Tải bản đầy đủ - 0trang
IX. Genetic Control of Hybrid Vigor in Sorghum
GENETIC CONTROL OF FLOWERING AND GROWTH I N SORGHUM
has been due to the fact that no mechanism to control hormone levels
had been recognized. If such a mechanism has now been recognized, there
is reason to revise some of the theories that are the basis of practices in
plant breeding. The theory that varieties of self-pollinated crops such as
sorghum are burdened with numerous, small metabolic deficiencies is now
Hageman et al. (1967) has postulated that hybrids are superior to parents in having better balanced metabolic systems. They suggested that the
fundamental metabolic systems involved in growth and yield needed to
be recognized and, particularly, the optimum levels of activity of each enzyme. The idea that the genetic control over growth is hormonal is probably not in conflict with their concept and they have recognized that “a
single enzyme, hormone, vitamin, or growth factor could be solely responsible for the enhanced growth of a hybrid.” Nevertheless, they have stated
that “the complexities of metabolism preclude a single factor from being
the universal underlying cause of hybrid vigor.”
Data in Table I11 (Quinby and Karper, 1946) show that plants heterozygous at locus 1 when locus 2 is homozygous recessive were later to
flower than either homozygous genotype. But when locus 2 was homozygous dominant, the genotype heterozygous at locus 1 was earlier to
flower than the later homozygous genotype. The heterozygous genotype
Malmalma2ma, was much later to flower than either homozygous genotype
and produced a much greater yield of heads. However, the difference in
duration of growth between genotypes was great and the influence of duration of growth and of hybrid vigor cannot be separated. However, the
heterozygous genotype Ma,maIMa,Ma, was only 3 days earlier to flower
than the homozygous genotype MalMalMa,Ma, but produced a yield of
heads 60% greater due largely to more heads per plant. Heterozygous
genotypes for maturity were different from homozygous genotypes in both
maturity and yield of heads and heterozygosity was important largely because of interaction among genes at different loci rather than between alleles within a heterozygous locus.
Differences, due to gene interaction, between pairs of hybrids that differ
only in being homozygous or heterozygous at one locus are shown in Table
IV. These data were presented previously (Quinby and Karper, 1948),
but, at that time, the maturity genotypes of most of the parents were not
known. Now that the maturity genotypes are known, it is possible to draw
conclusions regarding the genetic cause of hybrid vigor that could not be
drawn in 1948.
J. R. QUINBY
Effect of Heterozygosity at the ma1 Locus on Days to Flower and Yield of Heads ‘ , b
Planted Sune 20, 1944
Ma1 Malmazmazrnaama3Ma4 Mac
Planted June 9, 1942
M a l m a lMazMazma3maaM a rM a 4
Ma1MalMa?Mazma3maaM a 4 M a 4
Data from Quinby and Kaper (1946).
Plant populations were grown at Chillicothe, Texas.
It is assumed that the paired varieties listed in Table IV differ only in
the genes controlling maturity. Many plant breeders or geneticists, including Schuler (1954), are skeptical that two varieties could differ only at
one or two gene loci. Such skepticism is difficult to allay because the skepticism originates in the misconception that so-called quantitative characters
are, necessarily, complex in inheritance.
Part of the data from the 1948 paper are presented again in Table IV
with the yield figures converted to grams. Only the data from pairs of hybrids that differ in one allele are presented. Quinby (1967) has presented
evidence that multiple allelic series exist at the maturity loci; but, because
the information has no bearing on the point under discussion, the multiple
allelic designations are not shown.
Heterozygous Masmas in the HEGARI x TEXAS MILO hybrid as compared
to homozygous recessive ma3ma3in the EARLY HEGARI X TEXAS MILO hybrid resulted in little difference in days to flower but in a large difference
in grain yield. The same was true in the HEGARI x SOONER MILO and
EARLY HEGARI X SOONER MILO hybrids. Heterozygous Masmas as compared to homozygous dominant Ma,Ma, in TEXAS BLACKHULL
KAFIR x EARLY HEGARI
and TEXAS BLACKHULL KAFIR x HEGARI
pair of hybrids resulted in earlier flowering and greater grain yield. TEXAS
Maturity Genotype, Days to Flower, and Grain Yield of Parents and Hybridso.b
Variety or hybrid
T E X A S MILO
TEXAS M I L O
S O O N E R MILO
TEXASB L A C K H U L L
x E A R L Y IIEGARI
x E A R L Y K.4LO
.Ifa ~ r n a ~ M a 2a2Ma3Ma3&1
m a ~ m a ~ M a ~ m a da&
f a ~a&
malma1M a2Ma2Af aaMa3Ma4Ma4
Data from Quinby and Karper (1948).
* Plant populations were grown at Chillicothe, Texas, in 1941.
Significantly greater than yield of other member of pair at 0.01 level.
J. R. QUINBY
BLACKHULL KAFIR x KALO and TEXAS BLACKHULL KAFIR x EARLY KALO
hybrids are similar in maturity to the commercial hybrids in general use.
Heterozygous Mu2mu,, in the former hybrid as compared to homozygous
Mu2Mu2 in the latter caused a 37% increase in grain yield even though
the days to flower of the two hybrids differed by only one day.
AT O N E HEIGHT
Graham and Lessman (1968) presented data at the 1968 meeting of
the Crop Science Society that showed hybrid vigor due to heterozygosity
at the dw, height locus in sorghum. This hybrid vigor was not the point
of their presentation and the abstract contains no reference to the yield
of the plants heterozygous for height.
