weeks during the
sensitive phase.
Likewise, higher temperature (above the CSP)
during TGMS seed multiplication at low-temperature locations can result in reduced percentage
seed set, thus seriously affecting seed yield.
Aside from chemical remedies, the real solution to these problems lies in developing stable
TGMS lines adapted to reasonable fluctuations in
temperature during hybrid seed production and
TGMS seed multiplication.
Rather than using a strictly temperature-sensitive sterility system, it is desirable to use TGMS
lines slightly influenced by photoperiod because
they are flexible regarding temperature fluctuations and are more stable for fertility-sterility expression.
E E E E E E E E E E
P P
E E E E E E E E E E P
E E E E E E E E E E
P P
E E E E E E E E E E
P P
E E E E E E E E E E
P P
E E E E E E E E E E
P P
E E E E E E E E E E
P P
E E E E E E E E E E
P P
15 cm
30 cm
15 cm
45 cm
Fig. 24. Field layout for two-line hybrid seed production. E = EGMS line and P =
pollen parent line.
63
CHAPTER 11
Two-line rice hybrids:

maintenance of genetic
seed purity standards
Compared to three-line hybrids, the development

of two-line hybrids was mainly limited by their
purity. The purity of two-line rice is determined
by the uniformity of the fertility transformation of
EGMS lines, which is probably controlled by both
major and minor genes. These minor genes are not
always homozygous during the selection and
evaluation of EGMS lines, which thus segregate
during seed reproduction, which leads to the difference in the critical temperature of fertility alteration among individual plants within lines.
When the environmental temperature is above the
critical sterility point (CSP), the whole population of a TGMS line shows complete male sterility and looks uniform. However, once the temperature decreases, a difference among individual
plants appears. Plants with a lower CSP are sterile,
whereas plants with a higher CSP show partial fertility or even complete fertility. During hybrid
seed production, plants with a low CSP are sterile
and produce hybrid seeds, but plants with a high
CSP become fertile and produce self seed at low
temperature. This causes a mixture of hybrid seeds
with self seeds, which results in a nonuniform hybrid rice production field. Therefore, purification
of EGMS lines is critical to the production of twoline hybrid seed. The following methods are recommended for purification of EGMS lines:
1. Use nucleus seeds directly from the research
center for seed reproduction of EGMS lines.
2. Apply anther culture for the purification of
EGMS lines.
3. Introduce a recessive morphological marker
into the EGMS line or a dominant morphological marker gene into the pollen parent
line by gene transformation or conventional
backcross breeding methods. The pseudohybrids can be identified and eliminated in

the nursery, thereby increasing the purity of
hybrid rice in the field.
1. Using nucleus seeds of EGMS

for seed production
The nucleus seeds of EGMS lines produced must
be used for only five generations since there can
be a gradual change in CSP levels.
The procedure for purifying EGMS lines is as
follows:
Select about 100 plants with typical morphological characters of the original line
from a population with less purity and plant
them in pots.
Transfer the pots into a glasshouse or phytotron with the controlled temperature of
CSP until heading, when the selected plants
develop into the secondary rachisbranch
primordial differentiation stage (stage III).
Select the plants with 100% sterility. When
the plants begin heading, investigate pollen and spikelet fertility under the microscope.
Ratoon the selected plants and adjust the
controlled-temperature to CFP till heading.
The self-pollinated seeds harvested are
called nucleus seeds.
Plant nucleus seeds from each selected plant
again in rows. Compare their agronomic
traits and fertility with those of the original
line and harvest the seeds from the rows that
are identical to the original line in bulk.
65
2. Using anther culture for

purifying EGMS lines
Applying anther culture to purify EGMS lines by
producing instant dihaploids that are completely
homozygous for the EGMS trait can be highly
useful for the proper maintenance of any given
EGMS line. Such an approach will be highly useful for hybrid seed production.
EGMS lines that are genetically homozygous
and uniform in morphology have a better photoand thermoperiod response to fertility alteration.
Those showing a significant difference in fertility
among individual plants are difficult to purify effectively by conventional methods. Such lines
should undergo anther culture for purification.
Choose a suitable sowing date to make
EGMS plants head at the fertile phase. Alternatively, use the ratooned plants that passed the
evaluation test in the growth chamber for purification. Make the ratooned EGMS plants tiller as
much as possible to increase their panicle number
for anther culture. Appendix II contains the detailed procedure for anther culture.
3. Transferring a recessive
marker gene into EGMS lines
A recessive morphological marker is used
to distinguish the
pseudo-hybrid
plants. The marker
must have (1) singlegene control, (2) a
clearly visible phenotype, (3) no obvious
negative effect on
other useful traits, and
(4) no influence on
EGMS trait expression.
The phenol reaction gene (ph gene) is
one such marker that
could be incorporated into TGMS
lines. Paddy grains of
varieties possessing
this gene, when
treated with solutions
of phenolic com66
March 1991
IGM 19
(with pgl)
July 1991
Oct. 1991
pounds (such as phenol, catechol, hydroquinone,

pyrogallol, and tyrosine), become uniformly
black. A monogenic recessive gene controls the
expression of this trait. If the pollen parent has
this gene in the homozygous dominant form, the
selfed seed of the TGMS line mixed with the hybrid seed can be identified easily by the appearance of black grains after staining treatment. The
true hybrid F1 seed material will appear brownish
white after staining. These two classes of grains
can then be separated easily by a color-sorting
machine (Virmani and Maruyama 1995). This
method can also be used by seed certifying agents
to determine seed purity percentage.
The backcross method is also used to transfer
the recessive marker gene into EGMS lines. The
procedure is illustrated in Figure 25 using the pale
green leaf marker gene (pgl) as an example.
4. Insertion of a dominant marker

gene into the pollen parent
A dominant morphological marker is used to distinguish the pseudo-hybrid plants and it must have
the following characteristics: (1) be controlled by
a single gene, (2) have a visible phenotype, (3)
have no negative effect on any other useful trait,
8902S
F1
8902S
F2 (plants with pale green leaves)
March 1992
B 1F 1
(out of 22 lines, 15 showed fertility

transformation and M2S was the best
among them)
Aug. 1992
B 1F 2
(36 individual sterile plants were selected

from 661 plants with pale green leaves)
March 1993
B 1F 3
(22 lines with uniform characters and better

fertility were selected from 36 lines)
Aug. 1993
B 1F 4
(22 lines were sown in intervals. After 1.5

months, the best oneM2Swas selected)
M2S indica P(T)GMS line with pale green
leaves
Fig. 25. Procedure for transfer of recessive marker gene pgl into EGMS lines.
and (4) have no influence on the restoration ability of the pollen parent line.
The protocol for transferring the herbicide
(Basta) resistance gene (Bar) into a pollen parent
using genetic transformation and/or backcrossing
is described below.
a) Screen the pollen parents used widely in
the commercial hybrids for their regeneration ability and those possessing good regeneration ability are selected.
b) For gene transformation, use mature seeds
and immature embryos as explants. The Bar
gene is inserted by a particle gun or
Agrobacterium-mediated transformation.
c) Transplant about 50 transgenic plants (T0)
into the soil using appropriate agronomic
management procedures.
d) Spray herbicide (Basta) on the regenerated
plants to kill the sensitive plants. Conduct
molecular analysis (such as Southern blotting) of the resistant plants.
e) Harvest separately seeds (T1) from T0 plants
that were identified as single-copy inserted
transgenic plants.
f) Sow T1 seeds on a line basis. The seedling
number of each line should be more than
50. Meanwhile, sow the original pollen parent line as a control. In addition, EGMS
lines should be sown timely using routine
agronomic management practices.
g) Spray herbicide Basta on the T1 plants in
the nursery and select the surviving seedlings from the lines that show a 3:1 segregation ratio for resistance to sensitivity.
Then transplant them in the field as a single
plant per hill. The ideal transgenic plants
will be the ones that show no significant
difference from their original pollen parent
in phenotype.
h) At the heading stage, cross the transgenic
plants (pollen parent + Bar gene) with the
EGMS line and their control (pollen parent
line) with the EGMS line at the same time
and harvest their F1 hybrid seeds.
i) Noncrossed panicles from the transgenic
plants should also be bagged, then harvest
self and crossed seeds (T2) plant by plant.
j) Raise 24 F1 plants per replication, replicated

three times. Measure agronomic traits such
as plant height, tiller number, resistance to
diseases and pests, and yield. The perfect
transgenic lines should have nonsignificant differences with their control in F1 characters.
k) Sow T2 seeds derived from T1 plants (nearly
50 plants are required for each T1-derived
plot). Spray Basta in the nursery, select one
plot in which all seedlings are resistant to
Basta from each line, and transplant them.
l) Harvest the selfed seeds from the Basta-resistant plants; these will give a transgenic
pollen parent possessing the Bar gene.
m) In the next year, use these transgenic Bar
lines to test the production potential of the
two-line transgenic hybrid. If the production potential of the hybrid is acceptable,
produce the bulk F1 seed. Carry out the normal hybrid trial to ascertain transgenic restorer lines with practical production potential.
n) Seed the bulk seeds of transgenic two-line
hybrids in the seedbed and spray Basta on
the seedbed to kill the selfed EGMS seedlings and other nonhybrid plants. Only true
hybrid seedlings carrying the Bar gene will
survive in the seedbed, which can be transplanted to the field.
Special note: Any field experiments with
transgenic plants need proper approval and permission from the government and must be done in
an isolated area for testing. Transgenic plants and
their derived seeds are forbidden to serve as parents of a cross, animal feed, and human food without government approval. This approach, therefore, cannot be used freely by plant breeders. As a
result, it has limited practical application.
The purity standards of two-line hybrids are
based on grade, varietal purity, seed purity, germination percentage, and percent moisture content. The standards for China appear in Table 16.
The criteria of breeder seeds and foundation
seeds of three-line hybrids appear in Tables 17
and 18.
67
Table 16. Quality standards for rice seed according to China national standards issued in 1996.
Type
Grade
Varietal purity
(%)
Seed purity
(cleanliness) (%)
Germination
percentage
Moisture
content (%)
Conventional
rice
Basic seed
Certified seed
99.9
98.0
98
85
13.0 (indica)
14.5 (japonica)
MS a
ML
RL
Basic seed
Certified seed
99.9
99.0
98
80
13.0
Hybrids
Class 1
Class 2
98
96
98
80
13.0
MS = male sterile, ML = maintainer line, RL = restorer line.
Table 17. Seed standards for the three parental lines.

