ll

WW

 

!

M
\
i

mmwwnu

f

W

A.

*IIIIIIIIIIIIIIWIIIIIIIIIIII

3 1293 10725 9206

   
 
  

   
 
 

LIBRARY

Michigan State
University

This is to certify that the
thesis entitled

GENETIC AND BREEDING STUDIES IN A CUCUMIS

SATIVUS L. X E. HARDWICKII R. POPULATION

presented by

NEIL MADISON COHEN

has been accepted towards fulfillment
of the requirements for

M . S Jegree in HORT I CULTURE

 

 

 

Major professor

Date 6-29-81

 

0-7639

flERDU; FINE§z
25¢ per day per Ital

RETURNING LIBRARY MTEklALS: '
Place in bookne turn to renove 3
charge from circulation records

 

 

 

 

 

GENETIC AND BREEDING STUDIES IN A CUCUMIS
SATIVUS L. X g. HARDWICKII R. POPULATION

 

BY

Neil Madison Cowen

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Department of Horticulture

1981

ABSTRACT

GENETIC AND BREEDING STUDIES IN A CUCUMIS
SATIVUS L. X g. HARDWICKII R. POPULATION

 

BY

Neil M. Cowen

The P P

l’ 2’ l
L. x g. hardwickii R. interspecific cross were evaluated for

 

F and F2 generations of a Cucumis sativus

 

ll traits: seedling bitterness; spine color; nodes to first
pistillate flower; percent (%) gycoecious nodes for nodes
l-20; percent (i) nodes with laterals; fruit number; number

of fruits on the main stem; number of fruits on the laterals;
fruit diameter; fruit length; and fruit yield. Significant
genetic variation was observed for all traits in the F2.

High heritabilities and high gains from selection were observ-
ed for all traits except fruits on the main stem. Non-addi-
tive types of gene action probably are involved in the

expression of most traits. It is concluded that g. hardwickii

 

may serve as a source of genes for increasing yields in

Q. sativus.

Guidance Committee:

This thesis is condensed into a format suited
and intended for publication in Euphytica.

 

ii

To my daughter Candace Marie

iii

ACKNOWLEDGEMENTS

The author extends his sincere appreciation to his
major professor, Dr. D. Helsel for her guidance and encourage-
ment throughout the course of this research, and also to
committee members, Drs. S. Honma and K. Payne for their
valuable suggestions. Further, he extends his sincere
thanks to Dr. L. R. Baker and Dr. J. F. Kelly for their
continued support and encouragement throughout the course
of this study.

Special thanks is given to Mr. Anand Nandgoankar and
Ms. Mary Hunsperger for their invaluable help in carrying

out this research.

iv

TABLE OF CONTENTS

Page
List of Tables vi
List of Figures \dii
SUMMARY 1
INTRODUCTION 2
MATERIALS AND METHODS A
RESULTS AND DISCUSSION 8
REFERENCES 39

APPENDIX 42

LIST OF TABLES

 

 

 

 

 

 

 

 

Table Page

1 Chi square test for goodness of fit

for several traits in the cross

Cucumis sativus x g. hardwickii l3
2 Broad sense heritabilities for various

traits in the cross Cucumis sativus x

g. hardwickii 1n
3 Test of gene action for several traits

in the cross Cucumis sativus x g.

hardwickii 15
A Potence ratios for 2 measures of

gynoecious sex expression in the

cross Cucumis sativus x g. hardwickii 15
5 Number of effective factor pairs by

which parents of the cross Cucumis
sativus x g. hardwickii differed for
11 traits l6

 

6 Gains per cycle from selection with
a 10% selection intensity for
various traits in the cross Cucumis

 

 

 

 

sativus x g. hardwickii 19
7 Transgressive segregates for various

traits in the cross Cucumis sativus

x g. hardwickii l9
8 Generation means for several traits

in the cross Cucumis sativus x Q.

hardwickii 2O

 

vi

Table

10

ll

12

13

1A

15

l6

17

Analysis of variance of nodes to
first pistillate flower for the cross
Cucumis sativus x g. hardwickii

 

 

Analysis of variance of percent
gynoecious nodes, nodes 1-20 for
the cross Cucumis sativus x g.
hardwickii

 

 

Analysis of variance of percent
nodes with laterals for the cross
Cucumis sativus x g. hardwickii

 

 

Analysis of variance of fruit
number for cross Cucumis sativus
x g. hardwickii

 

 

Analysis of variance of number of
fruits on the main stem for the
cross Cucumis sativus x g.
hardwickii

 

 

Analysis of variance of number of
fruits on the laterals for the cross
Cucumis sativus x C. hardwickii

 

 

Analysis of variance of fruit diameter

for the cross Cucumis sativus x g.
hardwickii

 

 

Analysis of variance of fruit length
for the cross Cucumis sativus x C.
hardwickii

 

Analysis of variance of fruit yield
for the cross Cucumis sativus x g.
hardwickii

