, ”I wag“???
. z ‘ .. '
:1? Ft " " “1‘94 2* $54!???”

‘7'
‘-

235;;‘14231 ,. ' , ‘ a" » ‘1 25‘?”
bf "i‘g’v’h 4. ‘9' .13: £3 ‘ '
'5' ~"‘f\‘§§"“ A “A W,

1

33%: gym 3k
,3 .3; Km; "’4
13" .
, .1.“
3:33.;

'1
532?

‘55 $29. L ._
M £53 ‘

'I

i 4 (z (I?L.:Hfi
”II-um x
M. -. jib)“ "Vt/‘3}! 1’”.

-‘£;.m"~' .,

47 " w

I‘ ‘23 .
¥. 0 -
iiziJ‘f‘; “r

6?

?I4
4’;

.) 31:32:}
‘1]:

" ‘If
$333,. 4'4'369“ If'
raj-Q’Hf/‘aft éygf‘: ‘1',
Vi. 'f' “(I’ll-f '1'”?
- . @fl
‘9" 1. 1“” ~ 502%;11;
: .U ‘g-f: 'fifr' "Hi-1:);
I : (r

44.1» "_ “I

‘1: Y!
14.577"
4:, 5443' {£33}
«.4 If"

n-opn "' 1
_ :63; .fzfir: :5? v J "'
:3‘ ‘ firr'ioeélsr'a ”:1th
am”? «gyfl
." 4.:
griffin g 14%;"?!
I," I‘I 7,

gu'afi'
y'd'lm ’
4%

o
.v
1

v I/ ,3¢. 2'
. .21,“ ’ "120»

V. Err-5 19%

2* 4;”

f:
‘I0 is;

o
“07’

 

aura/«3

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the
thesis entitled

POLLINATION, SEED SET AND POLLEN TUBE GROWTH INVESTIGATIONS
IN VIOLA PEDATA L.

 

presented by
Wayne Alan Becker

has been accepted towards fulfillment
of the requirements for

Masters degree in Horticulture

 

 

' a
"/ ‘i :1 /'/ I)
, // 1- I
' V I , ,v 'i— '
a 1’6: .. , » «fl.

Major professor

 

Date 8//&/ g g

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

___—___.i — _. _ —_ ._—.——_._.__44_._.~_fi

 

 

MSU

LIBRARIES
.——

‘—

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

 

 

POLLINATION, SEED SET AND POLLEN TUBE GROWTH

INVESTIGATIONS IN ZIQLA EEQAIA L.

BY

Wayne Alan Becker

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

Master of Science

Department of Horticulture

1988

ABSTRACT

POLLINATION, SEED SET AND POLLEN TUBE GROWTH
INVESTIGATIONS IN ZIQLA REQAIA L.

BY
Wayne Alan Becker

Plants of 2121; pggata collected from the wild in‘
Michigan, Arkansas, Tennessee and Wisconsin, produced very
low amounts 'of seed when grown in a greenhouse and self-
pollinated. The objectives of these investigations were:
1) to determine the reason(s) for low (self) seed set and:
2) to determine how seed set could be increased.

Pollen viability measured by cotton blue staining,
pollination and seed set studies, and pollen and pollen tube
observations using fluorescence microscopy were employed.
Cross-pollinations resulted in significantly greater
quantities of seed than did self-pollinations, and many
fewer pollen tubes gained entry to the ovaries in self--
pollinated pistils compared to cross-pollinated pistils..

These investigations indicate that a self-
incompatibilty system is operative in 21913 pedgtg, that
this is responsible for the low (self) seed set values, and

that cross-pollination can increase seed set.

DEDICATION

This work is dedicated to the memory of my mentor and
beloved grandfather, Gilbert J. Hoehn. He brought great joy
to his family and friends, made the best wild strawberry ice

cream, and was the horticultural envy of Ottoville, Ohio.

ACKNOWLEDGMENTS

I wish to extend my appreciation to the people who so
kindly assisted in this research and manuscript
preparation. To Dr. Lowell Ewart, a most patient major
professor, for introducing me to 2191; pgdata, and for
allowing me to investigate its wonders. To my committee
members, Dr. Joanne Whallon and Dr. Art Cameron, and to my
colleagues, Candice Shoemaker and Steve Krebs, for their
ideas, advice, and friendship. To my loving families,. for

their support and understanding.

iv-

TABLE OF CONTENTS

Page
List of Tables.......................................... v
List of Figures..................... ........ ............ vi
Introduction...... ...... . ............................... 1
Literature Review.................................. ..... 3

materials and Methads O O O O O O O I O O O O O O O O O O O O O O O O 0 O O 0 O O O O O O 10
Plant Material and Culture... ......... ............. 10
Experimental Methads O O O O O O O O O O O O O O O O O O O O 0 O ........ O 12

Results................................................. 19
Experiment 1. Anther Dehiscence................... 19
Experiment 2. Pollen Viability.................... 19
Experiment 3. Seed Set............................ 19
Experiment 4. Fluorescence Microscopy

Investigations........... ..... . 31

Discussion...................................... ....... . 40
Anther Dehiscence.................................. 41
Pollen Viability................................... 41
Seed Set........................................... 42
Fluorescence Microscopy............................ 45
Self-incompatibility in Viola pgdata.... ......... .. 55
Late-acting Self-incompatibility and

Ovarian Inhibition............................ 56

sumaWOOOOOOOOOOOOOOOOOCOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO. 59

Appendices
Appendix 1. Average seed set of self-pollinated
flowers, by plant............................. 61
Appendix 2. Average seed set of cross-pollinated
flowers, by plant............................. 62

Bibliography............................................ 63

Table

1.

9.

10.

11.

12.

LIST OF TABLES

Page

Anther dehiscence in flowers of Arkansas and
WisconSin plants..0.000000000000000000IOOOOOO0000. 20

Pollen viability measured by cotton blue
staining................................... ....... 20

Seed set without artificial pollination on
flowers from plants studied................. ...... 23

Seed set without artificial pollination on
emasculated flowers of Arkansas and Wisconsin
plantSOOOOOOOOOO0.0.000...OOOOOOOOOOOOOOOOOOOOOOOO 23

Effect of flower age at pollination on self

seed set for Arkansas and Wisconsin plants. ....... 23
Effect of wet vs dry stigmas on self seed set

of flowers on Arkansas and Wisconson plants ....... 25
Self seed set on flowers pollinated when

their stigmas became wet................... ....... 25
Self seed set of F1 plants................... ..... 32

Pollen and pollen tube characteristics in
self-pellinated pistils. O O O O O O O C O O O O O O O O O ......... 33

Pollen and pollen tube characteristics in
cross-pollinated pistils.......................... 34

Number of pollen tubes in the ovaries of self-
and cross-pollinated pistils...................... 37

Average ovule counts for plants from groups
studied.OOOOOOOOOOOOOOOOOOOOOOO0.0.000...O0.0.000. 39

vi

LIST OF FIGURES

Figure Page
1. 2191; negate plant with bicolor flowers.. ......... 5
2. Concolor flower with narrow petals.. ....... ....... 5
3. Concolor flower with wide petals ...... . ........... 6
4. Fruit capsule 25 days after pollination..... ...... 6

5. Dehisced fruit from which several seeds have
been ejectedOOOOOOOOOOOOOOOOO...OOOOOOOOOOOO..0... 8

6. Seed from fruit shown in Figure 5........ ..... .... 8
7. Pollen stained with cotton blue. 4OOX ..... ....... 21

8. Average seed set of self- and cross-pollinated
flowers by group.................................. 28

9. Self seed set distribution ................... ..... 29
10. Cross seed set distribution.......... ............. 29

11. Style (left) and upper ovary (right) of a pistil
with glow of fluorescing pollen tubes in style.

ZOXOOOOCOOOCCOOOOOIOOOO...OOOOOOOOOOOOOOOOOOOOOOO. 46

12. Pistil in Figure 11, split open; most pollen
tubes stop at top of ovary. 20X.................. 46

13. Pistil in Figure 12, slightly squashed; greater
number of tubes apparent. 20X.................... 48

14. Pistil in Figure 13, squashed further. 20X....... 48

15. Cross-pollinated pistil showing many pollen
tubes in the ovary. 20X..................... ..... 49

16. Self-pollinated pistil in which pollen tubes
have stopped at top of ovary. 20X......... ....... 49

vii

17.

18.

19.

20.

21.

22.

23.

24.

viii

Self-pollinated pistil in which a small
percentage of pollen tubes have entered ovary.

20x00...00.00.00.000...0....OOOOOCOOOOOOOOOOOOOOO.

Self-pollinated pistil in which two pollen tubes
can be seen entering ovules. ZOXOOOOOOOOOOOOOOOOO

Ends of pollen tubes in a self-pollinated
pistil. IOOXOIOOOOOOOOOOOOIOOOOOOOOOOOOOOO .......

Ends of pollen tubes in a self-pollinated
pistil. IOOXOOOOOO0.0000000000IOOOOOOOOOOOOO .....

Cross-pollinated pistil with pollen tubes
having grown along placenta to ovules. 20X.......

Cross-pollinated pistil with pollen tubes along
placenta and in many (detached) ovules. 20X ......

Pollen tubes entering ovules. 100x...............

Self-pollinated pistil of an F1 plant with many
pollen tubes, all stopping at top of ovary.

ZOXOOOOOOIOCOOOOOOOOOO...IOOOOOOCOOOOOOOOOOOOOO0..

51

51

52

52

53

53

54

54

W

2191; pedata L., the birdfoot violet, is a perennial
wildflower native to the eastern half of the United States..
Its aesthetic qualities indicate that it could become a more
widely utilized garden plant and/or be developed into a pot
plant. Presently it is sold as a vegetatively propagated
crop through specialty nurseries, which may propagate the
plants in-house and/or rely on plants collected from the
wild. Inconsistent availability is frequently a problem. It
would be of benefit to the ornamental horticulture industry
and to the gardening public if seed-produced cultivars could
be developed. In addition, 2‘ negate is listed as
endangered in several states. A greater knowledge of its
reproductive biology would aid in its preservation and.
reestablishment.

