SELF-AND CROSS-INCOMPATIBILITY IN
BLACK CHERRY (Prunus serotina)
DONOVAN C' ,*:, L
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
The author is grateful to his chairman, Dr. Earl S. Horner,
or the interest he took and aid he gave in the preparation of this
issertation. The kind assistance of the other members of the author
committee, Dr. Harry H. Griggs, Dr. Thomas E. Humphreys, Dr. Paul L.
fahler, and Dr. Stanley C. Schank, is also appreciated.
Special thanks is accorded to Dr. Ray E. Goddard, the author's
visor, for his guidance, encouragement, and help when it was most
needed; to Mr. Ray K. Strickland, who completed one phase of the
field work when the author was incapacitated; and to the University
f Tennessee Atomic Energy Commission Agricultural Laboratory and its
wo laboratory technicians, Carroll Shell and Louise Russell, for use
f facilities and aid in the microtechnique phase of the work.
Finally the author is indebted to the Tennessee Valley
ithority for allowing him time to do the research and for providing
id and equipment,
TABLE OF CONTENTS
ACKNOWLEDGMENTS ...................... ii
LIST OF TABLES ...... . . . ..... vi
LIST OF FIGURES .................. . . .viii
INTRODUCTION .................. . . 1
Early Work .. . . . . . . . . ... . 4
Attempts at Interpretation ............... 4
Systems of Incompatibility ... . . . . . 5
Heteromorphic Plants ...... . . . . .. 6
Homomorphic Plants ...... . . . . . .. 7
Effect of Polyploidy on Incompatibility . . . . . 8
Variations of Incompatibility . . . . . . . .. 11
Unilateral Incompatibility .. . . . . . 11
Somatoplastic Sterility ... . . . . . . 11
Biochemistry of Incompatibility . . . . . . 12
Incompatibility in Plums and Cherries . . . . ... 14
Immature Fruit Drop in Cultivated Cherries . . . . . 16
Techniques ............. .... ...... 17
Pollen Collection, Germination, and Storage . . ... 17
Artificial Pollination .... . . . . . . 18
Time of Flower Removal .... . . . . . . 19
Flower Killing and Fixation .. . . . . . 20
Slide Preparation .... . . . . . . . 20
MATERIALS AND METHODS . . . . . . . . . 21
Plant Materials . . . . . . . . . . 21
Principal Studies ...................... 21
Pollen Tube Length Study--Gainesville. . . . . 24
Fruit Set Studies . . . . . . . . . . 27
Techniques and Procedures . . . . . . . 28
Pollen collection ................ .. 28
Pollination techniques . . . . . . . . 30
Tagging for identification . . . . . . ... 32
Bagging for isolation ................... 32
Flower killing and fixation . . . . . . .. 33
Slide preparation ........ ....... .. 33
RESULTS AND DISCUSSION . . . . . . . . . . 35
Preliminary Trials ..................... 35
Date and Duration of Flowering . . . . . .. 35
Flower Location on the Tree . . . . . . . .. 36
Characteristics of Flowers and Racemes . . . . ... 36
Natural Crossing . . . . . . . . . . 36
Natural Selfing ................. ..... 37
Pollen Germination Tests . . . . . . . ... 38
Bagging . . . . . . . . . . . . . 38
Chromosome Count . . . .. 39
Pollen Tube Length Study . . . . . . . 39
Pollen Germination ..... ........... ... 39
Summary of results ..................... 47
Pollen Tube Extension ................ ...... 47
Summary of results ..... .. ........ ..... 56
Fruit Set Studies . . . . . . . ... . . 57
Prevention of Insect Pollination without Bagging--
Gainesville . . . . . . . . . . . 57
Selfing and Crossing--Gainesville . . . . . 57
Selfing and Crossing--Norris . . . . . . . . 59
SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS . . . .... 64
Summary and Conclusions .................. 64
Recommendations for Breeding and Tube Length Studies . .. 68
Recommendations for Future Work . . . . . . .. 71
APPENDIX . . . . . . . . . . . . . 73
BIBLIOGRAPHY . . . . . . . . . .. . 83
BIOGRAPHICAL SKETCH . . . . . . . .. . . 92
LIST OF TABLES
1. Effects of Polyploidy on Incompatibility in Several Auto-
tetraploids Arising under Controlled Conditions, by Species
and Author (from Lewis, 1949) ......... . .. 10
2. Pollination Treatments, Gainesville Studies, by Female
Parent . . . . . . . . . . . . 25
3. Pollination Treatments, Norris Fruit Set Study, by Female
Parent and Pollen ................. .. 29
4. Comparison of Pollen Treatments with Respect to Pollen
Germination, Tree Group A . . . . . . . 43
5. Comparison of Pollen Treatments with Respect to Pollen
Germination, Tree Group B . . . . . . . 44
6. Rank of Removal Intervals in Tree Groups A and B with
Respect to Pollen Germination .. . 46
7. Percent of Pollen Tubes in the Upper Third of the Style that
Reached the Middle Third, Over All Trees and Removal
Intervals, by Pollen Treatment and Tree . . . . .. 49
8. Percent of Pollen Tubes in the Upper Third of the Style that
Reached the Lower Third, Over All Trees and Removal
Intervals, by Pollen Treatment and Tree . . . . 50
9. Percent of Pollen Tubes in the Upper Third of the Style that
Reached the Middle Third, Over All Trees and Pollen
Treatments, by Removal Interval and Tree ... . . 54
10. Percent of Pollen Tubes in the Upper Third of the Style that
Reached the Lover Third, Over All Trees and Pollen
Treatments, by Removal Interval and Tree . . ..... 55
11. Percent Fruit Set among Cross-Pollinated Flowers, by Pollen
and Mother Tree, Gainesville . . ....... 58
12. Percent Fruit Set among Self- and Cross-Pollinated Flowers,
by Pollen and Mother Tree, Norris . . . . . 60
LIST OF FIGURES
1. Location of Gainesville, Florida, test trees . . . 22
2. Location of Norris, Tennessee, test trees .. . . ... 23
3. Black cherry chromosomes at metaphase, not well spread . 40
4. Well spread black cherry metaphase chromosomes . .. 40
5. Abnormal pollen tube, upper one-third style, Tree 1 normal
self, 7 days . . . . . . . . . . . 52
6. Abnormal pollen tube, middle one-third style, Tree 6 x
Tree 8, 5 days (lower tube) ... ............ 52
7. Embryo from selfed fruit, Norris . . . . . . . 62
8. Fence row that included study trees, Norris . . . . 74
9. Heavily fruiting tree, Norris .. . . . . . . .. 75
10. Typical tree, Norris ... ................. 75
11. Fruit set from a cross, Norris ... . ....... 76
12. Fruit set from a self, Norris . . .. . ..... 76
13. Fruit set from a self, Norris . . . . . . . 77
14. Typical bag, Gainesville .... . . . . . 78
15. Petal removal prior to pollination, Gainesville ....... o 79
16. Pollen tubes, upper one-third style, Tree 4 no pollen,
3 days . . . . . . . . . . . . 80
17. Pollen tubes, stigmatic surface, Tree 1 no pollen, 5 days . 80
18. Pollen tubes, stigmatic surface, Tree 4 no pollen, 5 days
19. Pollen tubes, lower one-third style, Tree 6 normal self,
5 days .
20. Pollen tubes,
21. Pollen tubes,
5 days .
lo.er one-third style, Tree 6 x Tree 2,
lower one-third style, Tree 6 x Tree 2,
lower one-third style, Tree 6 x Tree 4,
. . . . . . . . . . . ..
22. Pollen tubes, lower one-third style, Tree 6 x Tree 5,
5 days . . . . . . . . . . ...
23. Pollen tubes, lower one-third style, Tree 6 x Tree 5,
7 days . . . . . . . . . . .
24. Pollen tubes, lower one-third style, Tree 6 x Tree 7,
5 days . . . . . . . . . ..
25. Pollen tubes,
5 days .
26. Pollen tubes,
27. Pollen tubes,
5 days .
lower one-third style, Tree 6 x Tree 7,
junction ovary and style, Tree 7 x Tree 6,
upper one-third style, Tree 7 x Tree 8,
Black cherry (Prunus serotina, Ehrh.) is the largest native
American cherry species in height and diameter and the only one that
provides large amounts of material for woodworking. On better sites it
may reach over 100 feet in height and upwards of five feet in diameter.
From earliest colonial times cherry has been a favorite furni-
ture wood. Its dark reddish-brown heartwood finishes to a beautiful
and distinctive luster giving it a quality look. Famous English and
American furniture craftsmen such as Duncan Phyfe, Hepplewhite, and
Chippendale used cherry for many of their creations. Those which re-
main today still retain their beauty and utility.
Cherry has been used in many ways from agricultural implements
to wartime tank construction, but in recent years its use has become
more and more restricted to high-value products because of its popu-
larity and diminishing supply. As late as 1940 the two outstanding
uses were furniture and backing blocks for use in printing. Today
lumber and veneer for furniture manufacture lead the list.
It is not a slow-growing species but makes good growth only on
the best of sites. On the Allegheny Plateau where sites are best,
dominant and codominant trees average about 2.5 inches in diameter
growth per decade between the ages of 13 and 33 (Defler, 1937, cited by
Hough, 1960). Since these better sites are limited in number and the
demand for cherry has been high, the best stands have long since been
depleted in all but this small section of the natural range, To make
matters worse, regeneration has come from low-quality trees, leaving
less desirable crop trees for future use,
The present outlook is not bright. Demand for high-quality
material is rising despite competition from synthetics, while the sup-
ply is diminishing. Furniture manufacturers realize this and are pro-
moting regeneration with better planting stock. It is a late start,
but progress is being made.
If a sufficient supply of good-quality material is to be assured,
future crop trees must be descendants of superior-quality trees now
bearing seed. High grading has removed almost all of these except
those in a small area in the Allegheny Mountains of New York, Pennsyl-
vania, and West Virginia known as the commercial range. But a few iso-
lated individuals of high quality can still be located in relatively
inaccessible and protected areas in most of the range. Along with
better trees in the commercial range, these are the trees that tree
improvement workers are using as parents in the first phases of black
Any breeding program must be adjusted to fit the needs of the
species studies. It appears that special consideration must be given
self- and cross-incompatibility in the case of black cherry. Prelimi-
nary evidence indicates that in the Gainesville, Florida, area the
species is probably completely self-incompatible under natural condi-
tions, but also that under certain artificial conditions, it might be
selfed. Near relatives of black cherry, the sweet and sour cherries
and the plums (Prunus avium, Prunus cerasus, and Prunus domestic,
respectively) all have a similar incompatibility system--one that
1Removal of only higher quality trees during harvest operations,
prevents nearly all selfing and also some crossing in certain species
but that allows for selfing in others.
Since forest tree improvement programs usually include estab-
lishment of clonal seed orchards in which natural cross-pollination is
expected along with hand pollinations for production of progeny test
material, any incompatibility system in the species used must be un-
covered and cross-incompatible combinations pointed out. Also, such
a system must be known and overcome if any close inbreeding program is
to be successful.
The main purpose of this work is to determine if there is an in-
compatibility system in black cherry and if there is, how it compares
to the one in the other Prunus species. But since black cherry breed-
ing is in its early stages, new techniques will have to be adapted or
new ones developed to accomplish the primary goal. So as a secondary
goal several different studies aimed at adapting and developing tech-
niques have been incorporated. These include observations on flower
and flowering characteristics, natural and artificial pollination,
pollen collection, isolation techniques, and determination of presence
or absence of ploidy.
Most people give Kolreuter (1964, cited by East and Park, 1917)
credit for the discovery of what he called self-sterility in plants.1
He found three Verbascum phoeniceum plants that failed to set seed with
their own viable pollen but seeded readily when pollinated by V. blat-
aria, V. n phlomoides, and V. lychnitis Darwin (1876) later
verified that V. phoeniceum was self-infertile along with V. nigrumbut
found two related species, V. thapsus and V. lychnitis, that were self-
fertile. Probably the next work on the subject was Herbert's (1837,
cited by East, 1929). He found self-infertility very common among
hybrids of the Amaryllidaceae and also that the non-hybrids Zephyranthes
carinata and Hippeastrum aulicum were self-infertile. A little later
Gartner (1844, cited by East and Park, 1917) found several additional
examples in other species including Dianthus aponicus, D, barbatus,
and Lobelia fulgens,
Several observers, including Mowbray (1830, cited by East, 1929)
and Munro (1868, cited by East, 1929) reported self-infertility in
Passiflora. Munro's work was especially important. He found self-
infertile plants that were sometimes cross-fertile and sometimes
1Hereafter I shall use the term incompatibility when pollen and
ovules were known to be functional and infertility when the reason for
non-fertilization was at the time unknown.
cross-infertile with others of the same species and even the same gen-
eration. Some plants of P. alata crossed easily with their own progeny.
