Tải bản đầy đủ - 0trang
CHAPTER 4. SOYBEAN GENETICS AND BREEDING
HERBERT W. JOHNSOX AND RICHARD L. BERNARD
lion) was equal to that harvested for beans. The acreage harvested for
hay decreased steadily after 1940 and only about one-half million acres
have been harvested for hay annually since 1956. In contrast, the acreage
harvested for beans was nearly 14 million in 1950 and nearly 24 million
in 1960. An estimated 27 million acres were harvested for beans in 1961.
Soybeans are produced primarily for oil and protein. Seeds of
varieties produced in the United States average about 21 per cent oil
and 40 per cent protein on a dry weight basis. Nearly 90 per cent of the
soybean oil used in the United States is in human foods, primarily
shortening and margarine, and over 95 per cent of the protein used
domestically is in animal feeds.
The soybean has been used little as a research tool in genetics and
breeding research. However, research workers interested in improving
the species for economic use have made substantial contributions to
the literature on the genetics and breeding of soybeans. Notable breeding progress has been made in the United States and Canada in the past
25 years and this progress played a major role in the expansion of soybean acreage in the two countries. Breeding progress is becoming increasingly difiicult, however, because of an increased number of breeding
objectives, notably resistance to diseases, and because the gross gains
in the breeding of the introduced crop for a new production area have
been made. Future gains will be more difiicult than those of the past
and will require more refined techniques or procedures.
The purpose of this review is to summarize and present information
on the genetics and breeding of soybeans, particularly information relative to improving breeding procedures. In the attempt to make the
review as complete as possible, permission to use unpublished information was obtained from several individuals. This was done when the
unpublished information was considered to be especially pertinent in
rounding out the available information on a given subject or in a few
instances when the information was pertinent to a given subject and
would not be published elsewhere. The authors express their sincere
thanks to all the individuals contributing unpublished information.
The soybean belongs to the family Leguminosae, subfamily Papilionoideae, and the genus Glycinc L. The botanical classification of the
cultivated form has been controversial and the multiplicity of names
applied to it has created confusion as to its correct designation. However, Ricker and Morse (1948) contend that according to international
botanical rules the correct name of the soybean is Glycine max (L.)
Merrill, a viewpoint shared by most taxonomists.
SOYBEAN GENETICS AND BREEDING
The literature on the species situation within the genus Glycine has
been a grossly confused issue, and this confusion has complicated the
investigation of species in the genus. A recent concentrated attempt to
obtain seed of species of Glycine has met with some sucoess and a
taxonomic investigation by F. J. Hermann has greatly simplified the
species situation. He concluded that in addition to G. inux the genus
was made up of the following species: G. clandestina Wendl.; G. falcata
Benth.; G. ktrobeana (Meissn.) Benth.; G. tabacina (Labill.) Benth.; G.
tomentella Hayata; G. petitiana (A. Rich.) Schweinf.; G. javanica L.;
G. ussuriensis Regel & Maack; and a tenth species, G. sericea Benth. not
Willd., for which a new name is proposed in Dr. Hermann’s publication.
Seven subspecies or varieties are listed for G. javanica and one for G.
G. max and G. ussuriensis are known to have 40 chromosomes and
both behave as diploids. They are cross fertile and their hybrids usually
have normal fertility. Ramanathan (1950) listed G. javanica as having
20 pairs of chromosomes, the basic number of the genus being 10. H. L.
Weaver (personal communication) observed 20 pairs of chromosomes
in some types of G. javanica and 10 in others. He also observed 20 pairs
in G. falcata. Limited attempts in the United States to cross some of the
species (other than G. ussuriensis) listed above to G. mux have failed.
A type referred to frequently in the literature as G. gracilis was considered to belong to G. max in Hermann’s investigation of the genus.