 

vii

Page

A2

A2

A3

A3

A3

AA

AA

AA

“5

Figure

LIST OF FIGURES

Page

Frequency distribution of nodes to

first pistillate flower for the P1,

P2, F1, and F2 generations of the

cross Cucumis sativus x g. hardwickii 21

 

 

Frequency distribution of percent (%)
gynoecious nodes for nodes 1-20 for

the cross Cucumis sativus x g.

hardwickii 23

 

 

Frequency distribution of percent (%)

nodes with laterals for the P1, P2,

F1, and F generations of the cross

Cucumis sativus x g. hardwickii 25

 

 

Frequency distribution of fruit number

for the P1, F1, and F2 generations of

the cross Cucumis sativus x g.

hardwickii 27

 

 

Frequency distribution of fruits on

the main stem for the P1, P2, F1, and

F2 generations of the cross Cucumis

sativus x g. hardwickii 29

 

Frequency distribution of fruits on

the laterals for the Pl, F1, and F
generations of the<mxss Cucumis sativus x

g. hardwickii 31

 

Frequency distributions of mean fruit
diameter in cm. for the Pl, F1, and

F2 generations of the cross Cucumis

sativus x g. hardwickii 33

 

Frequency distribution of fruit

length in cm. for the P1, FL, and

F2 generations of the cross Cucumis

sativus x g. hardwickii 35

 

\dii

Figures

 

Page
9 Frequency distribution of fruit
yield in gm. for the P1. F1, and
F2 generations of the cross Cucumis
sativus x g. hardwickii 37

ix

SUMMARY

The P1’ P2, F1’ and F2 generations of a Cucumis sativus

 

L. x g. hardwickii R. interspecific cross were evaluated

 

for 11 traits: seedling bitterness; spine color; percent
(%) gynoecious nodes, nodes l-20; percent (%) nodes with
laterals; fruit number; number of fruits on the main stem;
number of fruits on the laterals; fruit diameter; fruit
length; and fruit yield. Significant genetic variation was
observed for all traits in the F2. High heritabilities and
high gains from selection were observed for all traits
except fruits on the main stem. Non-additive types of gene

action probably are involved in the expression of most

traits. It is concluded that C. hardwickii may serve as a

 

source of genes for increasing yields in g. sativus.

INTRODUCTION

 

Fruit yield in pickling cucumber (Cucumis sativus L.)
for mechanical harvest have typically been limited to l-2
fruits per plant (Miller & Hughes, 1969). Developing seeds
in the first fertilized fruit inhibit the develOpment of
subsequently fertilized flowers. This inhibition severely
limits the number of harvestable fruits per plant. This
inhibitory effect is presumed to be due to internal sink
competition. The developing fruit acts as a powerful sink
for translocated photosynthates, the strength of which pre-
cludes further fruit set and development due to the unavail—
ability of metabolites (McCollum, 193A; Neihuis & Lower,
1980).

Several approaches have been proposed for increasing
fruit set per plant: both by plant breeding techniques as
well as by physiological means. Breeding of parthenocarpic
or seedless pickling cucumbers has been proposed as one
method of increasing fruit set per plant (Pike & Peterson,
1969; Baker et al., 1973; Deena, 1973). The absence of
developing ovules in the fruit reduces the inhibitory effect
on subsequent fruit set. Deena (1973) and others (Connor &
Martin, 1970) further proposed that breeding for small mature
fruit size and delayed fruit set would lead to higher per
plant fruit numbers. The physiological approach proposed by
a large number of researchers (Robinson et al., 1971;

Quebedaux & Beyer, 1972); Cantliffe, 197A; Elassar et al., 197A)

utilizes chemical growth regulators to induce parthenocarpic
fruit set in pickling cucumber cultivars which are not gene-
tically parthenocarpic.

Recently, Horst and Lower (1978) proposed using Cucumis

hardwickii R. as a source of genes for increasing fruit

 

yields per plant in pickling cucumber. 9. hardwickii is an

 

annual, monoecious, short day, unadapted species which
hybridizes readily with Q. sativus, producing fertile Fl's.
The fruit set and branching characteristics exhibited by

g. hardwickii are atypical of g. sativus. g. hardwickii

 

 

ordinarily has more and larger laterals than 9. sativus
as well as being capable of sequentially setting large
numbers of seeded fruit (Whitaker & Davis, 1962; Horst &
Lower, 1977, 1978; Robinson & Kowalski, 1978; Neihuis &
Lower, 1979, 1980).

The present study was undertaken to clarify the in-
heritance of,gene action involved in, and estimate herit—
abilities and gains from selection for a number of fruit
and growth habit characteristics in a Q. sativus x g.

hardwickii pOpulation.

 

MATERIALS AND METHODS

 

The material used in this study was MSU A1, a gynoecious
inbred line of g. sativus which was selected out of GYlA,
by Dr. Baker at MSU, and LJ 9OA30, an inbred line of C.

hardwickii.

 

Greenhouse Procedure

 

The P P F1’ and F2 generations of the cross MSU A1

1’ 2’
x LJ 9OA3O were sown in peat pots filled with VSP mix on
December 9, 1980. The germinating seedlings were transplanted
into raised beds filled with VSP mix on December 26, 1980.
Temperatures were maintained at 2A0: 30C during the day and
180: 30C at night. Standard cultural practices for fertility,
disease and insect control were followed for the duration of
the experiment.