Preliminary attempts to produce (self) seed from a
single specimen of y&,pgga§a collected in Michigan, 'and
from a group of plants collected in Arkansas, met with low,
inconsistent seed set. Such low seed yields would hinder
breeding efforts and commercial seed production.

This thesis reports initial investigations into the
sexual reproduction of 21 negate, centering on pollination,
pollen tube and seed set studies. The objectives of this

1

2

research were 1) to determine the reason(s) for low (self)
seed set and 2) to determine how seed set could be

increased.

't a ev'

2191a negate, the birdfoot violet, is a member of the
Violaceae family, section Nominium, sub-section Plagiostigma
and group Pedatae (22). It is described as monotypic, very
distinct and without any close relatives (30).
Distinguishing characteristics [e.g. the massive style and
its unique form, a complete absence of cleistogamous
flowers, the presence of adventitious buds on the roots, and
a unique chromosome number of 2N=56 (16)] distance x; peggga
from other members of the Plagiostigma sub-section. Gershoy'
notes that, phylogenetically, y; negate is probably most
closely related to the species of the Boreali-Americanae and
Adnatae groups (22).

The morphology of z; pggata has been described by a
number of authors (8, 13, 31). The following is a
compilation of their descriptions, along with observations
made by the author.

1‘ negate can be found growing in sunny areas on sandy,
well-drained, acidic soils. It is a stemless blue violet
with a rosette of leaves growing from a short, thick,
vertical rhizome. The leaves are deeply cut and palmately.
divided into 3-5 sections, each section being cleft or
toothed near the apex. y; pggata is one of the most

polymorphic violets and many additional names have been
3

4

assigned to the various forms. Different leaf forms may be
found on the same plant at the same time and/or at different
times of the year (31).

The flowers of y; pggata are stalked, solitary and
nodding, with a diameter of 2-4 cm (Figure 1). The calyx
consists of five sepals and the corolla of five petals.
Four of the five petals are arranged in two pairs, each pair
differing in shape. The fifth, lower petal is spurred and
beardless. There are two principle types of coloration. The
concolor type has all five petals one shade of blue (Figures
2, 3). In the bicolor type, the upper two petals are a dark
purple and the lower three are blue (Figure 1). Thin, dark-
purple markings may pattern any or all of the petals,
appearing as lines originating at the base of the petal(s).

The five stamens have very short filaments and form a
tube enclosing the ovary and part of the style. Orange
connective tissue extends from the distal end of the anthers‘
and closes tightly around the style.

The pollen is shed on the inside of the staminal tube,
falls into the groove of the lowermost petal and is retained
there. This groove leads to the spur of the petal
containing the nectar. The two nectaries are appendages of
the two lowermost stamina.

The stigmatic surface is cupped, with a lip on the
lower side, and ringed with papillae. The style is hollow,
with a very flexible hinged area just above the ovary. The

ovary is one-celled with many ovules.

  

  

 

Figure 1.

2191; pggata plant with bicolor flowers.

 

Figure 2.

Concolor flower with narrow petals.

I
:'..l.--

 

Figure 3. Concolor flower with wide petals.

 

Figure 4. Fruit capsule 25 days after pollination.

7
The fruit is an ellisoid capsule and dehisces along

three sutures (Figures 4, 5). The seed are light to dark
brown, smooth, round to ellipsoid, and approximately 1 mm in
diameter (Figure 6). The seed are forcibly ejected from the
dehisced capsule.

Beattie and Culver stated that the flowers of y;
negate failed to set seed in the abscence of insect visitors
in field and greenhouse trials (7). Beattie described 2;
pedaga as a self-compatible species and stated that insect'
pollination resulted in varying degrees of self— and cross-
pollination depending on the type of insect, plant spacing,
the number of flowers present on individual plants, and
other factors (6).

Gershoy reported that "the capacity for self-
fertilization is retained for at least five days in many
(1191;) species, as determined by hand-pollination of old
blossoms" (22). Kristofferson reported that thrips
pollinated 2; 3319919: flowers as a result of living in the
spur petals and the stigmatic cavity (26).

In a study of approximately 75 1121; species, including.
11 negate, Gershoy observed that, upon artificial self-
pollination, the pollen tubes reached the upper portion of
the ovary cavity, through the hollow style, within 14 to 16
hours. Gershoy also mentions that the species used in these
studies were grown for several generations using self seed
(21).

There is no consistent evidence of the occurrence of

 

l

Q11

Figure 5. Dehisced fruit from which several seed have been
ejected.

 

Figure 6. Seed from fruit shown in Figure 5.

9
self-incompatibility in the violets (Viola sp.) (21, 22,.

37). However, Emino, investigating low seed set among
twenty-nine inbred breeding lines of y, trigglg; Hortensis
L., reported the possible existence of a self-
incompatability system (19). According to Emino, cross-
pollinations usually resulted in greater seed set than did
self-pollinations, especially when low self seed set types
were used as male parents in crosses onto other lines.
Emino stated that the increased seed set in cross
pollinations was not consistent (i.e. for some inbred lines,
cross-pollinations with certain other lines did not result
in ’higher seed set compared to the self seed set of one
and/or both of lines involved) and that the problem of low.
seed set is therefore complex in nature. Pollen fertility,
differing chromosome numbers, and meiotic division did not
appear to contribute to low seed set among the inbred lines

studied.

W

W

The 2‘ pggata plants were obtained from four geographic
areas. One concolor plant was collected in Michigan (MI
group). Nearly one hundred were collected by the author in
March of 1986 in the vicinity of Hot Springs Village,
Arkansas (AR group). They were very heterogeneous in
appearance, approximately half being concolors and half
bicolors. Six concolor plants were obtained from a nursery-
in Nebraska, but had originally been collected in Tennessee
(TN group). Their genetic relationship is unknown. One
hundred plants (50 bicolor, 50 concolor) were purchased from
a nursery in Wisconsin (WI group), were native to that
state, and were more uniform in appearance than the Arkansas
plants. Plants of the WI group had been propagated by the
nursery both sexually and asexually so that they represented
a number of genotypes and their clones, and reportedly set
seed regularly when grown out in beds (Personal
Communication: Prairie Nursery, Westfield, WI).

The F1 plants were grown from seed produced from.
cross-pollinations between and among plants of the MI and AR
groups. No guidelines for the germination of y; pggata seed
were found in the literature. Attempts to germinate both

self and cross seed were made using recommendations for
10

11
other 2191; species including such treatments as 0.2% KNO3,

100-400 ppm 6A3, light, pre-chill and scarification (1, 3,
18).

All plants were propagated asexually by division when
possible to increase the number plants to provide more‘
flowers for pollinations. Each plant and its clones were
given the same identification number.

Since the literature was void of greenhouse cultural
requirements for y; negate, the following procedures were
utilized. The plants were potted up in 4 or 6 inch clay
pots, depending on the size of the plant. A soilless mix (2
peat : 1 coarse sand : 1 pea gravel) was used, along with an
inch of pea gravel in the bottom of the pot for adequate
drainage. The plants were kept in the greenhouse on raised
benches. In the winter the greenhouse temperature was set
at 20°C, day and night. In the summer, a fan and pad‘
cooling system was used to keep the greenhouse as cool as
possible during the day. Night temperatures were allowed to
drop as low as 10° C during the spring, summer and fall.
Supplemental lighting (high pressure sodium vapor) was
applied from November to March of 1986-87 to keep the
daylength at a minimum of 10 hours in order to encourage
growth and flowering. The plants were fertilized every
three months with a 300 ppm 20-20-20 soluble fertilizer.
Iron sulfate was applied to keep the growing medium at a pH
of 3 to 4. A fungicide (Benlate) was applied to control
Bhizggtgnia spp. root and stem rot, and the plants were.

12

watered only when the top inch of growing medium was dry.
Pesticides (Temik and Metasystox-R) were applied as
recommended to control thrips and red spider mites. Flowers
which bloomed within two days after pesticide spraying were

not used.

W

The availability of flowers on plants from the various
groups determined the extent of each group’s involvement in‘

the following experiments.

WW. To determine when anther

dehiscence occurred, observations in a different set of 5
flowers each from of the AR and WI groups were made at each
of 4 stages of flower development: 1.) petals still closed,
color well-defined, flower just about to open: 2.) petals
open approximately halfway: 3). anthesis: 4.) one day
after anthesis. The number and position(s) of dehisced

anthers was recorded.

BMW. To determine whether poor
pollen viability was a factor contributing to low (self)

seed set, 100 pollen grains each from 3 flowers per plant,
from twelve different plants, were scored for viability
using cotton blue stain (equal parts phenol crystals, lactic
acid, glycerine and distilled water) (24). Plants were

chosen which represented a range of average (self) seed set

values.

13
A mixture of pollen from all five anthers of each’

flower was collected on the day of anthesis. A quantity of
pollen was transferred onto a drop of stain on a slide, and
a coverslip was placed over the drop. Observations were
made after 10 minutes, using a compound microscope (Olympus
model BHS) at 10X. The pollen grains were scored either as
plump and darkly-stained (indicating viability) or as

misshapen and lightly-stained.

W. The following pollination and

seed set measurement techniques were used in all seed set
experiments.

Pollinatign. For self-pollinations, pollen from the‘
anthers and/or groove of the spur-petal was transferred to
the stigma with a blunt needle probe. The pollen was placed
into the stigmatic cavity and around its perimeter.

Cross-pollinations were made in a similar manner. Using
a fine forceps, the female parents were emasculated prior to
anthesis by either removing the spur petal and anthers
(Experiment 3b), or the spur petal alone (Experiment 3f).

Each flower was pollinated only once. The flowers were
tagged and pollinated either on the day of anthesis or one
day after anthesis unless otherwise indicated. All-
pollinations were made in the greenhouse which was equipped
with screens to exclude insects which might effect

extraneous pollinations.

Segd_§§§_uea§urem§n_. The flowers were scored for seed

14

capsule formation two weeks after pollination. Four weeks
after pollination, the capsules were collected, allowed to
dry at room temperature for aproximately one week, and the

seeds then counted.