B. ORIGINAND DISTRIBUTION
The origin of the cultivated form of the soybean is unknown. “The
soybean is native to eastern Asia” is a statement frequently transferred
from one publication to another. Nagata (1960a) recently reviewed the
literature on the subject and considered the distribution of soybeans and
other ethnobotanical principles in an interesting study of the origin of
the cultivated type. Although he concluded that the origin of soybean
culture still remains obscure, his results indicated that the origin was
in China proper, especially in north and central China. He based his
conclusions in part on the distribution of G. ussuriensis, which he considers to be the progenitor of the cultivated form. According to Morse
(1950) there is little doubt that G. max was derived from G. ussuriensis
since apparently no other wild plant found can possibly be its ancestor.
Nagata postulated that the cultivated form was introduced into Japan
1 Information from a manuscript entitled “A Revision of the Genus Glycine and
Its Immediate Allies” prepared by F. J. Hermann for publications as U . 5’. Dept. Agr.
Tech. Bull. 1268. Sincere appreciation is extended to Dr. Hermann for permission to
use the information prior to publication.
HERBERT W. JOHNSOS AND RICHARD L. BERNARD
via Korea and presented information to suggest that it was introduced
into Korea directly from north China sometime during the period 200 B.C.
to the third century. Hamada (1955) described preserved types stored
in the Shosoin Treasury since about the seventh century (along with
medicinal herbs introduced from China) that resembled the shortseason types currently grown in Kyushu and Loochoo Provinces of
Japan. Nagata (1960a) interpreted this to indicate that the short-season
types of Japan may have been introduced directly from central China
to south Japan. Additional detail on the ancient history of the soybean
may be obtained from Morse (1950).
According to Bening (1951) the first news of the soybean was
brought to the Western Hemisphere in the writings of Engelbert Kaempfer in 1712. Morse (1950) presented a detailed account of the modem
history of the soybean and recorded that the first published account of
the plant in the United States appeared in 1804. According to him not
more than eight varieties of soybeans were grown in the United States
prior to the numerous introductions by the U. S. Department of Agriculture beginning in 1898.
Introductions from Manchuria, China proper, Korea, and Japan have
played a predominant role in the soybean industry in the United States.
The early varieties and the germ plasm used in soybean breeding in
this country came from these introductions (see Section V, F).
Soybean flowers are normally about 6 to 7 mm. in length, and their
smallness imposes a limitation on the ease with which controlled pollinations can be made. Guard (1931) described the soybean flower as
having a tubular calyx terminating in five unequal lobes. The largest of
these is anterior, the next two lateral, and the smallest two, obliquely
posterior. The calyx is persistent, being intact on the ripe fruit, but
rapidly deteriorates and only fragments may be found on pods that
have been exposed to weather for a considerable length of time. The
corolla consists of five separate petals. The largest (standard) is posterior, the two next in size (wings) lateral, and the two keel petals
anterior. There is no fusion of the keel petals as in some other legumes.
The ten stamens are separate at first, but shortly before anthesis the
filaments of nine of them are elevated as a single structure by the
development of a basal region, leaving the posterior stamen separate.
Miksche ( 1961 ) recently reviewed the literature on morphological
studies with soybeans and presented the results of an interesting study
on the developmental anatomy of the plant. He studied organ and tissue
SOYBEAN GENETICS AND BREEDING
organization from dormant seed to floral initiation. Although this area
of work is beyond the scope of this review, Miksche’s results and the
literature cited by him are valuable for research workers interested in
The time of flowering of soybean plants depends largely on the
number of hours of darkness they receive each day. Other factors such
as temperature, nutrition, and light intensity and quality may influence
the response of soybeans to dark periods suitable for flowering; but in
the field the length of the dark period is usually the primary influence
in the induction of flowering. Plants of many varieties are completely
incapable of flowering unless they receive 10 or more hours of darkness
daily and plants of all varieties flower more quickly with daily dark
periods of 14 to 16 hours than with shorter ones (Borthwick and Parker,
1939; Parker and Borthwick, 1951). Since the length of the daily dark
period is a function of latitude, soybean varieties are adapted as a fullseason crop to narrow belts of latitude.
