Tải bản đầy đủ - 0trang
X. Hybridization and Exploitation of Hybrid Vigor
PREM l? JAUHAR AND WAYNE W. H A N N A
that time that a commercial method for producing 100% hybrid seed was needed,
because the varying amounts of selfed seed produced by the chance method did
not allow maximum expression of hybrid vigor at the low seeding rate for a commercially planted grain hybrid. However, Burton (1948, 1989) showed that up to
50% selfed plants in forage chance hybrids would not decrease forage yields at
recommended seeding rates, which are higher than those for grain hybrids.
Anand Kumar and Andrews (1984) found that research in the 1950s demonstrated the large yield increases possible with F, hybrids and that a crns system
was needed to produce hybrids on a commercial scale. Tift 23A, a crns inbred, was
made available to Indian pearl millet breeders in 1962 (Burton, 1965).Indian pearl
millet breeders pollinated Tift 23A with Bil-3B, an Indian inbred, to produce HB 1,
the first released pearl millet single-cross grain hybrid using the crns system. Hybrids using Tift 23A and Tift 18A as female parents and Indian inbreds as pollinators averaged 102% more grain production than the best available varietal checks
in India from 1964 to 1967 (Rachie and Majmudar, 1980). Hybrids such as HB 1,
using Tift 23A as the seed parent, eventually became susceptible to downy mildew
(Sclerospora graminicola Sacc. Schroet.) and ergot (Clavicepsfusiformis Loveless). This initiated a concentrated effort to develop inbreds resistant to these diseases for production of resistant hybrids. The research is ongoing today. Scientists
at ICRISAT (International Crops Research Institute for the Semi-And Tropics),
India, have been exploring new sources of cytoplasmic male sterility for hybrid
production (Sujata er al., 1994; Rai, 1995).
The first release in India of a top-cross hybrid was announced in 1996 by government authorities in Madhya Pradesh. The hybrid named “Jawahar Bajra Hybrid
1 (JBHl)” has high grain-yield potential, medium height, nonbristled compact
ears, and medium bold, globular grains. Both the hybrid and its top-cross pollinator are highly resistant to downy mildew. Similarly, Gujarat State Fertilizers Company Limited has developed a hybrid “Sardar Hybrid Bajra (SHB I),” which has
about 20% more yield, has better quality grain, and matures earlier than the existing hybrids (SATNews, 19961997).
Interest in producing pearl millet for grain in the United States and Australia has
increased. HGM 100 was the first commercial grain hybrid released in the United
States in the early 1990s (Hanna el al., 1993). The area planted to the crop was increasing in the southeastern United States until a new race of rust attacked the crop
in late plantings. Pearl millet’s high-quality grain, drought resistance, and flexibility in rotation and multiple cropping systems have caused interest in it as a grain
crop outside its traditional growing areas.
Gahi 1, the first commercial pearl millet forage hybrid-produced by harvesting
all the seed from a field planted to a mixture of four inbreds that flowered at the same
CYTOGENETICS AND GENETICS OF PEARL MILLET
time and gave high-yielding hybrids in all combinations-yielded 52% more than
Common and 35% more than Stan: Gahi 3 replaced Gahi 1 and was the first singlecross forage hybrid produced using crns (Burton, 1983).Subsequentsingle-crosshybrids, such as Tifleaf 1, and Tifleaf 2, and a three-way hybrid, Tifleaf 3, have increased animal gains because of improved forage yields, leafiness, quality, andor
disease resistance (Burton, 1983; Hanna et al., 1988; Hanna et al., 1997).
Over 20,000 accessions of cultivated pearl millet and its wild relatives are stored
in India and the United States. These accessions include landraces, improved
populations and breeding lines, and wild relatives from the primary, secondary,
and tertiary gene pools that are available to plant breeders.