The plants were trained on bamboo stakes to a height of
2 meters, and all laterals were allowed to develop to a
length of 2 nodes.

Seedlings were planted in a randomized complete block
design with 3 replications of the P1’ P2, and F1 generations,
and 12 replications of the F2 in each 21 blocks.

All female flowers were hand pollinated. Flowers
reaching anthesis during the day the pistillate flowers
were to be pollinated were used. This procedure was

continued throughout the entire pistillate flowering period

for all plants in the experiment.

Datavmwwecollected on all plants for 11 traits; seed-
ling bitterness; spine color; nodes to first pistillate
flower; percent (%) gynoecious nodes for nodes 1-20; percent
(%) nodes with laterals; total fruit number; number of
fruits on the main stem; number of fruits on the laterals;
fruit diameter (cm.); fruit length (cm.); and fruit yield
(gm.). Seedling bitterness was evaluated using the tasting
method described by Andeweg and De Bruyn (1959). Spine color
was evaluated on either develOping or mature fruit, and
characterized as either black or white spined. Gynoecious
nodes were defined as nodes bearing female flowers- Fruit
diameter and length measurments were made on randomly

selected fruits.

Statistical Procedure

 

Missing values for all characters were estimated
using Yate's pseudo approximations (Neter & Wasserman, 197A).

Tests of genotypic ratios and tests for linkage of
spine color, bitterness, nodes to first pistillate flower,
and percent gynoecious nodes for nodes 1-20 were performed
using the Chi square test for goodness of fit (Steel & Torrie,
1980).

Values for 05 and 0%, the genetic and error variances,
respectively, were calculated by equating estimated mean
squares, from the analysis of variance, with expected mean
squares and solving for the apprOpriate component of var-

iance.

n

Heritability was calculated using the formula:

H2=og/ (oE/r+cg) where 05 and a;

error variances, respectively, and r is the number of

are the genetic and

replications. This formula gives heritability in the
broad sense and on an entry mean basis.

Gains from selection were predicted using the formula:

2)l/2
G

per cycle, c the parental control value, K is the selection

GC= (cKo§)/ (cg/r+o with Gc being the genetic gain

differential in standard units, r equals the number of

replications, and 05 and GE are genetic and error variances,

respectively.
Tests for number of effective factors were performed

using the Castle-Wright formula (Wright, 1937). This value

2
G

effective factors, R equalling the range of the F

is expressed as K=R2/ 80 with K equallingtfluanumber of

2 population,

and cg representing the genetic variance. In the original
formula, R was defined as the range of the parents in a
mating, but Lawrance and Frey (1976) argued that the range

of the F segregates was a more appropriate estimate of R

2
when the parents did not represent the genotypic extremes
for segregating loci. An effective factor does not neces-
sarily represent a single locus, but rather may represent
a cluster of genes or even an entire chromosome.

Tests of gene action were performed for characters

measured, when possible. Mid-parent values were compared

with F2 means: significant differences of these values

suggest rmww—additive gene action,whereas nonsignificant
differences are suggestive of additive gene action (Mather
& Jinks, 1971; Rosielle & Frey, 1977).

Potence ratios were calculated for nodes to first
pistillate flower and percent gynoecious nodes for nodes
1-20, using the formula h3= XFl- XMP)/ (XHP JMP) where h3
is the potence ratio, X

equals the mean of the F genera—

Fl
equals the mid parent value (X

1

tion, XMP HP+ XLP)/ 2, and

EHP and ELF are the mean of the parents having higher and
lower expression of the character, respectively (Mather &
Jinks, 1971). Potence ratios explain the direction (positive
or negative) and degree (partial or complete) of dominance
demonstrated by the loci under consideration.

Transgressive segregates were defined as those plants
exceeding the high parent value by one standard deviation,
where such information was available. Transgressive segre-
gates for yield were defined as those plants equalling or
exceeding the mean of the F1 population. This definition
was adOpted since datavmnmeunavailable on one of the parents,
and Neihuis and Lower (1980) have shown heterosis for yield

in crosses of closely related 9. sativus inbred lines and

g. hardwickii line LJ 90A30. Transgressive segregates were

 

not defined for traits where datavwnwanot available for one
of the parents, and where heterosis has not previously been

observed.

RESULTS AND DISCUSSION

 

Significant variation was observed for all traits.
F tests for genotypic effects were significant at the

0.0005 level for all traits assayed.