Experimgn;__1§. To determine the extent of seed set on
flowers which were not artificially pollinated, observation
were made of 100 flowers each from the AR and WI groups, and
20 each from the TN and MI groups. The flowers were tagged
as buds and allowed to mature without further disturbance

until being scored for seed capsule formation.

fixpg;1mgn§__1§. To determine if agammospermy could be

responsible for any seed set, and to evaluate the efficacy
of the emasculation technique, 50 flowers each from the AR
and WI groups were emasculated, as buds, by removing the
spur petal and the anthers. The emasculted flowers were
covered with small glycine bags to prevent extraneous

pollination.

Experimgn;__gg. To determine the effect of flower age at
pollination on seed set, 40 flowers each from the AR and WI
groups were self-pollinated on one of the five days
beginning at anthesis. The pollinations for each day were

made between 10 am and 12 noon.

Experimgn;_1d. To determine when stigmatic fluid exuded and
to determine the effect of stigmatic fluid on seed set, the

following two methods were used. (Stigmas on which a drop

15

of stigmatic fluid could be seen are referred to as "wet";
"dry" stigmas had no visible stigmatic fluid.)

Experimgn§__3glil. Fifty flowers each from the AR and
WI groups were self-pollinated at each 10 am and 3 pm on the
day of anthesis. A record was kept of whether or not the
stigma was wet when pollinated.

Experimgg§_1gliil. Fifty flowers each from the AR and
WI groups were observed at 10:00 am and 3:00 pm on the five‘
days beginning at anthesis. If the stigma was wet, the
flower was self-pollinated. Flowers were not pollinated if

their stigmas remained dry for the five days.

fixpg;1mgn;__3g. To evaluate the self-compatability of each
of the plants in all groups, 10 flowers per plant were self-
pollinated. The average seed set per pollination was used

as a measure of each plant’s self-compatibility.

Bxpg;1mgn§_1£. To evaluate the cross-compatibility of each
plant, and to compare cross vs self seed set, random cross-
pollinations were made among and between many of the plantS'
from all four groups. Each plant was crossed separately
with each of at least 3 different male parents. The average
seed set per pollination was calculated from the total seed
produced by at least 3 flowers per plant, and was used as a

measure of each plant's cross-compatibility.

fixpg;1mgn§_1g. To evaluate the self-compatability of the F1
plants, and to determine likelyhood of inbreeding depression

as a factor contributing to low self seed set of the

16
parental plants, up to 10 flowers from each of the F1 plants

were self-pollinated. The average seed set per pollination

was used as a measure of each plant's self-compatibility.

Experimeg;_4a(11. To observe pollen and pollen tube growth

in pistils of flowers from plants of all groups, the
flowers were either self- or cross-pollinated and then
harvested after 24, 48, 72, 96, 120 or 144 hours.
Fluorescence microscopy, from a method described by Rho and
Baer with some modification, was used to observe pollen and'
pollen tubes in the pistils of the harvested flowers (25).
This procedure is based on the fact that the pollen tubes
contain callose, a material which selectively takes up
aniline and consequently fluoresces when illuminated by blue
or ultraviolet light. With this technique, the pollen tubes
fluoresce bright green against a dark background, especially
under ultraviolet light.

The pistils were carefully dissected from the flowers
and placed in 8N NaOH to clear and soften. After 24 hours,
the pistils were rinsed in distilled water for 6 to 24
hours. The pistils were then transferred to a solution of-
0.1% aniline w.s. dissolved in 0.1 N K3PO4 for at least 15
minutes, or until the stain had thoroughly penetrated the
pistil. A stained pistil was placed on a slide in a drop of
.stain, and then split open longitudinaly, using a thin,

sharp razor. The two halves were laid out so that the cut

17

surfaces faced upward.

Observations were made under a binocular microscope
(Olympus model BHS) equipped with a high pressure mercury
vapor lamp. The lamp, along with the proper blue and UV
barriers and filters (Olympus L-420, 0515), provided the
required light waves between 350 and 400 mu. Using a 2X
objective, the pistil was examined for pollen quantity and
germination, pollen tube growth and probable fertilizations.
A coverslip was then placed over the pistil and slight
pressure applied. Further observations were then made under
higher magnifications (10 and 20X).

The quantity of pollen in the stigmatic cavity of each
pistil was categorized as: 0.) no pollen: 1.) one to
approximately 100 pollen: and 2.) approximately 100+ pollen.
The quantity of germinated pollen was categorized as 0.) no
germinated pollen: 1.) one to approximately 100 germinated'
pollen; and 2.) approximately 100+ germinated pollen.

Pollen tube growth in each pistil was categorized as:
0.) no pollen tubes or tubes shorter than diameter of
pollen: 1.) tubes within the style; 2.) tubes at the bottom
of the style/top of the ovary; 3.) tubes in the ovary; 4.)
tubes penetrate ovules. The number of probable
fertilizations per pistil was determined by counting the

number of ovules into which pollen tubes had penetrated.

Experiment_4g(iil. To observe pollen and pollen tube growth

in pistils of the F1 plants, flowers were selfed, crossed or

back-crossed and harvested after 144 hours. Observations

18

were made using' the fluorescence microscopy technique

described above for Experiment 4a(i).

Experim§n§__4b. To estimate potential seed set, ovule
counts were obtained from many of the specimens prepared in
Experiment 4a. The average number of ovules per ovary for
each plant was used as a measure of its seed set potential.
The averages were calculated from counts of the total number~

of ovules in at least 3 different ovaries per plant.

5! !l !' J E J . T] HE!!! Hi ! E! !' !° 1
fizggnam (MSTAT Development Team, 1986) was used for all
statistical analyses. The "Manager", "Sort", and "AOV"

(ANOVA-l) programs were utilized.

u ts

W. There was little

observable difference between the AR and WI groups regarding
anther dehiscence (Table 1). Dehiscence usually began
before the flower opened (Stage 1; tight bud showing color)
and was completed one day after anthesis (Stage 4). Flowers
went through the 4 stages of development in 2-3 days. The.
anthers dehisced in a specific order, beginning with the
anther situated between the upper pair of petals. Next, the
two anthers adjacent to this upper anther would dehisce,

followed by the lower two anthers.

WW- Pollen viability as
determined by cotton blue staining was high in all tested

plants. Average counts ranged from 75 to 97 plump, darkly-
stained pollen out of 100 (Table 2). One pollen grain in
Figure 7 appears shrunken and lightly-stained, and is
typical of grains scored as non-viable; the pollen scored_

as viable appear round, plump and darkly-stained.

2W.

Experiment__3a. For flowers which were not artificially

pollinated, the average seed set for all groups combined was

19

20

Table 1. Anther dehiscence in flowers of Arkansas (AR) and
Wisconsin (WI) plants.

Avg. No, of Dehisced Anthezsl
Stage of Flower

Development AR WI Combined
1. Tight bud/color apparent 1.2 0.6 0.9
2. Open halfway 1.4 1.4 1.4
3. Open completely 3.8 3.8 3.8
4. 1 day after stage 3 5.0 5.0 5.0

 

IFrom observation of 5 flowers.

Table 2. Pollen viability measured by cotton blue staining
and average self seed set of selected plants.

 

W1

. Ave. Self

Plant l. Counts . Ave. Seed Set
1 90 95 94 93.0 3.2
8 84 89 91 88.0 0.3
16 96 91 93 93.3 0.0
27 - 74 68 83 75.0 2.9
61 98 97 96 97.0 1.1
65 72 88 75 78.3 0.7
75 86 97 91 91.3 0.0
113 73 77 81 77.0 0.5
168 96 79 83 86.0 6.4
187 97 96 94 95.7 15.4
216 89 95 79 87.7 1.2
230 96 89 94 93.0 8.1

 

iCounts of 100 pollen grains from 3 flowers.

21

 

 

   

Figure 7. Pollen stained with cotton blue. 400x.

22
0.5 seed per flower (Table 3). The WI group had the highest

average seed set (0.9 seed per flower) and the MI group had
the lowest (0 seed per flower). There was no significant.
difference (p a 0.1) in seed set between the groups.

No seed or capsules were set on flowers from the MI
group. The seed sets between flowers ranged from 0 to 4
seed on those of the AR group, from 0 to 9 seed on those of
the TN group, and from 0 to 13 for those of the WI group.
Seeded capsules formed on 5%, 10% and 20% of the AR, TN and
WI flowers respectively. Seedless capsules formed on 3%,

20% and 1% of the AR, TN and WI flowers, repectively.

Experimgn§__3p. No seed were produced on any of the
emasculated flowers from either group. Seedless capsules
formed on 8% the AR flowers, and no capsules formed on the.

WI flowers (Table 4).

Experimgn;__1g. For self-pollinations made on one of the
five days beginning at anthesis, there was no significant
difference (p - 0.1) in seed set between days for either the
AR or WI groups. Over the five days, the seed sets between
flowers ranged from 0 to 12 seed on those of the AR group
and from 0 to 37 seed on those of WI group. Seeded capsules
formed on 15% of the AR flowers and on 55% of the WI
flowers. Seedless capsules formed on 6 % of the AR flowers
and on 3% of the WI flowers.

For each day, there was a significant difference (p =

0.01) in seed set between groups (Table 5). For all days

23

Table 3. Seed set without artificial pollination on flowers

from plants studied.
No. of

No. of Total Total Seed Seed Capsules Ave. Seed

Group Flowers Seed Capsules with No Seed per Flwr.
MI 20 0 0 0 0.0
AR 100 12 8 3 0.1
TN 20 12 6 4 0.6
WI 100 89 21 1 0.9
0.5

All 240 113 35 8

 

iVarious plants from Michigan (MI), Arkansas (AR), Tennessee
(TN), and Wisconsin (WI) groups.

Table 4. Seed set without artificial pollination on
emasculated flowers of Arkansas (AR) and Wisconsin
(WI) plants.