Most germplasm is in the primary gene pool. Objectives need to be clearly defined to effectively select and use the best germplasm. Principal component and
cluster analyses can be used to help identify the genetic and phenotypic diversity
needed in a breeding and improvement program (Wilson et al., 1991). Weedy relatives in the primary gene pool (Hanna et al., 1988; Hanna, 1989) and wild relatives in the secondary (Hanna, 1990) and tertiary gene pools (Hanna et al., 1993)
are also potential sources of valuable genes (Hanna, 1987).
Hybrids usually out-yield open-pollinated cultivars (Andrews, 1987; Burton,
1983). However, since all cross combinations may not always produce superior
hybrids, inbreds with good general combining ability (GCA) and/or specific combining ability (SCA) need to be identified (Anand Kumar et al., 1992). Hybrids
maximize yields and can be most easily made using crns in the seed parent (Anand
Kumar and Andrews, 1984),especially if pollen-fertility restorer genes are present
in the pollinator of hybrids grown for grain. Lack of complete male fertility
restoration can result in poor grain yields and a higher incidence of smut and ergot diseases. Restorer genes are not needed (and probably undesirable) in pollinators of forage hybrids.
Most pearl millet hybrids are single crosses. A single cross between two elite inbreds with high SCA is probably the best way to maximize yield. In addition to using crns in one inbred to produce single-cross F, hybrids, single-cross hybrids can
also be made between two elite male fertile inbreds by taking advantage of naturally occumng protogyny in pearl millet. Protogyny can be used to make hybrids
in at least two ways: (1) equal quantities of seed of two or more inbreds, equal in
height and maturity, can be mixed, planted, and allowed to interpollinate; and (2)
elite male fertile inbreds can be planted in adjacent rows and seed harvested from
PREM P. JAUHAR AND WAYNE W. HANNA
only one inbred.The inbred from which seed is harvested should flower 3 or 4 days
earlier than the other inbred used to produce the hybrid. The use of protogyny to
produce hybrids will result in some selfed and sibbed seed. The effects of selfed
and sibbed seed can be overcome to some extent in the hybrid production field by
increasing the seeding rate to crowd out the weaker plants. Seed from selfing and
sibbing in grain hybrids may be more objectionable, especially when the hybrid
grain is mechanically harvested.
Seed yields can be increased in the hybrid seed production fields by producing
three-way hybrids. Two inbreds are used to produce a cms F, hybrid, which is used
as the seed parent and pollinated by a third inbred in hybrid production fields. The
commercial forage hybrid Tifleaf 3 is produced by pollinating crns F, Tift 8593
(Hanna, 1997) with inbred Tift 383 (Hanna et al., 1997). Twice as much hybrid
seed is produced on Tift 8593 as on crns inbred Tift 85D,A,, the seed parent of
Tifleaf 2. Forage yields of Tifleaf 2 and Tifleaf 3 are similar.
Inbred (crns or male fertile) X landrace hybrids may not maximize hybrid vigor but should increase yields and provide more genetic diversity in a hybrid population. These hybrids would maintain some of the agronomic characteristics of
landraces preferred by farmers and provide more genetic diversity for diverse environmental growing conditions. Mean grain yields of crns inbred X open-pollinated variety crosses have been equal to or superior to the open-pollinated variety
(Mahalskshmi et al., 1992).
Landrace X landrace crosses seem to have the most potential for improving
yield and reliability in harsh, variable climates. Ouendeba er al. (1993) showed
that the better-parent heterosis for hybrids among five West African landraces
ranged from 25 to 81% for grain yield.
Apomixis is a reproductive mechanism that bypasses the sexual process and allows a plant to clone itself through seed. In Pennisetum, a chromosomally unreduced egg cell develops into an embryo in an embryo sac derived from a vegetative nucellar cell. This type of apomixis is called apospory. In addition to the egg
cell developing into an embryo without fertilization by a sperm, pseudogamy or
fertilization of the central cell is needed for endosperm and seed development.
Apospory is the only type of apomixis confirmed in Pennisetum.