 

Bitterness, derived from the g. hardwickii parent,
segregated in the F2 (Cochran, 1937; Barnham, 1953;
Andeweg & DeBruyn, 1959), giving a good fit to a 3:1 ratio
(0.25 < P < 0.5; Table 1). Selection against seedling
bitterness should be exercised in any breeding program,
since seedling bitterness is associated with bitterness in
fruit tissue. Contrary to expectation (Cochran, 1937;
Hutchins, 19A0; Shanmugasundaram et al., 1971) spine color
segregated in a 9:7 ratio of black spine to white spine,
(0.90 < P < 0.95) (Table 1). This ratio is different from
the 3:1 or 15:1 ratio's reported previously; where
pleiotropic effects on mature fruit color and fruit netting
were reported as well (Hutchins, 19AO). The sole report
where a pleiotropic effect on mature fruit color (orange
mature fruit color associated with black spine color) was
not reported a 3:1 ratio was also observed (Cochran, 1937).

The black spine color derived from g. hardwickii can be

 

explained by a 2 gene epistatic model, therefore these
genes probably are distinct from those controlling spine
color in Q. sativus. Linkage was tested for, and they are

unlinked to the gene controlling bitterness.

Female expression was measured as nodes to first
pistillate flower, and percent gynoecious nodes for nodes
1-20. Both measures gave similar results: frequency
distributions were trimodal in the F2 (Figures 1 and 2);
number of effective factor pairs vans estimated at 2 (Table 5);
and potence ratios were nearly identical (Table A). For the
Chi square test, homozygous and heterozygous classes were

defined in terms of the parents and F respectively. In the

1
case of nodes to first pistillate flower, the ratio was
adjusted for misclassification of heterozygotes (Figures 1
and Table 1). When the Chi square tests were performed,
both measures approximated a 1:2:1 ratio (0.1 <P< 0.25 and
0.9 <P< 0.95, respectively). These results are similar to
those reported by Kubicki (1969) where he obtained a 1:2:1
ratio for nodes to first pistillate flower in the F2 of a
cross between a gynoecious inbred line and a monoecious
inbred line of Q. sativus. He explained his results on the
basis of segregation of alleles at the Acr locus. Because
both measures fit a 1:2:1 ratio, and have nearly identical
potence ratios, it is postulated that they are controlled
by the same locus. This follows intuitively as well, for
the following reason. A plant having a low number of nodes
to first pistillate flower will have a high percent gynoec-
ious expression for nodes 1-20. The converse is also true.
Further, plants having an intermediate number of nodes to

first pistillate flower will have intermediate gynoecious

expression.

 

Both measures of gynoecious expression had high herit-
abilities: 0.9A and 0.93 respectively (Table 1). The high
gains from selection (Table 2) also reinforces the conclusion
that gynoecious expression is simply inherited. Based on
this evidence, one would assume that the trait would be
responsive to selection. However, transgressive segregates
appeared at low frequencies or not at all (Table 3) indica—
ting that gains might be rapid at first, but would likely
level off soon.

Based on the estimate of number of effective factors
(Table 5) percent nodes with laterals is under the control
of a relatively small number of factors. Tests for types
of gene action showed highly significant differences
between the mid parent value and F2 mean, suggesting some
type of non-additive gene action involved in the expression
of the trait. Further, the F1 mean is equal to or greater
than the mean of the high parent (Figure 3), suggesting
dominance and/or epistasis is involved. The presence of
dominance or epistasis does not allow an
accurate estimate of the number of effective factors
(Wright, 1937). Percent nodes with laterals also had a
high heritability estimate (0.92, Table 2), and high gains
from selection (Table 6). Because of the involvement of
either epistasis or dominance the heritability value
overestimates narrow sense heritability, and actual gains
from selection would be less than predicted gains. Trans—

gressive segregates could not be defined for percent nodes

10

 

with laterals because the mean of the high parent plus one
standard deviation gave a value greater than 1.0, therefore

it was impossible for any individuals to fall in this class.

According to the test for number of effective factors
(Table 5), fruit number is also under the control of a small
number of factors. Narrow sense heritability for fruit
number in g. sativus, as reported by Smith (Smith, Lower &
Moll, 1978) using a random mating population derived from
18 inbred lines from various breeding programs in the U.S.,
was 0.17. Estimates of narrow sense heritability for fruit

number in a cross with C. hardwickii is given by Horton as

 

0.88 using parent offspring regression (Horton, 1980; Lower,
1980). Our estimate of heritability for fruit number is 0.96
(Table 2). Lower (1980) indicated that the variance for fruit
number is primarily additive with some additive x additive
epistasis. Therefore narrow sense heritability should not

be considerably less than our estimate of heritability. Gains
per cycle of selection for fruit number, using Sl recurrent
selection and a 20% selection intensity, have been estimated
by Lower (1980), for a gynoecious synthetic population and

a g. hardwickii introgressed exotic population. Estimates

 

of 0.37 and 0.A9 were calcualted for the respective populations.
The estimate of gain per cycle with S1 recurrent selection
and a 20% selection intensity would be 6.85. At a 10%

selection intensity, gain was estimated at 8.61 (Table 6).

ll

Number of fruits on the main stem is apprently much
more complex than any other character examined. The number
of effectivefEctors estimate was higher than that of any
other trait (a value of 18), including fruit yield (Table 5).
The test for type of gene action suggests the involvement of
non-additive gene action. Number of fruit on the main stem
had the lowest heritability of any trait examined; 0.67
(Table 2). Gains from selection were much lower than either
fruit number or fruit on the laterals (Table 6). Because non-
additive types of gene action are probably involved in the
expression of this character, narrow sense herit—
ability is overestimated by this heritability estimate, and
actual gains from selection will be less than predicted.
Further evidence for the involvement of non-additive gene
action in the expression of this trait is the extremely
large number of transgressive segregates (Table 7) which
amounted to more than 50% of the pOpulation.