No. of Total Total seed No. of Capsules

Group Flowers Seed Capsules with No Seed
AR 50 0 4 4
WI 50 0 0 0

 

Table 5. Effect of flower age at pollination on self seed
set for Arkansas (AR) and Wisconsin (WI) plants.

Flower Age No. of Total Ave. Seed
Group at Pollination1 Observations Seed per Flower
AR 1 40 23 0.6
2 4O 30 0.8
3 4O 30 0.8
4 4O 7 0.2
5 40 9 0.2
WI 1 40 251 6.3
2 40 252 6.3
3 40 215 5.4
4 40 253 6.3
5 40 188 4.7

 

iIn days: 1=day of anthesis.

24

combined, the average seed set was 0.5 seed per pollination
for the AR group and 5.8 seed per pollination for the WI

group.

Expg31m§n3_1§111. For flowers self-pollinated on the day of
anthesis at 10 am and 3 pm combined, there was no
significant difference (p s 0.1) in seed set between flowers.
which had wet stigmas when pollinated compared to those
which had dry stigmas when pollinated, for either the AR or
WI groups. Combined over time of pollination, only 7% of
the AR flowers and 8% of the WI flowers had wet stigmas on
the day of anthesis (Table 6).

Over time of pollination, the average seed set of the
AR group was 1.4 seed per pollination for wet stigmas and
0.8 seed per pollination for dry stigmas: the seed sets
between flowers ranged from 0 to 11 seed on those which had
wet stigmas and from 0 to 23 seed on those which had dry
stigmas. Over time of pollination, the average seed set of.
the WI group was 4.1 seed per pollination for wet stigmas
and 5.0 seed per pollination for dry stigmas: the seed sets
between flowers ranged from 0 to 16 seed on those which had
wet. stigmas and from 0 to 39 seed on those which had dry
stigmas.

For all flowers which had wet stigmas, seeded capsules
formed on 15% from the AR group and on 13% from the WI
group: seedless capsules formed on 42% from the AR group and
on 22% from the WI group. For all flowers which had dry

stigmas, seeded capsules formed on 19 % from the AR group

25
Table 6. Effect of wet1 vs dry2 stigmas on self

seed set of

flowers on Arkansas (AR) and Wisconsin (WI)

plants.
Time of Mouse Totaljeed
Group Pollination3 Wet Dry Wet Dry
AR 10 a.m. 2 48 0 55
3 p.m. 5 45 11 19
WI 10 a.m. 2 48 1 284
3 p.m. 6 44 32 175

Ave. Seed

pg: Flower
Wet Dry
0.0 1.2
2.2 0.4
0.5 5.9
5.3 4.0

 

iStigmatic fluid visible on stigma.
2No stigmatic fluid visible on stigma.
3On day of anthesis.

Table 7. Self seed set on flowers pollinated
stigmas became wet .

Flowgr Time of No. of Wet Total

when their

Ave. Seed

 

Group Age Pollination Stigmas Seed per Poln.
AR 1 10 a.m. 3 5 1.7
3 p.m. 7 o 0.0
2 10 a.m. 5 17 2.6
3 p.m. 13 5 0.9
3 10 a.m. 4 0 0.0
3 p.m. 7 1 0.1
4 10 a.m., 5 9 1.8
3 p.m. 3 o 0.0
5 10 a.m. 0 - ---
3 p.m. 1 o 0.0
Remained dry 2 0 0.0
50
WI 1 10 a.m. 0 - ---
3 p.m. 0 - —--
2 10 a.m. 6 20 3.3
3 p.m. 7 24 3.4
3 10 a.m. 7 28 4.0
3 p.m. 7 35 5.0
4 10 a.m. 2 16 8.0
3 p.m. 6 8 1.3
5 10 a.m. 0 - ---
3 p.m. 1 o 0.0
Remained dry 14 0 0.0
50
*Stigmatic fluid visible on stigma.
2At pollination, in days from anthesis: 1=day of anthesis.

26

and on 45% from the WI group: seedless capsules formed on

10% from the AR group and on 5% from the WI group.

Experimgn§__1gliil. For all days and times combined,
pollination on only wet stigmas did not result in

significantly different (p = 0.1) seed set as compared to
that resulting from the pollination on stigmas whether wet
or dry (from the results of Experiment 3c; Table 5).
Ninety-six percent of the AR flowers produced wet stigmas,
76% occurring after the day of anthesis. Seventy-two
percent of the WI flowers produced wet stigmas, all of which‘
occurred after the day of anthesis (Table 7).

The overall average seed set for flowers with wet
stigmas was 0.8 seed per pollination for the AR~ group and
3.6 seed per pollination for the WI group: the seed sets
between flowers ranged from 0 to 13 seed on those'of the AR
group and from 0 to 17 seed on those of the WI group.

For all flowers which had wet stigmas, seeded
capsules formed on 17% from the AR group and on 58% from the
WI group: seedless capsules formed on 15% from the AR group
and on 11% from the WI group. For day of pollination
combined over time of pollination, there was no significant'
difference (p a 0.1) in seed set between days for either the

AR or WI groups.

Experiment__1e. The average self seed sets between plants
ranged from 0 to 8.3 seed per pollination for those of the

AR group, from 0 to 18.7 seed per pollination for those of

27
the TN group, and from 0 to 21.2 for those of the WI group:

the MI plant averaged 3.2 seed per pollination (Appendix 1).
The overall average seed set was 3.2 16.38 seed per
pollination (Figure 8). No seed were set on 32%, 30% and 7%
of the plants from the AR, TN and WI groups respectively.

Between flowers, seed sets ranged from 0 to 13 seed on
those on the MI plant, from 0 to 23 seed on those of the AR
group, from 0 to 34 seed on those of the TN group, and from
0 to 48 seed on those of the WI group: 63% of the
pollinations resulted in no seed, and 18% in from 1 to 5
seed' (Figure 9). Seeded capsules formed on 40%, 21%, 23%
and 49% of the flowers for the MI, AR, TN and WI groups
respectively. Seedless capsules formed on 0, 37%,. 1% and
16% of the flowers for the MI, AR, TN. and WI groups.
respectively.

There was a significant difference (p - 0.01) between
the average seed set values of the AR and WI groups. The
average seed set values of the TN group were not
significantly different (p = 0.1) from either the AR or WI
groups. For these analyses, the averages were transformed
using the equation (Y + 0.5)1/2 (Little and Hills, 1978) to
reduce the heterogeneity of the variances, caused by the
fact that a large proportion of the averages were 0.0 seed
per pollination (Appendix 1). The single MI plant was

excluded from these analyses.

Experimgn§__31. The average cross seed set between plants

ranged from 5.4 to 35.4 seed per pollination for those of

28

 

 

 

 

 

4a... _ .. ..
42- F __

«5 36d 4..

(O

’D .30-

3% 2. ‘ i l ' l

'25: I

iv 12- _L .L

< 5" 4L -L _L
.. l 1

 

 

 

Ml AR TN WI ALL
Group
Group Self Averages Cross Averages
MI 3.2 $4.57 27.9 $18.06
AR 1.1 $3.15 20.6 1:14.69
TN 3.0 17.39 27.7 $20.01
W! 4.5 17.38 29.0 $18.57
ALL 3.2 $6.36 23.8 1.1 6.76

 

1Michigan (MI). Arkansas (AR). Tennessee (TN). Wisconsin (WI)

Figure 8. Average seed set of self- and cross-pollinated flowers by
group.

29

Number of Seed

 

 

   

65101520253 3540455 5 6065
X of Self-pollinations

Figure 9. Self seed set distribution.

 

Number of Seed

 

 

0 5 1'0 1'5 2'0 2'5 3'0 3'5 4'0 4'5 5'0 5'5 6'0 5'5
z of Cross-pollinations

Figure 10. Cross seed set distribution.

30
the AR group, from 9.7 to 47.7 seed per pollination for

those of the TN group, and from 7.0 to 53.6 seed per
pollination for those of the WI group: the MI plant averaged
27.9 seed per pollination (Appendix 2). The overall average
seed set was 23.8 $16.76 seed per pollination (Figure 8).
Seed were set on all plants from all groups.

For pollinations on flowers from all plants from all
groups combined, the individual seed set values ranged from
0 to 73 seeds: 17% of the pollinations resulted in no seed,
and 51% in 15 to 40 seed (Figure 10). For the MI, AR, TN
and WI groups, seeded capsules formed on 80, 81, 88 and 88%
of the flowers, respectively: seedless capsules formed on 0,
10, 0 and 1% of the flowers, respectively.

There was a significant difference (p z 0.01) in the
average seed set values between the AR and WI groups. The
average seed set values of the TN group were not
significantly different (p20.1) from either the AR or WI.
groups. For these analyses, the averages were transformed
as described for Experiment 3e, above. The single MI plant
was excluded from these analyses.

For the AR, TN and WI groups, each group’s self seed
set averages were lower than, and significantly different (p
- 0.01) from, their respective cross seed set averages.
This analysis compared the transformed average seed sets for
self and cross seed from Experiments 3e and 3f,

respectively. The single MI plant was excluded from these

analyses.

31
the AR group, from 9.7 to 47.7 seed per pollination for

those of the TN group, and from 7.0 to 53.6 seed per
pollination for those of the WI group: the MI plant averaged
27.9 seed per pollination (Appendix 2). The overall average
seed set was 23.8 116.76 seed per pollination (Figure 8).
Seed were set on all plants from all groups.

For pollinations on flowers from all plants from all
groups combined, the individual seed set values ranged from
0 to 73 seeds: 17% of the pollinations resulted in no seed,‘
and 51% in 15 to 40 seed (Figure 10). For the MI, AR, TN
and WI groups, seeded capsules formed on 80, 81, 88 and 88%
of the flowers, respectively: seedless capsules formed on 0,
10, 0 and 1% of the flowers, respectively.

There was a significant difference (p s 0.01) in the
average seed set values between the AR and WI groups. The
average seed set values of the TN group were not
significantly different (p=0.1) from either the AR or WI
groups. For these analyses, the averages were transformed
as described for Experiment 3e, above. The single MI plant
was excluded from these analyses.