Seed grade
Purity (%)
Cleanliness (%)
Germination (%)
Moisture (%)
Weed seeds (%)
A line
Breeder
Foundation
100
>99.9
>99.9
>99.0
>93.0
>90.0
<13.0
<13.0
0
0
B line
Breeder
Foundation
100
>99.9
>99.8
>99.0
>98.0
>96.0
<13.0
<13.0
0
0
R line
Breeder
Foundation
100
>99.9
>99.8
>99.0
>98.0
>96.0
<13.0
<13.0
0
0
Table 18. Criteria for field identification of parental multiplication.

Sterility and sterile
plants (%)
Restoring rate
(%)
Off-type plants
(%)
A Line
Breeder
Foundation
100.0
100.0
0
<0.01
B line
Breeder
Foundation
0
<0.01
R line
Breeder
Foundation
>85.0
>85.0
0
<0.01
Seed grade
68
CHAPTER 12
Future outlook for

two-line rice hybrids
The exploitation of hybrid vigor in rice has shown

the way to increasing rice production after yield
stagnation with the use of semidwarf inbred rice
varieties in irrigated ecosystems. The deployment
of three-line rice hybrids in China and elsewhere
in Asia has substantially increased the hopes of
sustaining Asian food security. Hybrid rice technology has caught the attention of rice farmers
outside China and, during the next ten years, several countries should have a large area covered
with rice hybrids. There is a continuous need to
reduce the cost and increase the efficiency of hybrid rice seed production. The discovery of the
EGMS system in China and later in Japan, at IRRI,
and in India has improved the chances of substantially reducing the cost of seed production by using two-line rice hybrids. These hybrids also help
to increase heterosis beyond the level of threeline rice hybrids. The two-line hybrids have already created an impact in China, with their area
reaching 2.6 million ha. Among the several twoline rice hybrids, Liangyou PeiJiu (Peiai64S/9311)
gave the highest average yields of 11 t
ha1 on 10-ha farms in two successive years. The
highest yield recorded was about 12.1 t ha1, a
new record in these areas, clearly revealing the
enhanced potential of two-line rice hybrids. Outside China, two-line rice hybrids are also being
developed at IRRI and in Vietnam, India, Korea,
the Philippines, Thailand, and Egypt.
Future research priorities include the following:
1. Development of stable EGMS lines. Stable
elite EGMS lines with a precise fertility alteration mechanism hold the key to success
in developing two-line commercial hybrid
rice. The underlying genetic mechanism of

fertility alteration needs to be understood
clearly to properly enhance the efficacy of
EGMS seed multiplication and hybrid rice
seed production. Breeding of TGMS lines
with a low CSP is important for developing
two-line commercial rice hybrids in the tropics. The genetic characterization of the loci
of the EGMS genes from different sources
in relation to closely tagged molecular markers is useful for marker-assisted selection.
2. Use of anther culture to develop and/or
purify elite EGMS lines. Anther culture techniques involving dihaploidization can be
used to expedite the development and/or
purification of EGMS lines possessing major genes and QTLs in influencing the
PGMS/TGMS trait.
3. Breeding for super high-yielding two-line
hybrids. Two-line hybrid rice technology
involving EGMS lines allows the choice of
a wider range of parental combinations and
avoids the negative effects of male-sterility-inducing cytoplasm. Rice scientists at
IRRI and in China can now use new plant
type (NPT) lines developed in the tropical
japonica and indica/tropical japonica background as male and/or female parents to develop hybrids with enhanced heterosis.
Two-line breeding technology can overcome the major problems of wide incompatibility and the narrow range of restorers
in exploiting indica-japonica heterosis.
4. Incorporation of hybrids with resistance to
biotic stress. Two-line rice hybrids possessing multiple resistance to diseases and in-
69
sects can be developed more expeditiously

than three-line hybrids since the desired resistance genes need to be incorporated in
two rather than three parental lines.
5. Abiotic stress tolerance. Hybrid rice technology so far has been used under the irrigated rice ecosystem. Environmental degradation resulting from salinity and water
shortages has posed a major threat to sustaining food production. Researchers at IRRI
and in Egypt, India, and China have found
that hybrid rice technology can be extended
to saline-prone irrigated conditions because
hybrids have performed exceedingly well
under moderate to high saline soil conditions. Since the high cost of seed is a major
constraint to the adoption of this technology by resource-poor farmers and two-line
hybrid breeding and the seed production
70
approach make it more cost-effective, this

technology can be adopted by resource-poor
farmers.
6. Quality. The negative influence of WA cytoplasm on certain quality parameters (such
as grain chalkiness) allows the alternative
use of EGMS-based two-line hybrid rice
technology to overcome such drawbacks.
A multidisciplinary approach in developing
superior EGMS lines and pollen parents can help
to develop two-line rice hybrids suitable for the
different ecological situations in which rice is
grown. Despite the promise that two-line hybrid
rice technology holds, it would be wise to have a
harmonious balance in using two-line and threeline hybrids and conventional rice varieties in an
appropriate manner in national rice production
programs.
Bibliography
Abdallah AA, Verkerk K. 1968. Growth, flowering

and fruit set of the tomato at high temperature.
Neerl. J. Agric. Sci. 16:71-76.
Agarwala SC, Chatterjee C, Sharma PN, Sharma CP,
Nautiyal N. 1979. Pollen development in maize
plants subjected to molybdenum deficiency. Can.
J. Bot. 57:1946-1950.
Agarwala SC, Sharma PN, Chatterjee C, Sharma CP.
1980. Copper deficiency induced changes in
wheat anthers. Proc. Indian Natl. Sci. Acad. Sect.
B 2:172-176.
Ahokas H, Hockett EA 1977. Male sterile mutant of
barley. IV. Different fertility levels of msgqci (cv.
Vantage), an ecoclinal response. Barley Genet.
Newsl. 7:10-11.
Ali J. 1990. Studies on gametocides in rice (Oryza
sativa L.). M.Sc. thesis submitted to the Indian
Agricultural Research Institute, New Delhi, India. 124 p.
Ali J. 1993. Studies on temperature sensitive genic
male sterility and chemical induced male sterility
towards development of two-line hybrids in rice
(Oryza sativa L.). Ph.D. thesis submitted to the
Indian Agricultural Research Institute, New Delhi,
India. 138 p.
Ali J, Siddiq EA, Zaman FU, Abraham MJ, Ahmed I.
1995. Identification and characterization of temperature sensitive genic male sterile sources in
rice (Oryza sativa L.). Indian J. Genet. 55(3):243259.
Ali J. 1996. Characterisation and genetics of temperature sensitive male sterile line SA 2 in rice. Abstract: PI-102. 2nd International Crop Science
Congress, New Delhi, India.
Anonymous. 1978. Studies on male sterility of rice
induced by male gametocide no. 1. Acta Bot. Sin.
230:303-313.
Atsumi S, Takeo N, Yasuo K. 1992. Sterility in rice
induced by chemical treatment. Biosci. Biotech.
Biochem. 56(1):1619-1622.
Batch JJ, Morgan, DG. 1974. Male sterility induced in

