Published May, 1952
Production of Homozygous Diploids of Maize from Monoploids’
Sherret S. Chase2
HOMOZYGOUS diploid lines of maize, the genetic equiv-
alents of advanced generation inbred lines, can be
produced directly from monoploid maize plants. The two
major problems involved are, first, the production and early
recognition of large numbers of the monoploid sporophytes
and, second, the derivation of self (and therefore, pre-
sumably homozygous) diploid progeny from these aberrant
plants. The first problem has been solyed in spite of the
low rate of natural occurrnce of monoploids (about one
per thousand diploids on the average) by a combination
of genetical, morphological, and cytological sorting meth-
ods and by the use of favorable seed and pollen parents.
The second problem has been partially solved by depend-
ence on spontaneous doubling of the chromosome com-
plement in occasional cells of the tassel and ear meristems
of the monoploid plants. Such doubling leads to the devel-
opment of diploid sectors in the tassels and ears and greatly
increases the chances for successful self-fertilization (1).
About 10% of untreated monoploids yield successful self
progeny, largely as a result of this spontaneous somatic
diploidization.
At Iowa State College approximately 100 homozygous
diploid lines have been developed by the monoploid
method during the past 2 years. The following is a dis-
cussion of (a) techniques used in isolating monoploid
plants in quantity; (6) techniques used in growing these
plants and in developing homozygous diploid lines from
them; and (c) some ways monoploids and monoploid
derivatives may prove useful in a breeding program.
ISOLATION OF MONOPLOIDS
Genetic Methods
Monoploid sporophytes of maize presumably develop by
parthenogenesis of otherwise normal eggs, or by apogamy,
or, rarely, as a result of androgenesis. Since most mono-
pIoids are derived from the seed (female) parent, any
such plants occurring in the progeny of a cross between
two dissimilar stocks will resemble the seed parent, within
the limits of the heterozygosity of that parent, and will
be unlike either the pollen (male) parent or the hybrid
progeny. On this account the search for monoploids is
greatly simplified if one looks for them among individuals
of progenies of crosses between markedly dissimilar stocks.
Further, in order that the monoploids be recognizable in
the seedling stage, it is necessary that the contrast be
evident in the seedlings. Therefore, stocks carrying unusual
dominant marker genes, or gene systems, which in the
hybrid seedlings produce easily recognizable phenotypes,
make the most satisfactory pollen parents or ‘markers’ for
general use.
Since, in at least the majority of cases, the endosperm
of kernels containin: monoploid er?bryos is a normal
Journal Paper No. 5-1708 of the Iowa Agricultural Experi-
ment Station, Ames, Iowa. Project No. 1201. Received for pub-
lication April 7 , 1951.
Research Associate Professor, Botany and Plant Pathology
Department, Iowa Agricultural Experiment Station, Ames, Iowa.
263
(3n) fusion product it should not differ in phenotype
from the endosperln of normal hybrid kernels. O n this
account endosperm markers can not be used to locate ker-
nels containing monoploid embryos. They can, however,
be used effectively to eliminate kernels resulting from
accidental self or cross pollinations. For this purpose,
as for seedling marker genes, unusual genes with easily
recognizable phenotypes make the most satisfactory mark-
ers for general use. Elimination of seedlings resulting
from accidental self or cross pollinations is desirable since
if these seedlings are not recognized early they are likely
to be classified as putative monoploids and carried along
until the determinations of chromosome number have been
made.
One additional general requirement for efficient pollen
stocks for this work is that they stimulate monoploid
occurrence. As noted elsewhere ( 2 ) , .certain stocks when
used as seed parents tend to produce significantly higher
frequencies of monoploids than other stocks, irrespective
of the pollen parent used, and sinrilasly certain stocks when
used as pollen parents tend to produce higher frequencies
of monoploids than others, isreApective of the seed parents
involved. The actual incidence of monoploids obtained in
the progeny of any particular cross is influenced by both
parents in spite of the fact that the pollen parent con-
tributes none of its genes to the monoploid.
Thus, the ideal marker stocks carry unusual dominant
genes which produce an easily recognizable phenotype in
the hybrid seedlings, dominant genes which produce an
easily recognizable phenotype in the hybrid endosperm,
and the requisite genes for stimulation of parthenogenesis.