For the AR, TN and WI groups, each group’s self seed
set averages were lower than, and significantly different (p
a 0.01) from, their respective cross seed set averages.
This analysis compared the transformed average seed sets for
self and cross seed from Experiments 3e and 3f,
respectively. The single MI plant was excluded from these

analyses.

32

Table 8. Self seed set of F1 plantsl.

No. of Total Ave. Seed
Plant Pollinations Seed per Pollination
301 3 0 0.0
302 4 0 0.0
303 2 0 0.0
304 8 0 0.0
305 8 0 0.0
306 6 3 0.5
307 10 2 0.2
308 5 0 0.0
309 10 0 0.0
310 3 5 1.7
All 59 10 0.2

 

iPlants 301 and 308 were the results of crosses between
Arkansas and Wisconsin plants. All others were the results
of crosses between various Arkansas plants.

33

Table 9. Pollen and pollen tube characteristics in self-
pollinated pistils.

. Heur§_frgn_zgllinatien

 

 

 

24 48 72 96 120 144

No. of 0bserv.s so 73 49 74 59 157

P0116? "o o 1 o o o o

Qnty. '2 1 2 6 o o o 2

2 98 93 100 100 100 98

Pollen 0 12 12 2 8 3 6

Germ. '3 1 6 8 1o 5 2 4

2 82 80 96 87 87 9o

Tube 0 12 12 2 8 3 6.

Length1'4 1 16 18 1o 7 3 2

2 28 38 43 27 32 29

3 38 18 6 5 9 11

4 6 14 39 53 53 52

Ave. Probable 5 0.1 0.3 2.8 4.4 3.4 1.7
Fertilizations

 

IAs percent of pistils observed at each time period.
2Categories are: 0 - no pollen: 1 - sufficient pollen for
100% fertilization; 2 - very large quantity of pollen.
3Categories are: 0 - no pollen germination: 1 8 sufficient
germination for 100% fertilization: 2 - germination of
very large quantity of pollen.
4Categories are: 0 a no pollen tubes, or tubes shorter than
diameter of pollen; 1 - tubes not to top of ovary; 2 =
tubes at top of ovary; 3 - tubes inside of ovary; 4 8
tubes enter ovules.
5Average number, per pistil, of ovules in which pollen tubes
have penetrated.

34

Table 10. Pollen and pollen tube characteristics in cross-
pollinated pistils. ~

 

 

.11 H0QI§_II9E_EQlliflé§iQD_________e

24 48 72 96 120 144

No. of 0bserv.s 55 95 59 so 23 17

Pollen 0 0 0 0 0 0 0

Qnty. .2 1 9 2 o o 4 o

2 91 98 100 100 96 100

Pollen 0 25 9 5 6 4 0

Germ. '3 1 11 7 o 6 17 o

2 64 84 95 88 78 100

Tube 0 25 8 5 6 4 o

Length1'4 1 31 15 3 2 4 o

2 4 6 o o o o

3 o 16 o 2 o o

4 4o 55 92 9o 92 100

Ave. Probable 5 2.0 4.6 21.1 20.0 24.0 22.8
Fertilizations

 

*As percent of pistils observed at each time period.
2Categories are: 0 a no pollen; 1 - sufficient pollen for
100% fertilization; 2 a very large quantity of pollen.
Categories are: 0 a no pollen germination: 1 a sufficient
germination for 100% fertiliation; 2 - germiantion of very
large quantity of pollen.
4Categories are: 0 - no pollen tubes, or tubes shorter than
diameter of pollen; 1 = tubes not to top of ovary: 2 8
tubes at top of ovary; 3 8 tubes inside of ovary; 4 =
tubes enter ovules.
5Average number, per pistil, of ovules in which pollen tubes
have penetrated.

35
to 100 germinated pollen; 80% of the selfed pistils and 77%

of the crossed pistils had 100+ pollen in their stigmatic~
cavities. For the 72 to 144 hour time periods combined, 5%
of the selfed pistils and of the crossed pistils had no
germinated pollen; 5% of the selfed pistils and of the
crossed pistils had from 1 to 100 germinated pollen; 89% of
the selfed pistils and 91% of the crossed pistils had 100+
germinated pollen in their stigmatic cavities.

Egllgn__§ub§__gzgw§h. (Tables 9, 10). For each the
selfed and crossed pistils at each time period, the
percentage of pistils which had none or very short pollen
tubes was the same as the percentage which had no germinated
pollen (pollen germination; category 0).

For selfed pistils at 24 hours, 16% had tubes within
the style (category 1), 28% had tubes at the end of the
style/top of the ovary (category 2), 38% had tubes in the
ovary (category 3), and 6% had tubes in ovules (category 4).
For crossed pistils at 24 hours, 31% were category 1, 4%
were category 2, 0 were category 3, and 40% were category 4.

For selfed pistils at 48 and 72 hours, 38% and 43%
respectively were category 2; 14% and 39% respectively were
category 4. For crossed pistils at 48 hours and 72 hours,
6% and 0 respectively were category 2; 55% and 92%
respectively were category 4.

For selfed pistils at 96, 120 and 144 hours combined,
29% were category 2 and 52% were category 4. For crossed

pistils at 96, 120 and 144 hours combined, 0 were category 2

36
and 92% were category 4.

WWW- (Table 11)- 0f
the selfed pistils at 120 and 144 hours combined, 37% had no
tubes in their ovaries, 52% had 1-10 tubes, 9% had 11-49
tubes, and 2% had 50+ tubes. For crossed pistils at 120 and
144 hours combined, 5% had no tubes in their ovaries, 5% had‘
1-10 tubes, 5% had 11-49 tubes, and 85% had 50+ tubes.

2:92ab1g_1§;tiligatigg§. (Tables 9, 10). At 24 hours,
the average number of probable fertilizations was 0.1 and
2.0 probable fertilizations per pistil for the selfed and
crossed pistils respectively. At 72 hours, the averages
were 2.8 and 21.2 probable fertilizations per pistil for the
selfed and crossed pistils respectively. At 96, 120 and 144
hours combined, the averages were 2.7 and 21.5 probable
fertilizations per ovary for the selfed and crossed pistils

respectively (Tables 9, 10).

W.

W All selfed.
crossed, and back-crossed pistils had 100+ pollen. One of
the selfed pistils had from 1 to approximately 100
germinated pollen in its stigmatic cavity; the remaining 16
selfed pistils and all of the crossed and back-crossed
pistils had 100+ germinated pollen.

£911;n__§gbg_g;gg§h. 0f the 17 selfed pistils, 2 had
tubes in their styles, 12 had tubes at the bottom of the
style/top of the ovary, and 3 had tubes which penetrated

ovules. The 3 crossed pistils and the 4 back-crossed_

Table 11. Number of pollen tubes in the ovaries
and cross-pollinated pistils.

Hours from

Pollination
Selfsl 120
144
Crossesl 120
144

37

No. of
Pistils

59
157

23
17

0

39
36

1-10

41
57

9
0

17

6

9
0

of

11-49

self-

50+

3
l 3

73
100

 

1As percent of pistils observed at each time period.

38
pistils all had tubes which penetrated ovules.

Number 9: tubes in the gvgry. 0f the 17 selfed
pistils, 14 had no pollen tubes in their ovaries and 3 had
from 1 to 10 tubes. 0f the 3 crossed pistils, 2 had from 11
to 49 tubes in their ovaries and 1 had 50+ tubes. All four
of the back-crossed pistils had 50+ tubes in their ovaries.

£19k§bl§__jg;tili;§§ign§. The selfed pistils had an
average of 0.5 probable fertilizations per pistil. The
crossed and back-crossed pistils had averages of 30.0 and~

21.5 probable fertilizations per pistil respectively. and

BEE§I1E§DE__AD- Among the plants from all groups combined,
the average ovule counts ranged from 28.7 to 80.0 ovules per
ovary (Table 12). For all plants from each the AR, TN and
WI groups, the averages were 42.3, 50.3 and 60.1 ovules per
ovary respectively. The average for the MI plant was 37.1
ovules per ovary. For all plants from all groups combined,
the average was 46.4 i9.48 ovules per ovary.

For the average ovule counts of the plants from each
group, there was a significant difference (p=0.01) between
the averages of the AR group and those of each the TN and WI.
groups. There was also a significant difference (p=0.05)

between the averages of the TN and WI groups.

ovule counts

Table 12. Average
studiedl.

No. of Ave.
Plt. Counts Ovules2 SD
1 60 37.1 5.39
2 43 50.4 9.22
3 6 50.0 9.36
6 5 37.8 10.43
8 11 42.7 9.06
12 12 41.9 3.87
15 14 36.9 4.05
16 11 46.5 5.87
18 4 46.0 3.56
23 4 41.8 9.57
24 9 43.0 10.49
29 6 36.0 6.16
30 5 39.2 6.38
34 14 41.1 10.50
36 5 30.0 3.54
42 12 54.7 9.23
44 3 36.3 5.51
47 . 3 29.3 4.93
50 16 42.9 5.41
53 3 36.0 10.58
54 13 30.8 3.00
57 4 37.0 3.92
61 19 33.4 5.15
65 10 47.1 6.94
67 20 51.4 7.84
68 16 46.9 6.49
73 9 34.9 5.78
75 8 46.1 4.76
77 3 28.7 3.51
82 23 40.7 5.83
113 4 52.8 13.15
114 12 80.0 9.05

39

for plants

Plt.

115
116
117
127
132
164
168
173
178
182
187
191
194
208
210
211
215
216
220
221
223
224
226
229
230
235
237
238
239
240
241

NO.

Counts Ovules

H
mueosbuumooo

NP
U9

P
so

p

H
upuquoqmommmmumu

Ave.