barley by photoperiod. Nature 250:165-167.
Berthelem P, Le Guen J. 1975. Rapport dactivite-station damelioration des plantes. INRA Rennes
(France): 1971-74. (As reported in Kaul, 1988.)
Borkakati RP, Virmani SS. 1993. Inheritance of a
thermosensitive genic male sterile mutant of indica rice. Rice Genet. Newsl. 10:92-94.
Brar DS. 1982. Male sterility in sesame. Indian J.
Genet. 42:23-27.
Chan YW, Cheah CH. 1981. Evaluation of ethrel as a
selective gametocide in the breeding of rice
(Oryza sativa L.). In: Proceedings of the 4th International SABRAO Congress, Malaysia. p 24.
Chen P. 1985. Studies on the effect of amidosulfonic
acid (a male sterility agent) to induce male sterility of rice. (Hua-nan-nung-yeh da hsueh-husehpao.) J. South China Agric. Univ. 6(2):62-68.
Chen PL, Chen B, Chen JL, Ho LK. 1986. Arsenic
content and castration effect in reaction to male
gametocide no. 2 in rice plants. (Hua-nan-nungyeh da hsueh-huseh-pao.) J. South China Agric.
Univ. 7(4):22-27.
Cheng YK, Huang, CS. 1978. Some properties of cytoplasmic male sterile rice. J. Agric. Res. Taiwan
27:267-290.
Cheng YK, Huang CS. 1980. Studies on the cytoplasmicgenic male sterility in cultivated rice (Oryza
sativa L.). II. Morphological-histological investigations on functional male sterility. J. Agric. Res.
China 29:67-80.
Cheng S, Sun Z, Si H, Zhuo L. 1995. Response to
photoperiod and temperature in photo-thermo period sensitive genic male sterile line Xinguang S
(Oryza sativa L. subsp. indica). Chinese J. Rice
Sci. 9(2): 87-91.
Cheng SH, Si HM, Zhou LS, Sun ZX. 1996. Classification of environmentally induced genetic male
sterile lines of rice based on their fertility responses
to photoperiod and temperature. J. Agric. Sci.
(Camb.) 127:161-167.
71
Chu CC, Wang CC, Sun CS, Hsu C, Yin KC, Chu Y.
1975. Establishment of an efficient medium for
anther culture of rice through comparative experiments on the nitrogen source. Sci. Sin. 18:659668.
Comstock RE, Robinson HF. 1952. Estimation of average dominance of genes. In: Heterosis. Ames,
Iowa (USA): Iowa State College Press. p 494516.
Daskaloff S. 1972. Male sterile pepper (C. annuum)
mutants and their utilization in heterosis breeding.
Proceedings of Eucarpia meeting. Capsicum
7:205-210.
Davenport CB. 1908. Degeneration, albinism and inbreeding. Science 28:454-455.
Dell B. 1981. Male sterility and anther structure in copper deficient plants. Ann. Bot. 48:599-608.
Duc G. 1980. Effect of environment on instability of
two sources of cytoplasmic male sterility in faba
beans. FEBS Lett. 2:29-30.
Duvick DN. 1966. Influence of morphology and sterility on breeding methodology. In: Frey KJ, editor.
Plant breeding. Vol. I. Ames, Iowa (USA): Iowa
State University Publications. p 85-138.
East EM. 1908. Inbreeding in corn. Rep. Connecticut
Agric. Exp. Stn. for 1907. p 419-428.
East EM. 1936. Heterosis. Genetics 21:375-397.
Fisher JE 1972. The transformation of stamens to ovaries and of ovaries to inflorescences in I. aestivum
L. under short day treatment. Bot. Gaz. 133:7885.
Ikehashi H, Araki H. 1984. Varietal screening for compatibility types revealed in F1 sterility of crosses
in rice. Jpn. J. Breed. 34(3):304-312.
Ikehashi H, Araki H. 1986. Genetics of F1 sterility in
remote crosses in rice. In: Rice genetics. Manila
(Philippines): International Rice Research Institute. p 119-130.
Jan CC. 1974. Genetic male sterility in wheat (Triticum aestivum L.): expression, stability, inheritance and practical use. Ph.D. thesis. University
of California, Davis, Calif., USA.
Jia JH, Zhang DS, Li CY, Qu XP, Wang SW,
Chamarerk V, Nguyen HT, Wang B. 2001. Molecular mapping of the reverse thermo-sensitive
genic male sterile gene (rtms 1) in rice. Theor.
Appl. Genet. 103(4):607-612.
Jinks JL. 1983. Biometrical genetics of heterosis. In:
Frankel R, editor. Heterosis: reappraisal of theory
and practice. Berlin (Germany): Springer-Verlag.
Kaul MLH. 1988. Male sterility in higher plants. Monograph on Theoretical and Applied Genetics 10.
Berlin (Germany): Springer-Verlag. 1,005 p.
Kempthorne O. 1957. An introduction to genetic statistics. London (UK): John Wiley and Sons, Inc.
72
Kitaoka M, Ota K, Ogawa M. 1991. The synergistic

effects of ethephon and isourea in male sterility in
rice. Jpn. J. Crop Sci. 60(3):349-354.
Li X, Pan XB, Chen ZX, Yu HX, Cheng ZK, Giu MH
1995. Anther culture of photosensitive male sterile rice and the breeding of Miai 64S. J. Jiangsu
Agric. Coll. 16(4):7-12.
Li XL, Xie R, Deng XH, He GH, Yang ZL, Zuo YS,
Wu LJ. 1994a. Comparison of the effect of anther
culture in a photoperiod (temperature) sensitive
genic male sterile line and a wide compatibility
restoring line in rice. Southwest China J. Agric.
Sci. 8 (Special Issue):88-93.
Li RH, Cai HW, Wang XK. 1994b. A specific esterase band found in Annong 1-S. Int. Rice Res.
Newsl. 19(3):6-7.
Li ZK, Luo LJ, Mei HW, Shu QY, Tabien R, Zhong
DB, Ying CS, Stansel JW, Khush GS, Paterson
AH. 2000. Genetic basis of inbreeding depression and heterosis in rice (Oryza sativa L.). In:
Paper presented at the International Rice Research Conference, 31 March-3 April 2000. Los
Baos (Philippines): International Rice Research
Institute.
Long JD, Weed MB, Fitzgerald DJ, Barrier GE. 1973.
Chemical control of pollination. In: Proceedings
of the American Society of Agronomy, Madison,
Wis. (USA). p 69.
Lu X, Wang J. 1988. Fertility transformation and genetic behavior of Hubei photoperiod-sensitive
genic male sterile rice. In: Hybrid rice. Los Baos
Lu XG, Virmani SS, Yang RC. 1998. Advances in
two-line hybrid rice breeding. In: Virmani SS,
Siddiq EA, Muralidharan K, editors. Advances in
hybrid rice technology. Proceedings of the Third
International Symposium on Hybrid Rice, 14-16
November 1996, Hyderabad, India. Los Baos
Lu XG, Zhang ZG, Maruyama, K, Virmani SS. 1994.
Current status of two-line method of hybrid rice
breeding In: Virmani SS, editor. Hybrid rice technology: new developments and future prospects.
Manila (Philippines) International Rice Research
Institute. p 37-39.
Luo Z, Zhong A, Zhou Q. 1988. Effects of a new non
toxic chemical gametocide on rice. In: Hybrid
rice. Manila (Philippines): International Rice Research Institute. p 147-156.
Martin JA, Crawford JH. 1951. Several types of sterility in Capricum frutescens. Proc. Am. Soc. Hort.
Sci. 57:335-338.
Maruyama K, Araki H, Kato H. 1991. Thermosensitive

genetic male sterility induced by irradiation. In:
Rice genetics II. Manila (Philippines): International Rice Research Institute. p 227-235.
Mei G, Wang X, Wang, M. 1990. Genetical analysis
of the photoperiod sensitive male sterility of
Nongken 58s and its derivatives. J. Huazhong
Agric. Univ. 9(4):400-406.
Mei CS, Zhang JY, Wu GN. 1988. A preliminary report on increasing of green pollen plants in indica
rice. Jiangsu J. Agric. Sci. 4(2):45-48.
Mei MH, Dai KK, Xu CG, Zhang Q. 1999. Mapping
and genetic analysis of the genes for photoperiodsensitive genic male sterility in rice using the original mutant Nongken58S. Crop Sci. 39:1711-1715.
Mou TM, Li CH, Yang GC, Lu XG. 1998. Breeding
and characterizing indica PGMS and TGMS lines
in China. In: Virmani, SS, Siddiq EA,
Muralidharan K, editors. Advances in hybrid rice
technology. Proceedings of the Third International
Symposium on Hybrid Rice, 14-16 November
1996, Hyderabad, India. Manila (Philippines): International Rice Research Institute. p 79-88.
Murty UR. 1995. Breeding two line hybrids in Sorghum bicolor (L.) Moench. Cereal Res. Comm.
23:397-402.
Oard JH, Hu J. 1995. Inheritance and characterization
of pollen fertility in photoperiodically sensitive rice
mutants. Euphytica 82:17-23.
Pan XG, Zhu DY, Yin JH, Ding XH 1993. A study on
the technological procedures for breeding dualpurpose genic male-sterile rice. Acta Agric.
Jiangxi, China 5(Supplement):1-5.
Parmar KS, Siddiq EA, Swaminathan MS. 1979.
Chemical induction of male sterility in rice. Indian J. Genet. Plant Breed. 39:529-541.
Perez AT, Chang TT, Beachell HM, Vergara BS,
Marciano AP. 1973. Induction of male sterility in
rice with ethrel and RH-531. SABRAO Newsl.
5:133-139.
Peterson PA. 1958. Cytoplasmically inherited male
sterility in Capsicum. Am. Nat. 92:111-119.
Reddy OUK, Siddiq EA, Sarma NP, Ali J, Hussain
AJ, Nimmakayala P, Ramasamy P, Pammi S,
Reddy AS. 2000. Genetic analysis of temperature-sensitive male sterility in rice. Theor. Appl.
Genet. 100:794-801.
Rerkasem B, Jamjod S. 1997. Boron deficiency induced male sterility in wheat (Triticum aestivum
L.) and its implications for plant breeding.
Euphytica 96:257-262.
Rick CM, Boynton JE. 1967. A temperature sensitive
male sterile mutant of the tomato. Am. J. Bot.
54:601-611.
Rudich, J, Peles A. 1976. Sex expression in watermelon as affected by photoperiod and temperature. Sci. Hort. 5:339-344.
Rundfeldt H. 1960. Untersuchungen zur zuchtung des