The 2-dwarf parent used in their crosses was TEXAS MILO and the
3-dwarf parent was CALIFORNIA 38 MILO. Both parents are
Mulma2mu3Ma4for maturity but CALIFORNIA 38 MILO is recessive dw,
whereas TEXAS MILO is dominant Dw,,for height. Because both are MILO
varieties, they are in similar genetic backgrounds.
The pertinent data are shown in Table V. Plants heterozygous Dwz dw?
produced more grain than plants of either homozygous parent.
Effect of Heterozygosity at the diup Height Locus
I N MII.Oa*h
%dwarf X %dwarf, FI
2-dwarf X 3-dwarf, FI
3-dwarf parent (dw~dwzDiuaDwa)
Data from Graham and Lessman (1968).
Plant populations were grown at Lafayette, Indiana, in 1963 and 1965.
Significantly above either parent at 0.01 level.
The information presented leads to the conclusion that heterozygosity
at one height or one maturity locus, due to interaction between loci or
epistasis, results in greater grain yield; and this is true regardless of the
dominant or recessive condition at the homozygous locus.
GENETIC CONTROL OF FLOWERING AND GROWTH I N SORGHUM
Assuming that maturity genes control hormone levels, it appears that
heterozygous genotypes, due to gene interaction, produce levels of auxin
and gibberellin that are different from the levels produced by homozygous
genotypes. Some heterozygous combinations apparently produce levels of
hormones more favorable to growth than any homozygous combination,
but some heterozygous combinations would produce hormone levels more
favorable than others. Quinby (1963) presented data from two high- and
one low-yielding hybrid. Patanothai and Atkins ( 1971 ) presented growth
curves from two hybrids showing medium and high heterosis, but a growth
curve for the low heterosis hybrid was not shown because that hybrid was
not significantly superior, in the characters measured, to midparental
Sorgnum Genotypes as Experimental Subjects
Sorghum hybrids came into use in 1957 after a female parent was produced using cytoplasmic male-sterility (Stephens and Holland, 1954).
Prior to that time, farmers grew true-breeding varieties. Plant breeders,
for about 40 years, worked at producing improved varieties. Because
farmers saved mutations and plant breeders were interested in genetics,
the sorghum species now includes a number of varieties or genotypes that
are useful experimental subjects. If the hypotheses presented here are taken
seriously, they will need to be confirmed or refuted. Strains will be identified in this section that are in similar genetic backgrounds, but differ at
only one or two loci that affect growth. Physiologists working on the flowering process have, for the most part, neglected to use varieties, and have
thus been working without a check.
A tropical sorghum variety reached the United States in 1879 and was
called “MILLO MAIZE” (Karper and Quinby, 1947). In spite of its tall
height and late maturity, farmers grew the variety and by 1910 had selected
earlier maturities and short statures from the original variety. All the varieties that originated in this way are MILOS and differ from one another
only in maturity, height, or pericarp color. It was determined about 30
years ago that four mutations at maturity loci, two mutations at height
loci, one mutation for pericarp color, and one mutation for Periconia rootrot resistance had been preserved (Quinby, 1967).
In the process of studying the genetics of duration of growth, eight maturity genotypes were produced by selection from a cross between two va-
J . R. QUINBY
rieties of the genotypes Malmarrna3Ma, and malMa2Ma3Ma4.A linkage
between height and maturity was broken and the resulting eight genotypes
are all in the same genetic background and are recessive at three height
loci. When the eight genotypes are grown in long nights, they flower at
about the same time (Miller et al., 1968a); but in the short nights of temperate zones in the summer, they flower at different times as shown in
Table I. Subsequently, a second mutation at the third maturity locus was
found in a 3-dwarf MILO variety (Quinby and Karper, 1961) and named
RYER MILO. The new variety was identified for maturity as being
Malma2ma,RMa, and was given the designation of 4 4 M . 38M was then
obtained as a segregation product out of a cross between 44M and SM60.
The list of MILO maturity genotypes now includes the ten strains shown
in Table I.
LOCUS FOURFOR MATURITY
The only recessive known at locus 4 was identified in HEGARI. Because
all the MILO maturity genotypes are dominant at locus 4, 4-recessive maturity genotypes in pure MILO could not be obtained. But several useful
maturity genotypes using recesives from BLACKHULL KAFIR, MILO, and
have been produced. These genotypes are KMH-1
(Ma,ma,ma,ma, ), KMH-2 ( ma,mu,ma3ma,), MH-3 ( mulma2nza:,Rma4).
C. PAIRSOF VARIETIES
AT ONE MATURITY
Fourteen pairs of genotypes that differ at only one maturity locus exist
.among the MILO maturity genotypes shown in Table I. In addition, KALO
and EARLY KALO differ at locus 2; and HEGARI and EARLY HEGARI at lOCUS
3. The origins of KALO and EARLY KALO and of HEGARI and EARLY HEGARI
have been presented previously (Quinby, 1967). Maturity loci 1 and 2
have been assumed here to control the synthesis of auxin and loci 3 and
4, the synthesis of gibberellin. These pairs of genotypes might be used to
verify or refute these assumptions.
In an effort to make tropical germplasm readily available to plant breeders of sorghum in temperate zones, more than 100 tropical varieties have
now been converted to temperate zone adaptation. Most tropical varieties
are dominant at all four maturity loci and the conversion to temperate