OF h o r n s m Pennisetum SPECIES
Apomixis is relatively common in the polyploid species of Pennisetum, especially those in the tertiary gene pool. Apomixis has been reported in polyploids
CYTOGENETICS AND GENETICS OF PEARL MILLET
(triploid and higher) of both the x = 8 and x = 9 chromosome groups. Only x =
7 chromosome species have been reported in the primary and secondary gene
pools, and all are sexual. Likewise, tertiary gene pool species with the x = 5 and
x = 7 chromosome groups and diploids with x = 8 or x = 9 have been reported
to be sexual. Jauhar (1981a) listed at least nine species that have been reported to
reproduce by apomixis. Additionally, F! squamulatum, F! polystachyon, and t!
macrourum have been reported to be apomictic (Dujardin and Hanna, 1984).
Apomixis may have played a role in building and maintaining new genome
combinations in Pennisetum. Hanna and Dujardin (1991) summarized some of
their research, which showed how apomixis was used in crosses among two sexual and three apomictic species in the x = 7 and x = 9 chromosome groups from
the primary, secondary, and tertiary gene pools to develop and maintain more than
20 new chromosome and/or genome combinations. These were developed from
sexual X apomictic crosses, parthenogenesis of a reduced gametophyte, and fertilization of an unreduced egg. Hussey er al. (1 993) and Bashaw ef al. (1 992)
showed that facultative apomictic F! fiaccidum hybridized with Cenchrus
setigerus, P. massaicum, F! mezianum, and P. orientale, as n + n and/or 2n + n
hybridizations, produced new genome combinations.
B. GENETICSOF APOMIXIS
The genetics of apomixis is difficult to study because sexual and apomictic
counterparts are usually not available within the same species. Therefore, crosses
need to be made between sexual and apomictic plants from different species. Genetic studies on apomixis are made more complex by facultative apomixis, lack of
F, segregatingpopulations, and the limitation of having to use the apomictic plant
as pollen parent in crosses.
Asker and Jerling (1992) summarized the current status of the genetics of
apomixis. Most researchers agree that it is probably under relatively simple genetic control. Both dominant and recessive gene actions have been reported. Crosses between sexual and apomictic Penniserum species indicate a major dominant
gene and some modifiers (Hanna et al., 1993).
APOMIXIS FOR EXPLOITATION
Apomixis has tremendous potential for revolutionizingfood, feed, and fiber production around the world because it makes possible true-breeding hybrids through
seeds. Apomixis not only would fix hybrid vigor but also could make possible
commercial hybrids in seed-propagated crops lacking an effective male-sterility
system for producing hybrids. The opportunities apomixis offers for developing
PREM P. JAUHAR AND WAYNE W. H A N N A
superior hybrids and simplifying hybrid production have been previously discussed (Hanna and Bashaw, 1987; Hanna, 1995).
Probably more progress has been made in transferring the apomictic mechanism
from wild I! squamulatum to cultivated pearl millet than in any other grain crop.
The mechanism has been transferred to the BC, generation where high levels of
apomixis have been maintained (Hanna et al., 1993; and unpublished data). However, a problem encountered has been the loss of 80-90% of the seed set postanthesis. Efforts are under way to transfer apomixis from Tripsacum dactyloides
(L.) L. to maize (Savidan er al., 1993; Kindiger et al., 1996) and from Elymus rectisetus (Nees in Lehm.) to wheat (Carman and Wang, 1992).
The greatest impact of apomixis may be realized by cloning and inserting the
gene(s) controlling apomictic reproduction into various sexual species by molecular methods. To be useful, a transferred gene must express itself and be stable in
an alien genome. The gene(s) controlling apomixis needs to be mapped before it
can be cloned and used in other species. Molecular markers linked to apomixis are
being developed in Pennisetum (Ozias-Akins et al., 1993; Lubbers et al., 1994).
XII. GENETICS OF QUALITATIVE TRAITS
Numerous qualitative traits have been reported for pearl millet. Comprehensive
reviews on the genetics of qualitative traits in pearl millet have listed at least 145
mutants (Koduru and Krishna Rao, 1983; Anand Kumar and Andrews, 1993).