Kopfkohls (B. olerecea L. var. Capitata). Z.
Pflanzenzuchtung 44:30-62. (As reported in Kaul,
1988.)
Rutger JN, Schaeffer GW. 1989. An environmentally
sensitive genetic male sterile mutant in rice. In:
Annual Meeting of the American Society of
Agronomy, Las Vegas, Nev., USA. Agron. Abstr.
p 98.
Sawhney VK. 1983. Temperature control of male sterility in a tomato mutant. J. Hered. 74:51-54.
Sharma RK, Reinbergs E. 1976. Male sterility genes
in barley and their sensitivity to light and temperature intensity. Ind. J. Genet. Plant Breed. 3 2:408410.
Shen YW, Kai QH, Gao M. 1993. A new thermo sensitive radiation induced male sterile rice line. Rice
Genet. Newsl. 10:97-98.
Shen Y, Gao, M, Cai Q. 1994. A novel environmentinduced genic male sterile (EGMS) mutant in indica rice. Euphytica 76:89-96.
Sheng XB. 1992. Genetics of photoperiod sensitive
genic male sterility of Nongken 58S (Oryza sativa L.). Chinese J. Rice Sci. 6(1):5-14.
Shi MS. 1981. Preliminary report of breeding and utilization of late japonica natural double purpose line.
J. Hubei Agric. Sci. 7:1-3.
Shi MS. 1985. The discovery and the study of the photosensitive recessive male sterile rice (Oryza sativa L. subsp. japonica). Sci. Agric. Sin. 2:44-48.
Shi MS, Deng JY. 1986. The discovery, determination
and utilization of Hubei photosensitive genic male
sterile rice (Oryza sativa L. subsp. japonica). Acta
Genet. Sin. 13(2):107-112.
Song MT, Kim, JK, Choe ZR. 1990. Effects of chemicals on inducing grain sterility of rice. Korean J.
Crop Sci. 35(4):309-314.
Stevens MA, Rudich J. 1978. Genetic potential for
overcoming physiological limitations on adaptability, yield and quality in tomato. Hort. Sci. 13:673678.
Stuber CW, Lincoln SE, Wolff DW, Helentjaris T,
Lander ES. 1992. Identification of genetic factors
contributing to heterosis in a hybrid from two elite
maize inbred lines using molecular markers. Genetics 132:823-839.
Subudhi PK, Borkakati RP, Virmani SS. Huang N.
1997. Molecular mapping of a thermosensitive
genetic male sterility gene in rice using bulked
segregant analysis. Genome 40:188-194.
Sun ZX, Min SK, Xiong ZM. 1989. A temperature
sensitive male sterile line found in rice. Rice
Genet. Newsl. 6:116-117.
Takeoka YI, Ito M, Yamamura S, Saka H. 1990. Morphogenetic alterations of rice floral organs as affected by spray of chemical hybridizing agent
HGR-626. Jpn. J. Crop Sci. 59(3):528-534.
73
Tan ZC, Li YY, Chen LB, Zhou GQ. 1990. Studies on

ecology and adaptability of dual-purpose line An
Nong S-1. Hybrid Rice 3:35-38.
Virmani SS, Maruyama K. 1995. Some genetic tools
for hybrid breeding and seed production in self
pollinated crops. In: Sharma et al, 1995 editors.
Genetic research and education: current trends
and next fifty years. New Delhi (India): Indian
Society of Genetics and Plant Breeding. p 708728.
Virmani SS, Voc PC. 1991. Induction of photo- and
thermo-sensitive male sterility in indica rice.
Agron. Abstr. 119.
Virmani SS. 1992. Transfer and induction of
thermosensitive genic male sterile mutant in indica rice. In: Proceedings of the Second International Symposium on Hybrid Rice, 21-25 Apr.
1992. Manila (Philippines): International Rice
Research Institute.
Virmani SS, Viraktamath BC, Lopez MT. 1997b.
Nucleus and breeder seed production of
thermosensitive genic male sterile lines. Int. Rice
Res. Newsl. 22(3):26-27.
Virmani SS, Viraktamath BC, Casal CL, Toledo RS,
Lopez MT, Manalo JO. 1997a. Hybrid rice breeding manual. Manila (Philippines): International
Rice Research Institute. 151 p.
Wang B, Wang JZ, Wu W, Zheng HG, Yang ZY, Xu
WW, Ray JP, Nguyen HT. 1995. Tagging and
mapping the thermosensitive genic male sterile
gene in rice (Oryza sativa L.) with molecular
markers. Theor. Appl. Genet. 91(6-7):1111-1114.
Wang X, Que RF. 1981. Induction of pollen sterility by
ethylene and gametocides in rice. Acta
Phytophysiol. Sin. 7:381-383.
Wang X, Que, RF, Xu MF. 1981. Preliminary study on
the effect of AWN on the induction of male sterility in rice. Zheijiang Nongye Kexue (Zheijiang
Agric. Sci.) 3:135-139.
Wang X, Yu M, Tao LX, Yao F. 1991a. The male
sterility in rice plant induced by CRMS. I. Effect
of CRMS on male sterility in rice plants. Chinese
J. Rice Sci. 5(2)49-55.
Wang, X, Yu M, Tao LX, Yao F. 1991b. The male
sterility in rice plant induced by CRMS. II. Physiological effects of CRMS on inducing male sterility in rice. Chinese J. Rice Sci. 5(3)104-108.
Wang, X, Yu M, Tao LX, Yao F. 1991c. The male
sterility in rice plant induced by CRMS. III. Effect
of CRMS on free proline content in anthers of rice
and its possible relation to abortion of anther. Chinese J. Rice Sci. 5(4):169-174.
74
Wang XM, Wang MQ, Mei GZ, Wu HG, Duan WJ,

Wang WJ. 1991. Photoperiod conditioned male
sterility and its inheritance in rice. In: Rice genetics II. Proceedings of the Second International
Rice Genetics Symposium, 14-18 May 1990.
Manila (Philippines): International Rice Research
Institute. p 217-226.
Wang FP, Mei MH, Xu CG, Zhang Q. 1997. pms1 is
not the locus revelant to fertility difference between the photoperiod-sensitive male sterile rice
Nongken 58S and normal rice Nongken58. Acta
Bot. Sin. 39:922-925.
Wei BG, Sun GC, Chang JW, Jiang CX. 1994. The
determination of photoperiod-temperature sensitive male sterile soybean. Proceedings of Third
National Youth Symposium on Crop Genetics and
Breeding. Edited by China Agricultural Scientech
Press. p 185-189.
Xi DW, Cheng WJ, Yi DL, Yu DR, Ning ZL, Deng
XX, Li M. 1997. Genetic analysis on the TGMS
line in rape of B. napus Xiangyou 91S. In: Lu RL,
Cao XB, Liao FM, Xin YY, editors. Proceedings
of the International Symposium on Two-Line
System of Heterosis Breeding in Crops, 6-8 September 1997, Changsha, China. p 215-220.
Xiang YW, Zhang ZX, Zhang AZ, Zhou XM, Wie R.
1993. Selection and breeding of photosensitive
genic male sterile rice by anther and somatic cell
culture. J. Southwest Agric. Univ. 15(5):462-465.
Xiao J, Li J, Yuan LP, Tanksley SD. 1995. Dominance is the major genetic basis of heterosis in
rice as revealed by QTL analysis using molecular markers. Genetics 140:745-754.
Xiao GY, Yuan LP, Fu, XQ, Deng XP. 1997. Studies
on effect of water temperature on male fertility of
the thermo-sensitive genic male sterile (TGMS)
lines in rice under simulated low air temperature
conditions in high summer. In: Lu RL, Cao XB,
Liao FM, Xin YY, editors. Proceedings of the
International Symposium on Two-Line System
of Heterosis Breeding in Crops, 6-8 September
1997, Changsha, China. p 64-69.
Xie R, Deng XH, Li XL, Luo GC, Wu LJ, He XY.
1993. Studies on fertility transition in the anther
culture-derived progeny of photothermo-sensitive
genic male sterile rice. J. Southwest Agric. Univ.
15(5):450-453.
Yamaguchi Y, Ikeda R, Hirasawa H, Minami M,
Ujikara A. 1997. Linkage analysis of
thermosensitive genic male sterility gene, tms2, in
rice (Oryza sativa L.). Breed. Sci. 47:371-373.
Yang RC, Wang NY, Mang K, Chau Q, Yang RR,
Chen S. 1990. Preliminary studies on application
of indica photo-thermo genic male sterile line
5460S in hybrid rice breeding. Hybrid Rice 1:3234.
Yang YQ, Wang L. 1990. The breeding of

