COUPLED WETH

ENTERCRQSSNG
A MEANS OF

NEUTRON RRADEAT‘ON A5
ETIC VARIABIUTY W. NAVY

ENCREASMG GEN
OLUS WLGARES L.)

BEANS (PHASE

RECURRENT

Degree 0‘ pk. D.

MICHKGAN STATE UNIVERSITY
Antonio M. Pinchinat
1964

Thesis *or We

LIBRARY

Michigan State

University

 

This is to certify that the
thesis entitled
RECLIL‘K’LLZJT II-ETLRCRLL-SSIN} COLIILZD MITH lIiITTIiC-Ii
IRR-XDIATICEI‘J AS A I-L‘JJAIIS OF 115035713133 3515mm

VnRIABILITY IN MVY BEANS (ma-3mm: VL’L-GARIS L.)

presented by
Antonio Michael Pinchinat

has been accepted towards fulfillment
of the requirements for

,9
M degree in Waco

Major professor

 

Date M11341 /7>é4(
/ / '

0-169

ROSA”: e om

ROOM 53E ONLY

ABSTRACT
RECURRENT INTERCROSSING COUPLED WITH NEUTRON
IRRADIATION AS A MEANS OF INCREASING GENETIC
VARIABILITY IN NAVY BEANS (PHASEOLUS VULGARIS L.)

By Antonio M. Pinchinat

The hypothesis has been prOposed that a combination
of irradiation and recurrent intercrossing would generate
more variability in a complex trait than would either
method alone. To test it, 10 homozygous lines of navy
beans were crossed in all possible combinations. Each
of the resulting 45 first cycle intercrosses was
simultaneously selfed for seed increase and out-crossed
to 4 unrelated hybrids. These second cycle intercrosses
were also selfed afterwards to produce enough seed for
the subsequent treatments. Then each group, including
the original parents, was split into two lots: one to
be neutron—irradiated and the other kept as control.

They were grown in bulk for two years and as single plant
progenies the third year.

The statistical analysis of the recorded data
indicated that the combination of recurrent intercrossing
and irradiation increased both variances and coefficients
of variability of yield and its components more than
did either alone. Intercrossing alone rated better than
the radiation treatment‘EEEIES. But only a few isolated

significant changes in treatment means emerged. Apparently

the neutron dosage applied was not high enough to produce
drastic effects.

,The second cycle intercrossing alone or with
radiation excelled the first cycle intercrossing in
increasing mean seed weight (Z) means, and without
radiation, in increasing seed number per pod (Y)
variances. Their effects otherwise were similar.

Bean yield per plant (W) was seen to be strongly
correlated with number of pods (X), average seed number
per pod (Y) and mean seed weight (Z), as postulated by
the geometric concept of yield. Component X showed
negative correlations with both Y and Z, which in turn
were positively correlated with each other. The nature
of these correlations does not preclude the existence
of different gene systems governing the components.
Correlation patterns did not sensibly vary with treatments.

Hybridization alone or with radiation increased
regression values of yield on its components more
consistently than did irradiation alone. Combined with
radiation, hybridization made yield more dependent on
number of seed per pod and hence would facilitate yield
improvement through manipulation of the components. This
combination induced enough variability to permit isolation

of transgressive recombinant lines for further selection.

RECURRENT INTERCROSSING COUPLED WITH NEUTRON
IRRADIATION AS A MEANS OF INCREASING GENETIC

VARIABILITY IN NAVY BEANS (PHASEOLUS VULGARIS L.)

 

by

i}
[Y]
\ (’\
\‘\'
' (J

Antonio N: Pinchinat

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Crop Science

1964

s

ACKNOWLEDGEMENTS

I wish to express my sincere gratitude to
Dr. M. w. Adams and Dr. S. T. Dexter for their
guidance and helpful advice throughout the course
of my studies at this University.

I am also grateful to Professor H. M. Brown
and Dr. J. B. Grafius for their suggestions in the
statistical processing of the data and to Dr. C. M.
Harrison for his critical review of the paper.

Sincere thanks are also extended to Dr. A. F.
Yanders, Dr. F. C. Elliott, and Dr. L. W. Mericle
for their interest in this study and to Dr. T. S.
Osborne for his invaluable cooperation in the

irradiation of the experimental materials.

TABLE OF CONTENTS

Page

LIST OF TABLES........................... i
INTRODUCTION............................. 1
REVIEW or LITERATURE..................... 3
MATERIALS AND METHODS.................... 12
RESULTS.................................. 20
DISCUSSIONS.............................. 40
CONCLUSIONS.............................. 44
LITERATURE CITED......................... 46

APPENDIX TABLESOOOOOOOOOOOOOOOOOOOOOOOOOO 53

Table

10

11

12

13

14

LIST OF TABLES

Neutron spectrum of graphite reactor
With U235 plate (Univ. Of Tenn.).............

Correlation coefficients between bean
yield (W) and its components (X,Y,Z)
and among the components.....................

Test of significance of differences
between mean correlation coefficients........

Regression of bean yield on its
components...................................