50.1
50.5
54.2
36.7
45.0
49.3
46.3
49.0
41.0
44.8
50.7
48.2
56.9
55.5
57.3
53.8
59.0
52.1
59.7
46.7
50.3
56.2
56.8
58.0
61.6
51.7
40.9
48.5
35.0
48.3
65.7

from groups

SD

6.28
4.75
7.69
3.22
4.36
9.98
4.12
4.62
7.94
8.49
9.75
5.11
7.85
9.88
2.52
6.41-
8.72
8.87
5.65
9.72

11.91

7.08
9.58
7.62
7.06
6.78
6.09
2.52
2.65
5.62

14.22

 

1Plant 1, Michigan; 2
241,

2 Wisconsin.
Average number of ovules per ovary.

-82, Arkansas; 113-117, Tennessee; 127-

215.935.5198

The low (self) seed set of x; peggrg plants in these
experiments was difficult to explain a priori. The
literature contains references to 2; pggara grown from seed
(7, 13: personal communication, Prairie Nursery, Westfield,
WI), and from self seed in particular (21). Many if not'
most other 2191; species produce abundant self seed,
especially those in which cleistogamy and chasmogamy are
operative (e.g. yr riginiana Rehb., y; rgighgnbgghiaga Jord.
and 2‘ hilt; L.) (4, 37). The seed set on cleistogamous
flowers (necessarily self seed) is much greater than that on
chasmogamous flowers (cross and/or self seed) for species in
which both systems are operative.

The {1191; are considered out-crossing species,
particularly those which are strictly chasmogamous. The
characteristics of their floral morphology and of their
insect pollinators promote cross-pollination and limit'
self-pollination by excluding self pollen from reaching the
stigma. However, when self-pollination does occur, it is as
likely as cross-pollination to result in seed set (4).

Excluding Emino's work involving inbred lines of y;
trigglgr Hortensis L. (19), the literature contained no

references regarding the possible existence of self-

40

41
incompability in the Viola. Therefore, the first

experiments undertaken (Experiments 1, 2, and 3a-3e)‘
centered on pollen, pollination and seed set in order to
determine the reason(s) for low seed set and to determine
how seed set could be increased. Only self-pollinations
were made, since there was no a prigri reason to believe
that cross-pollinations would give significantly different
results. Because flowering could not be controlled, the
self-pollination of available flowers was more readily made
than cross-pollination, without the necessity of having a
flower at the same stage of development on each of two

different plants at the same time.

Anrngr_ggni§§gngg. Anther dehiscence is indicative of pistil
receptivity in the flowers of many species (20). Since, in
the flowers of y; pggata plants in these studies, the anthers
dehisced over a period of approximately 2-3 days (beginning
approx. 1 day before anthesis and ending 1 day after),
pollination during this period might prove most effective for

seed set.

Bgllgn__yigb111§y. Using cotton blue stain, the high .ul off
percentages of plump, darkly-stained pollen indicate high

percentages of viable pollen for the plants studied. Even if
these results overestimate the percentage of viable pollen,
it is still unlikely that low pollen viability limited the
self seed set, since large amounts of pollen (usually several

hundred) were transferred to the stigma upon artifiicial

42

pollination. In addition, the results of the fluorescence
microsc0py investigations show that large.numbers (100+) of
germinated (i.e. viable) pollen were found in the stigmatic
cavities of 90% of the selfed pistils. Therefore, it is not
likely that (low) pollen viability contributed significantly

to the problem of low self seed set.

figgg_§gr. For all groups, seed set was very low on flowers
which were not artificially pollinated and left undisturbed.
As is true of the 2191; in general, the physical arrangement
of the anthers and their appendages confines the dehisced
pollen unless disturbed. After approx. 6 days, depending on
weather conditions, the appendages loosen their grip on the
style, allowing the pollen to fall into the groove of the
spur petal. Some pollen may then spill forward and effect
pollination (4). It is possible that air drafts or thrips
effected self- and/or cross pollination, resulting in the‘
observed low seed set. The total lack of seed set on
emasculated flowers strongly suggests that agamospermy is
not operative in these plants of y‘ pedarg. On these
flowers, which were not artificially pollinated and left
undisturbed, the seed produced was the result of
fertilization. The emasculation technique appeared to be
very effective at preventing self-pollination, assuming that
seed set would result had self-pollination occurred.
Seedless capsule formation on 4 of the 50 emasculated

AR flowers would indicate that pollination had probably

occured prior to or during emasculation of these flowers. In

43
many 2191; species, pollination without resultant seed set

may stimulate capsule formation (22). It is also possible
that what appeared to be seedless capsules were actually
double ovaries which, because of their relatively large size,
look like small capsules.~

That seed set on flowers self-pollinated on one of the
five days beginning at anthesis did not vary significantly
between days suggests that the receptivity of the flowers
did not change during this period. Dehisced pollen confined.
to the anther cone remains viable for 5-8 days in other
investigated 2191; species (4, 22). Assuming that pollen
remains viable in 1‘ pgdgra flowers for a similar length of
time, it is unlikely that a concurrent severe drop in pollen
viability and sharp rise in pistil receptivity was
responsible for the observed constant seed set between the
five days.

The role of stigmatic fluid in stigmatic receptivity is
discussed by Beattie for three 2191; species. He found that
pistils would exude stigmatic fluid, usually on or after day
3 from anthesis. Beattie postulates "that the primary‘
function of the mucilage (stigmatic fluid) was to increase
the receptive area for pollen grains" thus "increasing the
efficiency of the pollination mechanism" (4). The 2; peggra
flowers in these investigations exuded stigmatic fluid with
irregularity during the five days beginning at anthesis. A
flower's stigma might appear "wet" with exudate on any one

of the days, but most frequently on days 2-4. The exudate

44
would not be detectable on the following day. During.

artificial pollination, more pollen appeared to adhere to
wet stigmas than to "dry" stigmas which had no exudate.
However, pollination of wet stigmas did not result in
increased seed set as compared to that resulting from
pollination of dry stigmas. It is possible that stigmatic
fluid was present in the cavity of dry stigmas, but not
visible. Because the artificial pollination technique placed
pollen inside the cavity, sufficient pollen may have adhered
to dry stigmas to result in seed sets comparable to that
resulting from pollination of wet stigmas.

Inbreeding depression is defined as a decrease in.
fitness and vigor, resulting from inbreeding imposed on
individuals that are normally outbreeding. It is the
consequence of increased homozygosity for deleterious
recessive genes and the breakup of balanced polygenic
systems. Inbreeding depression often results in decreased
self seed set, and may be indistiguishable from post-zygotic
self-incompatibility systems (34). Cross- pollination
between inbred lines often results in vigorous F1 plants,
which usually set much more self seed than their inbred
parents. Since the F1 hybrid plants of yr pgdata in these
studies had very low average self seed set values, or set no
seed at all, despite their presumed heterogeneity and hybrid
vigor, inbreeding depression seems an unlikely explanation
for the low self seed set of the parents.

While Experiments 3a-e were being conducted, several

45
cross-pollinations had been attempted so as to obtain seed.
for other, unrelated, inheritance studies for flower color
(bicolor and concolor types). These crossed flowers
produced very large amounts of seed as compared to the
amount being produced on selfed flowers in Experiments 3a-e.
These large cross seed sets prompted the inclusion of
Experiment 3f in which as many plants as possible were
scored for cross-compatability.

The highly significant difference in seed set between
self- and cross—pollinations had not been anticipated; the
literature contained no well-documented cases of self-
incompatablity in the 21913 (19, 22). The difference_
between self and cross seed set suggested that a type of

self-incompatibility system might be operative in yr pggara.

Elggrggggng§__migrggggpy. The large difference between

self and cross seed set led to the search for differences in
pollen and pollen tube characteristics between self— and
cross-pollinated pistils (Experiment 4a). The fluorescence
microscopy technique employed proved to be very useful for
observing pollen, pollen tube growth and ovules.

The observations of pollen and pollen tube growth had
to be made with great care so as to achieve the most
accurate measurements. First, the stained pistil was‘
examined whole without a coverslip, the stigmatic cavity
examined for pollen, and the style examined for tube growth
(a "glow" in the style, but this was often not seen due to

the thickness of the stylar wall) (Figure 11). The ovary was

46

 

Figure 11. Style (left) and upper ovary (right) of a pistil
with glow of fluorescing pollen tubes in style.
20X.

 

Figure 12. Pistil in Figure 11, split open; most pollen
tubes stop at top of ovary. 20X.

47
then cut open longitudinally, and the sections laid out so

that the cut surface faced upward. Figure 12 shows a pistil
examined at this stage, without a coverlip. Only a few tubes
appear to be entering the ovary. A coverslip placed on the
specimen and slight pressure applied. Figure 13 is of the
same pistil as Figure 12, which has been slightly flattened,
and many more tubes can be seen in the style. Finally, the
pistil was flattened even further, revealing many more tubes
in the style and several more tubes in the ovary (Figure 14).
The flattening process causes the pollen tubes and the ovules
to be "squashed" together, making them appear to be in closer
proximity than they actually were in the unflattened
preparation.

Observations on pollen quantities and germination
showed that these were not the limiting factors with regard
to the level of seed set. Pollen quantities and germination.
were high, and nearly the same, for both self- and cross-
pollinated pistils. Large numbers of pollen tubes appeared
to grow at the same rate, through the hollow style, to the
top of the ovary, in both the self- and cross- pollinated
pistils. At that point, the similarities ceased.

In the cross-pollinated pistils, large numbers of
pollen tubes continued on into the ovary and could be seen
entering many of the ovules (Figure 15). In approximately
one-third of the self-pollinations, no pollen tubes went
beyond the top of the ovary (Figure 16). Although pollen

tubes could be found entering the ovaries of the remaing'

48

 

Figure 13. Pistil in Figure 12, slightly squashed; greater
number of tubes apparent. 20X.

 

Figure 14. Pistil in Figure 13, squashed further. 20X.

49

 

Figure 15. Cross-pollinated pistil showing many pollen tubes
. in the ovary. 20X

 

Figure 16. Self-pollinated pistil in which pollen tubes have
stopped at top of ovary. 20x.

50
two-thirds of the self-pollinated pistils, they were far

fewer in number than in the cross-pollinated pistils
(Figures 17, 18).