Number of fruit on the laterals had a low estimate of
number of effective factors (Table 5). However, because

of the abnormal distribution in the F non-additive types

2’
of gene action are probably involved (Figure 6). Number of
fruit on the laterals had a high heritability: 0.95 (Table 1).
Gains from selection were also high (Table 6) and explained

82% of the gains in selection for fruit number.

12

Table 1.

Chi square test for goodness of fit for several

traits in the cross Cucumis sativus x g.
hardwickii

 

 

 

 

 

Frequency/ClassZ
Expected 2
Trait Gen. P F P Ratio P
l l 2
Bitter- Fl 63 0:1
ness
F2 68 18A 1:3 0.529 0.25 <P<0.5
Spine
Color Fl 63 0:1
F2 108 1A0 7:9 0.00A 0.9 <P<0.95
Nodes to
lst
pistillate Fl 18 A5
flower
F2 90 105 57 1.56:1.AAzl 3.15
% 0. <P<0.25
gynoecious
nodes,
nodes 1-20 Fl 63
F2 62 129 61 1:2:1 0.151 0.9 <P<0.95
ZPl, P2, and F1 are defined in terms of the respective
generations or as indicated in the text.

13

Table 2. Broad sense heritabilities for various traits
in the cross cucumis sativus x g. hardwickii

 

 

 

 

Trait H2

Nodes to first pistillate flower 0.9A
% gynoecious nodes, nodes 1-20 0.93
% nodes with laterals 0.92
Fruit number 0.96
Fruits on the main stem 0.67
Fruits on the laterals 0.95
Fruit diameter 0.89
Fruit length 0.89
Fruit yield 0.9A

 

Z r = 12

1A

'Table 3. Test of gene action for several traits in the
cross Cucumis sativus x g. hardwickii

 

 

 

'Traits P

P MP F t value

 

iNodes to lst
pistillate flower 2.30

% gynoecious nodes,
nodes 1-20 93.02

% nodes with
laterals 0.07

No. fruit on
main stem 1.A9

31.6 16.95 11.12 6.26**

0.08 A6.55 38.3 2.50**

0.85 0.A6 0.63 5.0A**

0.0 0.7A 3.A1 5.56**

 

**

significant at the 0.01 level

2 all values x 100

Table A. Potence ratios for 2 measures of gynoecious
sex expression in the cross cucumis sativus
x g. hardwickii

 

 

 

 

Measure h3 Description

Nodes to first

pistillate flower -0.528 Negative, incomplete dom.
% gynoecious nodes,

nodes 1-20 -0.51 Negative, incomplete dom.

 

15

ct

Table 5. Number of effective factor pairs by which parents
in the cross Cucumis sativus x g. hardwickii for

 

 

 

 

11 traits
Trait Number of effective factors (K)
Bitterness l
Spine color 2
Nodes to lst pistillate flower 2
% gynoecious nodes, nodes 1-20 2
% nodes with laterals 2
Fruit number 3
Fruit on the main stem 18
Fruit on the laterals 2
Fruit diameter 7
Fruit length 11
Fruit yield 6

 

l6

Mean fruit diameter and mean fruit length showed
normal distributions in the F2 (Figures 7 and 8). Both
had higher number of effectivefectorsestimates than for
most characters examined: 7 and 11 respectively (Table 5).
Because of the normal distributions obtained in the F2,
additive gene action is suggested to be primarily responsible
for their control. Fruit diameter and fruit length had
lower heritability estimates than most characters assayed;
values were 0.88 and 0.89 respectively (Table 2), and
moderate gains from selection (Table 6).

Number of effective factor estimates for fruit yield
is low: a value of 6. Based on the work of Neihuis and
Lower (1980) there is significant heterosis for yield in
crosses of Q. sativus with LJ 90A30. This suggests either
dominance,and/or epistasis is involved in the expression of
the character. The heritability and gains from selection
estimates for yield are exceptionally high (Tables 1 and 2).
Because of the presence of either epistasis or dominance
gains from selection will be less than predicted (gains

estimated @ 260% of (_3_. sativus parent mean for S selection;

1
Tables 6 and 8).

With the exception of fruits on the main stem, high
heritabilities and gains from selection were observed from
all traits. As indicated, non-additive types of gene action

were probably involved in the expression of the traits

assayed, therefore actual gains from selection will likely

17

be less than predicted gains. Further, the estimates of
heritability overestimate narrow sense heritability.
Selection for high yielding, multiple fruited,
multiple branching lines with the desired degree of female
expression and mature fruit size is potentially possible
in populations derived from the cross. No estimates of
genotypic correlations have been made. This information
would be of great value in the development of a selection
scheme for this population, where it may be necessary to
use indexed selection to Optimize gains from selection
for more than one trait. Because of the probable involve-
ment of non—additive types of gene action and the
occurance of heterosis for yield, one can capitalize on
these potential increases primarily in a hybrid product.
Therefore the use of some type of interpOpulation improve-

ment scheme could be justified.