Comparison of treatment error variances
for number of pods per plot (X)..............

Comparison of treatment error variances
for mean seed number per pod (Y).............

Comparisons of treatment error variances
for mean seed weight (2)000000000000000000000

Comparisons of treatment error variances
for seed yield per plot (W)..................

Comparisons of differences between
treatment means for number of pods per

plOt (X)...O.O.OOOOOOOOOOOOOOCOOOOOOOOOOOOOCO

Comparisons of differences between treatment
means for mean seed number per pod...........

Comparisons of differences between treatment
means for mean seed weight (Z)...............

Comparisons of differences between treatment
means for seed yield per plot (W)............

Coefficients of variation of number of
pOds per plot (X)...eooeooooooooooeooeoooeooe

Coefficients of variation for mean seed
numer per pOd (Y)oooooooeoooooooooooooooocoo

Page

15

21

22

24

26

27

28

29

31

32

33

34

35

36

Table Page

15 Coefficients of variation for mean seed

weight.(2)....o..............................

37

16 Coefficients of variation for seed
Yield per plOt (w)oooeooooooooooooooooooooooo 38

INTRODUCTION

Genetic studies on beans (Phaseolus vulgaris L.) have

 

been relatively limited even though the plant has agronomic
features advantageous to such investigations. Almost
wholly self-pollinated, the species possesses an inter-
mediate haploid chromosome number (n=11) and a well defined
set of fitness components (2,5,9). Its life cycle is
comparatively short (85-110 days) and each plant produces
ample quantities of seed.

The first commercial artificially induced navy bean
mutant, Sanilac (12,13), was developed at Michigan State
University through X-irradiation and subsequent backcrossing.
Irradiation thus offers a means of inducing variability in
an elite population without degrading it, as is often the
case with wide outcrossings, by the introduction of disparate
germ plasm.

_Theoretically, recurrent cycles of intercrossing are
expected to result in a gradual reduction in size of
initially intact gene blocks. The present experiment was
aimed primarily at enhancing genic recombination in a self-
pollinated species by combining neutron irradiation and
recurrent intercrossing. The effectiveness of this technique

might be reflected in measurable changes in some fitness

components or yield factors, and permit the isolation of

lines improved in one or more performance characteristics.

REVIEW OF LITERATURE

Hybridization

The success of hybridization as a breeding method rests
on the release of variability in a population through gene
recombination. Hagedoorn (18) remarks that the variability
caused by crossing even closely related subspecies may be
enormous. Most published reports on genetic recombination
have dealt with cross-fertilizing species. Dobzhansky and
his group (10) reached the conclusion that genetic variance
of viability in natural pepulations of Drosophila prosaltans
scarcely depends on newly arisen mutants. Most of the
variance traced back to recombination within the accumulated
store of genetic variants. A single generation of inter-
crossing between sets of second chromosomes produced an
unexpectedly large amount of genetic variability as
compared to the wild population. Similar results were
obtained with several other Drosophila species (41,42,43).
Milkman (30) believes that recombination in the chromosomes
contributed by randomly chosen pairs of parents in
Drosphila melanogaster could essentially generate the full
range of phenotypic variation characterizing the population.

Crossing early- and late-heading timothy plants
enabled Van Dijk (47) to enlarge the diversity of forms of
this grass. In a long-term study of composite-crosses with

barley, Harlan (23) demonstrated that steady improvement in

yield and other agronomic traits can be made in later
generations, presumably through increased recombination.
Donnelly and Clark (11) found inter- as well as intra-
specific crosses in.222ig_capable of generating potentially
useful variants for plant and seed types as early as the

F2 generation.

Hybridization and the Breakage of Linkage Blocks
Mitchell (31) attempted to:
(1) Break down initial gene sequences in X-chromosomes
of Drosphila melanogaster
(2) Accumulate new linkages and
(3) Promote continued change of the new linkages
through recombination in successive generations.
The mating system he devised allowed genotypic
variability to depend mainly on the rate of recombination
of sex-linked genes.
4 He found that:
(a) the only major source of increased variability
in the populations was through accumulation
of linkages of genese derived from the X—
chromosomes of the two parental inbred lines
(b) recombination raised the frequency of non-
fecund pairs and lowered the mean survival of

the offspring of fecund pairs

(c) accumulation of linkages in the X-chromosomes

increased the variance of developmental rate,
but did not create any significant change in
means.

Multiple inversion heterozygosity, he suggested,
resulted in an accelerated rate of release of potential
variability in the structurally homozygous portion of the
genome.

If increased recombination were to reduce variance
where linkages are predominantly in coupling phase, the
reverse should take place in the repulsion phase of linkage,
according to Adams (1). He also reasons that successive
intercrossings should progressively disrupt homeostatic
linkage segments in self-pollinated species. To support
his arguments, he drew some figures from papers by Hanson
(19,20,21). The reported theoretical considerations
indicated that four generations of recurrent intercrossing
in plants would cut intact segment lengths from 1/2 to 1/8
of the original chromosome lengths. And, the calculated ;
average recombination frequencies per chromosome, under

these conditions, would be at least 38%.