The effect of changes in natural photoperiods on the maturity of
soybeans occurs primarily prior to flowering. Rates of development in
subsequent stages also are influenced by photoperiod but the periods
from about seed set or end of flowering to maturity are similar for all
soybean varieties regardless of maturity. Natural photoperiods over a
wide range of latitude also are similar during the latter stages of
development of the soybean. When the natural photoperiods are manipulated to create substantial differences, differences among varieties in
their response to photoperiod in stages of development after flowering
are readily observed (Nagata, 1960b; Johnson et al., 1960). This is discussed in greater detail in the review by Cartter and Hartwig in this
Soybean plants normally produce many more flowers than pods that
finally mature. Shedding of 75 per cent or more of the flowers is not
uncommon, and even under the most favorable conditions the loss of a
substantial portion of the flowers can be expected. Flower and pod
shedding apparently are not due to a lack of viable pollen (Van Schaik
and Probst, 1958b) or to lack of fertilization (Kato .et d.,1955).
Soybeans are completely self-fertile and the amount of outcrossing
under natural conditions is about 0.5 per cent for plants in adjacent
rows and 1 per cent for plants grown in close contact (Weber and
HERBERT W. JOHSSOS ASD RICHARD L. BERNARD
Soybean breeders and geneticists have become increasingly concerned in recent years with techniques for increasing the amount of
“natural” outcrossing in soybeans. A search for male-sterile, female-fertile
types has been unsuccessful, and various other approaches have been
chemical referred to in the literature as FW-450 and two apparently
related chemicals ivere evaluated by Casas (1961) as selective gametocicles. Although pollen viability was reduced by the chemicals, the flowers failed to open properly and an actual decrease in outcrossing resulted.
Similar results were obtained with F\17-450by Hanson (personal communication ) . Casas obtained 5.2 per cent outcrossing of normal plants
in cages containing honey bees compared to only 0.6 per cent for
plants outside the cages.
\\‘eber and Hanson (1961) obtained a four- to sixfold increase in
outcrossing of plants from seed irradiated with different dosages of
X-rays and thermal neutrons. Outcrossing of the untreated checks was
approximCitely 1 per cent. From a theoretical consideration of the
amount of outcrossing required to have practical utility in intermating
populations, they concluded that the figure should be higher than the
approximately 4 to 6 per cent which they obtained.
Athow (personal communication) has observed as much as 16 per
cent outcrossing of plants infected with tobacco ringspot virus. A small
percentage of the seed from infected plants normally give rise to virusfree normal plants (-4thow and Bancroft, 1959) and the possibility of
utilizing the normal plants as pollen parents in a virus-infected population caged with bees is intriguing. The virus-free plants could be used
to establish noinial lines after the desired number of generations of
Crossing soybeans is tedious. The procedure followed in emasculating is standard. but the time of emasculation, collection of pollen, and
pollination varies greatly. This variability depcnds to a large extent on
environment and to a lesser extent on the personal preference or needs
of indib idual workers.
The small size and fragileness of soybean flowers make it necessary
to use extreme care in emasculating. The only instrument used is a small
pair of forceps. The inner surfaces should be flat without corrugations
and the spring end should have light tension. Flowers which would
normally open the morning following the day the cross is made are
used as females. These are in the bud stage with the color of the petals
readily visible. The lobes of the calyx are removed by grasping them
SOYBEAN GENETICS AND BREEDING
individually with the forceps and pulling downward. The corolla is
then grasped with the forceps at a right angle to the axis and removed
with a slow pull, working the forceps gently from side to side during
the process. The corolla and all the anthers may be removed in one
motion; however, the keel and/or anthers often are not removed with
the first pull. The keel can be removed easily and the points of the
forceps can be used to remove the anthers. However, removal of the
anthers in this manner frequently results in injury to the remaining
parts of the flower and poor success in crossing. When a genetic marker
can be used to distinguish F1 plants from plants of the female parent,
most experienced operators make no attempt to remove the anthers with
the points of the forceps.