thermosensitive male sterile rice R59TS. Sci.
Agric. Sin. 23(2):90.
Yang RC, Liang K, Wang N, Chen S. 1992a. A recessive gene in indica rice 5460S for thermosensitive
genic male sterility. Rice Genet. Newsl. 9:56-57.
Yang RC, Wang NY, Mang, K, Chau Q, Yang, RR,
Chen S. 1992b. Preliminary studies on application of indica photo-thermo genic male sterile line
5460S in hybrid rice breeding. Hybrid Rice 1:3234.
Yang ZP. 1997. Inheritance of photoperiod genic male
sterility and breeding of photoperiod sensitive
genic male sterile lines in rice (Oryza sativa L.)
through anther culture. Euphytica 94:93-99.
Yin HQ. 1999. Scientific progress made in the two line
system of hybrid rice research and production in
China. Hunan Agric. Res. Newsl. 6(3):4-8.
Yu SB, Zi JX, Xu CG, Tan YF, Gao YJ. 1997. Importance of epistasis as the genetic basis of heterosis
in an elite rice hybrid. Proc. Natl. Acad. Sci. USA
94:9226-9231.
Yuan LP. 1997. Exploiting crop heterosis by two-line
system hybrids: current status and future prospects. In: Proceedings of the International Symposium on Two-Line System of Heterosis Breeding in Crops, 6-8 September 1997, Changsha,
China.
Yuan SC, Zhang ZG, He HH, Zen HL, Lu KY, Lian
JH, Wang BX. 1993. Two photoperiodic reactions in photoperiod-sensitive genic male sterile
rice. Crop Sci. 33:651-660.
Zhang M, Liang C, Huang Y, Chen B. 1994. Fertility
expression of photoperiod temperature sensitive
genic male sterile rice in Guangzhan and the response to photoperiod and temperature treatments.
J. Trop. Subtrop. Bot. 2(4):100-107.
Zhang Z, Zeng H, Yang J, Yuan SC, Zhang D. 1994.
Conditions inducing fertility alteration and ecological adoption of photoperiod-sensitive genic
male sterile rice. Field Crops Res. 38(2):111-120
Zhang ZG, Yuan SC, Zen HL, Li YZ, Li ZC, Wei
CL. 1990. Preliminary analysis on fertility
changes and adaptability of HPGMR at different
altitudes in Yuanjiang. J. Huazhong Agric. Univ.
9(4):348-354.
Zhang KT, Fu HY. 1982. Effect of high temperature
treatment on male sterility in sorghum. Acta Genet.
(China) 9:71-77.
Zhang NY, Xue QZ. 1996. Development of photoperiod sensitive genic male sterile lines using anther
culture in rice (Oryza sativa L.). J. Zhejiang Agric.
Univ. 22(5):474-480.
Zhang Q, Shen BZ, Dai XK, Mei MH, Saghai-Maroof

MA, Li ZB. 1994. Using bulked extremes and
recessive class to map genes for photoperiod sensitive genic male sterility in rice. Proc. Natl. Acad.
Sci. USA 91:8675-8679.
Zhang Q, Hua J, Yu S, Xiong L, Xu C. 2001. Genetic
and molecular basis of heterosis in rice. In: Khush
GS, Brar DS, Hardy B, editors. Rice genetics IV.
Proceedings of the Fourth International Rice Genetics Symposium, 22-27 October 2000, Los
Baos, Philippines. New Delhi (India): Science
Publishers, Inc., and Los Baos (Philippines): International Rice Research Institute. p 173-188.
Zhang ZG, Yuan, SC. 1989. Effect of twilight on two
photoperiod reactions in Hubei photoperiod sensitive male sterile rice. Chinese J. Rice Sci.
3(3):107-112.
Zhang ZG, Lu XG, Yuan LP. 1992a. Reflections on
identification of light and temperature properties
of fertility transformation of photoperiod sensitive
male sterile rice. Hybrid Rice 6:29-32.
Zhang ZG, Yuan, SC, Xu CZ. 1987. The influence of
photoperiod on fertility changes of Hubei photoperiod sensitive genetic male sterile rice. Chinese
J. Rice Sci. 1(3):137-143.
Zhang ZG, Yuan SC, Zen HL, Li YZ, Zhang DP.
1992b. Studies on model of photo-thermo reaction of fertility alteration in photosensitive genetic
male sterile rice. J. Huazhong Agric. Univ. 11:1-6.
Zhang ZG, Yuan SC, Zen HL, Li YZ, Li ZC, Wei CL.
1990. Preliminary analysis on fertility changes and
climatic adaptation of HPGMR at different altitudes in Yuan Jiang. J. Huazhong Agric. Univ.
9:348-354.
Zhang ZG, Yuan SC, Zen HL, Li YZ, Li ZG, Wei
CL. 1991. Preliminary observations of fertility
changes in a new type temperature sensitive male
sterile rice IVA. Hybrid Rice 1:31-34.
Zhangxing Z, Chunnong H. 1980. A study of male
sterility induced in rice (Oryza sativa L.) by 3-(pchlorophenyl) 6-methoxy-S-triazine-2,4-(1H, 3H)
Dione. J. Huazhong Univ. 4:81-87.
Zhu DY, Sun ZX, Pan XG, Ding XH, Shen XH, Wan
Y, Pan H, Yin JH, Alejar MS, Torrizo LB, Datta
SK. 1988. Use of anther culture in hybrid rice
breeding. In: Virmani SS, Siddiq EA,
Muralidharan K, editors. Advances in hybrid rice
technology. Proceedings of the Third International
Symposium on Hybrid Rice, 14-16 November
1996, Hyderabad, India. Manila (Philippines): International Rice Research Institute. p 265-281.
Zhou TB, Xiao HC, Lei DY, Duan QZ. 1988. The
breeding of indica photoperiod male sterile lines.
J. Hunan Agric. Sci. 6:16-18.
75
Authors
S.S. Virmani is plant breeder and deputy head,

Plant Breeding, Genetics, and Biochemistry Division, IRRI; Sun Zongxiu is a research professor at
the Key Laboratory for Rice Biology of the Ministry of Agriculture and at the China National Rice
Research Institute, Hangzhou; Mou Tongmin is a
professor in the Agronomy Department of
Huazhong Agricultural University in Wuhan,
Hubei; Jauhar Ali is national coordinator of the
Hybrid and Molecular Rice Breeding Program of
Iran at the Rice Research Institute of Iran in Rasht;
Chang Xiang Mao was a plant breeder at IRRI and
is now senior rice breeder cum chief of the International Project Office at the Rice Research Institute of the Guangxi Academy of Agricultural Sciences in Nanning City, Guangxi, China.
76
Glossary
A line the male sterile parent involving cytoplasmic or cytoplasmic genetic male sterility. It is also known as a CMS line.
adaptability the ability of a genotype to adjust
to a given environment and give a reasonably good yield.
allo-plasmic lines CMS or restorer lines that
have different cytoplasms.
anther terminal part of the stamen that contains
the pollen grains (male gametes).
anthesis the action of opening of a flower or
spikelet of rice.
apiculus a small acute point or tip of a rice spikelet; extension of the lemma or palea.
apomixis a kind of asexual reproduction
through seed in which the embryo develops
from a maternal cell without fertilization. The
resulting seed has the same genetic constitution as that of the seed parent.
apparent heterosis subjective superiority of a
hybrid over its parents or a check variety
based on visual observation.
augmented design a statistical design used for
evaluation of genotypes in which the check
varieties are replicated and the test entries are
not replicated but are allotted randomly to
the blocks.
auricles the small paired ear-like appendages
on either side of the base of the rice leaf blade
that may not be present in older leaves.
awn a bristle-like extension of varying length
originating from the lemma of the spikelet.
B line the fertile counterpart parent of the male