Intercrossing, however, does not always increase
recombination. Results obtained by Stephens (46) from
certain crosses in cotton bear on this point. In an
interspecific cross, crossing-over in the non-homologous
terminal region differentiating the two species dropped
from 20 to 8%. Yet this drop was compensated for by slight,
but general, increases in other marked segments of the

chromosome.

Effects of Ionizing Radiations on the Genetic Material

That ionizing radiations can cause hereditary changes
in living matter has been known since 1927. Their mutagenic
action was demonstrated by Muller (33), Stadler (44,45),
and Goodspeed and Olson (16) among others. Sparrow and
his associates (40) have compiled a bibliography on the
effects of ionizing radiations on plants, for the period
1896-1955. At the latter date, Bacq and Alexander (6)
gathered and published the fundamentals of radiobiology.
Sparrow (39) in 1961 surveyed the current types of ionizing
radiation and their cytogenetic effects. Apart from
so-called point mutations, high energy radiations induce
a wide array of chromosome aberrations. Breakage and
translocation of fragments of chromatids and chromosomes

may lead to new gene associations and rare or totally

unexpected variants. Deletions, or losses of chromatin,

may induce lethal, aberrant, or fully viable novel types,
depending on the metabolic importance of the lost portion.
Cytologically the aberrations may appear as bridges, rings,
fragments or other configurations, although minute deletions
may escape direct detection.

Abnormal configurations followed X-irradiation of
Lilium anthers by Mitra (32). After X-ray treatment of
Drosophila melanogaster, Schacht (37) recovered a total of
20 translocations: 17 between chromosomes II and III,
and 3 between Y and II. More important, he also demonstrated
18 induced cross-overs: 11 singles, 6 doubles and 1 triple.
Similar irradiation - enchanced crossing over was noticed
by Wittinghill and Davis (48) and Scossiroli (38). In
X-rayed roots of onion seedlings, Cohn (8) identified two
classes of chromosome breaks based on average time of
restitution. Some breaks underwent rapid restitution
(15 min. or less), some remained open for a relatively
longer period (4 hrs. or more). The longer the breaks
remained open, the greater the opportunity for chromosomal
interchange. Heinz (24) reported a wide range of trans-
locations in tetraploid Dactylis seed bombarded with

neutrons for one-half hour.

Comparing Effects of Some Current Ionizing Radiations

Reviewing some studies with X—rays and neutrons (fast
and thermal), Elliott (14) concluded that the two types
of radiation differ sharply in some of their effects, at
least in barley. Neutrons, he notes, allow:

(1) more uniform seedling height and survival

of N1 plants,
(2) higher survival of low fertility N1 progeny, and
(3) higher frequencies of chromosomal aberrations
and mutations.

Bacq and Alexander (6) reported the damage caused by densely
ionizing radiations in the various stages of spermato-
genesis of Drosophila virilis. In all stages, fission
neutrons showed a higher relative biological efficiency
(REE) than X—rays of comparable energies. Kulik (26)
reported that treatments with gamma rays excelled thermal
neutron irradiation in inducing potentially useful mutants
in tomatoes. But, for the production of chromosome aberra-
tions, he deemed neutrons as superior to gamma rays. For
a given level of chromosome damage under conditions of
aeration, Evans and his co-workers (15) found fast neutrons
above 10 times more effective than gamma rays. In the
absence of air, the estimated effect ratio ran as high

as 18:1, in favor of neutrons. Biological effectiveness

of fast neutrons for radiation damage in field peas reached
40 times that of gamma rays, according to Hvostova and
Nevzgodina (25). Studies of the genetic effects of beta
rays, have been very limited compared with those of X-rays
and neutrons (34).

Expressing his reasons for suggesting the use of fast
neutrons, Dr. Osborne (35)remarked:

(1) Current knowledge (up to 1961) puts neutrons
ahead of gamma rays in producing gross
chromosomal changes

(2) With thermal neutrons - in which most ionization
results from capture by atomic nuclei and
subsequent disintegration - the chemical
composition of the target is extremely important.

(3) With fast neutrons - where ionizations result
mostly from collision, chemical composition is
of little importance, and moisture and storage
effects are trivial.

The foregoing survey of the literature exposes the
great potentiality of ionizing radiations, especially
neutrons, as a means of inducing genetic variability.
Coupling irradiation with intercrossing should enhance
genic recombination, assuming independent action of the

two breeding procedures.

10

As Lerner (28) points out for polygenic characters,
neither the number of loci nor the magnitudes of the effects
produced by allelic substitutions can be ordinarily known.
The usual available information consists of individual
phenotypic measurements for one or more generations and of
pedigree relationships between the members of the populatioh.
From these, means and statistics of higher order, such as
variances and covariances, can be computed as needed.

Yield (W) in navy beans, has been regarded as the
product of 3 components: pod number per plant (X), average
seed number per pod (Y), and mean seed weight (Z). Camacho
and his collaborators (7) reported negative correlations
between the yield components in kidney beans and, as
expected, a positive correlation between yield and each
component. Considering the genotypic correlations, an
increase in X seemed to cause a decrease in Y, and an increase
in Y corresponded to a reduction in Z. If linkage were to
be invoked from the results, he commented, genetic association
would be close between X and Y and between Y and Z, but
weak between X and Z.