In Oncidium, Scott (1865) showed that stylar tissue was pene-
trated freely by tubes from self pollen of self-infertile plants,
though fertilization did not occur. Muller (1868, cited by East, 1929)
corroborated Scott's findings and also showed each self-infertile
Oncidium flexuosum plant crossed with any other plant.
Attempts at Interpretation
Most work prior to the 20th century dealt with discovery of new
cases. Many were reported in addition to the ones already mentioned
including those by Bidwell, Bernet, Rawson (Darwin, 1876, cited by East,
1929); Lecoq (1862); Focke (1890, 1893, cited by East and Park, 1917);
and others. Work in the early 1900's uncovered many new examples of
incompatibility, including those reported by Backhouse (1911); Gardner
(1913); Raves (1921); Sutton (1918); Detjen (1916); Heribert-Nilsson
(1916, cited by East, 1929); Wada (1923, cited by East, 1929); Dahlgren
(1922, cited by East, 1929); but was more concerned with interpretation
of the phenomenon (East, 1929). Several noteworthy examples appeared.
Jost (1907) duplicated Hildebrand's (1866) earlier experiments
on Corydalis cava. Where Hildebrand had found absolute self-
infertility, Jost observed some self-fertility. Jost found that the
different behavior of these plants and of Secale cereal and Lilium
bulbiferum after self- and cross-pollination, was a difference in
pollen tube growth. In Secale, tubes penetrated the micropyle in eight
hours after cross-pollination but in 24 hours only reached the base of
the pistil after self-pollination.
Correns (1912) showed that pollen of self-incompatible Cardamine
pratensis germinated on stigmas of self-pollinated flowers but produced
only short tubes that did not penetrate the stigma. However, cross-
pollination produced pollen tubes that penetrated the upper ovary after
Correns (1912) and Compton (1912) suggested, as Jost (1907) had
previously, that diffusible substances which stimulated or retarded
pollen tube growth after cross- or self-pollination, respectively, were
present in the pistil. Compton felt that self-incompatibility might be
controlled by agents similar to those which govern immunity or suscep-
tibility in plants and animals.
Results of these and many earlier experiments were explained
when East and Mangelsdorf (1925) presented a new interpretation of self-
incompatibility in Nicotiana. They called it the oppositional factor
system. Filzer (1926) independently discovered the same system working
with Veronica syriaca. This work cleared up some puzzling aspects of
incompatibility and paved the way toward discovery of several additional
Systems of Incompatibility
There are six known systems of incompatibility. Two are found
in heteromorphic plants and four in homomorphic types.
The first system (I) was first described by Darwin (1877, cited
by Dnyansagar, 1963) in the genus Primula and is known as distyly. Two
different flower types are borne on different plants. Thrum plants have
short styles and high anthers; pin plants have long styles and short
anthers. Compatible pollinations occur only between the two different
types. Genetic control is by one gene with two alleles, with sporo-
phytic determination of the pollen reaction. This type control has
been found in all distylic species so far analyzed (Dnyansagar, 1963)
including Primula sinensis (Bateson and Gregory, 1905), P. acaulis
(Gregory, 1915), P. officinalis (Dahlgren, 1916, cited by Lewis, 1949),
P. viscosa and P. hortensis (Ernst, 1936, cited by Lewis, 1949), P. ob-
conica (Lewis, 1949), Pulmonaria agustifolia (Darwin, 1876, cited by
Lewis, 1949), and Fagopyrum esculentum (Dahlgren, 1922; Garber and
The second system (II) is a more complex form of heterstyly char-
acterized by short-, medium-, and long-styled flowers. Each flower has
anthers at two different levels, each different from that of the stigma.
Darwin (1877) showed that compatible pollinations are between stigmas
and pollen from anthers at the same level. Fisher and Mather (1943) and
Fisher and Martin (1947) found that genetic control was by two loci,
each with two alleles which they called Ss and Mm. Long-styled plants
are ssmm; medium-styled, ssM ; and short-styled, S mm or S M (Lewis,
Under the homomorphic class are two subclasses, gametophytic and
sporophytic. Each includes two of the six incompatibility systems,
In the gametophytic subclass, mating type of pollen grains is
determined gametophytically and alleles of incompatibility genes act
individually in the style. Stylar pollen tube inhibition is the result
of incompatible matings (Crowe,1964).
In one system (III) genetic control is by multiple alleles at
one locus as in Oenothera (Crowe 1964). Prell (1921) first proposed
this type and East and Mangelsdorf (1925) used it to describe incompat-
ibility in Nicotiana sandarae. It was also used by Crane and Lawrence
(1931) to describe incompatibility in sweet cherry (Prunus avium),
The other gametophytic system (IV) is controlled by multiple
alleles at two loci as in Secale (Crowe, 1964). Present in several
grasses (Lundquist, 1956, 1961a, 1961b, 1964a, 1964b; Hayman, 1956), it
is also found in at least two members of the Solonaceae, Solanum pin-
natisectum and Physalis ixocarpa (Pandy, 1957, 1960). Pandy (1962)
concluded that the two-locus system is a derivative of the one-locus
system and suggested that the second locus is a duplication arising
from a structural chromosome change.
Under the sporophytic subclass, mating type of pollen grains is
determined sporophytically and alleles of incompatibility genes express
either dominance or individual action in both male and female parts.
Site of the incompatibility reaction differs in the two systems included
(Crowe,1964). In the first system (V) genetic control is by a single
locus with multiple alleles. Pollen tubes are inhibited in the style
as in Parthenium. In the second (VI) genetic control is by one locus
with multiple alleles and incompatibility is expressed between gametes
after fertilization as in Theobroma. Type V is common in dicotyledons,
but Type VI has been found only in Theobroma and Callistemon among the
Effect of Polyploidy on Incompatibility
An increase in chromosome number of an organism produces a
drastic upset in both the genic balance and the physiological processes
they control. In self-incompatible species studied, the effect of dou-
bling the chromosome number on incompatibility ranged from none to a
change to self-compatibility (Lewis, 1949). Effects on these species
are summarized in Table 1.
Lewis (1949) stated that the key to variable behavior of tetra-
ploids is found in tetraploid Oenothera organensis. In its style the
four S alleles, even when different, all operate without interaction to
inhibit haploid pollen carrying the same alleles. In diploid pollen
genic balance is unchanged if both alleles are alike; the pollen behaves
like haploid pollen. However, complex interactions occur between unlike
alleles, and diploid grains may show two different kinds of genic inter-
action depending upon the pair of alleles involved (Lewis, 1949).
With certain pairs of alleles, neither functions efficiently be-
cause of competitive interaction. In these cases there is a much re-
duced incompatibility reaction, and pollen tubes are only partially
inhibited even in styles carrying both alleles. The result is either
weak self-incompatibility or complete self-compatibility depending upon
the strength of the original reaction (Lewis, 1949).
With other pairs of alleles there is no competitive interaction,
but one allele is dominant to the other. Such pollen functions like it
is homogenic for the dominant allele. Plants producing only this type
pollen are completely self-incompatible (Lewis, 1949).
Competition and dominance in the diploid pollen grain explain
most effects of polyploidy found in other species. Effects in species
that show no change can be explained by dominance or lack of inter-
action, either dominant or competitive. Species that show a change to
self-compatibility must exhibit competitive interaction (Lewis, 1949).
Effects of Polyploidy on Incompatibility in Several
Autotetraploids Arising under Controlled
Conditions, by Species and Author
(from Lewis, 1949)
Species tetraploid Effect Author
Antirrhinum molle Colchicine Self-compatible Straub, 1941
Campanula ersicifolia Spontaneous Self-compatible Gairdner, 1926
Petunia xillaris?)* Colchicine Self-compatible Stout & Chandler
Pyrus communis Spontaneous Self-compatible Crane & Lewis
P. malus Progeny of Self-compatible Johansson, 1945
Solanum (3 species) Colchicine Self-compatible Livermore &
Trifolium reopen Colchicine Self-compatible Atwood, 1944
Ananas sativus Colchicine No effect Kerns & Collins
Brassica campestris Colchicine No effect Howard, 1942
B rapa Colchicine No effect Howard, 1942
Oenothera organesis Colchicine Slight increase Lewis, 1943
O rmbipetala Colchicine No effect Hecht, 1944
Raphanus sativus Colchicine No effect Howard, 1942
Taraxacum kok-saghyz Colchicine No effect Bannan,. 1946
* The species used was probably the garden form, P. violacea.
Variations of Incompatibility
This occurs when pollen tubes of a self-compatible species fail
to penetrate the style of a self-incompatible species, but the reverse
cross is compatible. Such results have been recorded by Mather (1943)
with Petunia, McGuire and Rick (1954) with Lycopersicon, Harrison and
Darby (1955) with Antirrhinum and Pandy (1962) with Solanum. Lewis and
Crowe 0958) believe this interspecific incompatibility may be con-
trolled by S alleles and that incompatibility substances in the pollen
are superimposed upon an essential pollen growth substance. Further,
these growth substances in pollen from self-compatible species are not
protected by superimposed groups and are inactivated by "antibodies" in
the style. Martin (1964, 1967) and Grun and Aubertin (1966) have
pointed out that other genes account for unilateral incompatibility in
some crosses. And Grun and Aubertin (1966) feel that mechanisms other
than pollen tube inhibition may cause the phenomenon.
This is found mainly in interspecific crosses and results in
decreased endosperm growth and eventual seed death in most cases.
Cooper and Brink (1940) found that in the cross Nicotiana rustic x N,
glutinosa death was caused by endosperm starvation due to an overgrowth
of the nucellus stopping the food supply. The nucellus in this case
blocked the gap adjacent to the chalazal pocket through which nutrients
reach the endosperm.
Cooper and Brink (1940) stated that Bradbury's (1929) selfed
sour cherries (Prunus cerasus) exhibited one or more of the
post-fertilization features of somatoplastic sterility. They felt that
this might have been the result of inbreeding,
Biochemistry of Incompatibility
Linskens (1963) lists two types of self-incompatible plants from
an experimental point of view:
1. Those in which the incompatible reaction occurs on the
2. Those in which the reaction occurs during growth of pollen
tubes through the style.
In Type I the incompatibility barrier is probably the cuticle of
the stigma (Christ, 1959, cited by Linskens, 1963). In all cases inves-
tigated, pollen from plants with a stigma cuticular layer also had a
cutin-breaking enzyme system (Linskens, 1963). Kroh (1956, cited by
Linskens, 1963) and Tatebe (1959, cited by Linskens, 1963) obtained
normal seed set in incompatible pollinations of this type by bypassing
the cuticular layer. Kroh also showed that pollen from this type plant
placed on a cross-compatible plant for a short time then transferred to
a self penetrated the self stigma. Linskens (1963) stated that this is
evidence for a cutinase enzyme system linked to the incompatibility
barrier and that the cutinase can be activated irreversibly.
In Type II incompatibility, pollen tube growth is inhibited
within the style. Schlosser (1961, cited by Linskens, 1963) found
abnormal pollen tube behavior in incompatible crosses when compared
with compatible crosses. In about 50 percent of tubes studied the
generative nucleus did not divide, and the vegetative nucleus dis-
appeared just after germination without fulfilling its function,
There was also a disturbance of the normal carbohydrate metabolism
evidenced by a large deposit of fibrils on thickened walls, branched
tube tips, and an increased number of callose plugs (Linskens and
Esser, 1957; Schlosser, 1961; Tupy, 1961; all cited by Linskens, 1963).
Inhibited tubes also have a high respiration rate the first hours of
growth (Linskens, 1955, cited by Linskens, 196,3). And the fluorescence
patterns of style extracts taken after incompatibility reactions show
characteristic patterns after chromatographic separation; unidentified
substances that inhibit tube growth appear (Linskens, 1963).
Linskens (1963) believes pollen tube inhibition may be caused by
an antigen-antibody type reaction between a specific pollen protein and
a stylar protein with homologous specificity. Direct evidence to
support this conclusion comes from Lewis (1952) who found antigens that
were specific to different pollen genotypes in Oenothera and by Linskens
(1960) who found the same thing with protein in Petunia styles. Also
Linskens (1958, 1959, cited by Linskens, 1963) showed that a complex is
formed between pollen and stylar protein after an incompatibility
reaction. Selection of specific antibodies occurs during pollen devel-
opment as a result of intensive metabolic exchange between sporogen and
female tissue in the same flower. After pollination, antigen of the
pollen tube meets already selected antibodies in female tissue (Linskens,
1962, cited by Linskens, 1963).
Lewis (1960) showed that the S gene responsible for production
of the protein causing incompatibility has two cistrons. One controls
specific grouping of protein and the other controls the half-molecule
carrier responsible for protein activity in pollen and style.
According to Lewis mutations and recombinations are all changes in the
Linskens (1963) believes that S protein action occurs on the
pollen tube surface. Lewis' (1960) data support this hypothesis. He
found that S protein diffused out of both macerated and intact pollen.