These consisted of chlorophyll deficiencies (26%), plant pigmentation (1 8%), earhead characters (14%), pubescence and plant form (each 7%), seed characters and
reproductive behavior (each 6%), foliage striping and sterility (each 4%), leaf
characters and disease resistance (each 3%), and earliness (1%) (Anand Kumar
and Andrews, 1993). Other mutants have been described and not included in the
preceding reviews. Some of these include a naked flower mutant (Desai, 1959) and
a “spreading” mutant (Goyal, 1962). Most mutants are controlled by one or two
loci and dominant or recessive gene action.
Recently described qualitative characters include phylloid (Wilson, 1996), narrow leaf (Appa Rao et al., 1995), brown midrib (Gupta, 1995), and xantha terminalis (Appa Rao et al., 1992) mutants controlled by the phm phm, In In, bm, bm,,
and xt xt genes, respectively. Hanna and Burton (1992) showed that two plant-color mutants, red (Rp,)and purple (Rp,), are allelic; and RpI is dominant over Rp2
and normal green, whereas Rp, is dominant over normal green. Uma Devi et al.
( 1996) observed linkage of semidwarf phenotype to interchange homozygosity.
Most of the mutants have potential for mapping and various genetic and physiological studies. Some appear to have direct application in commercial cultivars.
Dwarf genes, especially the d , locus, has been widely used to produce high qual-
CYTOGENETICSAND GENETICS OF PEARL MILLET
ity shorter forage hybrids and dwarf grain hybrids that can be mechanically harvested. The early genes have been effectively used to produce early grain hybrids.
Forage quality could be rapidly increased with the brown midrib bm,gene, which
can reduce lignin by 20% in the plant (Cherney et al., 1988). The trichomeless or
tr locus could potentially have an effect on improving drought resistance, disease
and insect resistance, and palatability. Loci controlling disease resistance are being used in both commercial grain and forage hybrids.
Linkage relationships have been established for only a few of these mutants
(Minocha et al., 1980b; Hanna and Burton, 1992, and summarized by Koduru et
al., 1983; and Anand Kumar and Andrews, 1993). Minocha et al. (1980a) used trisomics to map genes to chromosomes 1,2,4,5,and 6 . Liu er al. (1994) placed 181
RFLP markers on a molecular map. The length of the linkage map for seven linkage groups was 303 cM, with an average map distance of 2 cM between loci.
Xm. GENETICS OF QUANTITATIVE TRAITS
Burton (195 1, 1959) conducted some of the first quantitative genetic studies on
various plant characters and yields of pearl millet. Virk (1988) published a comprehensive review on quantitative studies conducted on pearl millet. Both additive
and nonadditive genetic variances are important in pearl millet. However, the nonadditive component tends to be more important, indicating the opportunity to successfully take advantage of hybrid vigor for both grain and forage production.
This, in fact, has been the case in pearl millet (see Section X).
Efforts have been made to identify qualitative characters linked to quantitative
characters affecting forage yield. Burton et al. (1980) showed that three recessive
mutants, T13 orange node, T18 early, and T23 stubby head, increased forage
yields 34, 38, and 22%, respectively, when heterozygous in an F, hybrid. In another study involving crosses between nonlethal genetic markers and exotic pearl
millet lines, the Rp, gene was associated with 1861% heterotic chromosome
block heterosis (HCB), and the tr was associated with 1 7 4 % HCB heterosis
(Burton and Werner, 1991). A similar approach used to identify HCBs in Burkina
Faso landraces identified up to 5 1% HCB heterosis associated with the R p , locus
in certain crosses (Burton and Wilson, 1995).
XIV. CONCLUSION AND PERSPECTIVES
With world population currently growing at the alarming rate of more than 2%
per year, meeting the ever-expanding need for food will be difficult in the near fu-