Seedling Markers (4)
Purple plant color has, of the types of seedling markers
tried SO far, proved to be the most satisfactory ( 7 ) . This
plant color is both distinctive and unusual; consequently
marker stocks which in crosses yield a purple hybrid prog-
eny can be used widely without special regard for the
genetic nature of the seed parent. In combination with
the majority of stocks likely to be used as seed parents,
purple ( A B P1 R) or brown ( a B PI R ) marker stocks
yield a purple colored hybrid progeny. This phenotype
develops first in the older portions of the seedling roots
of the hybrids. If the germination of the marked kernels
i s carried out on trays so that the seedling roots can be
checked readily the elimination of the plants showing a
purple color in the roots can be completed within a period
of froin 6 to 10 days after starting. Pink and red light
induced colors develop in the roots of many types of
maize; consequently one should germinate the kernels
carrying the purple plant color genes in the dark.
Classification for purple plant color can also be made on
seedlings grown in sand beds. Under these circumstances
the classification is carried out from 1 2 to 16 days after
the kernels are planted, as soon as the purple plant
color becomes well developed in the sheaths of the seed-
ling leaves. The sand bench method is less satisfactory
than germination on a tray in a germinator, since the
process takes longer, ties up more space, is more tedious
264 AGRONOMY JOURNAL
for the worker, is less precise, increases the difficulty of
obtaining good root tips for cytological study, and increases
the shock of transplanting for the seedlings saved.
Purple plumule ( A Pzil Pu,), a spot of color at the tip
of the plumule of dormant and germinating seedlings,
makes a satisfactory marker in certain crosses. A disad-
vantage is that this dominant gene system has a rather
wide distribution among maize of commercial interest.
There ate also apparently a number of modifiers which
tend to suppress the purple plumule phenotype. Conse-
quently, seed stocks to be crossed by a purple plumule
marker should first be checked carefully for plumule color
and a preliminary check of the hybrid made to determine
the suitability of the marker for the particular cross under
consideration.
In some instances the genetic classification for purple
plumule can be carried out in the dry kernel stage; gener-
ally this is not satisfactory. Classification during , germi-
nation on trays is more certain. The checking is started
as soon as the plumule color is evident and continued
daily until the elimination of purple plumuled seedlings
is completed. In some hybrids there is a tendency for the
color mark to fade early, and in others colors with other
genetic bases may develop in the plumule during the ger-
mination period.
Purple and red scutellum, dependent on the genes A ,
C, R, i, S,, s, and at least two of the genes S, S , and S,,
and differentiated into purple and red by the alleles Pr
and PY, make excellent markers in certain crosses. As with
purple plumule, preliminary tests are necessary to deter-
mine their suitability in any given cross. In hybrids in
which scukllum color is well developed there is a real
possibility of carrying out effective preclassification in the
dry kernel stage, prior to tray germination.
Purple coleorhiza (Pc, Pc, Pc, pc4) can be used effec-
tively in some crosses. Again, the hybrids and seed parent
should be checked carefully before extensive crossings are
made. Classification can, and must, be carried out during
the earliest stages of germination.
Combinations of purple plant color, purple plumule,
purple scutellum, and purple coleorhiza in a single marker
stock are possible and afford obvious advantages.
Endosperm Markers ( 4 )
Purple and'red aleurone colors ( A , A , A , C R i Pr or
p r ) are probably the most satisfactory endosperm markers
to use in conjunction with either purple plant color or
purple or red scutellum. In fact, for scutellum color, the
genes. known to be basic to aleurone color are all required,
in addition to the special scutellum color genes, for expres-
sion of the colored scutellum phenotype. And several of
the genes necessary for purple plant color are also neces-
sary for purple or red aleurone. Purple aleurone is easily
recognizable in most hybrid kernels, though red pericarp
when present may mask the character. In many stocks genes
wliich modify the expression of the character occur. In
any doubtful cases the elimination of kernels lacking aleu-
rone color can be delayed until the pericarp is loosened
during germination. At this stage the latter can be peeled
off easily and the presence or absence of color in the aleu-
rone layer determined.
Starchy (nonsugary) endosperm (Sa) makes an excel-
lent endosperm marker when sweet corn is being used .as
the seed parent and when the surrounding maize in the
crossing plots is also sweet. (Since most purple aleurone
stocks carry the nonsugary gene these two endosperm
markers may frequently be used in conjunction.)
Y e l l o w endosperm ( Y ) is useful as a marker when the
seed stocks have white endosperm and when the surround-
ing maize, from which contaminant pollen is likely to
come, is also3 white.
Source (seed) Stocks
Some data on monoploid frequencies in maize have
been published ( 2 ) . These and other data developed since
(to be published) indicate that there are marked differ-
ences between stocks in respect to the frequencies with
which monoploids occur among their marked progenies.