This means that S protein either acts on the tube surface or outside in
stylar tissue. But since incompatible and compatible tubes in the same
style at the same time do not influence one another, it is likely that
S protein acts on the tube surface (Linskens, 1963).
Incompatibility in Plums and Cherries
M. B. Crane (1925) first reported self- and cross-incompatibility
in sweet cherry (Prunus avium). He showed that varieties involved could
be assigned to groups within which all but a very small percent of self-
and cross-pollinations failed. In one test over 18,000 selfed flowers
produced only 21 fruits. Other workers including Wellington (1926),
Tufts, Hendrickson, and Philp (1926), Sachoff (1931, cited by Crane
and Brown, 1937), Schanderl (1932, cited by Crane and Brown, 1937),
Gardner (1913), Schuster (1922), Florin (1924, cited by Crane and Brown,
1937), Einset (1932), Kobel and Steinegger (1933, cited by Crane and
Brown, 1937), Wenholz (1936), Sprenger (1908, cited by Crane, 1925),
and Hooper (1924) also showed that sweet cherry selfing is very rare.
Later Crane and Lawrence (1931) showed that incompatibility in
sweet cherry was comparatively simple with self-incompatibility the
rule and cross-incompatibility common, always reciprocally expressed.
But in the plum (Prunus domestic) incompatibility occurring in one way
of a cross and compatibility in the other was common. Darlington
(1930) pointed out that the degrees of fertility and complex results in
work with Prunus domestic were undoubtedly due to its hexaploid makeup
and manner of chromosome pairing. Bradbury (1929) found sour cherry
varieties both self- and cross-compatible, but selling produced a lover
percent set than did crossing.
Cytological studies by Evert (1922, cited by Crane and Brown,
1937) established the basic chromosome number in Prunus as 8 and that
all varieties of the sweet cherry were diploids (2x = 16). Kobel's
(1927, cited by Crane and Brown, 1937) work agreed with this. The sour
cherries, P. cerasus and Dukes, P. cerasus x P. avium are tetraploid
(Kobel, 1927; Darlington, 1928) as is P. cantabrigiensis (Faberge,
unpublished, cited by Crane and Brown, 1937). Black cherry (P.
serotina) is listed as tetraploid by Darlington and Ammal (1945).
Crane and Lawrence (1931) concluded that sweet cherry incompati-
bility was the oppositional factor type discovered in Nicotiana by East
and Mangelsdorf (1925) and discussed earlier in this paper under
incompatibility system III.
Roy (1938) compared (1) normal self-pollination, (2) self-
pollination in which styles were treated with growth promoting
phenylacetic acid before pollination, and (3) cross-pollination with a
compatible variety in sweet cherry. No significant differences were
observed between (1) and (2). Tubes which penetrated the style became
arrested in stylar tissue, and many appeared to swell up at their ends.
Tubes in (3) grew much more rapidly than in (1) and (2).
Earlier Afify (1933) studied pollen tube growth in cherries and
plums and distinguished between different pollen genotypes by behavior
of pollen upon germination and growth of tubes in the style. In sweet
cherries he found five types of pollen, those that: (1) failed to
germinate, (2) produced very short tubes which finally bent upwards and
ceased growth, (3) produced tubes that traveled about one-fourth the
stylar length, (4) produced tubes that traveled about one-half the
stylar length, and (5) traveled the length of the style and effected
So in Prunus there is a system of incompatibility which is
seemingly the same as that in several other genera. It is variable
according to chromosome number but is characterized by viable pollen
which germinates and produces tubes that penetrate the stigma, and by
slowed growth of tubes carrying the same alleles as those found in
Immature Fruit Drop in Cultivated Cherries
All cultivated varieties of cherry characteristically exhibit
three waves of fruit drop. The first occurs shortly after peak
flowering and is followed shortly by a second. The third drop is often
called the June drop since it usually occurs in that month.
Considerable work has been done to uncover the reason or reasons
for these drops including that of Detjen (1926) and Tukey (1933), but
Bradbury (1929) gives the best discussion of the problem. In her work
with sour cherry she made the following observations:
1o The first drop cannot be attributed to lack of pollination or
to failure of pollen tubes to reach the ovarian cavities. From 82 to 99
percent of first drop fruits had been pollinated and, in 76 to 98 per-
cent, pollen tubes had reached the ovarian cavities.
2. Degeneration probably begins in ovules of first drop fruits
before normal fertilization time. Both ovules were shriveled and
pollen tubes were growing at random in upper ovarian cavities,
30 Some unpollinated fruits were in the first drop, but the
large number of pollinated fruits in the first drop and unpollinated
fruits that were not in this drop indicate that pollination or lack of
it does not fully account for fruit development past the first drop.
4. A large part of the second drop is not due to lack of
fertilization. Embryos were present in 41 percent of dropped fruits
and 95 percent either contained embryos or pollen tubes in ovarian
cavities or ovules.
5. Partial embryo development has usually taken place in third
drop fruits of sour cherry.
Bradbury concluded that unfavorable nutritional conditions
probably play a major part in bringing about arrested development and
dropping of sour cherry fruits. Tukey (1933) agreed since he success-
fully cultured embryos from dropped fruits. Stage of development is
influenced not only by foods but also by pollination, fertilization,
and embryo development. Finally, Bradbury concluded that physio-
logical conditions in fruits leading to abscision layer formation
possibly are the same whether brought about by unfavorable nutritional
conditions or by lack of pollination or fertilization.
Pollen Collection, Germination, and Storage
Schuster (1925) tried combing sweet cherry anthers off the
flower and was successful, but found brushing flowers over a screen an
even better method. Anthers snapped off and fell through the screen.
Schuster dried the collected anthers in open dishes until all pollen
was liberated and dry.
All work with cherry pollen germination testing has been with
sweet and sour horticultural varieties. Highest germination counts
were observed in various sugar solutions. Pfundt (1910) used 20-30
percent, Crane and Brown (1937) used 8-12, and Raptopoulas (1939),
10-25 percent sugar. These workers all got 50 percent or better
germination (Raptopoulas, 1939).
No reference was found that gave a method of pollen storage for
any species of Prunus, whether successful or not. Fresh pollen was
evidently used in all work.
East and Mangelsdorf (1926) disagreed with the old idea that
stigmas are receptive only when they are secreting a substance assumed
to promote pollen germination. They pollinated as early as three days
before normal flower opening in Nicotiana without abnormal effect.
Crane (1925) did the same with Prunus avium. Schuster (1925) normally
pollinated cherry as soon as stigmas showed drops of liquid on their
surfaces but pointed out that pollination was successful 48 hours or
more before usual flower opening. Stigmas were either receptive before
normal opening or some pollen remained on the stigma and germinated
later when stigmas were receptive, He found that a flower past matu-
rity showed a slight reddening of stylar tissue just above the ovary.
East (1919) showed that selfing could be successful in Nicotiana
either by pollinating buds or by pollinating late in the flowering
season. He concluded that an inhibitor was formed during the life of
the flower, but its production coincided with flower opening, Bud
pollination avoided the inhibiting effect and allowed tubes to grow
normally before being slowed by the inhibitor. The longer period for
tube growth allowed fertilization to be accomplished. Smith (1926)
showed that end-of-season fertility in Nicotiana is due primarily to the
last few flowers remaining on the stem longer than usual,
The perfect flowers in species of Prunus make emasculation a
consideration to remove any doubt about penetration of self-pollen
tubes, Schuster (1925) emasculated as does Sharp (1964) with Prunus
avium, Sharp simply pinches off stamens between his thumb and
forefinger. This is done mainly to avoid mixing pollens since Prunus
avium is highly self-incompatible. Schuster (1925) used a camel's-hair
brush to apply pollen. Sharp (1964) uses a pencil eraser.
Schuster (1925) reported that emasculating and leaving flowers
exposed was preferred to bagging. He found that when bagged, flower
pistils often came in contact with the bag surface and were either
heat- or frost-killed depending upon temperature. He also felt that
insects would not be attracted to flowers without petals and did some
work without bagging. But Schuster (1925) also showed that bagging
added to flower longevity provided no contact with the bag itself
occurred. So bagging9 by prolonging flower life, could aid in ob-
taining fertilization from incompatible matings.
Time of Flower Removal
Roy (1938), in his histological study of pollen tube growth in
Prunus, removed flowers four days after pollination. Whether in selfs
or compatible crosses more than half of all pollen tubes penetrating
the stigma failed to grow more than 0.5 mm. A few did reach 5.0 mm
in selfs and 6.5 mm in compatible crosses.
Flower Killing and Fixation
Roy (1938) used Flemming's and Karpechenko's fluids in a mixture
of equal parts of 95 percent alcohol, glacial acetic acid, and lactic
acid. For work where fine detail and lifelike appearance are not
required, Jensen (1962) suggests the use of FAA--formaldehyde, alcohol,
acetic acid--as a killing and fixing agent. Its action is fast and
material can be stored in it some time before use.
Standard techniques for slide preparation are found in any one
of several manuals or texts. Jensen's (1962), though much smaller than
Johansen's (1940) standard text, adequately covers the process and
gives more up-to-date information on new equipment. Jensen's suggested
variation of replacing tertiary butyl with normal butyl alcohol in the
dehydration phase is appropriate when temperatures may drop below 60
F, the freezing point of tertiary butyl.
Fine points of the process including use of new equipment were
reviewed with C. Shell (personal communication, 1967) and L. Russell
(personal communication, 1967) of the UT-AEC Laboratory in Oak Ridge,
Tennessee. They gave the author instruction in the entire process from
dehydration through staining and coverslipping.
MATERIALS AND METHODS
A genetically controlled incompatibility system is more likely
to show up in crosses among close relatives than among non-relatives.
So a parent-offspring grouping was sought as test material. One of the
most likely arrangements of this type is an old tree with young flower-
ing trees under or very near it. A careful search around Gainesville,
Florida, located such a situation in an area uncut for some time which
had what appeared to be one or more offspring-parent groupings present,
Two older trees--possibly siblings themselves--and six younger ones
were selected as study trees. Locations of trees with respect to each
other and Gainesville, Florida, are found in Figure 1.
Six companion trees were selected in the Norris, Tennessee, area
to partially duplicate the Gainesville study. These were all young
trees found in a single fence row. Locations of these trees with re-
spect to each other and Norris, Tennessee, are found in Figure 2,
One way that incompatibility has been shown in Prunus is to
observe lengths and rates of growth of pollen tubes after different
types of pollination. Another method is to make pollinations, then
check results by observing fruit set. Both methods were used.
4- -m o ,m
--- c u------ ------ (
E 0 E -
to t 0 o to 0 ) to 0
HI H -
/~^ ^ ~^ V '<> ~'~~ ~*~*^ ^' *- -^- ^s &
C~ ~ (.cJ+ fE):A-^ St
Pollen Tube Length Study--Gainesville
Two groups were made of the eight selected trees because of
differences in flowering times. A breakdown of treatments for each
Gainesville tree group is given in Table 2. On group A the entire
scheme was reproduced three times to allow for removing flowers at 3-,
5-, and 7-day intervals after pollination. On group B two reproduc-
tions were used for 5- and 7-day removal intervals.
Pollination of young buds might be effective in avoiding an
incompatibility reaction. Such was the theory of East (1934), who be-
lieved that an inhibitor was formed within 24 hours of flower opening
in Nicotiana. Presumably, if germination occurred, tubes from such
crosses would be well down the style before an inhibitor was formed and
could possibly effect fertilization.
The no-pollen treatment, besides giving data on natural selfing,
was also meant to serve as a yardstick by which other pollinations
could be evaluated. This was required since flowers were not
emasculated. Hopefully, self tubes could be differentiated from non-
selfs in crosses based on their performance in natural-self checks.
This was not considered too difficult since a cross would be made in
an individual flower some time before pollen release in the same
Some question arose as to whether results of the tube length
portion of this work could be analyzed statistically. Unless most of
the applied pollen which would eventually germinate did germinate within
a short interval, tube length analysis would be confused. Tubes from
applied pollen would be indistinguishable from those of self pollen
from the same flower. Other factors which might affect an analysis
Pollination Treatments, Gainesville Studies,
by Female Parent
Group A Group B
Female parents Female parents
Pollen treatment 1 3 4 5 6 7 8
No pollen + + + -
Early self + + + -
Normal self + + + + + + +
Pollen 1 + + + + + +
Pollen 2 + + + + + + +
Pollen 3 + + -
Pollen 4 + + -+ + + +
Pollen 5 + + +
Pollen 6 + + +
Pollen 7 -+ + +
Pollen 8 + + +
- No treatment
1. No pollen: flowers bagged before opening, no pollen applied.
2. Early self: self pollen applied on flowers that would not
normally open for at least 24 hours.
3. Normal self: self pollen applied on flowers that would open
within one to two hours.