sterile A line of a cytoplasmatic or cytoplasmic genetic male sterility system that is used
as a male parent to maintain the latter. It is
also known as a maintainer line.
backcross method a breeding method in which
the F1 hybrid is again crossed with either of
its parents. The resulting progeny is called a
backcross progeny.
backcross nursery breeding nursery in which
male sterile plants identified among the testcrosses (CMS elite lines) are crossed with
the respective male parents to transfer cytoplasmic male sterility into the nuclear genotype of the elite line.
boot a rapidly growing panicle enveloped by
the flag-leaf sheath. In tissue culture, this refers to the panicle collected when the distance
between the collar of the flag leaf and subtending leaf is about 7 to 8 cm.
booting bulging of the flag-leaf sheath because
of the growing panicle inside.
border rows the recommended number of rows
of the male parental line grown on all sides of
the hybrid seed production plot to minimize
contamination by outcrossing with stray pollen.
bract a leaf from the axis from which a flower
arises.
breeder seed breeder seed is the seed of the highest genetic purity and is produced by the
agency sponsoring a variety; it is used to produce foundation seed.
77
C
caryopsis a small one-seeded dry indehiscent
fruit with a thin membranous pericarp adhering so closely to the seed that fruit and seed
are incorporated in one body, forming a
single grain, as in wheat and barley. In rice,
brown rice is the caryopsis.
certified seed seeds used for commercial crop
production produced from foundation, registered, or certified seeds under the regulation
of a legally constituted agency. In hybrid rice,
it is F1 seed produced directly from CMS
restorer lines grown as per certification standards.
CHA (chemical hybridizing agent) any chemical that is used to induce male sterility in
plants.
check variety any popular or high-yielding variety widely grown in a region.
chemical mutagen any chemical used to induce
mutations artificially.
chemical hybridizing agent (CHA) this is any
chemical formulation, that is, auxins, antiauxins, growth regulators, arsenicals, oxanilates,
ethylene-releasing compounds, halogenated
aliphatic acids, etc., that has the ability to
selectively sterilize the male gametes without affecting ovular fertility and is not phytotoxic.
CMS the CMS line is governed by genetic factors present in the mitochondria of the cytoplasm responsible for inducing selective male
sterility. But its pistil is normal and it can
produce seeds when pollinated by any normal plant.
combining ability the ability of a genotype (inbred, pure line, or synthetic) to transfer its
desirable traits to its progeny: generalaverage performance of a strain in a series of
crosses; specificdeviation from performance predicted on the basis of general combining ability of parental lines.
correlation coefficient a measure of the degree
of association between two variables that is
computed as the ratio of the covariance of the
two variables to the products of their standard errors. Its values vary between 1 and
+1.
covariance the mean of the product of the deviation of two varieties from their individual
means.
78
critical difference a statistical parameter computed to test whether the observed differences
between the means of entries are significant
or not.
critical fertility point (CFP) the critical temperature or photoperiod experienced by the
EGMS line during the sensitive stage results
in maximum pollen and spikelet fertility.
critical sterility point (CSP) the critical temperature or photoperiod experienced by the
EGMS line during the sensitive stage results
in complete pollen and spikelet sterility.
cross fertilization the fertilization of an egg
nuclei (ovule) of one parent by the pollen of
another parent.
cytoplasm all the protoplasm of the cell except
the nucleus.
cytoplasmatic heredity/inheritance the transmission of characters from parent to offspring
through the cytoplasm of the germ cell.
D
daylength - number of light hours in a day.
diallel mating a mating design in plant breeding in which a set of parents is crossed in all
possible combinations.
dihybrid a hybrid of two different genes; heterozygous for two pairs of alleles.
diploid (2n) an organism having two chromosomes of each kind.
disomic a plant having one or more chromosomes duplicated, but not the entire genome.
diverse having or capable of having various
forms or qualities.
dominance intra-allelic/intragenic interaction
with complete suppression of the effects of
one allele by another.
E
effective accumulated temperature (EAT) the
total effective temperature in centigrade received by the plant from seeding to flowering. It is useful for predicting flowering.
EAT = Mean daily temperature (C) temperature higher than 30 C temperature of
lower limit (18 C)
emasculation the process of removal of anthers
from the florets so as to make the plant male
sterile.
elite line an improved breeding line or a variety.

endosperm the nutritive tissue of the ripened
ovary. It consists of the aleurone layer and
the starchy tissue, and serves as the source of
food for the germinating embryo.
environmental genic male sterility (EGMS)
male sterilityfertility transformation controlled by environmental factors such as temperature and photoperiod.
epistasis the interaction of different genes in the
expression of a trait.
F
F1 abbreviation for the first filial generation,
usually the hybrid between two homozygous
parents.
fertility restoration an ability of a genotype to
restore fertility to its progeny when crossed
to a CMS line.
fertilization fusion of the nuclei of male and
female gametes.
flag leaf the uppermost leaf (of rice plant) originating just below the panicle base.
flag-leaf clipping a method of cutting 1/2 to
2/3 of the flag leaf from its tip in CMS and
restorer lines to facilitate easy pollen dispersal.
floret a unit of the spikelet, which includes the
lemma, palea, and the flower.
flower, rice the reproductive organ consisting
of lemma, palea, two lodicules, six stamens,
and the pistil.
foundation seed seed stock produced from
breeder seed by or under the direct control of
a breeder or a research station. Foundation
seed is the source of certified seed, either directly or through registered seed.
G
GA3 a form of gibberellic acid that is sprayed on
CMS lines to obtain good panicle exsertion.
gamete a mature reproductive male or female
germ cell, sperm, or egg specialized for fertilization.
gametic (tissue or generation) having n number of chromosomes (haploid), in contrast to
zygotic tissue with 2n (diploid).
gametocide organic or inorganic chemicals used
for killing the functional sexual parts (pollen, ovule) of the plant. These may be selective for male or female parts.
gametophytic in this system, the sterility/fertility reaction is imparted to the pollen by the
genetic constitution of the pollen itself and
is controlled by a single gene, which may have
a large number of allelic forms.
genetic purity trueness to type; seeds/plants
confirming to the characteristics of the line/
variety/hybrid as described by the breeder.
genetic shift change in the genetic makeup of
the line/variety/hybrid if grown over a long
period, particularly in areas outside its adaptation.
genic male sterility the type of male sterility
governed entirely by the nuclear genes. It may
be transmitted by either the male or female
parent.
germination the resumption of growth by the
embryo and development of the young plant
from the seed. Germination, precisely, is the
emergence and development from the seed
embryo of those essential structures that, for
the kind of seed being tested, indicate the
ability to develop into a normal plant under
favorable conditions in the soil.
grain the ripened ovary and its associated structures.
H
heading (flowering), rice growth stage of the
rice plant marked by the emergence of the
panicle from the boot followed by anthesis.
heritability broadly, the proportion of observed
variance that is inherited, the remainder being due to environmental effects. Strictly, the
proportion of variance caused by the additive effect of genes.
heterobeltiosis refers to the phenomenon in
which an F1 hybrid obtained by the crossing
of two genetically dissimilar parents shows
superiority over the better parent in one or a
combination of characters.
heterosis refers to the phenomenon in which an
F1 hybrid obtained by the crossing of two genetically dissimilar parents shows superiority over mid-parental values in one or a combination of characters.
heterosis (standard) refers to the phenomenon
in which the F1 hybrid obtained by the crossing of two genetically dissimilar parents
shows superiority over the best standard check
prevailing at that time in one or a combination of characters.
79
heterosis breeding a method of breeding to develop an F1 hybrid obtained by the crossing

of two genetically dissimilar parents.
heterozygote an individual having different alleles for any gene pair and producing two
kinds of gametes.
heterozygous hybrid for any gene pair, with different alleles for the gene being considered.
hill a group of rice plants directly adjacent to
each other because the seeds or seedlings were
planted together. A hill may also consist of
only one plant.
hybrid the product of a cross between genetically dissimilar parents.
hybrid rice the F1 seed of rice bred for commercial use.
hybrid vigor increased vigor of the hybrid over
its parents in one or more characteristics.
hybridization a breeding method in which two
varieties are crossed to produce new variability and desired recombinants. The hybrids are
allowed to self-pollinate and the segregating
populations are handled by an appropriate
method.
I
inbred an individual resulting from the mating
of closely related parents or by selfing.
inbred line a nearly homozygous line produced
by continued self-fertilization.
inbreeding the interbreeding of closely related
individuals occurring naturally (as in a closed
population) or as a deliberately chosen system of breeding and serving especially to preserve and fix desirable characters or to eliminate unfavorable characters from a suitably
selected stock but tending to bring about an
unwanted decline (as in size, vigor, or fertility) through the fixation of undesirable and
often recessive characters when the initial
stock is in any way defective.
indoor growth cabinets small indoor chambers
wherein temperature, humidity, and light are
artificially controlled.
intersubspecific hybrid a cross between different subspecies of a crop. For example, in rice,
hybrids between indica and japonica lines are
considered as intersubspecific hybrids.
isolation the separation of one group from another so that mating between or among groups
is prevented.
80
isolation (barrier) the separation between two

groups can be provided by topography surface features or artificial/natural obstacles to
a height of at least 2.5 m for rice.
isolation-free method a method of producing
hybrid seed for experimental purposes without isolation but by providing crop barriers
of 24 rows of the restorer lines.
isolation (space) separation is provided by keeping a certain distance between two groups. A
space isolation of 50100 meters is ideal for
hybrid rice seed production.
isolation (time) separation is provided by growing two groups at different times of the crop
season so that one group is already mature
(stopped providing pollen) when the other
group reaches flowering. Generally, a period
of 21 days difference in flowering is sufficient for rice.
isoplasmic these are the CMS or restorer lines
differing in nuclear genetic constitution but
having common cytoplasm.
L
leaf number total number of leaves developed
on the main culm of a plant, which is a characteristic feature of each variety.
lodicules the two scale-like structures adjoining the base of the palea that control the opening of the lemma and palea during anthesis.
M
maintainer line a pollinator variety is used to
pollinate a CMS line and produce progenies
that remain male sterile. If there is no maintainer line, the male sterile line cannot be
maintained and multiplied generation after
generation.
male sterility absence or nonfunction of pollen
in plants.
mature grain stage (rice) stage occurring during the ripening phase when the inside of the
grain is at first watery but later turns milky in
consistency.
milling yield the estimate of the quantity of
head rice and of total milled rice that can be
produced from a unit of rough rice. It is generally expressed in percentage.
multilocation trial yield trials conducted in
different locations to study the adaptability
of varieties/hybrids over environments.
N
nuclear genes genes located on the chromosomes.
nucleus a small quantity of genetically pure seed
produced under the strict supervision of the
plant breeder.
O
off-type the plants/seeds of the same crop deviating significantly from the characteristics of
the variety/hybrid as described by the breeder.
one-line breeding this uses apomixis as a means
to fix the heterosis of F1 hybrids into truebreeding hybrids and is also known as the
one-line breeding method.
outcrossing rate the extent of cross pollination
measured on the basis of seed set to the total
number of spikelets.
outdoor growth cabinets the small cabinets located outside where temperature and humidity are artificially controlled while light provided is natural.
ovary the bulbous basal portion of the pistils
containing one ovule.
overdominance superiority of the heterozygote
Aa over either homozygote AA or aa.
P
panicle the terminal component of a rice plant
that bears the rice spikelets.
panicle development the growth stage of the
rice plant in which the spikelets become distinguishable and the panicle extends upward
inside the flag-leaf sheath.
panicle exsertion growth stage of the rice plant
marked by the emergence of the panicle from
the boot.
panicle exsertion rate the extent to which the
panicle is exserted out of the flag leaf.
panicle initiation (rice) growth stage that starts
when the primordium of the panicle has differentiated and becomes visible.
partial restorer a pollinator variety used to
pollinate a male sterile line to produce F1 male
fertile progenies, which produce partial seed
set upon selfing.
pedigree the record of the ancestry of an individual or a cultivar.
pedigree nursery a nursery consisting of segregating families in different generations derived from different crosses.
PGMS photoperiod sensitive genic male sterile