The ends of these tubes, and of many of the tubes found
in the ovaries of self-pollinated pistils, fluoresced
poorly, appeared thin and/or varied in thickness, and often
formed large, rounded segments (Figures 19, 20). In
comparison, the pollen tubes found in cross-pollinated.
pistils fluoresced brightly and were of uniform appearance
throughout their length.

The pollen tubes appeared to grow along the ovary wall
to the placenta between the funiculi and into the ovules
(Figures 21, 22, 23). No difference was observed in the
appearance of self vs cross pollen tubes as they penetrated
ovules.

The observed number of ovules in the ovaries probably
underestimates the actual number to some degree due to the
loss of some ovules during specimen preparation. From a
comparison of the average ovule counts to the self and cross.
seed set data, it is evident that many ovules do not result
in seed, and that self-pollinations made use of far fewer
available ovules than did cross-pollinations.

It was evident that a type of barrier was severely
restricting the growth of pollen tubes into the ovaries of
self-pollinated pistils, or that a barrier was selectively
permitting the entry of pollen tubes into the ovaries of

cross-pollinated pistils (28, 29, 34). The difference in

51

 

Figure 17. Self-pollinated pistil in which a small
percentage of pollen tubes have entered ovary.
20x.

 

Figure 18. Self-pollinated pistil in which two pollen tubes
can be seen entering ovules. 20X.

52

 

Figure 19. Ends of pollen tubes in a self-pollinated pistil.
100x.

 

Figure 20. Ends of pollen tubes in a self-pollinated pistil.
100x.

53

 

Figure 21. Cross-pollinated pistil with pollen tubes having
grown along placenta to ovules. 20x.

 

Figure 22. Cross-pollinated pistil with pollen tubes along
placenta and in many (detached) ovules. 20X.

 

Figure 23. Pollen tubes entering ovules. lOOX.

 

Figure 24. Self—pollinated pistil of an F1 plant with many
pollen tubes, all stopping at top of ovary. 20X.

55
the number of pollen tubes gaining entrance to the ovaries

is reflected in the number of pollen tubes entering ovules
(probable fertilizations) and in the average seed sets for
self- and cross-pollinated pistils. For self-pollinations
(averaging 3.19 seed per pistil), an average of 2.41 ovules
per pistil had pollen tubes entering them.' For cross-
pollinations (averaging 23.82 seed per pistil), an average
of 21.37_ovules per pistil had pollen tubes entering them.
Pollen tube growth in self-pollinated pistils of the F1
plants also showed a very strong self-incompatability
response (Figure 24); this is reflected in their low or
nonexistent self seed sets. Crossed and back-crossed pistils-
of these plants readily allowed pollen tubes to enter the

ovaries.

-' o ' ' ' V t . Based on comparisons of
self and cross seed set, and on observations of pollen tube
growth in selfed and crossed pistils, a genetic self-
incompatibility system is proposed to be operative in z;
negate. Inhibition of self pollen tubes growth occurs at
the top of the ovary thereby preventing fertilization or
greatly reducing its frequency.

The fact that 23% of the cross-pollinations resulted in.
0 or 1-5 seed (Figure 10) could be explained if the plants
involved were effectively of the same self-incompatibility
genotype. It is probable that each the AR, TN and W1 groups
contained some plants with identical genotypes, due to

asexual reproduction. The fact that self-pollinations do

56 .
result in some seed set, and that pollen tubes can be found

entering some ovaries and ovules of selfed pistils,
demonstrates that the self-incompatability barrier is not
complete. In fact, the average self seed set values of the
plants studied were as high as 21.2 seed, yet 83% of the
averages were between 0 and 5.9 seed. Repeated
observations of pollen tube growth in self-pollinated pistils
of individual plants showed that, on occasion, a number of
tubes (most often 1-10, but sometimes more: see Table 11)
would gain access to the ovary. The incomplete nature of the
proposed self-incompatability system, and how it may be
affected by such factors as the (micro-) environmental
conditions and plant nutrition, could explain the variable
seed set resulting from self- and cross-pollinations on a
individual plant. In addition, the fact that self seed
yields contain a larger proportion of poorly-filled seed
compared to cross seed suggests that a post-fertilization SI

system may be operative in y; pgdgra.

:-a ,. :e f- 1 ouos i- 0 4, =1 11_s .,,

Seavey and Bawa, in a review of late-acting (ovarian)
self-incompatiblity in angiosperms, present evidence for
such systems in many species (34). They propose that late-.
acting self-incompatibility systems are not as rare as is
currently believed:

Brewbaker, in pointing out the occurence of,
hollow styles in many species with self-
incompatibility (SI) mechanisms in the ovary,
suggested that a lack of intimate contact in such
such styles might allow incompatible pollen tube

57

growth to proceed further than is usual in
gametophytic SI systems. Inhibition of pollen tube
growth in the ovary would indicate that such
inibition may at times operate by chemical
mechanisms similar, if not identical, to those that
operate in the style of most gametophytic SI
systems (9).

Such a late-acting pollen tube inhibition
appears to account for the observations of Brock in
three species of L111Qm. Self pollen tubes were
seen to reach the base of the style in all three,
and in two species, L; gand1dgm and L;
firgg1r§1gngm, the ovary enlarged following self-
pollination. However, no evidence of endosperm
development was found in 312 ovules of L; gang1dgm
examined, and only 11 of 468 ovules of L;
grgy1r§1angm possessed developing endosperm.
Ovaries of the third species, L; pardal1ngm, did
not enlarge following self-pollination (12). Thus,
although Brock speculated that some form of post-
fertilization inhibition might be operating, as had
previously been found in gasrgr1§ (33), it appears
more likely that incompatibile pollen tubes were
routinely rejected before reaching the ovule. The
rare occurence of developing ovules in L; gand1dgm
may be an indication that the rejection response is
not absolute (12).

Stout and Chandler found similar ovarian
responses in W115 1811;125:211 and H3.
g1§r1n§, although it is possibile that self pollen
tubes interacted with ovules (35). Brewbaker and
Gorrez found that the delayed arrest of self pollen
in H; thnnberg11 occurred before the micropyle was
reached (10).

In ngg1gggg sariva self pollen tubes do not
penetrate into the ovary as far as cross pollen
tubes, therefore reaching fewer ovules, and are
also less likely than cross pollen tubes to enter
the micropyle of an ovule that has been reached
(11, 15, 32). Because self-fertilization does
occcur, though much less frequently than cross-
fertilization, u§d1gagg £9212; is considered to be
partially (15) or weakly (14) self-incompatible.
The fact that 13 g1rrg studies of pollen tube
growth correlate with 13 v1vg cross-pollination
(2), but do not correlate with 13,2129 observation
for self-pollination (36), supports the conclusion
that the maternal ovarian tissues are retarding the
rate of growth of the self pollen tubes and/or
restricting their access to the ovules. (34).

The observations of pollen tube growth in flowers of 1;.

peggrg in these investigations suggest a SI system similar

58

to those found in the 1.1111119. Wis. and 142919.839
species, except that self pollen tubes were inhibited at the

top of the ovary in y; pggara. This is the point at which
pollen tubes appear to come into close contact with _the
maternal tissue, thus allowing inhibition by a chemical
mechanism similar to those found in gametophytic SI species.
The high percentage of poorly-filled self seed, like that
found in ned1§agg 5:111; (34), suggests the existence of a
similar post-fertilization SI system in y; REQAL3-

It is suggested that further research into this
proposed self-incompatability system for y; peggra [e.g.
genetics, (micro-) environmental effects] is needed before a
breeding program can be established for the development of
y; pgdgrg cultivars. Investigations into the population
genetics of y; pedarg, and into the evolution of the
proposed self-incompatability system, would prove
interesting and informative, as would the search for self-
incompatability in other 1191; species. The transfer of
self-incompatability to other 11Q13 species may be possible,.
either through interspecific hybridization or genetic
engineering, and prove useful in breeding and seed

production efforts.

fiBEEQI!

The results of these investigations indicate that a
self-incompatibility system is operative in 2121; peggra,
that this is responsible for the overall low (self) seed set
of the plants studied, and that seed set can be greatly
increased through cross-pollination.

The site of the self-incompatability reaction was found‘
to be in the area where the hollow style gives way to the
ovary cavity. Either the pollen tubes in self-pollinated
pistils are greatly restricted from entering the ovary, or
tubes of cross-pollinated pistils are selectively permitted
to enter. Since some self seed was obtained, the self-
incompatability reaction is not complete. Self-compatible
and/or semi-compatible genotypes may exist.

Pollen quantities and viability, the timing of
pollination, and the presence/abscence of visible stigmatic
fluid do not appear to be critical factors limiting (self)
seed set. It was found that agamospermy is highly unlikely-
to be operative in this species.

Further research into the genetics and expression of
the proposed self-incompatability system is suggested. In
addition, investigation into seed germination requirements

for this species is highly recommended.

59

60
It is suggested that 1191; pgdgtg would make a welcome

addition to the garden, and possibly to the pot-plant
market, if suitable seed-produced cultivars can be.

developed.

APPENDICES

61

Appendix 1. Average seed set1 of self-pollinated flowers by

plant.