4. Crosses: cross-pollen applied to flowers that would open
within one to two hours.
include stigma receptivity and loss of an exceptional number of
observations. But the necessary elements of statistical design were
included so that the material could be analyzed if conditions were
favorable. Both tree groups were factorial experiments within a
completely randomized design. Each treatment was replicated twice.
After the different day intervals had passed, flowers were
removed, killed, and fixed in formalin-acetic-alcohol. Later, serial
slides were prepared from longitudinally cut pistils, and data were
taken on tube characteristics.
Since computer analyses would likely be used, all data were
taken on IBM data sheets. Each flower from each treatment unit was
identified as to its parent, pollen number, removal interval, location
(bag) number, and flower number. Actual counts of tube sections were
made on the upper one-third of the style and/or just above the stig-
matic surface, number within the upper one-third of the style (Group I),
number in the middle one-third of the style (Group II), number in lower
one-third of the style (Group III), and number in the ovary. Addition-
ally, numbers in Groups II and III were expressed as percentages of
Group I to exert a standardizing influence on the data. So, a total of
six variables appeared on each flower card.
Tube sections--pieces of tubes--were counted rather than actual
tubes, because longitudinal microtoming of styles produced sectional
pieces of the meandering tubes. These pieces were considered propor-
tional to actual pollen tube members and were the only means of
relating tube characteristics in one flower with those in other
Data from the six categories mentioned previously were
transformed. The first four were transformed by adding 0.5 to each
value, then taking its square root. The last two categories were ex-
pressed as percent values so were converted to are sin values before
Each of the six variables for each tree was subjected to a
separate analysis of variance using the Least Squares Analysis program
of the Virginia Polytechnic Institute Computing Center. This program
provides for unequal classes and will give unbiased estimates of main
effects with or without interaction.
Effects showing significant differences were then subjected to
Duncan's multiple range tests to pinpoint individual differences.
However, results of the analyses will not be presented. Reasons are
given in the results section; they pertain to unequal opportunity for
tube penetration and lack of time for penetration.
Fruit Set Studies
The Gainesville fruit set study served as a check on the tube
length study in that every pollen treatment in the latter was included
in the former. The procedures for both tube length and fruit set
studies were nearly identical except that some additions were made in
the latter. For Tree Group A each pollen treatment was placed on
flowers at three different locations rather than two. Also, both
bagged and unbagged non-hand-pollinated flowers were included. The
depetaled unbagged flowers served as a check on attraction of insects.
The divergence of the Gainesville fruit set and tube length
studies came when flowers were removed from the tube length study.
Fruit set flowers were left on the trees and collected and counted as
fruit. Analysis was to be based on success or failure of pollen treat-
ments to set fruit.
The Norris fruit set study was intended to strengthen its Florida
counterpart and possibly uncover intraspecies differences due to lati-
tude or other factors. A breakdown of treatments is given in Table 3.
Techniques and Procedures
Pollen collection.--To avoid the possibility of collecting
insect-contaminated pollen in the Florida studies, cuttings were
brought inside, flowers already opened removed, then pollen collected
from those flowers which opened afterward. Anthers were dehisced and
pollen dried in a refrigerated desiccator.
Trials proved that the best collection method involved gently
rolling the cherry flower raceme between two pieces of ordinary window
screen. With a little practice the collector can retain an almost pure
batch of anthers. Actual equipment included several prepared screens
made by mounting a piece of screen between two pieces of plywood with a
circular hole in the center of each. They looked like this:
Pressure was applied by hand with a piece of screen shaped like a prism
Pollination Treatments, Norris Fruit Set Study,
by Female Parent and Pollen
Female self 1 2 3 4 5 6
1 + + + + + +
2 + + + + + +
3 + + + + + +
4 + + + + + +
5 + + + + + +
6 + + + + + +
- No treatment
1 21 inches
I inches I 4 inches
The collection sequence was:
1. Flowering branches collected.
2. Smaller twigs cut off and immediately placed in water.
3. Opened flowers removed and discarded.
4, Other flowers allowed to open.
5. Racemes with most flowers opened removed and anthers
collected by the screen-on-screen pressure method.
6. Anthers placed in cotton-stoppered bottles in a refrigerated
Pollen was not tested prior to use, because the period of the
delay plus time required for pollinations might well have been longer
than the flowering period of some or all of the trees. But a later
check proved it viable,
In Norris whole racemes were bagged and flowers allowed to open.
When needed, flower-bearing limbs were removed, still bagged, and
placed in water. Racemes were then removed and used as applicators.
Pollination techniques.--Since black cherry has perfect flowers,
emasculation was considered. However, two problems made this imprac-
tical. One is the small size and delicacy of the flowers; pinching or
cutting off the anther-bearing portion is possible, but often the style
is broken or cut. The second is that success leaves a weak, unprotected
style. Breakage of the style is the usual result either during the
emasculation process or afterward due to jostling in a bag. This rules
out emasculation as a practical device except possibly in greenhouse
work, Here flowers would not be subjected to the frequent and some-
times extremely rough treatment common in unprotected trees.
In the Gainesville studies petals were removed with tweezers
when they showed a lifting or swelling. This is about one to two hours
before normal opening. At this time anthers are tucked under and
around the stigma and style with the stigma usually easily accessible.
Then it is a simple matter to touch the stigma with a pollen-laden
applicator. In this case the applicator was a small dowel covered with
sterile waxed paper. The matter of finding enough flowers at the same
stage of opening offered no problem. Once the process was fully under
way on a raceme, at least 8-10 flowers could be found at the same stage,
The pollination sequence in Gainesville was:
1. Opened flowers removed and discarded,
2. A treatment tag selected at random and attached to the limbo
3. Pollinator's hands and tools cleaned with ethyl alcohol,
4, Petals removed from flowers that would normally open within
one to two hours (evidence for this was swollen petals) 1
5. Appropriate pollen applied to the stigma of each flower.
6. Flowers bagged.
In Norris, pollinations were made on flowers that were previ-
ously bagged and allowed to open. Pollen was applied directly by
brushing flowers with an opened raceme of flowers from the male parent,
A slight variation used here for one treatment is explained
under principal studies.
Pollen was kept available by placing cut stems of bagged flowers in a
rack of milk bottles filled with water. The pollination sequence was
much the same as in Gainesville except that upon bag removal prior to
pollen application, immature and overmature flowers were removed.
Tagging for identification.-Tags were designed as follows,
(tree number) (days left on tree
Pollen Tube Length Study Tag
Fruit Set Study Tag
All were prepared before work on a given tree began. Treatments were
assigned at random by drawing a tag from a bag after a flowering limb
was selected. In the pollen tube length study, different day intervals
were color coded with flags to permit easy identification and removal,
Bagging for isolation.--Crimped synthetic sausage casings were
used for bags. One was placed over all flowers covered by a single
treatment tag, Then cotton was placed around the limb near the base of
and inside the bag and a tying wire, known as a "Twist-em," used to tie
the bag around the cotton. A breathing quality of the bag material
along with the cotton allowed some air into the bag but not insects.
Limbs which were not strong enough to support the bag--many were in
this category--were supported by tying limbs to other limbs or in some
cases by attaching a small dowel to the bag and limb, then attaching
the dowel top and bottom to a higher limb. Theoretically, this second
method would keep the bag from being pulled off. However, several were
Flower killing and fixation.--All flowers to be sectioned and
mounted on slides were killed, fixed, and stored in formalin-acetic-
alcohol (FAA) of the following formula:
70% ethyl alcohol -- - - - 90%
Formalin -------------- 5%
Glacial acetic acid - - - - 5%
Slide preparation.--Material was processed as follows:
1. Pistils removed from flowers.
2. Groups of pistils run through alcohol series to paraffin.
3. Material placed under vacuum in warm paraffin.
4. Material embedded in paraffin.
5. Blocks cooled, then cut longitudinally at 20 microns on
6. Serial sections mounted.
7. Material dried thoroughly then run through xylene for
paraffin removal then through alcohol series to water.
8. Material stained with lacmoid and martius yellow,
9. Material destined slightly in alcohol, then placed in
10, Slides coverslipped.
Under number 2 above, the alcohol series consisted of the following
1. 70% alcohol -- 150 ml H20, 250 ml 95% alc., 100 ml butyl
2. 85% alcohol -- 75 ml H20, 250 ml 95% ale., 175 ml butyl
3. 95% alcohol -- 225 ml 95% ale., 275 ml butyl alcohol
4. 100% alcohol -- 125 ml 200% alc., 375 ml butyl alcohol
5. 50-50 paraffin and butyl alcohol
6. Paraffin Paraplast mixture
Under 7, 8, and 9 of the process the series was the following:
1. xylene -- 3 minutes
2. xylene -- 2 minutes
3. absolute alcohol -- 5 minutes
4. 95% ethyl alcohol -- 5 minutes
5. 70% ethyl alcohol -- 5 minutes
6. water -- 5 minutes or more
7. lacmoid and martius yellow -- 10 minutes
8. water -- 15 seconds
9. absolute alcohol -- 30 seconds
10. xylene -- 3 minutes or more
The stain was a mixture of:
5 mg. martius yellow
5 mg. lacmoid
12 mg. water
enough NH4 to rise pH to 8
RESULTS AND DISCUSSION
Since very little breeding work, if any, had been done with
black cherry prior to this study, especially with respect to controlled
pollinations, new techniques often had to be worked out. Methods used
by breeders of cultivated Prunus species were considered and used when-
ever possible. But familiarity with Prunus serotina came only after
development and trials of new techniques. All trials except the
chromosome count were carried out in the Gainesville, Florida, area,
Date and Duration of Flowering
Observations showed that both of these are largely dependent
upon weather conditions, mainly temperature. In the Gainesville,
Florida, area black cherry can be expected to flower sometime in late
February or early March. Cherries in the Norris, Tennessee, area
flower about one month later. Observations over a three-year period
showed much variation among individual trees with some flowering
several days or even one or two weeks before others in the same area.
Normally a tree flowered over a two-week period. This interval was
shortened to as little as five days when a hot period followed a cool-
to-cold first part of February in Gainesville.
Flower Location on the Tree
Flowers occurred with near equal frequency from top to bottom
and on all sides of a tree regardless of aspect. Careful observations
of naturally set fruit showed heavy fruiting all over an open-grown
tree. This indicated that the experimental design need not include
replication in blocks to offset variations in flowers due to location
on the tree.
Characteristics of Flowers and Racemes
Prunus serotina has perfect flowers which are about 1/4 inch in
diameter when fully opened. The 1/8- to 1/4-inch pistil is surrounded
by 15-18 stamens in a perigynous arrangement. Imminent flower opening
is evidenced by swollen or lifted petals. At this stage about one to
two hours in full sun remains to full opening.
Normally 50-70 flowers are borne on a raceme, but occasionally
very long racemes containing 80-90 flowers are encountered. Flowers
always begin to open at the base of the raceme, but some buds may be
bypassed by more rapidly opening flowers above. An individual raceme
may flower over a 2- to 5-day period.
Black cherry is insect pollinated. Although the principal
pollinator is the honeybee in Gainesville, other insects, including
other types of bees and several flies, were seen on flowers.
Temperature seems to highly influence insect activity, espe-
cially that of bees, One Florida cherry flowering season was cool, and
bee activity was limited to a short period each day. The resulting
crop was estimated at less than half that of years when temperatures
during flowering were in the 80 F to 900 F range during much of the
day. A cool season, then, is good for controlled pollination because
it spreads the flowering period out, but bad for natural pollination
because of limited bee activity.
Stigmas are exposed to cross-pollination before anther dehis-
cence in single flowers. Often flowers are open for one to two hours
before anthers show any sign of liberating pollen.
Nothing is known of what happens when a honeybee transfers self
pollen within a black cherry tree. What is known, however, is that
self pollen can easily be transferred from flower to flower or from
anther to stigma on the same flower by contact alone. If such trans-
fers produced compatible crosses, then black cherry crops would not
vary directly with honeybee activity as is the case in the Gainesville,
The almost complete lack of selfing in black cherry in Gaines-
ville was first uncovered during trials of pollination techniques.
Unopened flowers, bagged and left, consistently failed to set fruit.
Artificially selfed flowers produced the same result. But crosses,
especially mixed-pollen crosses, set some fruit in almost every case.
An exception to the natural selfing rule in black cherry occurred on
one occasion during a second season of preliminary trials in Gaines-
ville; three flowers on one tree did set fruit when isolated. This is
the only instance of natural selling within a bag that occurred in
Gainesville studies over three years in thousands of flowers. Though a
highly suspicious result, there was no reason to doubt that natural
selfing had occurred.
Pollen Germination Tests
Tests on agar, agar with 10 percent sucrose, and distilled water
showed that black cherry pollen germinated well on all three media.
Germination on each one was 50 percent or better after three days of
incubation. Tube growth was regular with most tubes of similar length.
