Tải bản đầy đủ - 0trang
IV. Developing Varieties with Multiple Resistance
DISEASE AND INSECT RESISTANCE IN RICE
1969 7 0 71 72 7 3 74
71 72 73 74
FIG. 4. Change in the proportion of F, populations and entries in the replicated yield
trials at IRRI with resistance to important insects and diseases (Khush, 1977). Reproduced
by permission of the New York Academy of Sciences.
60% in the wet season of 1974. Similarly, only 17% of the entries in replicated
yield trials of 1969 wet season were resistant to bacterial blight. The proportion
increased to 98% in the 1974 wet season (Fig. 4). Progress with tungro, grassy
stunt, brown plant hopper, and green leafhopper has been equally dramatic
The main thrust of the program, of course, has been to combine resistance to
all those diseases and insects with improved plant type. Equally rapid progress
has been made in this direction. About 87% of the entries in the replicated yield
trials of the 1969 wet season were either susceptible to all six diseases and
insects (blast, bacterial blight, tungro, grassy stunt, brown plant hopper, and
green leafhopper) or resistant to only one of them (Fig. 5). Only 2% of the
entries were resistant to three diseases and insects. The proportion of entries
with multiple resistance gradually increased, and in the 1974 trials, 90% were
resistant either to five diseases and insects or to all six. Six of the multiple
GURDEV S. KHUSH
Resistonce to a wrnber of d i m s and insects (%)
FIG. 5. Change in proportion of entries in annual replicated yield trials with multiple
resistance to important diseases and insects (blast, bacterial blight, tungro, grassy stunt,
brown plant hopper, and green leafhopper) at IRRI. Each year’s trial consisted of about
185 entries (Khush, 1977). Reproduced by permission of the New York Academy of
resistant lines were named varieties and recommended for cultivation in several
countries. Table XVII shows the disease and insect reactions of all IRRI named
varieties and indicates the progressive increase in the levels of resistance of the
newer rice varieties. The breakthrough in developing the improved plant-type
germ plasm with multiple resistance to major diseases and insects was achieved
through a well-planned breeding program and through a liberal exchange of ideas
and materials between IRRI scientists and rice scientists in other rice-growing
countries. Salient features of the breeding methodology and procedures employed and international cooperation are discussed in the following section.
A. BREEDING METHODS AND PROCEDURES
I . Choice of Parents
A major strength of the breeding program at IRRI has been the well-stocked
germ plasm bank. From 256 accessions in 1961, the germ plasm collection
DISEASE AND INSECT RESISTANCE IN RICE
increased to 6900 entries a year later, and to 23,560 entries in 1972 (Chang et
al., 1975). It now has about 40,000 accessions from 73 countries of the world.
Another major strength has been the presence of highly competent pathologists
and entomologists on the Institute staff. As soon as a serious disease or insect
problem was identified, scientists started developing screening techniques and
evaluating germ plasm for resistance. The identified sources of resistance were
immediately introduced into the crossing program. The donor parents generally
had poor plant type characterized by tall stature, droppy leaves, weak stems, and
consequently, low-yield potential. The first step therefore was to transfer those
sources of resistance to an improved plant type (short stature, erect leaves,
sturdy stems, high tillering). This “conversion” was achieved by crossing the
donor parents with an improved plant-type parent, growing large Fz populations
(2000-5000 plants), and selecting improved plant-type segregates. The selected
plants were examined for grain quality, and F3 progenies were grown from those
with good grain quality. The F3 progenies were evaluated for the resistance trait
under study, and several resistant selections (generally up to 100) with good
grain quality, improved plant type, and appropriate growth duration were saved
for further evaluation in the F4 and F5 generations. Through repeated evaluations, a few (2-10) true-breeding, improved plant-type selections with resistance
to given traits were selected. Between 1965 and 1969, IR8, several selections of
IR262 (IR262-43-8 in particular), and IR24 were used as improved plant-type
parents. Several donor parents were used for each disease and insect.
2. Crossing Program
During the period when emphasis was on germ plasm conversion, single crosses
were made. Poor progenies were obtained in a multiple cross if more than one
parent had poor plant type. Since very few improved plant-type parents were
available in the initial stages, only single crosses were feasible. About 200 to 400
such crosses were made each year. By the end of 1970 a large number of
improved plant-type breeding lines with resistance to one or two diseases and
insects became available. To combine resistance to all the major diseases and
insects together, the crossing program was expanded and a large number of
multiple crosses were made employing those breeding lines. A large number of
single crosses between those lines were made each season. The following season
either the two F l s were crossed with each other or an Ft was crossed with a
third breeding line. A fairly large number of F, seeds (300-400) were obtained
from multiple crosses. This allowed the gametic variability of single-cross F, s to
In producing single-cross F, s, each breeding line was crossed with a number of
other breeding lines. Thus, a set of single-cross F1 progenies were available for
making double crosses or topcrosses in the next season, and appropriate combinations could be selected t o combine the resistance to given diseases and insects.