Local environmental conditions, including weather and insects,
determine the time of day when pollen is collected and crossing is
most successful. Generally, in the central and northern parts of the
United States, flowers that have opened the day the cross is to be made
can be used to furnish pollen and crossing can be done successfully
throughout the day. However, in the southern States it is extremely difficult to get viable pollen from open flowers. In this area pollen flowers
are collected early in the morning an hour or two before they would
have opened and stored in a cool, dry place for use later in the day.
A desiccator is usually used to ensure dry storage conditions. Pollinations
in mid to late afternoon are usually the most successful. In the southern
States the percentage of success from pollinations made in the forenoon
is extremely low.
When the flowers are ready for pollinating, the tips of the forceps
are inserted in the back of the keel of the pollen flowers and the pistil
and column of anthers removed. This is used as a brush to deposit pollen
on the exposed stigma of the emasculated flowers. Generally three or
four pollinations can be made with a single flower.
From one to three flowers at a node may be in the right stage for
crossing and all other flowers and buds should be removed. The flowers
or node should be tagged for identscation of the cross. The flowers
should be checked about a week after the pollinations are made and all
newly developed flower buds removed. Soybean flowers may continue
to develop after a cross has been made and when harvesting the crossed
seed it is sometimes difficult to determine whether the pod developed
from the emasculated flower or from one that developed later. Pods
resulting from a cross can readily be distinguished when small by the
absence of the calyx lobes on the base, and this distinction can usually
be made when the pods are mature.
Some published information indicates that flowers should be emascu-
HERBERT W. JOHNSON AND RICHARD L. BERNARD
lated one day and pollinated the next and that a leaf should be pinned
around each crossed flower for protection. Most agronomists in the
United States use no type of covering or protection for crossed flowers.
They emasculate the flowers desired for one cross and pollinate them
Even when the best available techniques are employed by experienced operators, the percentage of successful crosses varies greatly from
time to time. Too much or too little moisture, low night temperatures,
insects, manipulation of the photoperiod, and various other factors have
been observed to influence the success of crosses. The ideal environment
for crossing soybeans is unknown, but the success of the crossing program often can be increased greatly by a well-timed application of an
insecticide or supplemental irrigation.
Although the standard procedure of crossing soybeans has been used
for some large undertakings in recent years, much time and effort are
required in the actual crossing operation and in obtaining the desired
flowering plants over a sufficient length of time. Techniques for storing
pollen and means for speeding large-scale crossing operations would
therefore facilitate current research programs.
Hanson (personal communication ) materially increased the number
of pollinations that could be done per day by doing some of the operations in the laboratory in the morning when pollinations are least successful. Flowers were collected at the appropriate time in the morning
and the anthers and pistil were separated from the floral parts in the
laboratory. Up to 30 anther rings were stored in a 00 gelatin capsule
by sticking them around the edges of the capsule. The capsules were
stored over a mixture of 25 ml. of concentrated sulfuric acid and 75 ml.
of water in a refrigerator to maintain the desired humidity.
Kuehl (1961) recently obtained useful data on a number of questions
of importance in crossing soybeans: (1) Germination of pollen in a
30 per cent sucrose solution containing 120 p.p.m. of boric acid was
found to be a good indicator of the germination or effectiveness of
pollen in crosses; ( 2 ) pollen was stored successfully in a calcium chloride
desiccator at 3.3"C. for approximately 1 month. Storage at -20" was
less successful. The stored flowers were dry and brittle but storage for
about 30 minutes over water in a closed container restored moisture to
the tissues and induced the anthers to dehisce. ( 3 ) Pollen first became
viable about 10 hours prior to natural anthesis; and (4) the stigma of
emasculated flowers was most receptive to pollen on the day preceding
the morning of normal anthesis and remained receptive for 2 days
In a report of a detailed study of the time required for various
SOYBEAN GENETICS AND BREEDING
stages of development from flower bud differentiation in the plant to
differentiation of the embryo in the seed, Kato et ul. (1954) presented
photographs interpreted to indicate fertilization on the day of flowering.