l8

Table 6. Gains per cycle from selection with a 10% selection
differential for various traits in the cross
Cucumis sativus x g. hardwickii

 

 

 

2 Selection Scheme

 

Trait Mass S1 S2
Nodes to lst pistil—

late flower 7.0 lA.61 18.08
% gynoecious nodes,

nodes 1-20 22.A8 AA.97 55.75
% nodes w. laterals 0.21 0.A3 0.53
Fruit number A.3 8.61 10.6A
Fruits on the main

stem 0.73 l.A5 1.88
Fruits on the

laterals 3.56 7.11 8.8
Fruit diameter 0.32 0.63 0.79
Fruit length 0.95 1.9 2.37
Fruit yield 3A3.36 - 686.73 8A9.76

 

Zone parent selected after pollination

Table 7. Transgressive segregates for various traits in the
cross Cucumis sativus x g. hardwickii

 

 

 

Trait Definition Number %

 

Nodes to first

pistillate flower <l.6 1 0.39
% gynoecious nodes,

nodes 1-20 <97.0 0 0.00
Fruit on the

main stem >2.08 1A5 57.5A
Fruit yield >1530.A gm. A 1.6

 

l9

 

Table 8. Generation means for several traits in the cross
Cucumis sativus x C. hardwickii

 

 

 

Generation mean

Trait P1 P2 F1 F2

 

Nodes to first
pistillate flower 2.30 31.6 9.22 11.12

% gynoecious nodes,y

 

nodes 1-20 93.02 0.08 22.86 38.3
% nodes w.

laterals 0.07 0.85 0.88 0.63
Fruit number l.A8 ___z 17.56 6.58
Fruits on the

main stem l.A8 0.0 A.53 3.Al
Fruits on the

laterals 0.0 --—Z 13.03 3.17
Fruit diameter 2

(cm.) 5.56 --- A.56 A.5A
Fruit length

(cm.) 11.09 ---z 8.26 8.08

Fruit yield
(gm.) 262.71 ---z 153.u0 587.71

 

ya11 values x 100

Zmeans unavailable

20

Figure 1.

Frequency distribution of nodes
to first pistillate flower for
the P1, P2, F1: F2 generations
of the cross Cucumis sativus x

 

g. hardwickii

 

 

 

 

 

 

 

 

 

 

 

 

 

01
P1
”d
P!
”d
10
o t I I I I fi
10 N S Q C
“an.
”I
3‘ F1
‘0. fix IR
1 r I I f 1 1
‘0 fl so an O
man.-
“1
fl- '3
”1
Q:
0%; I II I n“
r 1 r fi' 1 V ' I Y 1
10 an & 40 D
"mum
Figure 1.

22

F ‘-¢n_
I

Figure 2. Frequency distribution of percent
gynoecious nodes for nodes 1-20 for
the P1: P2, F1, and F2 generations
of the cross Cucumis sativus x g.
hardwickii

 

 

23

 

 

 

d

 

 

 

8.1

 

 

F1

81

 

 

 

Bi

8.4

 

 

 

$fi”

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3.

Frequency distribution of percent
(%) nodes with laterals for the
P1, P2, F1, and F2 generations of
the cross Cucumis sativus x g.
hardwickii

 

 

25

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3.

26

‘r‘uI‘
Ii

 

Figure A. Frequency distribution of fruit
number for the P1, F1, and F2

generations of the cross Cucumis
sativus x g. hardwickii

27

PDQ

 

 

 

 

10"

 

F1

 

10‘

 

 

Figure A.

28

 

Figure 5. Frequency distribution of fruits
on the main stem for the P1, P2,
F1, and F2 generations of the cross
Cucumis sativus x g. hardwickii

 

 

29

 

 

 

 

 

 

 

 

 

.- .‘r—
Id .1
01 " ‘. ll
~°" I m”:
ao~ :-
”‘I ”d
0 fl 0 V fi
0 0 O I
M.-- “I‘-
”a
'1
h.”‘
o 1 1 I fl
0 0 fl 1'

 

 

 

 

Figure 5.

30

Figure 6. Frequency distribution of fruits
on the laterals for the P1, F1,
and F2 generations of the Cucumis
sativus x Q. hardwickii

31

 

 

'1

 

 

 

 

 

 

 

a.
”I
0 f I f t
O 5 10 15 fl
“a...“
”u
F1
201
o w H§
T t r
5 1° 15 ”
“GIN“
100-4
'2
”.7
”a
“q
20..
0 fit I V Y
0 5 10 15 20

“MI.“

Figure 6.