After three days, tubes of average length were long enough to have
reached the ovary in normal pollinations. This regularity of tube
growth and fairly short period of extension set the pattern for later
Bagging is not only a means of isolation but also a way to keep
developing and mature fruits available and labeled, even though they
may become detached for one reason or another. It avoids the need for
constant attention over the ripening period to be sure that successes
are not lost before being counted.
Individual flower life varied within bags, with temperature an
important factor. Under cool conditions an individual flower would
last as long as nine days when pollinated and bagged. At the six- or
seven-day point pistils began to darken and wither. Very hot air
temperatures reduced these figures to as low as five- and three-day
periods, respectively, and prolonged contact with the bag usually meant
death to individual flowers.
Since Schuster (1925) in his work with sweet cherry concluded
that petal removal would deter insects, the possibility that isolation
was not needed was considered. However, the probability that unbagged
fruit would be eaten by birds or otherwise lost seemed high and bags
Ploidy influences incompatibility in Prunus. Darlington and
Ammal (1945) listed Prunus serotina as a tetraploid, and several tetra-
ploids in Prunus are self-compatible. But trees in the Gainesville
area, with the exception mentioned earlier, were all self-incompatible
in early trials. For this reason chromosome counts were made. Seed
for this study came from the Norris, Tennessee, area.
Root tip cells undergoing mitosis exhibited very long and fat
chromosomes--much of the cell was filled with them. By focusing up and
down on the microscope, each chromosome and/or broken parts were traced
Many squash preparations were examined, but in only five cells were
chromosomes dispersed enough to make counting practical (Figures 3 and
4). Counts ranged from a low of 22 to a high of 25 chromosomes plus
segments. One end of these short pieces was in almost every case very
near what appeared to be the point of breakage of a much longer
chromosome. Since there was no way to be sure of a count and continued
squashing produced no better separation, these trees cannot categori-
cally be classified as diploids. However, the arrangement of the short
segments indicates that they are diploid. They are definitely not
Pollen Tube Length Study
The main purpose of this study was to supply a basis for compar-
ison of pollen treatments with respect to tube penetration. Lack of
germination was not expected for any pollen treatment including self,
since early trials had shown black cherry pollen would germinate well
Figure 3. Black cherry chromosomes
at metaphase, not well spread.
Figure 4. Well spread black cherry
on plain water, However, flowers on the same raceme showed highly dif-
ferent amounts of germination with the same pollen. This occurred even
though techniques used were meant to give individual flowers an equal
chance to respond to various pollen treatments. Some factors which
might have influenced this result are:
1. Differences in receptivity among flowers treated the same,
2. Interference with liberation of self pollen because of direct
damage caused by petal removal or indirect damage resulting from
exposure of immature flower parts.
3. Interference due to a reaction between applied and self
4. Failure to distribute pollen uniformly.
Although there is no way to be sure, a combination of Factor 1 and
Factor 2 seems the most likely reason for results obtained for the
1. Flower sections showed no evidence of physical damage to the
stigmatic surface; any damage must have been elsewhere.
2. Early trials using cross-pollens on unemasculated flowers
with petals removed showed that crosses consistently produced fruit,
3. Some early-pollinated flowers failed to extend anthers,
4, Pollen applicators were always kept covered with fresh
pollen and were checked regularly to be sure that pollen was being
left on the treated flower,
If some treated flowers on each raceme were a day or more away from
opening, applied pollen may not have germinated readily and self pollen
would have been liberated later. Self pollen would not have had an
opportunity to penetrate deeply if at all. Disregarding Factor 1, all
flower group means with the possible exception of early selfs should
have been at least equal to no-pollen-self means since no flowers were
emasculated. However, where the no-pollen treatment was included, it
was either first or second among treatments in number of tubes present,
There are only two possible explanations for this behavior (1) the
pollen application process interfered with release and/or growth of
self pollen, or (2) flowers that received the no-pollen treatment were
closer to the pollen-release stage as a group than were flowers in
other treatments. The second alternative is a weak explanation except
in the case of early self flowers which were selected on the basis of
immaturity. However, there is the direct evidence for the first alter-
native listed under reason 3 above. All early self flowers extended
anthers later than more mature flowers as expected, but some failed to
extend anthers after seven days. Any flower selected should have
opened within that period. The consistent last position of the early
self treatment is probably the result of both non-receptivity and
flower damage directly or indirectly (premature exposure) caused by
the pollination process,
Ranks of pollen treatments are shown in Table 4 and Table 5
Statistical comparisons were run but are not considered valid since
amount of pollen applied could not be controlled.
The fact that the no-pollen-self ranked high when used indicates
that not only was pollen transferred within individual flowers success-
fully, but that it germinated well.
Cross and normal self treatment results are confused because of
the likelihood that pollen from within the flower was also included in
counts. But the rankings did produce some patterns that give
Comparison of Pollen Treatments with Respect to Pollen
Germination, Tree Group A
Tree 1 2 3 4 5 6
1 PO P2 Ph P3 NS ES
3 NS PO P2 P1 Ph ES
4 PO P2 P1 P3 NS ES
* Rank 1 had highest germination.
ES Early self
NS Normal self
PO No pollen
P1 Pollen 1
P2 Pollen 2
P3 Pollen 3
Ph Pollen 4
Comparison of Pollen Treatments with Respect to Pollen
Germination, Tree Group B
Tree 1 2 3 4 5 6 7
5 P7 NS P2 PI P6 P8 Ph
6 P7 PI NS P2 P8 P5 Ph
7 P6 P8 P2 NS P1 P5 -
8 NS P7 P2 P6 P5 P1 Ph
- No treatment
* Rank 1 had most germination.
NS Normal self
P1 Pollen 1
P2 Pollen 2
Ph Pollen 4
P5 Pollen 5
P6 Pollen 6
P7 Pollen 7
P8 Pollen 8
indications of performance by some treatments. In tree group A, pollen
2 ranked second or third, and on group B it was third three times and
fourth once. Pollen 3 was only better than the applied selfs, and
pollen h was last in three of five cases. Pollen 7 always ranked high
and pollen 5 low.
Removal intervals were intended to give pollen tube growth rate
data, but the receptivity problem voided this possibility. Comparative
rankings are found in Table 6. The longer interval should have allowed
maximum pollen germination. It should also have included more germi-
nated pollen grains from flowers that became receptive later than
expected. But five out of seven times the 5-day interval produced more
germination. The 3-day interval was always last when included. Indi-
vidual flowers were again unpredictable. Quite often single flowers
from the 3-day interval showed more germination than others from longer
intervals. No definite reason can be given at this time for these
results. The following possibilities could well be responsible:
1. By chance, more flowers included in the 7-day interval were
pollinated too early and failed to promote germination as did most of
the early self.
2. A high percentage of flowers from the 5-day interval were
close to or had reached peak receptivity, with the result that more
pollen grains germinated.
3. Many flowers from the 3-day interval became receptive on
the second or third day and had no chance to reach a high level of
Rank* of Removal Intervals in Tree Groups A and B
vith Respect to Pollen Germination
Removal intervals (days)
Tree 3 5 7
1 3 1 2
3 3 1 2
h 3 1 2
5 1 2
6 2 1
7 1 2
8 2 1
- No treatment
* Rank 1 had most germination.
The consistent last position of the 3-day interval is easy to under-
stand and accept. But there seems to be no logical way to explain the
average second position of the 7-day interval.
On tree 3, pollen 2 from the 3-day interval ranked second com-
pared to seventeenth for pollen 2 from the 7-day interval, although the
3-day interval was the worst overall. On tree 6, pollen 8 from the 5-
day interval was superior to the same pollen from the 7-day interval.
And on tree 8, the normal self from the 5-day interval was superior to
the same treatment from the 7-day interval. Clearly, some factor
unaccounted for influenced the results.
Summary of results.--Among pollen treatments the no-pollen
treatment ranked first twice and second once in the three cases it was
used. Among removal intervals the 5-day period produced more germina-
tion than did the 3- or 7-day intervals, although the expected order
was 7, 5, 3 (highest to lowest). Other noteworthy results included:
1. Early selfing proved unsuccessful. It ranked last whenever
2. Self pollen applied just prior to flower opening showed no
consistent pattern. It ranked first on two trees, next to last on two
others, and in varied positions on the other three trees.
3. Some cross pollens showed definite patterns. Pollens 2 and
7 always ranked high and pollens 4 and 5 always low.
4. Pollens varied in ability to germinate, but there was no
barrier to germination.
Pollen Tube Extension
In this phase the style was divided roughly into three equal
parts, and counts were made of tube sections in each part. Since no
control was possible over amount of pollen deposited on each stigma--
and, consequently, the number of tubes possible--an equalizing effect
was used; number of tubes in the middle and lower thirds of the style
was expressed as a percent of the number in the upper third. These
percentages were converted to arc sin values and analyzed statistically.
However, these findings are not considered a fair estimate of pollen
performance. Reasons are presented later.
Treatments for groups A and B are compared within and over all
trees with respect to pollen tube penetration into the middle third of
the style in Table 7. It is highly probable that self-pollen tubes
were included in counts of cross-pollen treatments. Therefore, no-
pollen treatment results could not be used to distinguish between self-
and cross-pollen tubes. Results did indicate that no-pollen-self tubes
failed to penetrate the lower third of the style. Because of these
findings, data for middle-third penetration by cross pollens are not
considered reliable, while the following lower-third data are thought
to be more accurate.
Treatments for tree groups A and B are compared within and over
all trees with respect to pollen tube penetration into the lower third
of the style in Table 8, Tube penetration into the lower third of the
style was much reduced over penetration into the middle third. Only
1.4 percent of the tubes reached this level. The most probable reason
for this lack of tube extension is insufficient time for tube travel.
Seven days was the maximum time allowed before flower removal and
fixation. This interval was selected for three reasons: (1) tubes
extended fully within three days in incubated pollen germination tests,
(2) pollen tubes had extended into the ovary by this time in Prunus
Percent of Pollen Tubes in the Upper Third of the Style that
Reached the Middle Third, over All Trees and Removal
Intervals, by Pollen Treatment and Tree
Tree PO P1 P2 P3 P4 P5 P6 P7 P8 ES NS Av*
1 9.1 32.8 7.1 0 0 0 10.6
3 0 12.8 5.3 8.0 2.3 3.5 5.3
4 0 1.8 3.0 8.5 0 0 2.2
5 4.9 5.5 2.2 0 3.7 37.0 4.2 6.4
6 53.5 9.8 46.1 31.1 28.4 47.8 27.0 35.9
7 15.4 0 0 0 12.3 4.6 0 7.1
8 18.7 0 0 0 50.0 0.5 0 10.2
Av.* 3.6 33.1 7.5 8.3 29.4 30.1 12.0 21.3 36.0 1.3 20.1 20.7
- No treatment
* Weighted average
ES Early self
NS Normal self
PO No pollen
P5 Pollen 5
P6 Pollen 6
P7 Pollen 7
P8 Pollen 8
Percent of Pollen Tubes in the Upper Third of the Style that
Reached the Lower Third, over All Trees and Removal
Intervals, by Pollen Treatment and Tree
Tree PO P1 P2 P3 P4 P5 P6 P7 P8 ES NS Av,*
1 0 5.2 0 0 0 0 0.5
3 0 0 1.1 0 0 0 0.3
4 0 0 0 0-0 0 0
5 3.3 0 0 0 0 0 0 0.9
6 2.7 0.4 1.1 3.7 1.8 0.6 1.2 1.8
7 0 0 0 0 9.1 0 0 3.5
8 .7 0 0 0 50.0 0 0 3.6
Avo* 0 2.6 0.6 0 0.6 3.6 9.6 1.3 0.4 0 0,8 1.4
- No treatment
* Weighted average
ES Early self
NS Normal self
PO No pollen
avium studies (Roy, 1938), and (3) flowers begin to deteriorate quickly
after seven days. However, tube travel continued past this stage and
probably was not completed until the style was well into decay.
Support for this view comes from the fruit set phase of this work.
Flowers in that phase treated exactly as those in the tube length phase
produced some fruit from nearly every cross. So tube penetration must
have continued in those flowers after flower removal in the tube length
study. Because of the lack of time for tube extension, no statistical
analysis is presented here.
The first visual evidence of possible incompatibility in black
cherry was found in this phase. One pollen tube from the normal self
on tree 1 produced a swollen tip or head (Figure 5). In studies of
Prunus avium (Roy, 1938), this condition was considered good evidence
for incompatibility. Another tube from the cross of tree 8 with tree 6
turned around and began to grow back up the style (Figure 6). This is
also evidence of incompatibility in P. avium (Roy, 1938). No other
examples of this type were found.