Because flowers normally open in the early daylight hours and the exact
time of collection of flowers was not given, the results can be interpreted
to indicate that fertilization takes place in about 10 hours or less after
natural pollination. If this same time sequence prevails in flowers used
in crosses 15 to 20 hours before anthesis would have occurred, there
would seem to be little likelihood of self-pollination even in nonemasculated flowers if viable pollen is used since the pollen of the
crossed flower would not be viable until 5 to 10 hours after the cross
was made (Kuehl, 1961).
111. Genetics of Qualitative Characters
Although the soybean has never been the subject for extensive
research by geneticists, the mode of inheritance of simply inherited
characters has been studied from time to time by agricultural workers
interested in the soybean as a crop. Reviews which include the results
of much of this work have been published by Owen (1928a), Woodworth (1932, 1933), Matsuura (1933), Morse and Cartter (1937), Weiss
(1949), Williams (1950), and Johnson (1961).
Genes reported for the soybean and the traits that they influence are
given in Table I. The list includes all genes reported except those
presented as only tentative suggestions and those apparently based on
expressions of previously reported gene pairs. Accordingly, in this
review, del (Stewart and Wentz, 1930) is considered to be the same
as t; de2 (Woodworth and Williams, 1938) to be p2; f ( Takahashi, 1934)
to be nu; h (Ting, 1946) to be t; lo (Domingo, 1945) to be o (oval leaflet); 90 (Stewart, 1930) to be o (reddish-brown seed coat); and y2
(Morse and Cartter, 1937) to be g. Table I includes the gene symbols
(where duplicate symbols have been assigned the ones used in the more
recent review papers are presented here), a brief descriptive phrase for
the contrasting traits, and the major reference( s ) identifying the gene
pair and assigning the symbols. A soybean strain carrying the specified
gene(s) is listed for each of the more uncommon traits. A few genes
such as df, st, and yl have been lost whereas others such as A, BZ, E, e,
L, 1, S, s, sh, sp, and w2 are probably present in available germ plasm
but have not yet been reidentified.
The variation among soybean varieties in pigmentation of various
parts of the plant, especially the seed, has provided the material for
of the I’h(wotylw.: and tlw 1iefrrcwc.c.s 13st:il)lishing the h l o d ~of
1nlicrit;inc.c nncl Assigninq thv S ~ m l m l s
‘4 1,ist of Cerics 13t*imrt(dfor thc Soy1)cwi Ilic.111diiig;i &scription
Bl 4 B ,
h , or h , o r 12,
c , c,
c , or c2
Appww’cI p ~ i l w s ~ x * n ~ ~ ~ ~
Erwt p r i l ~ w c . ~ w x
1,c:nf abscission :it matririty
Dc4aycd nlwission ( T206)a
Bloom on srcd coat ( T 4 )
Slrarp prilwsccncc: tip
Cr;ick(d s ( d coat ( T217)
Entirc sccd coat
1hist;incc. to frogoy lcafspot
Yollow cotyicdons in srecl
Green cotylcdons (T38)
Indeterminate stem (T10)
Determinate stem ( T 6 )
Fasciated stem (T173)
Normal iron utilization
Inefficient iron utilization (T203)
Kurasn\vn, 1036; Morsr and C:irtt(,r, 1937
Woodworth, 10.32, 1933
Napii, 19%; hf;itsurira, 1933
Athow and Probst, 19S2
Woodworth, 1921; Owm, 1 9 2 7 ~ ; Veatch und
Stewart, 1927; Wooclworth, 1932, 1933
Woodworth, 1932, 1933
Nagai, 1926; Takngi, 1929; Woodworth, 1932,