Inbred lines, single cross, double cross and topcross hybrids,
and open pollinated and synthetic varieties have been
tested. Monoploids have been found among the marked
progenies of all of these. In general those stocks which
have been subject to the greater degree of agronomic selec-
tion are the more favorable sources of monoploid plants.
In making the source,/marker crosses the standard meth-
ods of controlled cross pollination may be followed.
Handling and Germination of Marked Seed
During harvest marked ears showing a high degree of
pollen contamination are discarded and occasional con-
taminant kernels are thrown out as recognized. During
processing a fungicide such as Arasan (used dry at a rate
of about 0.5 gm per 1000 gm of seed) is applied. Seed
treatment has been found to be important since hyb'rid
seedlings developing from diseased kernels often fail to
express the expected color phenotype during the early
stages of developme nt .
For germination a high humidity germinator held at
approximately 25 C is satisfactory. Seedlings showing the
hybrid phenotype 'are discarded promptly as soon as the
marker phenotype is expressed. Daily checking is neces-
sary. The process of discarding recognizable hybrids should
be completed before the putative monoploids show the
'first seedling leaf. Then after root samples are taken, the
seedlings are transplanted (21/-inch pots are satisfactory
at this stage) for holding until the cytological examina-
tions for chromosome number are completed. Counting
of chromosomes is necessary to establish definitely the
monoploid nature of suspected seedlings. In some cases
the genetic screen is very efficient, in others only about
1 in 10 of the putative monoploids on checking proves
to be truly monoploid. The others are either maternal
diploids, accidental selfs or outcrosses, mutant hybrids,
hybrids in which the marker phenotype has been masked
by modifying genes, or diseased hybrids.
During the holding period it is possible to restrict fur-
ther the group of putative monoploids. If the fully devel-
oped first seedling leaf (first leaf above the plumule) is
less than one-half the length of the comparable leaf of
seedlings of the source stock the seedling in question is
probably monoploid. If the leaf is of approximately the
same length as those of the source stock the seedling is
almost certainly diploid. The latter may safely be
discarded.
At the time the putative monoploids are recognized,
actively growing root tips, one or two per plant, are
placed in small bottles in about 5 ml of saturated aqueous
solution of paradichlorobenzene, monobromobenzene (6),
 CHASE: PRODUCTION OF HOMOZYGOUS DIPLOIDS OF MAIZE FROM MONOPLOIDS
or 8-oxyquinoline (9). This pretreatment produces a super-
contraction of the metaphase chromosomes and conse-
quently makes the subsequent smear preparations more
satisfactory. After 2 or 3 hours the aqueous solution is
poured off and acetic alcohol (1 part glacial acetic acid;
3 parts absolute ethyl alcohol) added. After 1 to 3 days
storage in this solution the root tips are ready for smearing.
If sand bench is used instead of germinator tray, for
' progenies marked with the purple plant color phenotype,
the classification is made at a later stage, usually about
the three-leaf stage. In order to obtain good root tips from
these putative monoploids it is usually necessary to take
the tips after the plants have become reestablished follow-
ing transplanting to the small pots.
T o prepare a smear the meristematic tip of the root is
placed in a small drop of acetic-orcein on a clean slide
and cut into three or four or more small pieces. These
are then mashed thoroughly to separate the individual
cells. The visible pieces of debris are next removed from
the drop of dye and the preparation covered with a No. 1
coverglass. The slide is then heated over an alcohol flame
several times within a period of 5-10 minutes, inverted
on a piece of absorbent paper placed over a smooth glass
plate, and pressed firmly on the back. This pressing results
in a flattening of the cells. Considerable manual pressure
is necessary to insure adequate spreading. The cells are
then ready for examination under the microscope. Mono-
ploids of maize have 10 chromosomes pet cell, diploids 20.
CULTURE A N D POLLINATION
Culture of the Monoploid Plants
As the cytological data on chromosome number become
available the true monopIoids may be transferred to 8-inch
pots, ground bed, or to the field. Use of the greenhouse
during the fall, winter, and spring followed by field plant-
ings during the season makes possible a fairly continuous
schedule of monoploid production.
The ideal stage for the second transplanting seems to
be that at which the fourth leaf appears. If the seedlings
are held much longer in the small pots they are apt to
be stunted; if transplanted sooner loss of plants through
root rotting is increased. Furthermore, most plants geneti-
cally too weak to survive will die before the four leaf
stage.
Greenhouse temperatures of from 70 to 8oo F during
the winter months and an extended day length of 14 hours
have proven satisfactory for growth and fairly normal
development of the plants. Since the floral primordia of
corn plants are differentiated at a very early stage, control
of day length is particularly important during the seedling
and early plant stages.