Seed Seed Seed Seed
Plt2 Set Plant Set Plant Set Plant Set
1 3.2 59 0.0 139 1.0 200 2.5
2 0.9 60 0.7 140 0.0 201 5.7
3 0.0 61 1.1 141 0.0 202 11.9
4 0.6 65 0.7 143 0.0 203 3.7
5 3.2 67 0.0 148 0.0 204 2.4
6 0.0 68 0.8 149 2.1 205 7.9
8 0.3 71 0.3 151 0.0 206 0.7
10 0.2 72 6.2 156 6.4 208 2.5
11 1.6 73 1.2 157 2.9 209 21.2
14 0.0 74 5.0 159 2.9 210 10.1
15 0.0 75 0.0 160 2.0 211 18.2
16 0.0 76 5.0 161 4.5 213 1.3
18 0.0 77 1.2 162 0.5 214 0.0
19 2.5 79 1.7 163 4.0 215 14.0
20 0.0 80 0.0 164 4.2 216 1.2
21 0.1 81 1.2 165 2.9 218 1.3
23 0.7 82 1.3 166 3.8 219 15.3
24 8.3 85 0.8 167 2.6 220 0.2
25 0.1 86 0.1 168 6.4 221 0.6
27 2.9 113 0.5 172 1.7 222 3.7
28 1.0 114 0.2 173 2.5 223 0.2
29 1.5 115 0.0 174 2.8 224 4.1
30 4.5 116 3.3 175 3.2 225 1.2'
31 0.9 117 0.1 178 5.5 226 0.2
32 0.0 118 0.0 180 6.9 227 0.8
33 1.6 119 6.6 181 1.6 228 4.7
34 4.9 120 0.7 182 6.3 229 13.7
35 0.0 121 0.0 184 1.7 230 8.1
36 0.0 122 18.7 185 0.5 231 5.2
37 0.9 123 0.7 186 4.3 232 8.0
39 1.3 124 1.3 187 15.4 233 0.8
40 0.2 125 2.7 188 10.4 234 1.2
41 0.0 126 2.4 189 9.4 235 1.0
42 0.3 127 1.4 190 1.1 236 17.4
44 1.4 129 1.1 191 10.9 237 6.8
47 0.0 130 0.0 192 5.6 238 3.3
48 0.0 131 0.5 193 2.9 239 13.7
50 0.5 132 4.4 194 18.8 240 9.9
53 0.0 133 1.4 195 2.9 241 16.2
54 1.5 134 0.9 196 1.5 242 2.2

56 0.0 135 2.7 198 1.0

57 2.5 136 3.7 199 3.2

 

3Average of 10 self-pollinated flowers for each plant.

2Plant 1, Michigan; 2—86, Arkansas; 113-122, Tennessee;

Wisconsin.

123‘242,

62

Appendix 2. Average seed set1 of cross-pollinated flowers

by plant.

No. Total Ave. No. Total Ave.

Pit2 X’s3 Seed Seed Set Plant X’s Seed Seed Set
1 30 836 27.9 114 3 32 10.7
2 37 1124 30.4 115 5 132 26.4
3 10 158 10.8 116 6 286 47.7
4 58 882 15.2 117 6 99 16.5
5 9 263 29.2 118 5 123 24.6
8 16 491 30.7 119 7 234 33.4
12 21 473 22.5 121 3 29 9.7
14 7 83 11.9 122 8 310 38.8
15 10 202 20.2 127 4 84 21.0
16 23 500 21.7 130 3 56 18.7
23 7 103 14.7 178 4 69 24.8
24 3 76 25.3 181 4 154 38.5
25 3 93 31.0 187 28 972 34.7
27 16 384 24.0 188 4 147 36.8
29 6 122 20.3 189 4 35 8.8
30 7 248 35.4 190 4 109 27.3
32 8 140 17.5 191 4 127 31.8
33 3 60 20.0 193 3 71 23.7
34 35 916 26.2 194 13 464 35.7
35 3 40 13.3 205 4 50 12.5
36 7 38 5.4 206 3 131 43.7
37 9 173 *19.2 211 3 136 45.3
39 4 62 15.5 215 9 289 32.1
41 10 128 12.8 216 5 140 28.0
42 6 92 15.3 218 3 78 26.0
44 19 309 16.3 220 3 66 22.0
50 9 88 9.8 223 4 150 37.5
54 41 699 17.1 224 6 292 48.7
56 10 243 24.3 225 3 91 30.3
61 14 299 21.4 226 3 52 17.3
67 16 289 18.1 229 9 226 25.1
68 6 146 24.3 230 4 165 41.3
71 8 98 12.3 231 4 116 29.0
72 5 60 12.0 232 4 150 37.5
73 5 150 30.0 235 7 220 31.4
75 15 381 25.4 236 10 390 39.0
76 4 107 26.8 237 7 134 19.1
80 20 572 28.6 239 3 21 7.0
82 5 80 16.0 240 4 146 36.5
85 5 129 25.8 241 7 375 53.6
113 5 105 21.0 242 10 200 20.0

 

‘Average of at least three crosses with pollen from at least
three different male parents.

2Plant 1, Michigan; 2-85, Arkansas; 113-122, Tennessee; 123-
242, Wisconsin.

3Number of cross-pollinations made.

BIBLIOGRAPHY

10.

11.

12.

13.

51911991269111

Atwater, B.R. 1980. Germination, dormancy and morphology
of the seeds of herbaceous ornamental plants. Seed.
Sci. & Technol. 8:523-573.

Barnes, 0.x. and R.W Cleveland. 1963. Genetic evidence
for nonrandom fertilization in alfalfa as influenced by
differential pollen tube growth. Crop Sci. 3:295-297.

Baskin, J.M. and C.C. Baskin. 1972. Physiological

ecology of germination of 11919 r91199§9911. Amer. J
Bot. 59(10):981-988.

Beattie, A.J. 1969. The floral biology of three species
of 21919. New Phytol. 68:1187-1201.

1970. Insect visitors to three species of violet
(21919) in England. Entomol. Mon. Mag. :7-11.

 

1976. Plant dispersion, pollination and gene flow
in Viola. Oecologia (Berl.) 25:291-300.

Beattie, A.J. and D.C. Culver. 1979. Neighborhood size
in Viola. Evolution 33(4):1226-1229.

Brainerd, E. 1921. Violets of North America. Vermont'
Agr. Expt. Sta. Bul. 224.

Brewbaker, J.L. 1957. Pollen cytology and self-
incompatibility systems in plants. J. Hered. 48:271-277.

Brewbaker, J.L. and 0.0. Gorrez. 1967. Genetics of self-
incompatibility in the monocot genera Ananas (pineapple)
and 999:9r19. Amer J. Bot. 54:611-616.

Brink, R.A. and D.C. Cooper. 1938. Partial self-
incompatibility in M9919999 9991_9. Proc. Nat. Acad.
U.S.A. 24:497-499.

Brock, R.D. 1954. Fertility in L111gm hybrids. Heredity
8:409-420.

Brumback, B. 1985. Wildflowers. In Perennial Plt. Assoc.
Newsletter 2:Winter.

63

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

64

Busbice, T.H., R.Y. Gurgis and H.B. Collins. 1975.
Effect of selection for self-fertility and self-
sterility in alfalfa and related characters. Crop Sci.
15:471-475.

Cooper, D.C. and R.A. Brink. 1940. Partial self-
incompatibility and the collapse of fertile ovules as
factors affecting seed formation in alfalfa. J. Agric.
Res. 60:453-472.

Darlington, C. and A. Wylie. 1955. Chromosome atlas of
flowering plants. George Allen and Unwin Ltd., London.

de Nettancourt, D. 1977. Monographs on theoretical and
applied genetics 3. Incompatibility in angiosperms.
Springer-Verlag, Berlin.

Ellis, R.H., T.D. Hong and E.H. Roberts 1985. Handbook
of seed technology for genebanks Vol. II. Compendium of‘
specific germination information and test
recommendations. International Board for Plant Genetic
Resources, Rome.

Emino, B.R. 1967. Cytology and seed set studies in the

pansy, 21919 9r19919r Hortensis L. MS Thesis, Michigan
State Univ., E. Lansing.

Frankel, R. and E. Galun. 1977. Monographs on
theoretical and applied genetics 2. Pollination
mechanisms, reproduction and plant breeding. Springer-
Verlag, Berlin.

Gershoy, A. 1928. Studies in North American violets I.
General considerations. Vermont Agr. Expt. Sta. Bul.
279.

1934. Studies in North American violets III.
Chromosome numbers and species characters. Vermont Agr.
Expt. Sta. Bul. 367.

Headstrom, R. 1978. The families of flowering plants.
A.S. Barnes and Co. Inc., Cranbury, NJ. ‘

Johansen. D.A. 1940. Plant microtechnique. McGraw Hill,
New York.

Kho, Y.O. and J. Beer. 1968. Observing pollen tubes by
means of fluorescence. Euphytica 17:298-302.

Kristofferson, K.B. 1923. Crossings in Melanium violets.
Hereditas 4:251-289.

27.

28.

29.

30.

31.

32.

33.

34.

3s.

36.

37.

65

Kroon, G. 1972. Some aspects of the pollination

mechanism of 21919 9r19919r L. and V1019 z W1t9r99ki999
Gams. Acta Bot. Neerl. 21(6):630-632.

Mulcahy, D.L. 1984. The relationships between self-
compatibility, pseudo-compatibility, and self-
incompatibility. In W.F. Grant, ed., Plant
Biosystematics, Academic Press, Canada.

Mulcahy, D.L., and G.B. Mulcahy. 1983. Gametophytic
self-incompatibility reexamined. Science 220:1247-1251.

Russell, N.H. 1965. Violets (21919) of central and
eastern United States: an introductory survey. SIDA
Contrib. Bot. 2(1):1-116.

Russell, N.H. and W.R. Bowen. 1960. A hypothesis to
account for unusual leaf variation in 21919 999999 L.
Proc. Iowa Acad. of Sci. 67:189-194.

Sayers, E.R. and R.P. Murphy. 1966. Seed set in alfalfa
as related to pollen tube growth, fertilization
frequency, and post-fertilization ovule abortion. Crop
Sci. 6:365-368.

Sears, E.R. 1937. Cytological phenomena connected with.
self-sterility in the flowering plants. Genetics 22;130-
181.

Seavey, S.R. and K.S. Bawa 1986. Late-acting self-
incompatability in angiosperms. Bot. Review. 52(2):195-
219.

Stout, A.B. and C. Chandler. 1933. Pollen tube behavior

in 3999r999111§ with special reference to
incompatibilities. Bull. Torrey Bot. Club 60:397-416.

Straley, C. and B. Melton. 1970. Effect of temperature
on self-sterility and 19 21rr9 pollen growth of
selected alfalfa clones. Crop Sci. 10:326-329.

Valentine, D.H. 1962. Variation and evolution in the
genus 11919. Preslia 34:190-206.