line. The genic male sterile plants that respond to the photoperiod or duration of
daylength in terms of pollen fertility and sterility behavior.
phenotypic acceptability breeders shorthand
to record their observations on overall acceptability of breeding lines or populations. This
can be done using an acceptability score of
19. For example, 1 = excellent plant type
and absence of diseases. Promote to the next
level of testing and spread to other breeding
programs. 3 = very good appearance. Promote
to next level of testing. 5 = fair appearance,
but has a few essential shortcomings (too
early maturity, etc.). Use as parent in hybridization block. 7 = poor appearance, but has a
few important traits that make it suitable as a
donor. Make a few crosses. 9 = poor. Discard.
photoperiod duration of daylength.
pistils the female reproductive organ consisting
of the ovary, style, and stigma.
plant growth substances natural and synthetic
compounds that elicit growth and developmental or metabolic responses. These substances are usually not metabolites in the
sense that they are not intermediates or products of the pathways they control, and they
are active at very low concentrations.
planting ratio the ratio in which the male and
female parental lines are planted to make a
crossing block in hybrid seed production or
maintenance of the CMS line.
plumule the leaves of the young plant in any
embryo It is enclosed by the coleoptile.
pollen a mature reproductive male germ cell
(microsporocyte) specialized for fertilization.
pollen fertility/sterility the ratio of fertile/strerile
pollen grains to the total pollen grains
counted in 34 fields under a microscope and
expressed in percentage. Fertility/sterility of
pollen grains is determined by their stainability with 1% IKI stain.
Pollen fertility/sterility gradiation
% sterile pollen
020
2140
4170
7190
9199
100
Category
% fertile pollen
Fully fertile
81100
Fertile
6180
Partially fertile
3160
Partially sterile
1130
Sterile
120
Completely sterile
0
81
pollen load the amount of air-borne pollen per

liter per hour at peak anthesis on a specified
day.
pollen parent male parent of a cross combination.
pollination transfer of pollen from the anther to
the stigma of a flower.
progeny offspring; individuals resulting from a
mating.
pure line a line that has been rendered almost
homozygous by repeated self-pollination over
generations.
purity the composition by weight of the sample
being tested and, by inference, the composition of the seed lot; the identity of various
kinds of seeds and inert matter constituting
the sample.
R
random mating a system in which every individual plant in a population has an equal
chance of becoming pollinated by any other
individual.
randomization allotting treatments to different
plots without any bias.
recurrent selection a method of breeding designed to concentrate favorable genes scattered among several individuals by selecting
in each generation among the progenies produced by random mating of the selected individuals (or their selfed progenies) of the previous generation.
replication repeating the experiment under identical conditions with the objective of reducing the experimental error.
restorer line a pollinator variety is used to pollinate the male sterile line to produce F1 progenies that are male fertile and thus produce
seeds on selfing.
retestcross a cross made beween a cytoplasmatic
male sterile line and a test variety (identified
to be a restorer in the testcross) to recheck the
potentialities of the F1 to give normal seed
set upon selfing.
retestcross nursery breeding nursery to evaluate the retestcross F1s and corresponding male
parents.
ripening phase (syn. maturity phase, grain-filling phase) the period from pollination to
harvest.
82
rogue a variation from the standard of a variety

or strain. Roguing is the removal of undesirable individuals to purify the stock.
row ratio the proportion of seed parents and
pollen parents planted to maintain cytoplasmic male sterile lines or to produce F1 hybrid
seed in a seed production plot.
S
secondary tillers tillers arising from primary
tillers.
second leaf the first differentiated leaf with blade
and sheath.
seed the fertilized and ripened ovule of a seed
plant comprising an embryonic plant accompanied by a store of food (as endosperm or
perisperm), enclosed in a protective seed coat,
and capable under suitable conditions of independent development into a plant.
seed dormancy the ability of mature seeds to
delay their germination after reaching physiological maturity.
seed parent a female parent of a cross combination.
seed viability in general, the state of being alive;
ability of the seed to germinate and produce
normal seedlings.
seedbed the bed on which rice seeds are sown,
consisting of soil (wetbed method) or banana
leaves, plastic sheets, or concrete floor
(dapog method).
seeding sequence the order of seeding the parental lines based on their growth duration so
that they reach flowering at the same time.
seedling (rice) from seed germination to early
tillering; a juvenile plant.
self-fertilization fusion of male and female gametes from the same individual.
source nursery breeding nursery where all the
genetic material, including sources imparting cytoplasmic male sterility, genotypes
with specific traits useful for a hybrid breeding program, and elite rice lines showing high
general and specific combining ability, is
maintained for use in a hybrid breeding program.
spikelet the basic unit of the rice inflorescence
consisting of the two sterile lemmas, the
rachilla, and the floret.
spikelet fertility the number of filled spikelets
to the total number of spikelets on a panicle.
sporophytic in this system, sterility/fertility is

imparted to the pollen by the mother plant
upon which the pollen is borne and the genotype of the pollen has no bearing per se. It
may be controlled by more than one gene with
multiple alleles.
staggered planting planting the restorer line on
different dates to maintain a uniform and regular supply of the pollen to the spikelets of a
cytoplasmic male sterile line that continues
to bloom for a longer period.
stamen the male reproductive organ consisting
of the anther and the filament.
sterile failing to produce or incapable of producing offspring.
stigma the apex of the pistil of a flower, upon
which pollen is deposited at pollination.
stigma exsertion rate the proportion of spikelets with exserted stigma (either on one or on
both sides) to the total number of spikelets in
a panicle.
supplementary pollination a method of shaking the male parent at the time of peak anthesis so as to disperse pollen grains to increase
the seed set on a CMS line. This is particularly necessary when the wind velocity is less
than optimum (23 m sec1).
synchronization (anthesis) refers to the simultaneous opening of the spikelets of the seed
and pollen parents.
synchronization (flowering) refers to the simultaneous flowering of seed and pollen parents
despite having different growth durations.
T
TGMS (thermosensitive genic male sterile) line
the genic male sterile plants that respond to
the temperature in terms of their fertility/sterility behavior.
testcross a cross made between a cytoplasmic
male sterile line and a test variety to identify
maintainers and restorers.
testcross nursery breeding nursery where F1 progenies of cytoplasmic male sterile lines and
test varieties are screened for pollen sterility/
fertility and spikelet fertility to identify
maintainers and restorers.
thermosensitivity sensitivity of a genotype to
varying temperature regimes in terms of pollen or spikelet sterility/fertility.
three-line breeding a breeding strategy to develop hybrids uses three important lines
CMS lines (A lines), maintainer lines (B lines),
and restorer lines (R lines)in a two-step seed
production system as follows: (1) CMS line
multiplication from an A B cross in the field
through natural outcrossing and (2) hybrid
(A R) seed production.
tiller a vegetative branch of the rice plant composed of roots, culm, and leaves, which may
or may not develop a panicle.
tillering growth stage of the rice plant that extends from the appearance of the first tiller
until the maximum number is reached.
topcross a cross between a selection, line, clone,
etc., and common pollen parent is called the
topcross of a tester parent.
two-line breeding breeding methodology in
which only two lines, a male sterile line (photosensitive, thermosensitive, or chemically
induced) and a pollen parent, are used to produce F1 hybrids.
U
uniformity the extent of similarity between the
individuals of a population.
V
variance the mean squared deviation of varieties from their mean.
vegetative phase the period from germination
to panicle initiation.
viability the ability to grow and develop.
vigor the capacity for natural growth and survival, as of seed, plants, or animals.
volunteer plants unwanted plants growing from
the seed (may or may not be the same crop)
that remains in the field from a previous crop.
W
wide compatibility the ability of a genotype to
produce normally fertile progeny when
crossed with both indica and japonica testers.
wide hybridization a process of crossing between distantly related species.
83
Appendix I
Identifying male sterility