GURDEV S. KHUSH
That also allowed the rapid determination of the combining ability of the
breeding lines. A breeding line that yielded poor progenies in a number of cross
combinations was assumed to be a poor combiner and was removed from the
3. Handling Segregating Populations
The pedigree method of breeding was employed almost exclusively in handling
the segregating materials. Selection work was based on comprehensive records on
the disease and insect reactions of each line and, in the case of F4 and advanced
generation lines, on the reaction of the ancestral lines as well. The bulk method
of breeding was not used because it does not permit concurrent screening for a
number of diseases and insects. The backcross method was not used for lack of
suitable recurrent parents. A few backcrosses were made in the crosses with
Oryza nivara for the program on resistance to grassy stunt. IR8, IR20, and IR24
were used as recurrent parents. After 3 to 4 backcrosses, we obtained breeding
lines similar to IR8, IR20, or IR24, but they lacked resistance to other important diseases and insects, such as tungro and brown plant hopper, and were again
entered in the crossing program. Now varieties and breeding lines with multiple
resistance are available, and we are using the backcrossing program to incorporate resistance to white-backed plant hopper.
For traits under polygenic control, the pedigree method is not as suitable. At
IRRI, the diallele selective mating system, originally suggested by Jensen (1970),
is being tried for combining minor genes for resistance to stem borers and whorl
maggot from several sources. This method involves: (1) crossing a number of
moderately resistant parents (generally 5-6) in all possible combinations, (2) intercrossing the F1s so obtained in all possible combinations, ( 3 ) screening the
double-cross F1 progenies for resistance, and (4) intercrossing the selected plants
that have better resistance than either parent. This crossing, screening, and
selection process is continued until the minor genes from different sources are
accumulated and the intensity of the trait is built up. We are in the third cycle of
this type of recurrent selection program for stem borer resistance and the second
cycle of selection for the whorl maggot program. The success of the method is
difficult to prognosticate at this stage.
4. Screening Segregating Populations
The success of the disease- and insect-resistance breeding program depends to a
large measure upon the fidelity, speed, and practicability of the screening
technique. Various greenhouse and field screening methods employed at iRRI
have been tailored to accommodate large volumes of breeding materials. About
50,000 pedigree rows are grown each year at IRRI. Most are screened for
DISEASE AND INSECT RESISTANCE IN RICE
reaction to all the major diseases and insects. If possible, the breeding materials
are exposed t o artificially created or naturally occurring disease and insect
epiphytotics. Previously, breeding nurseries were grown with full insecticide
protection; however, since 1970 most of the nurseries are grown without
insecticide treatment to allow a buildup of huge populations of plant hopper,
leafhopper, and stem borers. Rant hoppers and leafhoppers in large numbers also
insure the spread of virus diseases in the nurseries. Sometimes artificially virusinoculated plants are planted around the borders of nurseries t o provide a source
of inoculum. The insect populations are manipulated by applying selective
insecticides. At IRRI farm, Diazinon has no toxic effect on the brown plant
hopper but it kills all predators and other natural enemies of t h s insect. By
judicious application of Diazinon, an outbreak of brown plant hopper has been
induced in the IRRI nurseries. This has also led to an increased incidence of
All nurseries are artificially inoculated with bacterial blight in the field and
tested for reaction to blast in the blast nurseries. Data on green leafhopper and
brown plant hopper reactions are also obtained from greenhouse tests. Selected
materials are planted at other locations in the Philippines under disease and
insect pressures different from those at IRRI.