At the time of transplanting and at weekly intervals
thereafter pot-grown plants should receive some ndtrient.
At Ames the dosage pet plant per week on the average
has been 0.5 gm ammonium nitrate plus 0.5 gm potassium
acid phosphate (mono).
Chromosome Doubling and Pollination of Monoploids
A monoploid as a monoploid produces little or no good
pollen ; however, monoploid plants occasionally produce
diploid cells, presumably through mitotic misdivision. If
these aberrant cells are in the meristems which give rise
to the reproductive tissues, the chances that functional pol-
265
len and eggs will be produced are greatly increased. If
part of the tassel is derived from such a diploid cell, that
part of the tassel will (assuming no genetic abnormalities
are involved) produce diploid anthers of normal size filled
with good pollen. If a sector of the ear is diploid, the
ovules of that part may be expected to set seed following
pollination with normal pollen. I n practice it has been
possibIe to get successful diploid progenies from from one-
fourth to one-half of the monoploid plants which have
been self pollinated with pollen from a diploid sector of
the tassel. Fortunately even a single diploid anther pro-
duces a large number of pollen grains. Since inany of the
diploid sectors developed involve only a single anther or
a single floret it is important to save all the pollen pos-
sible. In the greenhouse this is easy, for in still air the
pollen tends to remain within the anthers and may be
gathered directly from them. In the field it is more diffi-
cult to save the good pollen and the chances for accidental
cross pollinations are considerable unless the plants are
grown at a distance from diploids or are carefully guarded
during the pollinating period. In greenhouse and field
it is a good practice to cut back the ear tips before anthesis
so that good pollen when available will not be wasted for
lack of exposed silks.
The percentage of monoploids producing no good pol-
len may be decreased by artificially increasing the rate of
chromosome doubling. Injection of 0.5 ml of colchicine
solution (an aqueous solution of 0.05:% colchicine and
10% glycerine) into the scutellar node of the young mono-
ploid seedlings appears promising as a method for increas-
ing the rate of chromosome doubling, as indicated by pre-
liminary data (unpublished). The fact of natural doubling
of the chromosome complement encourages the belief that
the rate of doubling may be increased effectively by special
treatmenti.
Culture of Monoploid Progenies
Some special care is needed to assure successful estab-
lishment of diploid progenies from the monoploids which
do produce self seed. Since most of these plants will pro-
duce only one or a few good kernels, the chances of infec-
tion with fungus spores are greatly increased over that
to be expected in fully set ears. Seed treatment with a safe,
effective fungicide is recommended and also germination
of the seedlings on blotters, then transfer and establish-
ment in small pots, before planting in the field or green-
house. One should not consider a line successfully estab-
lished until it has passed through one diploid generation
and an ample supply of self seed has been obtained. To
insure that the lines developed are free of chromosome
abnormalities the plants of the first diploid generation
should be checked for pollen abortion. For maintenance
and increase of the homozygous diploid lines the same
techniques may be followed as are used with inbred lines.
BREEDING SYSTEMS
At present the two principal and related objectives of
the monoploid program at Iowa State College are the
production of monoploid-derived homozygous lines for
use in hybrid combinations as the equivalents of inbred
lines, and the development of elite monoploid-derived
synthetic varieties to be used in a program of recurrent
selections (3, 5 , 8).
2 66 AGRON03dY JOURNAL
The difference in time between this newer method and
the inbreeding method for producing tested pure breeding
lines from untested material will not b- the net saving in
breeding time, since with ordinary inbreeding it is possible
to test the lines being inbred during the breeding process,
whereas under the monoploid method the testing for yield
characteristics must, at least at present, be delayed until
the homozygous diploids are produced. AS compared with
inbreeding the use of monoploid intermediates promises a
saving of from 1 to 3 years in the combined breeding and
testing period. At present data are ‘being developed for
comparisons of monoploid-derived homozygous lines and
inbred lines developed from the same heterozygous sources.
The expectation is that the homozygous diploids will be
at least as good in combining ability as comparable inbred
lines. In terms of general vigor there is reason to believe
that the homozygous diploids as a group will prove some-
what better than unselected inbreds since passage through
the monoploid state as a sporophyte must entail rigorous
selection against deleterious genes or combinations of
genes.