32

 

Figure 7. Frequency distribution of mean
fruit diameter in cm. for the

Pl, F1, and F2 generations of
the cross Cucumis sativus x g.
hardwickii

33

 

1O

 

 

 

 

 

 

H
F“ T I i A
3.0 4.0 5.0 0.0 7.0
“MW
. j'
I F1 L2
L 1 f fl '
3.0 4.0 5.0 6.0 7.0
“MIG!”
F2
[—5 ”an .1
I I I j
u ‘0 5.0 5.0 7.0

 

Figure 7.

3A

Figure 8.

Frequency distribution of mean
fruit length in (cm.) for the
Pl, F1, and F2 generations of
the cross Cucumis sativus x g.
hardwickii

35

 

P1

 

 

 

 

 

.11
. H r—aI ‘II HIP-L. r——':'I nn
:- an m;
“”h’
'1
.1
.1
. T T j
u no 1”
“w“
.1
F2
.-
'1
O "I r firman i
I. 100 1”
“U.“
Figure 8.

36

Figure 9. Frequency distribution of fruit
yield in gm. for the P , F1’
and F generations of the cross
Cucumis sativus x C. hardwickii

 

37

 

 

 

 

 

 

38

.1
.1
. ‘ .1 ‘l f j I
an inn an: an
nan-hm
.1
F1
.d
o n MM mun n
T U I T
on run an: an
haw-hm
8c.
.d n
.1 l
o 41f‘1
an van am am
haw-bu
Figure 9.

 

REFERENCES

 

Andeweg, J.M., and J.W. De Bruyn. 1959. Breeding of non-
bitter cucumbers. Euphytica 8:13:20.

Baker, L.R., J.W. Scott, and J.E. Wilson. 1973. Seedless
Pickles - A New Concept. MSU Ag. Expt. Sta. Farm Sci.
Res. Report No. 227.

Barnham, w.s. 1953. The inheritance of a bitter principle I
in cucumbers. Proc. Amer. Soc. Hort. Sci. 62:AA1-AA2.

Cantliffe, D.J. 197A. Promotion of fruit set and reduction
of seed number in pollinated fruit of cucumber by chlor-
flurenol. HortScience 9(6): 577-578. A

Cochran, F.D. 1937. Breeding cucumbers for resistance to
Downy mildew. Amer. Soc. Hort. Sci. Proc. 35:5Al-5A3.

Connor, L.J. and E.C. Martin. 1970. The effect of delayed
pollination on yield of cucumbers grown for machine
harvests. JASHS 95:A56-A58.

Denna, D.W. 1973. Effects of genetic parthenocarpy and
gynoecious flowering habit on fruit production and
growth of cucumber Cucumis sativus L. JASHS 98:602-60A.

 

Elassar, G., J. Rudich, D. Palevitch, and N. Kedar. 197A.
Induction of parthenocarpic fruit development in cucum-
ber by growth regulators. HortScience 9(3):238-239.

Horst, E.K. and R.L. Lower. 1977. A phytotron study of the
effects of photoperiod and temperature on flowering and
growth response in Cucumis sativus L. and g. hardwickii.
(Abs.) HortScience l2(3):235.

 

 

Horst, E.K. and R.L. Lower. 1978. Cucumis hardwickii: A
source of germplasm for the cucumber breeder. Curcurbit
Genetics Cooperative Report 1:5.

 

Horton, R.R. 1980. An estimate of heritability of fruit
number from a cross between a picling cucumber inbred
(Cucumis sativus L.) and an inbred of g. hardwickii R.
CGC Rpt. 3:10-11.

 

 

Hutchins, A.E. 19A0. Inheritance in the cucumber. Journal
Agr. Res. 60:117-128.

39

 

Kubicki, B. 1969. Investigations on sex determination in
cucumber (Cucumis satiVus L.) Iszultiple alleles of
locus ACR Genetica Polonica 10:23-67.

 

Lawerance, P.L. and K.J. Frey. 1976. Inheritance of grain
yield in cat species crosses, (Avena sativa L. x A.
sterilis L.) Egyptian J. of Gen. and Cytol.5:A00-A09.

 

Lower, R.L. 1980. Presentation to the Pickling Cucumber
Improvement Committee of Pickle Packers International.

Mather, K. and J.L. Jinks. 1971. Biometrical genetics.
Cornell Univ. Press, Ithaca, N.Y.

McCollum, J.P. 193A. Vegetative and reproductive responses
associated with fruit development in the cucumber.
Cornell Univ. Agr. Exp. Sta. Mem. 163.

Miller, C.H. and G.R. Hughes. 1969. Harvest indices for

pickling cucumber in once over harvested systems. JASHS
9A:A85-A87.

Neihuis, J. and R.L. Lower. 1979. Interspecific grafting
to promote flowering in Cucumis hardwickii. CGC Rpt.
2:11-12.

 

Neihuis, J. and R.L. Lower. 1980. Influence of recipricol
donors scions of Cucumis sativus and g. hardwickii.
CGC Rpt. 3:17-19.

 

 

Neihuis, J. and R.L. Lower. 1980. Heterosis estimates for
several characteristics in a cross between a gynoecious
inbred of Cucumis sativus L. and g. hardwickii R. CGC
Rpt. 3:20-21.