Every effort is made to eliminate the susceptible materials in the early
generations. Screening begins in the F1 generation of multiple crosses. For
example, consider a double cross between four parents; A is resistant to bacterial
blight, B is resistant t o grassy stunt, C is resistant to brown plant hopper, and D
is resistant to green leafhopper. All these traits are controlled by single dominant
genes, whch segregate independently of each other. About 400 seeds from the
double cross A/B//C/D are obtained. The most logical system for screening the
progenies would be to germinate and inoculate all 400 seedlings with grassy
stunt in the greenhouse. Half of the seedlings would be susceptible and will be
eliminated. The remaining 200 would be transplanted in the field and inoculated
with bacterial blight; half of the 200 that would be susceptible would be rogued
out. Seeds from the remaining 100 plants would be harvested individually. Two
small seed samples would be taken out from each and the progeny tested for
resistance to brown plant hopper and green leafhopper. Those carrying the
brown plant hopper resistance gene (50%) and the ones carrying the green
leat3opper resistance gene (50%)would be identified. F2 populations would be
grown only from those carrying both genes (25-30 plants). Thus, by judicious
and timely screening, the original F, sample of 400 would be reduced to 25 to
30 plants, and F2 populations would be grown from these plants. All these F2
populations would be segregating for the four resistance genes. They could be
subjected to appropriate disease and insect pressures. Agronomically desirable
plants with multiple resistance would be selected and rescreened in the F, and
F4 generations to obtain true breeding lines.
GURDEV S. KHUSN
A large number of multiple crosses between breeding lines of improved plant
type, known combining ability, and resistance to a number of diseases and
insects are made each season and screened according to the outlined procedure.
New parents with different sources of resistance are constantly included in the
crossing program. This integrated varietal development program has resulted in
superior germ plasm with resistance to all major diseases and insects. Newer lines
with different genes and gene combinations for resistance should continue to
come from the program.
B. INTERNATIONAL COOPERATION
International and interdisciplinary cooperation has been the key ingredient of
the varietal development program at IRRI. Liberal exchange of ideas and
materials between different programs, cooperative testing for disease and insect
resistance in the Philippines at the Bureau of Plant Industry Stations, and in
several other countries such as India, Sri Lanka, Bangladesh, Thailand, and
Indonesia has contributed greatly to the development of germ plasm that is
resistant to diseases and insects.
The pedigree of IR28, IR29,and IR34 (Fig. 6 ) illustrates this international and
interdisciplinary approach. To develop these high-yielding, good grain quality,
and multiple disease- and insect-resistant varieties, eight varieties from six different countries were used in the crossing program. The seeds of these varieties and
40,000 others were supplied by scientists from those countries. This germ plasm
was evaluated by pathologists and entomologists for disease and insect resis-
BL T GLH
EL BEGS BPH QLH
BL EBT GS BPH Wli
FIG. 6. Pedigree of IR28, IR29, and IR34. The progress in combining together the
resistance to six major diseases and insects from several parents is indicated. BL = Blast
disease; BB = bacterial blight disease; T = tungro virus disease; GS = grassy stunt virus
disease; BPH = brown plant hopper; and GLH = green leafhopper (Khush, 1977). Repre
duced by permission of the New York Academy of Sciences.
DISEASE A N D INSECT RESISTANCE IN RICE
tance, respectively. The breeders combined the identified sources of resistance to
diseases and insects with improved plant type. The segregating populations were
tested at IRRI and the Philippine Bureau of Plant Industry Station at Maligaya,
and in Indonesia.
The seeds of the improved germ plasm as well as of the entries in the germ
plasm bank are shared with scientists all over the world. Up to the end of 1975
more than 95,000 seed samples of breeding lines were supplied to requesting
parties in 80 countries of the world. The breeding lines are used as parents in the
crossing programs, and some have become named varieties. To date 40 breeding
lines from IRRI have been named varieties in other countries. The recently
expanded international testing program will facilitate the exchange and dissemination of germ plasm between the various rice improvement programs.
V. Stability of Resistance
There is growing support for the contention that the resistance governed by
polygenes-also referred to as general resistance or horizontal resistance-is more
lasting than resistance governed by major genes (specific or vertical resistance).
When the program on breeding for disease and insect resistance in rice was
initiated at IRRI, little was known about the genetics of resistance. Available
donor parents were used as sources of resistance and we developed the improved
plant-type breeding lines and varieties with multiple resistance to as many as
four diseases and four insects (Table XVII) within a short period of 7 to 8 years.
Disease and Insect Resistance Reactions of IRRI Named Varieties
Disease and insect reaction'
'S = Susceptible; MS = moderately susceptible; MR = moderately resistant; R = resistant.
GURDEV S . KHUSH
During this period at IRRI, by investigating the mode of inheritance of resistance to some diseases and insects, it was found that resistance to most diseases
and insects with the exception of stem borer, is controlled by the major gene.