The consistently good overall performance of tree 6 as both male
and female is worth discussion. Some tubes from every pollen used on
this tree reached the lower style, even the normal self. No self on
any other tree reached this depth. Good performance by all pollens on
it indicates that tree 6 was highly receptive. It might even have a
genetic advantage in this respect over the other trees. But the most
important feature here is the penetration of self pollen. On P.
avium, Roy (1938) found that self-pollen tubes usually became arrested
in the style, and most of them appeared to swell up at their ends.
Afify (1933) found that self tubes were short and often bent upward
f *vy C M.
?*'' / ^
Figure 5. Abnormal pollen tube,
upper one-third style, Tree 1
normal self, 7 days.
Figure 6. Abnormal pollen tube,
middle one-third style, Tree
6 x Tree 8, 5 days (lover
before ceasing growth. But Roy (1938) also very rarely obtained one
or two fruits after self pollination. On tree 6 with self pollen,
tubes did not have swollen ends nor were any turned upward. However,
both of these characteristics might have appeared had tubes been
allowed more time to grow. On the other hand, one or more might also
have penetrated the ovary. But fruit set study results indicate that
fertilization probably would not have occurred.
Flower removal intervals for groups A and B are compared within
trees and over all trees and pollen treatments with respect to pollen
tube penetration into the middle third of the style in Table 9.
In tree group A, the pattern of tube penetration with respect to
removal intervals was the same as that in pollen germination: the 3-
day interval was low; the 5-day, high; and the 7-day, intermediate.
But tubes penetrated the middle third on every tree in the 7-day
interval, whereas tree 4 had no penetration in the 5-day interval, and
trees 3 and 4 had none in the 3-day interval. The pattern here as in
germination seems to reflect the stage of receptivity in the trees,
although there is no explanation for the 7-day period being inter-
mediate. Tree group B showed an overall advantage for the 7-day
interval though it ranked first on only one tree. Over all trees in
both groups the 7-day interval produced the highest percentage of tubes
reaching the middle third of the style. This was caused by the very
large number of penetrations observed on tree 6 in the 7-day interval.
Intervals for groups A and B are compared within trees and over
all trees and pollen treatments with respect to pollen tube penetration
into the lower third of the style in Table 10. Only the 5-day interval
produced tubes at this depth in tree group A, and these were at very
Percent of Pollen Tubes in the Upper Third of the Style
that Reached the Middle Third, over All Trees and
Pollen Treatments, by Removal interval and Tree
Removal interval (days)
Tree 3 5 7 Av.*
1 3.6 14.0 5.4 10.4
3 0 10.8 6.1 5.1
h 0 0 3.5 2.1
5 6.3 5.0 6.1
6 30.6 38.2 34.5
7 7.1 6.5 6.9
8 15.4 4.1 9.8
Av., 0.4 19.9 24,8 20.7
Percent of Pollen Tubes in the Upper Third of the Style
that Reached the Lover Third, over All Trees and
Pollen Treatments, by Removal Interval and Tree
Removal interval (days)
Tree 3 5 7 Av.*
1 0 0.8 0 0.5
3 0 1.0 0 0.3
4 0 0 0 0
5 0.9 0 0.8
6 3.1 0.4 1.7
7 3.9 1.9 3.4
8 4.0 3.0 3.5
Av.* 0 2.4 0.5 1.4
low percentages. In group B, the trend set in the middle third of the
style was reversed, and the 5-day interval ranked first. Percent pene-
tration was much lower, but some penetration occurred in every tree-
interval combination except tree 5, interval 7.
Summary of results.--The several noteworthy results included:
1. Pollen tubes from the no-pollen-self treatment, so highly
ranked in pollen germination, penetrated to the middle third of the
style only on tree 1, with no tubes reaching the lower third.
2. Early self tubes reached the middle third only on tree 3,
with no further penetration.
3. Normal-self tubes reached the middle third on trees 3, 5,
and 6, and the lower third on tree 6. Tree 6 pollen showed exceptional
ability to penetrate tree 6 stylar tissue.
4. Cross-pollen results were clouded by the strong possibility
that self pollen tubes were included in the counts. Based on no-pollen-
self results it is highly probable that they were included in tube
counts from the middle third of the style but unlikely in lower third
5. The pattern of tube penetration with respect to time of
flower removal after pollination showed less penetration in the
shortest interval as expected, but also less extension in the longest
than in the intermediate interval.
6. Two instances of abnormal pollen tube development, evidence
for an incompatibility system similar to that in Prunus avium, were
Fruit Set Studies
Prevention of Insect Pollination without Bagging--Gainesville
On each tree in group A, 60 flowers were depetaled. Half were
bagged; half were not. No fruit set occurred among bagged flowers, but
one unbagged flower did set on tree 3. If this low set could be dupli-
cated over a large number of flowers, the results might be acceptable
for most breeding work. However, the one fruit that did set causes
some doubt as to the possible success of the procedure.
Selfing and Crossing--Gainesville
Both early and normal selfs were included on tree group A. Only
normal selfs were included on group B. Not a single flower of any self
set fruit. This paralleled findings in preliminary trials. After ob-
serving results of the pollen tube length phase of this work, lack of
fruit set in the early self is understandable. Very little germination
occurred, and in all but one instance tubes traveled no farther than
mid-style. Since normal selfs produced many tubes with some penetrat-
ing into the lower style, the only reason for lack of fruit set is the
The overall fruit set among cross-pollinated flowers was 5.9
percent (Table 11). Pollens varied from 1.7 to 12.0 percent. Mother
(female) trees varied from 1.7 to 14.2 percent. Some results followed
trends set in the tube length study, but others did not. Tree 6 was a
good pollen parent on each tree where it was used and a good mother
tree for all pollens except 5 and 7, which failed on all trees. No
tree crossed successfully with every other tree. Tree 1 crossed with
every tree except tree 8. Tree 8 crossed only with tree 6.
Percent Fruit Set among Cross-Pollinated Flowers,
by Pollen and Mother Tree, Gainesville
Mother Pollen parent (tree)
tree 1 2 3 4 5 6 7 8 Av.*
1 6.7 3.3 13.3 7.8
3 0 0 10.0 3.3
4 6.7 0 0 2.2
5 10.0 0 5.0 10.0 0 0 4.0
6 4o0. 10.0 25.0 0 0 10.0 14.2
7 10.0 15.0 0 0 20.0 0 7.0
8 0 0 0 0 10.0 0 1.7
Av.* 10.0 4.4 1.7 9.3 0 12.0 0 3.3 5.9
- No treatment
* Weighted average
With these data alone, only compatible crosses can be distin-
guished with accuracy. But there are some indications that tree 8
might be incompatible, or partially so, with one or more trees. It
crossed successfully with tree 6 both as a male and female but with no
other tree. However, it is not possible under the diploid system of
incompatibility found in other species of Prunus for tree 8 to be com-
pletely cross-incompatible with all but one of the study trees. The
reasoning here is that if tree 8 is completely incompatible with all of
the others, then all must have the same alleles for incompatibility.
If this were true, the others would also be completely incompatible
among themselves. The logical deduction is that poor pollen germina-
tion or some other influence prevented fertilization or fruit set.
Selfing and Crossing--Norris
Percent set by male and female parent for selfs and crosses is
found in Table 12, Of the 17,921 flowers allowed to self, 3,859 pro-
duced fruit for a 21.5 percent set. Every study tree produced selfed
fruit, Individual trees ranged from a low of 9.9 to a high of 42.3
percent. Out of 1,181 cross-pollinated flowers, 516 or 43.7 percent
set fruit, Male-female combinations varied from 0 to 80.0 percent set.
Tree 1 produced a lower set from crosses than from selfs, the only tree
to do so. It was pollinated just as the other trees began to flower.
Racemes from other trees used to cross pollinate may not have liberated
much pollen with the result that fewer flowers were pollinated.
However, this does not explain why these flowers did not self.
Percentages in Table 12 are based on counts that were probably
made just before what is called the third or June drop in cultivated
Percent Fruit Set among Self- and Cross-pollinated Flowers,
by Pollen and Mother Tree, Norris
Mother Pollen (tree) Av. for
tree Self 1 2 3 4 5 6 Crosses*
1 9.8 17.4 0 8.0 2.5 0 5.0
2 11.5 40.0 13.5 20.5 36.7 65.6 34.3
3 42.3 63.2 63.4 22.5 80.0 63.6 58.8
4 28.5 48.4 55.4 54.3 70.0 55.9 57.0
5 29.8 37.9 56.4 47.6 67.7 36.0 50.0
6 16.8 74.4 14.7 30.2 38.5 32.3 38.7
Av.* 21.5 54.1 45.6 33.9 32.0 46.8 46.9 43.7
- No treatment
* Weighted average
cherry varieties. At that time--two weeks after peak flowering--fruit
was well developed. Ten selfed fruits--two from trees 1, 3, 4, and 5;
one from trees 2 and 6--were removed, opened, and inspected. Three of
the ten--two from tree 4, one from tree 6--appeared normal. The others
were either empty or had shriveled seed coats; five of these showed
insect larvae damage. Normal appearing seeds were sectioned by hand
and observed under the microscope. A typical picture of a developing
embryo from this type seed is shown in Figure 7.
Of ten crosses examined, all were filled, but seven showed
insect damage. Developing embryos were similar to that in Figure 7.
Five weeks after peak flowering only ten selfed fruits remained
of the 3,859 that set, but 117 of the 516 crosses that set remained on
the tree. There is little doubt that the insect infestation caused a
large part of the fruit to drop, but the wide difference between self
and crosses is unexplained. These results are similar in one respect
to Bradbury's (1929) with sour cherry and Cooper's and Brink's (1940)
with Nicotiana hybrids. In both cases selfs and interspecific hybrids
showed restricted embryo development when compared to intraspecific
crosses. What is important is that black cherry did self and that
selfed fruit remained on the tree past the usual drop periods of other
species of Prunus.
With regard to selfing, the Norris study produced completely
opposite results from those in Gainesville. In Gainesville black
cherry exhibited an incompatibility system similar to the one in sweet
cherry. In Norris it was self-fertile, a characteristic also exhibited
by the sour cherry, Prunus cerasus, The latter is a tetraploid.
Figure 7. Embryo from selfed fruit, Norris.
Chromosome counts will be made on Norris area cherry again and
on those from Gainesville to accurately determine their ploidy, but
there could be another reason for successful selfing in Norris cherry.
It is possible that there are one or more self-compatible alleles in
the Norris population. If this is the case, then there are at least
two races of black cherry based on self-compatibility. The small
amount of evidence now available supports this alternative.
Presence of selfing coupled with lack of emasculation of cross-
pollinated flowers makes impossible the classification of crosses in
the Norris study. The higher percent retention of crossed fruits does
indicate that crossing was successful, however.
SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS
Summary and Conclusions
Black cherry is one of our most valuable timber species and is
now included in several tree improvement programs. Its breeding
habits, however, had not been studied closely prior to this work,
Consequently, this study is intended to supply black cherry breeders
with not only a description of incompatibility in the species but also
a working knowledge of breeding habits and techniques used in making
Lack of literature on breeding black cherry and techniques asso-
ciated with the process made necessary preliminary trials. These were
conducted prior to the major work and covered flowering habits, natural
selfing and crossing, pollen germination tests, isolation of flowers,
and a chromosome count. Techniques established for other species were
adapted for black cherry when possible and new ones developed when
The major work was subdivided into studies of pollen tube
length, or penetration into the style, and fruit set. Pollen tube
characteristics, including amount of germination and depth of penetra-
tion,were observed for selfs and crosses among trees selected because
of likelihood of close relationship. The objective was to obtain
microscopic evidence for self- and/or cross-incompatibility among study
trees based on behavior of pollen tubes in the style. Fruit setting
was used as a check on microscopic evidence and to uncover possible
regional differences between the Norris, Tennessee, and Gainesville,
Florida, areas. Pollinations that readily produced fruit were con-
Conclusions on the pollen germination phase of the tube length
study accompany the following noteworthy results:
1. Pollen moved from anthers to pistil of the same flower
without insect visitation and germinated well. Transfer occurred
either by direct contact or through air currents. Both are possible.
The former must have accounted for most of the transfer because of the
close proximity of anther and stigma and lack of air movement in bags.
2. There was considerable variation in results among flowers
receiving the same treatment. The best explanation for this is that
flowers exhibiting external indications of opening (swollen petals)
were not receptive to pollen of any type until shortly after the time
of normal flower opening.
3. Selfing of young buds was unsuccessful. This failure re-
sulted because of "2" above and also because petal removal had a
detrimental effect on the maturation sequence as evidenced by the fact
that anthers occasionally failed to extend.
4o Selfing just prior to flower opening varied from good to
poor among mother trees. This, too, was caused by "2" above.
5. Cross-matings also produced variable results; some males
were always good, some always poor, others intermediate in amount of
pollen germinated. Part of this variation can be attributed to "2"
above, but there was evidence that some pollens germinated better than
others in general. Since all pollens were treated alike, some must
have had a genetic advantage; they probably had less stringent
6. There was no barrier to germination of any pollen. When a
flower was receptive, any pollen including its own would germinate on
the stigmatic surface.