 

 

Neter, J., and W. Wasserman. 197A. Applied Linear Statis-
tical Models. Richard D. Irwin, Inc. Homewood, 111.

Pike, L.M. and C.E. Peterson. 1969. Inheritance of
parthenocarpy in the cucumber (Cucumis sativus L.)
Euphytica 18:101-105.

 

Quebedaux, B. and E.M. Beyer, Jr. 1972. Chemically induced
parthenocarpy in cucumber by a new inhibitor of auxin
transport. HortScience 7:A7A-A76.

Robinson, R.W. and E. Kowalski. 1978. Interspecific
hybridization of Cucumis. CGC Rpt. 1:A0.

Rosielle, A.A. and K.J. Frey. 1977. Inheritance of harvest
index and related traits in Oats. Crop Science 17:23-28.

A0

 

 

Shanmugasundaram, S., P.H. Williams, and C.E. Peterson. 1971.

Inheritance of spine color in cucumber. HortScience
6(3):213-21A.

Smith, O.S., R.L. Lower, and R.R. M011. 1978. Estimates of
heritability and variance components in pickling cucumber.
JASHS 103:222-225.

Whitaker, J.W. and G.N. Davis. 1962. Cucurbits, Botany,
Cultivation, and Utilization. Leonard Hill Lmt. London.

Wright, S. 1937. The results of crosses between inbred

strains of guinea pigs, differing in number of digits.
Genetics 19:537-571.

A1

 

APPENDIX

APPENDIX
Analysis of variance

Table 9. Analysis of variance of nodes to first pistillate
flower for the cross Cucumis sativus x g.

 

 

 

 

hardwickii
Source S.S. d.f. MS F test Prob
Block 11A3.8005 20 57.190023 0.983 >0.5
Genotype 31977.23 20 1598.8615 27.A79 <0.0005
Error 230Al.15 396 58.18A7
Total 57172.18 A36

 

Table 10. Analysis of variance of percent gynoecious nodes,
nodes l-20 for the cross Cucumis sativus x g.

 

 

 

 

hardwickii
Source S.S. d.f. MS F test Prob
Block 17087.01 20 85A.3503 1.302 0.1<P<0.5
Genotype 308638.91 20 15A3l.95 23.519 <0.0005
Error 259839.18 396 656.13A3
Total 585555.10 A36

 

A2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 11. Analysis of variance of percent nodes with
laterals for the cross Cucumis sativus x C.
hardwickii'

Source S.S. d.f. MS F test Prob

Block 2.A19996 20 0.12099979 1.782 <0.025

Genotype 28.157891 20 l.A07895 20.736 <0.0005

Error 26.751588 39A 0.067897A

Total 57.329A75 A3A

Table 12. Analysis of variance of fruit number for the cross
Cucumis sativus x g. hardwickii

Source S.S d.f. MS F test Prob

Block AA2.5535 20 22.127675 1.76A 0.1<P<O.25

Genotype 911A.2A67 17 536.1322 A2.73A <0.0005

Error A215.3989 336 12.5A58

Total 13772.20 373

Table 13. Analysis of variance of number of fruits on the
main stem for the cross Cucumis sativus x g.
hardwickii

Source S.S. d.f. MS F test Prob

Block 239.9169 20 11.9958A6 1.679 0.025<P<0.05

Genotype A60.3398 l7 27.078812 3.79 <0.0005

Error 1971.5876 336 7.1A397

Total 2671.8A3A 373

 

A3

Table 1A.

Analysis of variance of number of fruit on
the laterals.for the cross Cucumis sativus
x g. hardwickii

 

 

 

 

 

 

 

 

 

 

 

 

 

Source S.S. d.f. MS F test Prob
Block 327.1032 20 16.355161 l.Al6 0.05<P<0.1
Genotype 6353.7A68 l7 373.7A98 32.353 <0.0005
Error 3881.6015 336 11.552A
Total 10562.A5 373
Table 15. Analysis of variance of fruit diameter for the
cross Cucumis sativus x Q. hardwickii
Source S.S. d.f. MS F test Prob
Block 8.130383 20 0.A0651915 1.776 0.01<P<0.025
Genotype 56.13291 17 3.301936 1A.A27 <0.0005
Error 75.983912 332 0.22887
Total 1A0.2A72 369
Table 16. Analysis of variance of fruit length for the cross
Cucumis sativus x Q. hardwickii
Block 93.9172A8 20 A.695862 2.AAA <0.001
Genotype A97.9096 17 29.2888 15.2AA <0.0005
Error 637.8766 332 1.9213
Total 1229.7035 369

 

AA

Table 17. Analysis of variance of fruit yield for the cross

Cucumis sativus x g. hardwickii

 

 

 

 

Source S.S. d.f. MS F test Prob
Block 68A6028. 20 3A2301.A 2.696 <0.001
Genotype 60055287. 17 353266A.O 27.82 <0.0005
Error A266A05A. 336 126976.A

Total 109565369. 373

 

A5

   

"‘1111111111