A. VERTICAL RESISTANCE
Information on the stability of vertical or major gene resistance in rice is
meager. As discussed earlier, bacterial blight-resistant IR20 and IR26, which
have Xa4, have been widely grown in the tropics. Their resistance has held up
except in a small area of the Philippines where a strain of bacterium that is
moderately virulent to Xa4 has appeared. This strain has remained localized and
causes only slight damage to rice varieties with Xa4. Several bacterial blight-resistant varieties, such as Benong, Sigadis, Syntha, and Dewi Tara, have grown in
Indonesia for 10 t o 20 years. The resistant TKM6, MTU15, and CO 21 have
grown in India for many years. The occurrence of bacterial strains virulent to
varieties with host resistance has not been reported.
Several varieties resistant to green leafhoppers-Peta, Intan, and Bengawanwere widely grown in Indonesia and the Philippines for 30 to 35 years. Several
improved-plant-type varieties-IR5, IR8, IR20, IR26, and C4-63, which inherited GZh3 for resistance to green leafhopper from Peta-have also been grown for
several years. No clear-cut evidence for the origin of green leafhopper biotypes
that are virulent to GZh3, under the influence of host resistance has been found.
However, varieties resistant t o brown plant hopper became susceptible within
1.5 years of their introduction into the British Solomon Islands, because of the
appearance of a new brown plant hopper biotype. Similarly, within 2 years of its
large-scale cultivation in the Philippines, IR26 was attacked by new biotypes of
the insect in several localities. The germ plasm for resistance to brown plant
hopper comes from South India and Sri Lanka where biotypes of the insect are
virulent t o those varieties. These biotypes probably originated under the influence of host resistance.
The influence of host resistance on the insect populations of brown plant
hopper and green leafhopper is obviously different. The difference may be due
to the differential selection pressure exerted by the resistant varieties on insect
populations. The level of resistance to brown plant hopper conveyed by Bphl
and bphZ is sufficiently high that the insect cannot perpetuate itself on resistant
varieties. It either changes or is eliminated. On the other hand, the level of
resistance to green leafhopper conditioned by GZh3 is only moderate. The insect
feeds on resistant plants and reproduces, although at a much lower rate than it
can when feeding on susceptible varieties. Thus, it can perpetuate itself on
resistant varieties but chances for the origin of more virulent biotypes are
considerably lower than those for the brown plant hopper. Thus, the useful life
DISEASE AND INSECT RESISTANCE IN RICE
of the Glh3 gene may be considerably longer than that of either Bphl or Bph2.
The other genes for resistance to the green leafhopper-Glhl and Glh2-convey
higher levels of resistance comparable with those of Bphl for brown plant
hopper. When varieties having either gene are grown widely, new biotypes of the
green leafhopper might arise rapidly. Differences in the inherent capacity of the
two insect species to change under the influence of host resistance may also be
responsible for the differences in longevity of resistance to the two insects.
The strategy at IRRI to utilize major gene resistance in rice is twofold. The
short-term strategy aims at incorporating the known major genes for resistance
to different diseases and insects into the improved plant-type background,
combining these genes in different combinations, and sharing the resulting germ
plasm with other programs. IRRI is close to meeting this goal. The long-term
objective is t o identify several genes for resistance to each disease and insect,
particularly bacterial blight and the brown plant hopper. As soon as a new gene
is identified, it is transferred to an improved plant-type background.
When a number of genes become available it would be possible to adopt any of
the following approaches to utilize these genes for vertical resistance:
1. Release one gene for resistance and wait until it becomes ineffective; release
the second gene, and so on. This approach was adopted to control stem rust of
wheat in Australia between 1938 and 1950 (Watson and Luig, 1963). This
approach is being taken with respect t o brown plant hopper resistance. IR26 was
released during the brown plant hopper outbreak of 1973 in the Philippines.
This variety and IRI 561-228-3, another brown plant-hopper-resistant selection,
were grown widely in the Philippines in 1974 and 1975. Both cultivars have
Bphl for resistance. Toward the end of 1975 and in 1976, hopperburn on these
varieties was reported in two locations in the Philippines. IR36 and IR38 which
have bph2 for resistance to brown plant hopper, were hastily released by the
Philippine Government in March 1976. IR36 and IR38 are expected to hold for
a couple of years. By that time varieties with Bph3 and bph4 would be available.
2. Pyramid two, three, or even more major genes together in the same variety,
as suggested by Watson and Singh (1952). Several wheat varieties that combined
up to five genes for resistance to stem rust were developed.
Canadian breeders have adopted the same procedure for developing oats that
are resistant to crown rust (Knott, 1974). Several scientists, most notably Nelson
(1972), favor this approach. This approach depends upon the existence and
availability of several races or biotypes capable of distinguishing between genotypes with various numbers of resistance genes. Otherwise the breeding procedure becomes too lengthy. The availability of several biotypes of brown plant
hopper makes it feasible to pyramid genes for resistance to this insect.