If increased recombination shows up as variability, it
might reduce variance in the traits under investigation
where:

(1) linkages, if present, are predominantly

in the coupling phase, and

ll

(2) increased genetic variance leads to decreased
phenotypic variance due to a gain in homeostasis.
On the other hand, the breaking of repulsion phase linkages
would tend to enhance variances. Treatment with 15 Kr of
X-rays reduced means and increased variances in yield of
F2 peanut progenies and their parent population, according
to Gregory (17). The larger variances, often were

associated with the smaller means.

MATERIALS AND METHODS

Plant Materials

Ten parental lines of navy beans (Phaseolus vulgaris L.)
were used that had been previously field-grown for at least
six generations. They were tested for genotypic homogeneity
and only two off-type plants for maturity showed up, one in
each of two lines. They were all white-seeded, and two,
Sanilac and Michelite, are standard commercial varieties.
They differed in growth habit, maturity, and yield
components but all were fully inter-fertile.

The whole study comprised two stages. The first
stage, on which the present paper is based, involved the
ten parental lines (P), all their possible first cycle
intercrosses (I1), and second cycle intercrosses (12). The
second stage, currently underway, will comprise a third
(I3) and a fourth (I4) cycle of intercrossing, in addition
to the parental lines.

As each generation of intercrossing became available,
equal amounts of beans harvested from four plants at random
in each line were planted to produce selfed seed. Sufficient
material was thus obtained to allot sizable seed samples for

each of a control (C) and a radiation (R) series.

12

13

Hybridization

 

The breeding work took place in the Plant Science
greenhouse of Michigan State University, East Lansing.
Crossing the 10 pafients in all possible combinations gave

1

45 Fl's which after selfing became 45 Fz's. Simultaneously
:

each F1 was crossed to four unrelated ones to make up an
expected total of 180 second cycle intercross lines.
Actually 178 such lines were recovered and subsequently
selfed. Thus the original material subjected to neutron

irradiation in 1961 consisted of selfed seed of:

(1) 10 parents (P)
(2) 45 first—cycle intercrosses (II)
(3) 178 second-cycle intercrosses (I2)

About half of the seed of each line in each class served as
the control group. Both treatment series were field-planted
that same year, and their progenies grown in 1962 and again
in 1963. Both the first-cycle and second-cycle inter-
crosses have been grown for three generations after their
initial production. No differential viability in emergence
was noted in any year. In 1962, stands were thinned after
emergence to prevent undue competition and possible
elimination of weak plants. The 1963 nursery consisted

of single plant progenies taken at random from the field in
1962. It was assumed that no appreciable changes had

occurred in the parental material.

l4

Neutron Irradiation

 

In May 1961, Dr. S.T. Osborne, of the University of
Tennessee, irradiated the seeds in the graphite reactor of
the Oak Ridge National Laboratory, using an enriched
uranium plate beneath the sample cart. He calculated the
neutron flux and set the other environmental conditions
inherent to treatment. The material received a dose of fast
neutrons (1-2 Mev) equivalent to about 739 rep, for a total
exposure time of half an hour. Rather crude observations
had previously indicated that for an expected 70-80% seedling
emergence such a dose and rate would work satisfactorily.

To get reliable data on possible breakage of presumed
linkage groups large irradiated as well as control populations
were required.

Osborne and Lunden (36) have provided detailed information
on plant and seed irradiation at Oak Ridge. Table l, drawn
from their article, describes the composition of the neutron
source.

Although thermal neutrons constituted over half the
total flux, they contributed less than 3% of the effective

dose. This is, therefore, essentially a fast neutron source.

Table 1.

Neutron Spectrum of Graphite Reactor with

 

 

 

U235 Plan (Univ. of Tenn.)
Monitor Min. energy Neutrons Percentage Rep/hr. Percentage
foil perceived per cm2/ of total of total

(eV) sec. flux rep.
Au 2.5x10’ 4.2x108 54.5 43.2 2.9

4 8
Pu 2.5x10 1.9x10 24.7 280.8 19.0
Np 7.5x105 1.1x108 14.3 705.6 47.8

6 7
U 1.5x10 3.5x10 4.5 252.0 17.1

6 - 7
S 2.5x10 1.3x10 1.7 147.6 10.0

6 6
A1 8.6x10 2.4x10 0.3 46.8 3.2
Totals 7.7x108 (100.0) 1476.9 (100.0)

 

15

16

Field Layouts and Statistical Approach

 

All the lines were grown in the field in 1961, in single
row plots 10 feet long and 32 inches apart. Seedlings
averaged 4 inches apart in the row. This spacing, was
adequate to allow survival of individuals of differential
vigor. To minimize accidental outcrossings, the control
set was planted 10 days before the irradiated series. The
two treatment groups were separated in the nursery by a
25 yard wide strip of soybeans. Each row was harvested in
bulk. Two 12 plots were missing in the control group and
one in the radiation group. The other lines did not suffer
any loss.