Until data on comparative combining abilities of homo-
zygous diploids and related inbreds become available th:
merits of the monoploid method can not be accurately
gauged. If the average value of unselected monoploid-
derived homozygous lines should prove to be superior to
that of uriselected inbred lines from the same source in
regard to yield characteristics, the method will certainly
become a major breeding tool. If such lines prove to be
as good as inbreds the method will at least prove advanta-
geous for the solution of certain special breeding prob-
lems. The unique feature of the method is that it fixeJ
the genetic systems of individual gmzetes, in testable, repro-
ducible form; that is, as sporophytes.
The monoploid method may prove especially effective
in developing replacements for inbreds now used in well
established hybrids. In such cases a possible approach would
bz to use topcrosses, of the deficient inbred to stocks
carrying the required new characteristics (or backcrosses
of such topcrosses), as the source material and restrict the
yield testing to direct comparisons of the new homozygous
lines obtained with the inbred to be replaced in a given
specific hybrid combination, using the opposed single cross
hybrid as t.he tester stock.
It may be possible in such cases and in others where
one harvested the original seed of a heterozygous stock
or topcross series by ear lot, to carry out a preliminary
nonspecific yield trial at the same time the source/marker
crosses are made. If material to be simultaneously marked
for genetic separation of the monoploids in its progeny
and compared for yield characteristics is planted ear to
row in sorne standard yield trial pattern in an isolated
crossing block with interspersed rows of the marker stock
and the sol;:rce stocks are detasseled, all seed on the source
stock plants would then be pollinated by the marker. At
harvest the separate items in the yield test could be com-
pared and only the better ones saved as monoploid source
material. The monoploids derived from these kernels
would, of course, be uncontaminated by the marker parent
and should prove to be an especially favorable group as
regards yield genes.
Since the monoploid method fixes gametes it may also
prove useful as an adjunct to recurrent selection. Though
each cycle of selection and recombination would probably
take one year longer than is taken when S, plants are used
for making LIP the new synthetics, the greater accuracy
of testing warranted by the use of homozygous stocks, the
fact that the stocks recombined are the stocks tested, not
a segregating progeny thereof, and the reduced possibility
of carrying along ‘poor’ genes should result in a more
effective shift of gene frequencies in the desired direction.
Further, the synthetics developed can be exploited promptly
since at every stage homozygous lines are produced or
can be extracted in a minimum of time.
Under some circumstances direct utilization of mono-
ploids without resort to selfing may prove of value. When
an abundance of good pollen is used on monoploid silks
about one-third of the plants form some good seed, as
compared to about one-tenth when only self pollinations
are attempted. By testing the monoploids directly for cer-
tain agronomic characteristics such as, say, disease resistance
or strength of stalk and by then backcrossing the plants
to the source stock (or by intercrossing the elite monoploids
to the extent possible) it should be possible to build up
rapidly heterozygous stocks with high frequencies of
desired genes. This method has already proved effective
in building up stocks characterized by a high rate of
parthenogenesis.
Study of monoploid derivatives and o f other stocks indi-
cates that both parthenogenesis and monoploid fertility
are genetically conditioned. The data suggest that about
a ten fold increase in efficiency in the development of
homozygous diploids can be had by the use as source
material of stocks derived from elite monoploids. The
monoploids elite for this purpose are those which are
both highly self fertile and which yield progeny in which
the frequency of parthenogenesis is high.
Of interest is the fact that the open pollinated varieties
tested so far have comparatively low rates of partheno-
genesis. To utilize these sources it may prove advantageous
to backcross all the monoploids obtained during the first
selection to thz source variety and then to use this back-
cross generation or some derivative of it as the base for
the development of homozygous lines.
SUMMARY
The two major problems involved in the development
of homozygous diploid lines of maize from monoploid
sporophytes are (1) the production and early recogni-
tion of large numbers of monoploids and ( 2 ) the deriva-
tion of self (and homozygous) diploid progeny from these
monoploids. The first problem has been solved in spite
of the low rate of occurrence of monoploids (about 1/1000
on the average) by a combination of genetical, morphologi-
cal, and cytological sorting devices and by the use of favor-
able sFed and pollen parents. A partial solution of the sec-
ond problem has resulted by reason of the fact that mono-
ploids are subject to occasional spontaneous doubling of
the chromosome complement. About 10% of untreated
monoploids have yielded successful self progeny, largely
as a result of this diploidization. At Iowa State College
about 100 homozygous diploid lines have been developed
by the monoploid method during the past 2 years. The
techniques used in isolating monoploids in quantity, in
growing and deriving homozygous diploids from them,
and some of the ways monoploids and monoploid deriva-
tives may b- used in a breeding program are detailed.
BAKER, ET AL.: A UNIFORMITY TRIAL ON UNIRRIGATED BARLEY 267