3. Develop multiline cultivars, as proposed by Jensen (1952) and Borlaug
(1958). This approach was followed for crown rust resistance in oats in Iowa
(Browning et al., 1969), and a program to develop multiline stem-rust-resistant
GURDEV S. KHUSH
cultivars of wheat is under way at Centro Internacional de Mejoramiento de Maiz
y Trigo. The development of multilines involves an extensive program of gene
identification and backcrossing. As suggested in an earlier section, this approach
merits serious consideration for an international project on blast.
4. Develop resistant varieties with different resistance genes and recommend
them for different geographical regions of the country where the crop covers a
sizable area. As pointed out by Nelson (1972), this type of gene deployment is
essentially a geographical multiline. A formal plan for regional deployment of
genes is in effect for resistance t o crown rust in oats in Iowa (Frey et al., 1973).
This approach may be followed for either rice diseases or insects when enough
genes are identified.
B. HORIZONTAL RESISTANCE
At IRRI the search for horizontal or polygenic resistance in rice continues.
The Institute program on stem borer resistance deals with polygenic systems.
Polygenic variation has been noted for the resistance to brown plant hopper,
grassy stunt, and bacterial blight. However, there are practical difficulties in
exploiting this variation. One concerns the breeding system of the crop. In an
outcrossing species a number of cultivars, with minor genes that are desirable for
accumulation, can be mixed, planted, and allowed to interbreed for several
generations. Appropriate disease or insect pressure is applied to each generation,
and individuals with higher levels of resistance are selected for growing the next
Random mating permits the formation of new gene combinations at each
generation, and recurrent selection changes the gene frequency for higher levels
of resistance. The process cannot be followed with rice because of its self-pollinating nature. However, a usable source of male sterility that can be employed
for inducing high rates of outcrossing in a composite population with several
sources of polygenic resistance is being sought. Pending the availability of a male
sterile, the diallele selective mating system discussed in an earlier section is being
The second difficulty concerns the screening techniques. Most artificial screening techniques fail to detect polygenic differences. During the brown plant
hopper outbreak of 1973 at the IRRI farm, several selections with tolerance to
the insect were identified. They withstood the insect attack longer than the
susceptible varieties did, but were eventually killed. However, when they were
tested in the greenhouse, these differences could not be detected.
At IRRI, a breeding program on horizontal resistance to brown plant hopper
was initiated, using these sources and following the diallele selective mating
system. When the F1 progenies from the double crosses were ready for testing,
there were no brown plant hoppers in the field.
DISEASE AND INSECT RESISTANCE IN RICE
It is important to develop horizontal resistance to the brown plant hopper, and
efforts are being made at IRRI to screen the breeding population in other
countries, such as the British Solomon Islands, where field populations of the
brown plant hopper are always high.
Among cereal crops, rice is the host of the largest number of diseases and
insect pests. These cause serious yield losses annually.
The magnitude of losses caused by the diseases and insects is likely to increase
as the level of rice production per unit area increases.
The germ plasm resources for disease and insect resistance are vast, but only a
portion has been collected and evaluated for resistance.
Much germ plasm remains to be collected and catalogued. It should be
collected before it becomes extinct through the adoption of high-yielding
The germ plasm that has not been evaluated, especially the national germ
plasm collections, should be evaluated to identify more sources of resistance.
Different sources of resistance should be genetically analyzed to identify
diverse genes for resistance.
A systematic international survey of races or biotypes of major diseases and
insects should be carried out with the use of differential varieties.
Sources of resistance to all races or biotypes should be identified and genetically analyzed.
The different major genes for resistance should be utilized according to needs
of each program. Various alternatives are discussed.
Greater efforts should be expended in studying and utilizing horizontal resistance, although vertical resistance will continue to be useful for years to come.
Interdisciplinary cooperation among the pathologists, entomologists, and
breeders is essential for the speedy implementation of host resistance progcams.
International cooperation is essential in collecting and evaluating germ plasm,
studying the races and biotypes, identifying the diverse genes for resistance,
cooperative testing for disease and insect resistance, and liberal exchange of
improved germ plasm.
Abeygunawardena, D. V. W. 1967. Proc. Symp. Rice Dis. Their Control Growing Resistant
Varieties Other Measures pp. 171-179. Agric., For. Fish. Res. Counc., Tokyo.
Abeygunawardena, D. V. W., Bandaranayaka, C. M., and Karandawela, C. B. 1970. Trop.
A@. (Ceylonj 126, 1-13.