In 1962, each plot included 2 replications per treatment.
Isolation in space only was maintained. Each replication
consisted of single rows, 20 feet long and 32 inches apart.
To reduce plant competition, the rows were thinned to 30
plants each after seedling emergence. At harvest, 5 plants
were randomly chosen per replication to compose family
sets of 10 sublines per treatment for each original line.
Each plant subline had to have enough well filled pods to
guarantee a minimum of 40 seeds for each sample. For
each parent, however, 15 one-plant samples were retained
per line to provide for more check rows in the following
year's nursery. So, assuming no missing plots, the

material harvested would consist of:

17

(1) 150 parental samples (P) in both treatment

groups
(2) 450 I1 sublines in both treatment groups
(3) 1760 12 sublines in the control group

(4) 1770 I2 sublines in the radiation group.

Relatively few plots were missing or did not yield enough
seed to meet the requirements for sample selection. Each
randomly-selected plant sample was separately threshed, and
its yield (W) and yield components (X,Y,Z) measured. Mean
seed weight (Z) was calculated on the basis of 100 seeds
per sample. Number of pods per plant (X) was obtained by
counting pods with at least one viable seed. Average number
of seeds per pod (Y) was derived from the equation
w = X-YoZ/100. The samples were stored in a warehouse
at 40°F to await the next planting season.

In 1963, the field layout consisted of single-row
plots, 10 feet long and 32 inches apart. The seed was
sown about 3 inches apart in the row as in regular planting.
Control and radiation blocks alternated throughout the
nursery in a systematic fashion. In spite of slight
eptam damage in some Spots where herbicide concentration
was perhaps critical, stands were satisfactorily uniform.

Growth habit, maturity dates (starting from seeding date),

18

and plant appearance were recorded. A few isolated off-
type plants were set aside for further study.

A uniform 4-foot section containing about 15 plants
was harvested from each plot. This method was thought to
adjust for differential competition resulting from
occasional missing plants. Due to the large number of
samples that had to be handled, some modifications in
measuring yield and its components appeared necessary.
Total yield (W) was measured on a plot basis. Average
number of seeds per pod (Y) was arrived at by taking
20 random pods in each plot and counting the seed. The
procedure to get mean seed weight (Z) was not changed.
Number of pods per plant (X) was obtained on a plot basis,
by using the same equation W = X.Y.Z/100. Distribution,
means, ranges, correlations, variances and coefficients
of variation of the measured variables were tabulated
so as to compare the effectiveness of the:

(l) hybridization phase alone

(2) neutron irradiation treatment alone

(3) combination neutron-hybridization.

The simple correlations and analysis of variance routines
used were programmed by the University's Computer Laboratory

staff (3,4).

19

The alternate arrangement of the control and radiation
plots plus other precautions to insure high environmental
uniformity in the field aimed at minimizing experimental
error. The control parent (CP) lines being sufficiently
homozygous and homogeneous, no significant genotypic
differences were expected within them. Thus the error
variance value found for the control—parent treatment
could be taken as a measure of the effect of environment on
all the genotypes. A significant phenotypic difference
between this value and that of another treatment could be
ascribed primarily to the combined effect of genotype,
treatment, environment, and their interactions. Similarly,
any two treatments might be compared and appropriate
inferences drawn. Then, for sake of convenience the error

variances can be expressed as:

l. Control-parent (CP): 6’e2
2. Radiation-parent (RP): 6’e2 +6'RP
3. Controlulst cycle intercrossing (C11):6'e2 +6’R12

4. Radiation-lst cycle intercrossing

. 2
(R11). (a “($12

5. Control-2nd cycle intercrossing 2
(CI ): o’e +{EI
2 2

6. Radiation-2nd cycle intercrossing 2

(R12): o’e +4412

RESULTS

Treatment Effects on Correlations and Regressions

Correlated variation of two characters may be due to
similar actions on both characters by genes or chromosomes
on the one hand or by environmental influences on the
other (22). To evaluate the effects of treatments on the
correlations between yield (W) and its components (X,Y,Z)
and among the components themselves, the homogeneity chi-
square test, described by LeClerg st 31 (27), was used.
Table 2 summarizes the overall results.

Hybridization and radiation, singly or combined,
apparently did not provoke any appreciable changes in
correlation patterns. Three of the homogeneity chi-square
values look somewhat high and could be an indication of
real differences between treatment effects on the correlations
involved. To test this possibility, critical comparisons
of r-value for XY 6&2 = 12.50), YW 0K2 2 14.05) and YZ
(X2 = 15.10) are sorted out in Table I in the appendix
section. None of the t-values obtained exceeds the 1%
level of probability for significance. The individual
r-values in each correlation class can be considered
drawn from the same population and averaged.