Pollen tube extension results gave limited support to the case
for self-incompatibility and some cross-incompatibility but did not
exclude self-compatibility completely. Most interesting findings and
conclusions based on them include:
1. Pollen tubes from self pollen distributed within individual
flowers penetrated the middle third of the style on only one tree even
though much pollen germinated. In trees of this phase of the overall
study, self-pollen tubes got a late start and also grew more slowly
than did those from cross pollen. These are the reasons for lack of
natural selfing in the Gainesville trees. However, self tubes did
penetrate more deeply than do self tubes in other Prunus species.
2. Pollen tubes from self pollen applied on young buds also
reached the middle third of the style on one tree only. Selfing the
young bud offers no advantage in obtaining longer tubes that might
3. Pollen tubes from self pollen applied just prior to flower
opening reached the middle third on three trees and the lower third on
one. Self pollen, though not as fast growing as cross pollen, can
penetrate the lower style and possibly farther on self-incompatible
trees, Flowers were removed before these and tubes from crosses could
penetrate the ovarian cavity. Styles were beginning to decay after
seven days, but tubes were still growing. This is supported by fruit
4. Two instances of abnormal pollen tube development were
observed; one was a self pollination and the other a cross. This lack
of physical evidence of incompatibility plus the deeper penetration of
self tubes in Prunus serotina than in other Prunus species are evidence
that the incompatibility reaction is not as strong in Prunus serotina.
5. Natural self pollination between anthers and adjacent
stigmas of individual flowers reduced the effectiveness of this work.
In future pollen tube length studies, either emasculation must be
practiced or a radioactive marker used to label applied pollen.
Fruit set study results varied greatly between the Gainesville,
Florida, and Norris, Tennessee, test areas. The most important result
in this phase and probably in the entire study is that all trees in the
Norris phase selfed while no selling occurred in Florida. Further,
embryo development was normal in the three selfed fruits observed.
Since chromosome counts ruled out the Norris trees being tetraploid,
there might be one or more self-compatible alleles in the Norris
population. If so, the Norris and Gainesville population could be
different races. However, environmental effects must be studied be-
fore a firm conclusion is reached.
The overall fruit set percent among crosses in Gainesville was
5.9 compared to 43.7 in Norris. In Gainesville no tree crossed with
every other tree; in Norris all crosses were successful at least one
way. Selfing is a definite factor in these results, but since the
percent of successful crosses was much higher than that of selfs (43.7
compared to 21.5) in Norris, either technique was better in Norris or
Norris trees have a reproductive advantage over those in Gainesville.
One depetaled flower exposed to possible insect visitation set
a fruit, Although this is not enough evidence to eliminate depetaling
as a method of avoiding insect pollination, it causes some doubt as to
Recommendations for Breeding and Tube Length Studies
1. Determine whether the particular trees to be worked are
self-compatible or not. The easiest procedure is to bag several
racemes on each tree prior to flower opening and observe them for fruit
set. Self-fertile individuals should not be used as females in a cross-
breeding program at this time; no satisfactory and, at the same time,
practical emasculation method is available. Further, use of self-
fertile trees as males will introduce self-compatibility genes into a
self-incompatible breeding group.
2. If trees of a breeding group originated in several geographic
regions, expect a wide range of flowering times. Even local trees will
vary up to a week or ten days in initiation of flowering. Satisfactory
long-storage procedures will be available soon from Robert Farmer,1 but
until that time do not count on keeping pollen more than one month, It
can be successfully stored for that period in a refrigerated desiccator.
This should be long enough to cross early trees with late ones,
3. Leafing out is an excellent indicator of the approach of
flowering in black cherry; the two are closely correlated. Flowers
will not open until full leafout. Therefore, no close examination is
needed until a tree is in full leaf.
1Physiologist, TVA Forestry and Fisheries Laboratory, Norris,
4. Temperature is closely correlated with both initiation and
period of flowering, In 850 F and over weather, expect only five to
seven working days per tree. Daytime temperatures under 650 F will
hold up flowering; between 650 and 850 F, the flowering period will
last at least two weeks on each tree.
5. Use fresh pollen when parents are close together and are
flowering at the same time. Small branches containing bagged flowers
can be kept in water and the pollen used up to three days. Pollen
collection and short storage will be required in other situations.
6. Select stout upright limbs of at least 3/16-inch diameter
for bagging. Often a larger limb can be trimmed so that several racemes
close to the base will fit into a bag. The synthetic sausage casings
used in this study are cheap and easy to use, but a better type is made
of a special paper. It is large, light weight, and has a plastic view-
ing window. Forestry Suppliers, Jackson, Mississippi, handles it.
7. Use a camel's hair brush to apply collected pollen and also
fresh pollen, if desired. The brush will hold more pollen than stiff
applicators and offers little chance for flower damage. Have plenty of
them available since each pollen treatment on each tree will require
one. Also place only small amounts of pollen to be applied in an appli-
cation bottle because of the transfer of pollen from unemasculated
flowers. Discard this pollen after use on a single tree or clone.
8. Pollinate only those flowers that have recently opened. On
a raceme there are usually two or three distinct classes of flowers.
Early during flowering only freshly opened flowers and buds are present.
Within two or three days there will be overmature, freshly opened, and
buds. Five or six days after initial flowering, only overmature and a
few freshly opened flowers remain. Pollination is best when the three
classes are present; more flowers are at peak receptivity. Strip buds
and overmature flowers from the raceme; 15 to 25 flowers should be left.
Overmature flowers have brownish-green sepals compared to the bright
green receptive flowers. Also, look for shriveled petals on overmature
flowers. Beginning discoloration--reddish brown spot--on the style is
a sure sign of overmaturity.
9. Remove synthetic sausage casing bags as soon as pollinated
flowers show signs of overmaturity. Slit the bag longitudinally and
carefully open it to avoid dislodging flowers. Some flowers will fall
at this time and afterward on all racemes, even though pollinated.
Paper bags can be left on the tree without heat buildup and will pro-
tect developing fruit from birds.
10. Fruit will turn a bright red, then dark red, then black.
Begin harvesting at the dark red stage.
11. In tube length studies labelled pollen or some form of emas-
culation must be used because of presence of self pollen. Complete re-
moval of calyx and corolla is unsatisfactory because it leaves an
unprotected style. For a small-scale project anthers can be clipped
off with small scissors. This is a very tedious and time-consuming
job, but it will work. Labelling applied pollen with a radioactive
tracer will allow rapid pollination procedures. The more detailed
exposure to film and subsequent processing to get tube tracings can be
done in the laboratory when convenient. Satisfactory procedures are
listed in Jensen's (1962) Botanical Histochemistry under autoradiog-
12. Plan to remove flowers for killing and fixation after sev-
eral periods with the longest at least 10 days. Styles will begin to
dry up and wither six to seven days after opening and possibly sooner
in very hot weather, but tubes will not penetrate the ovary until seven
or more days after pollination. Flower removal after different inter-
vals will allow study of fresh tissue at all stages of tube penetration.
Recommendations for Future Work
Future work can be subdivided into two classes: work with
(1) self-compatible, and (2) self-incompatible types.
Since no seedlings known to be the result of self pollination
are available, there is no way to estimate the extent of inbreeding
depression likely to be encountered, This should be studied first
under class io If depression is moderate to low or negligible, produc-
tion of pure lines should be relatively easy for two reasons: (1) only
isolation is necessary to obtain selfs and (2) the interval from seed
to flowering tree is about five years, a short period for a forest tree.
Once lines are developed that exhibit desired characteristics, they
could be crossed to combine traits and hopefully obtain hybrid vigor.
A large crossing program dealing with self-compatibles would
require either a new and acceptable emasculation method or development
of male-sterile lines. Both of these approaches should be investigated.
For self-incompatible trees the most needed work is in the area
of cross pollination techniques. Such innovations as controlled insect
pollination, adaptation of normally insect carried pollen to a type
dispensed by air currents, and better means of isolation all should be
Work on pollen and seed storage is being done now, but accepta-
ble methods are not yet available. There is room for more of this
Figure Fence row that included study trees, Norris.
Figure 8. Fence row that included study trees, Norris.
Figure 9. Heavily fruiting
Figure 10. Typical tree,
Figure 11. Fruit set from
a cross, Norris.
Figure 12. Fruit set from
a self, Norris.
6 v r -.F
W- F r
Figure 13. Fruit set from a self, Norris.
Figure 14. Typical bag, Gainesville.
Figure 15. Petal removal prior to pollination, Gainesville.
Figure 16. Pollen tubes, upper Figure 17. Pollen tubes, stig-
one-third style, Tree U no matic surface, Tree 1 no
pollen, 3 days. pollen, 5 days.
Figure 18. Pollen tubes, stig- Figure 19. Pollen tubes, lower
matic surface, Tree 4 no one-third style, Tree 6
pollen, 5 days. normal self, 5 days.
Figure 20. Pollen tubes, lower
one-third style, Tree 6 x
Tree 2, 5 days.
Figure 21. Pollen tubes, lower
one-third style, Tree 6 x
Tree 4, 5 days.
Figure 22 Pollen tubes, lower Figure 23. Pollen tubes, lower
one-third style, Tree 6 x one-third style, Tree 6 x
Tree 5, 5 days. Tree 5, 7 days.
Figure 24. Pollen tubes, lower Figure 25. Pollen tubes, lower
one-third style, Tree 6 x one-third style, Tree 6 x
Tree 7, 5 days. Tree 7, 5 days.
Figure 26. Pollen tubes, junction Figure 27. Pollen tubes, upper
ovary and style, Tree 7 x one-third style, Tree 7 x
Tree 6, 7 days. Tree 8, 5 days.
Afify, A. 1933. Pollen tube growth in diploid and polyploid fruits.
Jour. Pom. Hort. Sci. 11:113-119.
Atwood, S. S. 1944. The behavior of oppositional alleles in poly-
ploids of Trifolium repens. Proc. Nat. Acad. Sci. 30:67-79.
Backhouse, W. 0. 1911. Self-sterility in plums. Gard. Chron. 50:599.
Bannan, M. W. 1946. Tetraploid Taraxacum kok-saghyz. II. Canad.
Jour. Res. 24:81.
Bateson, W., and R. P. Gregory. 1905. On the inheritance of hetero-
stylism in Primula. Proc. Roy. Soc. B. 76:581-586.
Bradbury, D. 1929. A comparative study of the developing and aborting
fruits of Prunus cerasus. Amer. Jour. Bot. 16:525-542.
Christ, B. 1959. Entwicklungsgeschichtliche und physiologische
Untersuchungen uber die Selbststerilitat von Cardamine pratensis L.
Z. Bot. 47:88-112.
Compton, R. H. 1912. Preliminary note on the inheritance of sterility
in Reseda odorata. Proc. Phil. Soc. Cambridge 17:7.
Cooper, D. C., and R. A. Brink. 1940. Somatoplastic sterility as a
cause of seed failure after interspecific hybridization. Genetics
Correns, C. 1912. Selbststerilitat und Individualstoffe. Festschr.
d. mat.-nat. Gesell. zur 84. Versamml. deutsch. Naturfoscher u.
Arzte, Munster i. W. pp. 1-32.
Crane, M. B. 1925. Self-sterility and cross-incompatibility in plums
and cherries. Jour. Genet. 15:301-322.
Crane, M. B., and A. G. Brown. 1937. Incompatibility and sterility in
the sweet cherry, Prunus avium L. Jour. Pom. Hort. Sci. 15:86-116.
Crane, M. B., and W. J. C. Lawrence. 1931. Sterility and incompati-
bility in diploid and polyploid fruits. Jour. Genet. 24:97-107.
Crane, M. B., and D. Lewis. 1942. Genetical studies in pears. III.
Incompatibility and sterility. Jour. Genet. 43:31-43.
Crowe, L. 1964. The evolution of outbreeding in plants. Heredity
Dahlgren, K. V. 0. 1916. Eine acaulis-Varietat von Primula officinales
Jacq. und ihre Erblichkeitsverhaltnisse. Svensk hot. Tidskr.
Dahlgren, K. V. 0. 1922. Selbststerilitat innerhalb Klonen von
Lysimachia nummularia. Hereditas 3:200-210.
Darlington, C. D. 1928. Studies in Prunus. I and II. Jour. Genet.
Darlington, C. D. 1930. Studies in Prunus. III. Jour. Genet.
Darlington, C. D., and K. E. Janik Ammal. 1945. Chromosome atlas of
cultivated plants. George Allen and Unwin, Ltd., London.
Darwin, Charles. 1876. Effects of cross- and self-fertilization in
the vegetable kingdom. Ed. 2, 1876. New York: D. Appleton.
Darwin, Charles. 1877. The different forms of flowers on plants of
the same species. London.