When rice plants start heading, male sterility is
identified with the following methods:
1. Visual inspection
At the complete flowering stage, observe the
color and plumpness of the anthers in the
male sterile plants directly with the naked
eye. Shake the panicles slightly to examine
the dehiscence of anthers. Pay attention to
detecting any pore dehiscence that occurs
at the basal part of the anthers. Male sterility in rice plants is expressed by anthers that
are whitish/pale yellow, shriveled, and
nondehiscent.
2. Seed set on bagging
When the plants just start heading but their
florets have yet to flower, cover the panicles
with glassine paper bags to check whether
self seed setting has occurred. In practice,
two panicles are bagged for each plant. After 25 days, observe the seed setting in the
bagged panicles. If no seed is set, the plant
is considered to be completely male sterile.
When a few seeds (i.e., 520%) are produced,
the plant is considered to be partially male
sterile. The spikelet fertility percent is calculated as the number of filled spikelets divided by the total number of spikelets (i.e.,
filled and unfilled spikelets) per panicle on
a per plant basis multiplied by 100.
3. Pollen viability study under the
microscope
Collect five spikelets at the time of flowering and squash the anthers in 12 drops of
1% acetocarmine or 1% IKI (iodine potassium iodide) stain. When the plant just starts
heading but its flowers have yet to flower,

sample the five apical spikelets and immerse
them in a tube with a screw cap containing
fixative solution (alcohol:acetic acid, 3:1)
and store them at room temperature. For
microscopic observation under the light microscope, rinse the spikelets with distilled
water, place the anthers from three or five
spikelets on a slide, and crush them on the
slide with a drop of 1% IKI stain. Observations for each microscopic field must have
more than 60 pollen grains and be averaged
for five different fields. The fertile pollen
grains will be spherical and darkly stained,
whereas the sterile pollen grains will be either unstained and spherical or unstained
and irregular in shape. Pollen fertility percent is calculated as the number of stained
spherical pollen grains with normal shape
and size to the total number of pollen grains
expressed in percentage. Many of the EGMS
lines in the completely sterile phase, that is,
under very high-temperature or long-photoperiod conditions, become pollenless and
only the anther bag remains.
Classifying male sterility according

to the morphology of sterile pollen
1. Typical abortion type
The pollens are irregular in shape; some are
triangular, some are shuttle shaped, etc.
They are unstained with IKI solution. Pollen abortion occurs mainly at the onenucleus stage. So, this type is also called
the uninucleate abortion type. The CMS
lines of the WA type correspond to this type.
85
2. Spherical abortion type

Pollens are spherical and unstainable with
IKI solution. Pollen abortion occurs approximately at the two-nuclei stage. So, this
type is also called the binucleate abortion
type. The Hong-Lian type CMS lines are
representative of this type.
3. Stained abortion type
Pollens are spherical, but partially or lightly
stained with IKI solution. Pollen abortion
occurs mainly at the three-nuclei stage, so
the trinucleate abortion type is its other
name. The boro-type CMS line are included
in this type.
4. No-pollen types
There are no pollen grains, indicating that
abortion occurs even before the pollen
mother cell formation stage. An example is
SA2.
5. Antherless type
Under extreme cases, this has been observed
in certain EGMS lines with no anthers at all,
leaving the female organs functional and
intact.
86
For most EGMS lines, fertility to sterility alteration behavior has been well documented, especially for different pollen types with the gradual
change in environmental conditions. When an
EGMS line is placed during its sensitive phase
under extremely high-temperature (e.g., >35 C)
or long-photoperiod (15 h) conditions, no-pollen
types and typical abortion types are observed.
While under relatively less higher temperature near
its CSP level (e.g., >32 C) or photoperiod >14 h,
the unstained spherical and abortive types are
observed. And with a further lowering of temperature or photoperiod, the fertile stained spherical
pollen types and unstained spherical pollen types
are observed. As the EGMS lines experience the
environment that approaches the CFP level, that
is, <24 C or <13.5 h, the fertile stained spherical
pollen grains are observed in a higher percentage.
This gradual change can be observed with changing environmental conditions within an EGMS
line by collecting regularly the pollen grains upon
flowering from subsequent flowering panicles.
Appendix II
Protocol for anther culture

This section is based on Xiang et al (1993), Li et
al (1994a, 1995), and Pan et al (1993).
1. Young panicles at the meiosis stage as
plant material for in vitro irradiation (0.5
to 3.0 KR). The panicles were pretreated
for 7 d under 610 h of irradiation, followed by incubating anthers on M8 medium with 2 mg L1 2,4-D, 2 mg L1 NAA,
1 mg L1 KT, and 5% sucrose.
2. Select and plate the anthers in N6 media
from the F1 hybrids of TGMS non-TGMS
or PGMS non-PGMS (Chu et al 1975) or
in M8 media (Mei et al 1988).
3. To obtain a higher success rate of useful
dihaploids, a large number of anthers must
be plated in N6 or M8 media for more than
5,000 petriplates, with 100 anthers on each
petriplate (6 cm). Callus initiation is <7%
and the regeneration rate is <20%.
4. The callus obtained must be tried for the
regeneration of plants.
5. Regenerated plants should be hardened in
1/2 MS liquid media followed by activated
charcoal-sterilized soil medium, adopting
standard procedures.
6.
Regenerants should be taken to the field

and raised under special fertility-conducive conditions.
7. Vigorous-looking plants that may be
dihaploids should be selected and their
seeds kept separate. Weak plants should
be treated with colchicine for chromosome
doubling since they may be haploids.
8. Seeds obtained should be sown and raised
under sterility-conducive conditions for
the selection of male sterile lines.
9. Sterile plants should be ratooned and
brought to low temperature or shorter photoperiod for inducing fertility to get self
seeds.
10. TGMS and PGMS lines should be screened
based on the criteria for ideal EGMS lines.
11. Four TGMS lines, 6442S, 1286S, HS-1,
and HS-5, and one PTGMS line, Lu Guang
2S, have been developed through anther
culture and extensively tested in Jiangxi,
Sichuan, and Fujian provinces of China.
87
Appendix III
Data to be recorded for hybrid rice experiments

Breeding nurseries
Characters
Days to flowering (d)

Vegetative vigor (score)
Plant height (cm)
Panicles m2 (no.)
Anther color and shape (ypl;ws)b
Pollen fertility (%)
Panicle exsertion rate (%)
Stigma exsertion rate (%)
Outcrossing rate (%)
No. of filled grains panicle1
Apparent heterosis (scale)
Grain yield (kg ha 1)
Grain type (scale)
Phenotypic acceptability (scale)
Reaction to pests/diseases
Weather data
Source
nursery
EGMS
nursery
Testcross
nursery
!
!
!
!
!
!
!
!
!
!
Evaluation trialsa
Backcross OYT
nursery
!
!
AYT
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
88
!
!
!
!
!
OYT = observational yield trial, PYT = preliminary yield trial, AYT = advanced yield trial, MLT = multilocational trial.
ypl;ws = yellow plumpy; white shriveled.
!= essential, o = optional.
MLT On-farm
!
!
!
!
!
PYT
!
!
!
!
!
!
o

2 Line Hybrid Rice Breeding

Uploaded by

Copyright:

Available Formats

2 Line Hybrid Rice Breeding

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2 Line Hybrid Rice Breeding

Uploaded by

Copyright:

Available Formats

Two-Line Hybrid

INTERNATIONAL RICE RESEARCH INSTITUTE

The International Rice Research Institute (IRRI) was established in 1960 by

Editing: Bill Hardy

APPENDIX III: DATA TO BE RECORDED

using environment-sensitive genic male sterility.

Hybrid rice and heterosis

Heterosis is a phenomenon in which F1 hybrids

How is heterosis measured?

Genetic basis of heterosis

Molecular basis of heterosis

Li et al 2000) except for a few that support the

Approaches for using heterosis

intrasubspecific hybrids yield nearly 15% to 20%

Methods of using heterosis

Why hybrid rice?

How is hybrid rice developed?

Male sterility systems

Male sterility can be defined as a condition in

Cytoplasmic genetic male sterility

Hybrid seed production using

Fig. 2. Classification of male sterility systems in rice.

Fig. 3. Schematic description of the cytoplasmic genIc

Environment-sensitive genic male sterility

Martin and Crawford (1951)

Maize, barley, oats,

Rerkasem and Jamjod (1997)

Spontaneous mutation, China

Induced (R), China

Shi and Deng (1986)

Any fertile line can be used as a pollen par-

Since there is no need for restorer genes in

ent (PP); therefore, the frequency of heterotic

the male parents of two-line hybrids, this

Disadvantages of the EGMS system

Table 3. Two-line rice hybrids released up to 2001 in China.

for example, <27 C (maximum) or <13 h

Seed setting (%)

Seed setting (%)

Fig. 4. Fertility of PGMS line N9044S in phytotron

Fig. 7. Low fertility of PTGMS line Miai 64S in phytotron

Seed setting (%)

TGMS lines are sensitive to the temperature

Fig. 5. Fertility of TGMS line W9046S in phytotron

Seed setting (%)

Fig. 6. High fertility of PTGMS line Xinguang S in

vis--vis a shorter photoperiod (i.e., 14 h at 30 C

Table 4. Photoperiod and temperature conditions in

Illumination time: 12.5 h (06001830), 14.0 h (05301930), 15.0 h

the same time, panicles on primary and secondary

Table 5. Results of two-factor variance analysis.

Table 6. Classification of EGMS lines based on statistical

* = significant at the 5% level and ns = nonsignificant at the 5% level.

bination of short daylength with low temperature

high-altitude areas and/or in a season when the

Chemically induced male sterility

Male sterility induction

Chen et al (1986), Ali (1990, 1993)

Other chemicals (with and without codes)

have been tested along with 3-line bred hybrids