Table 3, which sums up differences between mean

correlation coefficients, permits these generalizatiohs:

20

ea pa ucmuamacmam "..
enouuuouca eauhu cam-NH .nuOMUHODCw danku peHuaH “accuse n a new

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

oe.~ mm.a sm.~m no.0 se.m mm.HH a .Imuv .caeumuou
.mmooo .e><
eehmHo eeQNHo eeHmbo eeomoo eeHmNol eefiVMol mmv¢ 03HM>IH OOMH0><
He.m mo.ea e~.m oa.ma as.~ om.ma m
wuaocemmfiom
..mmH. ..sea. ..~ms. emo. ..mm~.u ..mom.u «was NH
..oaa. ..ooH. ..mes. ..mmH. ..-m.n ..~mm.u one HH Ame
soap
«mo. «as. ..mma. mao.n ..omm.u ..Hsm.u one a (manna
..emH. ..mHH. ..oom. ..m~H. ..mm~.u ..mmm.u sees NH
..oma. moo.: ..aom. ..Hma. Hm~.: ..~ee.n was HH
AUV
ems. mao.u ..sms. moo. ..vm~.n ..¢ma.n and a Houucou
an as 3x as Nx ex vooum asouo
COHNflHUh—HHOU emeQ u “aflgflwflue
eucesoeeou one 0coa< use AN.».xv
mucocomeou muH one ABV oaoww seem ceesuom mucowuwmmeou coausaouuou .N oHQwB

21

22

 

 

Table 3. Test of Significance of Differences
Between Mean Correlation Coefficients
Comparisonl Difference t—value2
XY - X2 .059 2.81"
XY — YZ .269 12.81"
XZ — YZ .210 10.00“
XW — 2w .917 43.67"
XW - Yw .948 45.14“I
ZW - Yw .031 1.48
XW — XY .715 34.05”

 

l. The corresponding

in Table 2

2. Def a 0Q,

" Significant at 1%

so t.01

mean r-values appear

= 2e576

23

(l) The three yield components (X,Y,Z) form strong

positive associations with seed yield (W).

(2) The phenotypic correlation XW is the strongest
of all. Statistically the average correlation

coefficients ZW and YW do not differ appreciably.

(3) A positive correlation YZ opposes the negative

ones: XY and X2, all three being significant.

In Table 4, however, the calculated values of the
regression of yield on each component, even without
elaborate statistical analysis, showed that the regression
values increase with intercrossing in both the control and
the radiation series.

These values are practically of the same magnitude for
both the first and second cycles of intercrossing within
each series, except for the control series in YW. The
non-significant values at CP and CI1 for YW contrast
sharply with the high value at C12.

Where it is not negligible, the regression value for
YW is overwhelmingly greater than the values for XW and ZW
at the same treatment level.

Radiation 225.25 increased the regression value Yw
over that of control, left XW unchanged and decreased ZW.

The combination hybridization-radiation increased the

regression value for Yw, decreased that of ZW, but left that

of XW practically unchanged.

aH um ucmuHMHceHm ...

nmouuueucw eauxu oanNH .nmouuue»:H oauao umHuHH .ycousm a m "G.

 

 

 

 

 

 

 

 

 

 

 

..mm.oa ma.m ..mo.me mN.HN ..HN.mmmN as.o ameH NH
..ee.NH Nm.m ..Nm.oH Nv.mH ..es.mNm mo.o oHv HH
:3
mo.o ma.o em.N mm.eH ..ON.HmH No.0 mmH a :oHsaHesm
..eN.oe Ho.e ..oe.mN mm.mH ..mN.eNmN ma.o esmH NH
..em.oH mm.N mo.o mH.Hu ..Ho.Nms oa.o va HH
ADV
am.e mm.m No.0 mm.Ha ..mm.HmH me.o emH a Houucoo
osz>1m enao> n 03Hm>tm 03Hm> m. 63Hm>nm, 05Ho> a qwoeum macaw
gm 3» 3x .m.o
Gownoeumou usecHu uneavsona
oncesoeeou mufl so caddy seem mo coaeneumem .¢ canoe

24

25

Effects of Treatments on Variancej Means
and Coefficients of Variability

Variances

Comparisons of population variances appear in Table 5,
for X, Table 6 for Y, Table 7 for Z, and Table 8 for W.
“The inference of significance in these tables, as in the
subsequent ones, is again drawn at the 1% level of
probability.

Hybridization as such (11, 12), generated excess
variance over the parents, both within the control and
radiation series, except for Y. For Y, no significant
difference was established within the radiation set; in
the control group only the second cycle intercrossing
variance exceeded parental variance.

In general, the second and the first cycle inter-
crossing variances did not differ significantly for Y,
where I2 exceeded Il in the control group.

Radiation plus the first cycle of intercrossing
(R11) produced higher variances over the control plus the
first cycle of intercrossing (C11) in both Y and Z; their
differences were negligible in X and W.