Defler, S. E. 1937. Black cherry; characteristics, germination,
growth, and yield. Unpublished thesis, N. Y. State College of
Forestry, Syracuse, N. Y.
Detjen, L. R. 1916. Self-sterility in dewberries and blackberries.
No. Carolina Expt. Sta. Tech. Bull. 11.
Detjen, L. R. 1926. Physiological dropping of fruits. Delaware
Agric. Expt. Sta. Bull. 143.
Dnyansagar, V. R. 1963. Self-incompatibility in angiosperms. Jour.
Bio. Sci. 6:52-72.
East, E. M. 1919. Studies on self-sterility, III. The relation
between self-fertile and self-sterile plants. Genetics 4:341-345.
East. E. M. 1929. Self-sterility. Bibliogr. Genet. 5:331-370.
East, E. M. 1934. Norms of pollen-tube growth in incompatible matings
of self-sterile plants. Proc. Nat. Acad. Sci. 21:225-230.
East, E. M. and A. J. Mangelsdorf. 1925. A new interpretation of the
hereditary behavior of self-sterile plants. Proc. Nat. Acad. Sci.
East, E. M., and A. J. Mangelsdorf. 1926. Studies on self-sterility.
VII. Heredity and selective pollen-tube growth. Genetics
East, E.M., and J. B. Park. 1917. Studies on self-sterility. I. The
behavior of self-sterile plants. Genetics 2:505-609.
Einset, 0. 1932. Experiments in cherry pollination. N. Y. Agric.
Expt. Sta. (Geneva) Bull. 617:13.
Ernst, A. 1936. Heterostylie-Forschung. Versuche zur genetischen
Analyse eines Organisations-und 'Anpassungs'-Merkmales. Z. Indukt.
Ewert, R. 1922. From Jahresbericht Bot. Versuchsanstalt Proskau.
Filzer, P. 1926. Die Selbststerilitat von Veronica syrica. Z. Indukt.
Fisher, R. A., and V. C. Martin. 1947. Spontaneous occurrence in
Lythrum salicaria of plants duplex for the short style gene.
Fisher, R. A., and K. Mather. 1943. The inheritance of style length
in Lythrum salicaria. Ann. Eugenics 12:1-23.
Florin, R. 1924. Korsbarstradens Pollinering. Sver. Pomol. Foren.
Arsskr. 1:1-33 (Stockholm).
Focke, W. 0. 1890. Versuche und Beobachtungen uber Kreuzung und
Fruchtansatz bei Bluten-pflanzen. Abh. nat. Ver. Bremen 11:413-421.
Focke, W. 0. 1893. Uber Unfruchtbarkeit bei Bestauburg mit eigenem
Pollen. Abh. nat. Ver. Bremen 12:409-416, 495, 496.
Garber, R. J., and K. S. Quisenberry. 1927. Self-fertilization in
buckwheat. Jour. Agric. Res. 34:185.
Gairdner, A. E. 1926. Campanula persicifolia and its tetraploid form,
"Telham Beauty." Jour. Genet. 16:341.
Gardner, V. R. 1913. A preliminary report on the pollination of the
sweet cherry. Ore. Agric. Expt. Sta. Bull. 116:1-40.
Gartner, K. F. 1844. Versuche und Beobactungen uber die
Befruchtungsorgane der vollkommeneren Gewachse und uber die
naturliche und kunstliche Befruchtung durch den eigenen Pollen.
pp. X plus (2) plus 644. Stuttgart E. Schweizerbartsche
Gregory, R. P. 1915. Note on the inheritance of heterostylism in
Primula acaulis. Jour. Genet. 4:303.
Grun, P., and M. Aubertin. 1966. The inheritance and expression of
unilateral incompatibility in Solanum. Heredity 21:131-138.
Harrison, B. J., and L. Darby. 1955. Unilateral hybridization.
Hayman, D. L. 1956. The genetic control of incompatibility in
Phalaris coerylescens. Austral. Jour. Bot. Sci. 9:321-331.
Hecht, A. 1944. Induced tetraploids of a self-sterile Oenothera.
Herbert, W. 1837. Amaryllidaceae. London: Ridgway and Sons. pp. VI
Heribert-Nilsson, N. 1916. Populationsanalysen und Erblichkeitsversuche
uber die Selbststerilitat, Selbstfertilitat, und Sterilitat bei dem
Roggen. Z. f. Pflanzenzuchtung 4:1-44.
Hildebrand, F. 1866. Uber die Nothwendigkeit der Insektenhilfe bei
der Befruchtung von Corydalis cava. Jahrb. wiss. Bot. 5:359-363.
Hooper, C. H. 1924. Pollination in relation to cherry orchards.
Jour. S. E. Agric. Coll. 30:244-246.
Hough, Ashbel F. 1960. Silvical characteristics of black cherry
(Prunus serotina). Sta. Paper 139. N. E. For. Expt. Sta. Upper
Howard, H. W. 1942. Self-incompatibility in polyploid forms of
Brassica and Raphanus. Nature 149:302.
Jensen, W. A. 1962. Botanical histochemistry. W. H. Freeman and Co.,
Johansen, D. A. 1940. Plant microtechnique. McGraw-Hill Book Co.,
Johansson, E. 1945. Befruktnings forhallanden has Apple, Paron,
Plommon och Korsbar. S. P. F. Medd. nr. 28 franStatens trad-
Jost, L. 1907. Uber die Selbststerilitat einiger Bluten. Bot. Ztg.
Kerns, K. R., and J. L. Collins. 1947. Chimaeras in the pineapple.
Jour. Hered. 38:323.
Kobel, F. 1927. Zytologische Untersuchungen an Prunoideen und
Pomoideen. Arch. Jul. Klaus-Stift. 3:1-84.
Kobel, F., and P. Steinegger. 1933. Die Befruchtungeverhaltnisse
schweizerischer Kirschensorten. Landwertschaftliches Jahrb. d.
Kolreuter, J. G. 1761-6. Vorlaufige Nachricht von einigen das
Geschlecht der Pflanzen betreffenden Versuchen und Beobachtungen,
nebst Fortsetzungen 1, 2u. 3. pp. 266. Ostwald's Klassiker, Nr.
41. Leipzig: Engelmann.
Kroh, M. 1956. Genetische und entwicklungsphysiologische Untersuchungen
uber die Selbststerilitat von Raphanus raphanistrum. Z. Indukt.
Abstamm.-u. VererbLehre. 87:365-384.
Lecoq, H. 1862. De la fecondation naturelle et artificielle des
vegetaux et de 1'hybridation. pp. 17-425. Paris: Maison
rustique. Quotes M. Riviere on Oncidium cavendishianum.
Lewis, D. 1943. The physiology of incompatibility in plants. III.
Autopolyploids. Jour. Genet. 45:171-185.
Lewis, D. 1949. Incompatibility in flowering plants. Biol. Revs.
Lewis, D. 1952. Serological reactions of pollen incompatibility
substances. Proc. Roy. Soc. B. 140:127-135.
Lewis, D. 1960. Genetic control of specificity and activity of the
S, antigen in plants. Proc. Roy. Soc. B. 151:468-477.
Lewis, D., and L. Crowe. 1958. Unilateral interspecific incompati-
bility in flowering plants. Heredity 12:233-256.
Linskens, H. 1955. Physiologische Untersuchungen der Pollenschlauch-
Hemmung selbststeriler Petunien. Z. Bot. 43:1-44.
Linskens, H. 1958. Zur Frage der Entstehung der Abwehr-Korper bei
der Inkompatibilitatsreaktion von Petunia. I. Ber. bot. Ges.
Linskens, H. 1959. Zur Frage der Entstehung der Abwehr-Korper bei
der Inkompatibilitatsreaktion von Petunia II. Z Bot. 48:126-135.
Linskens, H. 1960. Zur Frage der Entstehung der Abwehr-Korper bei
der Inkompatibilitatsreaktion von Petunia. III. Ber. bot. Ges.
Linskens, H. 1962. Die Anwendung der "Clonal Selection Theory" Auf
erscheinungen der Selbstinkompatibilitat bei der Befructung der
Blutenpflanzen. Pota. Acta. Biol. A6:231-238.
Linskens, H. 1963. Biochemistry of incompatibility. Proc. llth Int.
Cong. Genet. 2:629-635.
Linskens, H., and K. Esser. 1957. Uber eine spezifische. Anfarbung
der Pollenschlauche im Griffel und die Zahlder Kallosepfropfen
nach Selbstung und Fremdung. Naturwiss. 44:16-17.
Livermore, J. R., and F. E. Johnstone. 1939. The effect of chromosome
doubling on the crossability of Solanum chacoense, S. jamesii and
S. bulbocastanum with S. tuberosum. Amer. Potato Jour. 17:169.
Lundquist, A. 1956. Self-incompatibility in rye. I. Genetic control
in the diploid. Hereditas 42:293-348.
Lundquist, A. 1961 a. The nature of the two-loci incompatibility
system in grasses. I. Hereditas 48:153-168.
Lundquist, A. 1961 b. The nature of the two-loci incompatibility
system in grasses. II. Hereditas 48:169-181.
Lundquist, A. 1964 a. The nature of the two-loci incompatibility
system in grasses. III. Hereditas 52:189-196.
Lundquist, A. 1964 b. The nature of the two loci incompatibility
system in grasses. IV. Hereditas 52:221-234.
Martin, F. 1964. The inheritance of unilateral incompatibility in
Lycopersicon hirsutum. Genetics 50:495-469.
Martin, F. 1967. The genetic control of unilateral incompatibility
between two tomato species. Genetics 56:391-398.
Mather, K. 1943. Interspecific differences in Petunia. Jour. Genet.
McGuire, D. C., and C. M. Rick. 1954. Self-incompatibility in species
of Lycopersicon sect. Eriopersicon and hybrids with L. esculentum.
Mowbray, W. 1830. Trans. Hort. Soc. A letter to the editor.
Muller, F. 1868. Notizen uber die Geschlechtsverhaltnisse
brasilianischer Pflanzen. (A letter to Hildebrand.) Bot. Ztg.
Munro, R. 1868. On the reproduction and cross-fertilization of
Passifloras. Bot. Soc. Edinburgh 9:399-402.
Pandy, K. K. 1957. Genetics of incompatibility in Physalis ixocarpa.
Amer. Jour. Bot. 44:876-887.
Pandy, K. K. 1960. Self-incompatibility in two Mexican species of
Solanum. Nature 185:483-484.
Pandy, K. K. 1962. Genetics of incompatibility behavior in the
Mexican Solanum species, S. pinnatisectum. Z. f. Vererbungsl.
Pfundt, M. 1910. Einfluss der Luftfeuchtigkeit auf die Lebensdauer
des Blatenstaubes. Jahrb. wiss. Bot. 47:1-40.
Prell, H. 1921. Das Problem der unfruchtbarkeit. Naturwiss.
Wochenschr. N. F. 23:440-446.
Raptopoulas, T. 1939. Pollen germination tests in cherries. Jour.
Raves, A. N. 1921. Self-fertility and self-sterility in plums.
Jour. Roy. Hort. Soc. 46:353-356.
Roy, B. 1938. Studies on pollen tube growth in Prunus. Jour. Pom.
Hort. Sci. 16:320-528.
Russell, L. 1967. Personal communication.
Sachoff, T. 1931. Untersuchungen uber die Fruchtbarkeit der
Susskirschen, Sauerkirschen, Zwetschen und Pflaumensorten.
Scott, J. 1865. On the individual sterility and cross-impregnation of
certain species of Oncidium. Jour. Linn. Soc. London 8:163-167.
Schanderl, H. 1932. Untersuchungen uber die Befructungsverhaltnisse
bei Stein-und Kernobst in Westdeutschland. Gartenbauwiss 6:192-239.
Schlosser, K. 1961. Cytologische und cytochomische Untersuchungen
uber das Pollenschlauchwachstum selbststeriler Petunien. Z.
Schuster, C. E. 1922. Pollination of the sweet cherry. Oregon Agric.
Expt. Sta. Circular 27.
Schuster, C. E. 1925. Pollination and growing of the cherry. Oregon
Agric. Expt. Sta. Bull. 212.
Sharp, R. E. 1964. Personal communication.
Shell, C. 1967. Personal communication.
Smith, F. F. 1926. Pseudo-fertility in Nicotiana. Ann. Missouri Bot.
Sprenger, A. M. 1908. De Onvruchtbaarheid der Kersenboomen in Zuid-
Straub, J. 1941. Die Beseitigung der Selbststerilitat durch
Polyploidisierung. Ber. dt. bot. Ges. 59:296.
Stout, A. B., and C. Chandler. 1941. Change from self-incompatibility
to self-compatibility accompanying change from diploidy to
tetraploidy. Science 94:118.
Sutton, I. 1918. Report on tests of self-sterility in plums, cherries,
and apples at the John Innes Hort. Inst. Jour. Genetics 7:281-300.