Neutron irradiation alone (RP) did not differ from
the control-parent (CP), except for Y in which the neutron

treatment was associated with a higher variance. But,

26

 

 

 

Table 5. Comparisons of treatment error variance for number
of pods per plot (X)°
Treatment Eivalue of
Group Breeg Source of D.f.. M.sq. treatments treatments
variation comparison comparison
Control P Between lines 9~37861.20 C Il/P 2.13"
(C) Error 126 2882.75 12/9 1.88”
I1 Between lines 44 15459.78 11/12 1.13 L
Error 382 6130.23 R Il/P 1.93"
12 Between lines 175 14485.66 I2/P 1.78"
Error 1500 5428.25 11/12 1.09
Radiation P Between lines 9 26507.91 RP/CP 1.12
(R) Error 130 3235.99 RIl/CIl 1.02
I1 Between lines 44 10924.07 RIZ/CI2 1.06"
Error 373 6252.67
12 Between lines 176 12144.59
Error 1509 5753.52

 

o e

A .

. PuParents, I

l

": significant at 1%

. A harvested plot contained approximately 15 plants

=1st cycle intercross, I2=2nd cycle intercross

27

 

 

 

Table 6. Comparisons of treatment error variances for mean
seed number per pod (Y)

F-value of
orgiiatmefifieeaé fississisfi D.F. M.sq. §§§§§§i§§§ gggggggggg
Control P Between lines 9 1.37 C Il/P 1.18

(C) Error 126 0.17 Iz/P 1.41u
Il Between lines 44 0.73 I2/I1 1.20"
Error 382 0.20 R Il/P 1.15
12 Between lines 175 0.53 I2/P 1.15
Error 1500 0.24 IZ/Il 1.00
Radiation P Between lines 9 1.17 RP/CP 1.59
(R) Error 130 0.27 RIl/CI1 1.55"
Il Between lines 44 0.69 RIz/CI2 1.29"
Error 373 0.31
12 Between lines 176 0.52
Error 1509 0.31

 

A : P=Parent, I

1=1st cycle intercross, 12=2nd cycle intercross

": Significant at 1%

28

 

 

 

Table 7. Comparisons of treatment error variances for mean
seed weight (Z)
_:Treatment Source of ‘7:T?:atments E:value75f-
Group Breed variation D.f. M.sq. comparison treatments
comparison
Control P Between lines 9 35.35 C 11/? 1.41“
(C) Error 126 1.72 Iz/P 1.53u
Il Between lines 44 - 15.52 IZ/Il 1.09
Error 382 2.42 Il/P 2.04!‘
12 Between lines 175 12.44 Iz/P 1.94"
Error 1500 2.64 12/11 1.05
Radiation P Between lines 9 28.60 CP/RP 1.13
(R) Error 130 1.52 RIl/CIl 1.28u
Ii Between lines 44 15.90 RIz/CIZ 1.11::
Error 373 3.10
12 Between lines 176 12.24
Error 1509 2.95

 

£1:P=Parent, Il=lst cycle intercross, I2=2nd cycle intercross

": Significant at 1%

29

Table 8. Comparisons of treatment error variances for seed
yield per plot (W)°

 

 

 

groggeatmenggzgaA Source of Treatments F-value of
variation D.f. M.sq. comparison treatments
comparison
Control P Between lines 9 25225.12 C Il/P 2.55“
(C) Error 126 1960.28 12/2 2.60n
Il Between lines 44 7868.76 Iz/I1 1.02
Error 382 4995.96 R Il/P 1.79"
12 Between lines 175 9504.29 IZ/P l.88"
Error 1500 5099.49 IZ/Il 1.05
Radiation P Between lines 9 8940.27 RP/CP 1.45
(R) Error 130 2837.67 RIl/Cll 1.02
Il Between lines 44 5928.60 RIZ/CI2 1.05"
Error 373 5093.25
I2 Between lines 176 7963.73
Error 1509 5339.52

 

. A harvested plot contained approximately 15 plants
45 : P=Parent, I1=1st cycle intercross, I2=2nd cycle intercross
" : Significant at 1%

30

radiation plus the second cycle of intercrossing (RI2)
led in all cases to higher variances over the second

cycle of intercrossing alone (C12)’

rears.

Comparisons of treatment means are arranged in
Table 9 for X, Table 10 for Y, Table 11 for Z, and
Table 12 for W. Differences in magnitude are very limited
and generally not impressive.

Hybridization (11, 12) means did not exceed the parent
means (P), except for two cases. For Y, both I1 and I2
were greater than P in the control lot; for Z, only 12
exceeded P in the radiation set.

The second cycle intercrossing (12) and first cycle
intercrossing (11) means did not differ notably, except
for Z, where 12 exceeded both the control and the radiation
sets.

Radiation alone (RP) or combined with hybridization

(RI R12) produced means not statistically different from

1’
the control counterparts (CP, C11, C12) except for Y.

There the C12 mean was slightly greater than RIZ'

Coefficients of Variation
Comparisons of coefficients of variation appear in
Table 13 for X, Table 14 for Y, Table 15 for Z, and

Table 16 for W.

umouuueucw sanku ocmumH .enouuueusfi manhu uoHuHH .ucouwmum u Q

upceaa ma waoumEonuaee pesasucou road oeumo>umc 4 a a

 

 

 

 

NN.o Ho.o Hon Hr
oe.o mN.N HHmnHHu
om.o oo.m ousom
No.0 No.0 NHuHH MH.om me.mHm omoH NH
NN.o mm.H NHuo MH.Nm Nm.mHm mHv HH
oH.o Hm.H HHuo Nm.mo mm.mHm ooH o costwwor
mo.o oo.N NHIHH am.oo em.on oNeH NH
so.o mo.m ouNH NN.er om.on NNv HH
Hm.o Ho.o ouHH HN.NN oN.eHm oMH m Howwwoo
use: ecowps> oeeum macaw
coauoa>en (Henna d
osao>lu eucoHoMMHo QOEHHEQEOU unaccoum Henfinz “ceEudoua
Haxv “can use upon me
Hones: Hom asses useEuseHu seezuen neuceuemmav mo enemaudasou .m magma

31

mmouuuopca oau>u ocNuNH .mmouuuoeca oau>u Dean

aH so orooHHaronn..

. ucwummumu Q

 

 

 

 

9900.0 NHoO NHMINHU
He.N oo.o HHmnHHo
No.0 mo.o omuoo
. . N IH . . N
me o No 0 H H em 0 mN m omoH H
oo.H mo.o ouNH om.o Hm.m mHe HH
Amv
oN.H No.0 onHH sm.o «N.m oeH o roHuoHoom
mm.o Ho.o HHINH Nm.o He.m oNoH NH
..mH.m eH.o onNH Hm.o oe.m NNo HH
HOV
..mo.N mH.o onHH om.o NN.m omH o Hourroo
are: mcoanm> «wooum macho
coaumw>ep luomno cos won
oDHm>I# oucouomMNo somHHdeou choocmum Honesz u u e
Awe pom Hem Hones: omen
cmmE How enema pamEumme c003umn noucmnommap mo chmHHMQEOU .oH wanme

32

nH no uroUHchon 1..

 

 

 

 

mmouuumpcw eHuwu oanNH .mmowuueucH macho umHnHH .ucoumdnm " Q
me.H oH.o NHuuNHm
oN.H mN.o HHouHHr
HH.o oo.o omuoo
e e H IN 0 e N
..mm m mm o H H mm H Ne oH omoH H
..mo.m om.o ouNH HH.N eo.oH mHe HH
:3
«H.H HN.o ouHH a Hm.H mm.mH 04H m roHuoHoor
..mo.m mm.o HHuNH Ho.H Nm.oH mNoH NH
em.N ma.o ouNH so.H oH.mH NNv HH
ADV
Ho.o mo.o HHuo o mo.H No.mH omH a Hourroo
coo: mCOHuo>
cowuma>oo (Honno epoeHm asouw
osam>lp oucmuommwo COmHHMQEOU oumocmum Honesz ucoeumoue
ANV pnmao3 one» some
How ucmoE ucoEymoHp coozpon meucouemmao mo macawuoaeou .HH manme

33

nH pm rcooHHHron ...

mmOHUHdeH eaumu chan .mmouuueucH oau>u umHuHH .ucoumaua " s

mvcmaa ma >HonEonudam oocamucou #oHd poeme>umr < n H

 

 

 

oH.N oo.m HmuNHo
oa.o Na.m HmuHHo
Nm.o oN.N sonar
0 e HIIN e o N
we H or m H H No as oo HNm omoH H
Nm.N mm.HH ouNH mo.HN mH.on mHe HH
Amv
Ho.H No.o onHH om.om HH.0Hm 06H m roHooHoom
(mo.H MH.N HHuNH om.as mo.NNm oHoH NH
..me.m mH.oH oaNH mN.Na oo.on NNo HH
AUV
mo.H mo.NH ouHH oN.om mm.som omH o Houuooo
coo: mEOHpo> ooeum QSOHO
coapma>op (Henge Q m
03Hs>lw mucoHoMMwb COmHHMQEOU onmocmum HoQEDZ uses» 0H9

 

HABV uoHd Hoe mama» room

How mcmoE “cprmeuu cmezuen eeUCeHOMMHU mo mCOmHHMQEOU .NH oHQmB

34

nH so rcoUHMHrmHm

ee
O
Ir

 

 

35

 

mmouuuoucw maumu UCNuNH .mmouuuoucH mau>u umHsHH .uceummnm " q
vmma .cmo .H .02 .mm .Ho> .HDOU .m< Eoum szmno .>.u mo COHHMEHOMmcmHH " .
encode ma >HowMEHxOHQQm confimucou uoHQ topmo>umn 4 u H
mH.o H.o NHouNHa
Hm.o m.o HHm1HHo
Hh.o «EH mmlmu
Hm.o o.o NHuHH N.mN omoH NH
..Ho.N N.m ouNH m.mN mHv HH
a Ame
oomw.m m.v on H m m.HN ova m :oHHoHomm
mH.H N.H NHnHH H.mN oNoH NH
ee.H N.N onNH m.oN HNo HH
H Auv
emoa v.m an H U m.mm mma m Houpcou
ozam>lu oucouommao COmHHmeeou o.>.u mcoaum> doeeum msonw
(Homno .02 I, DcoEumeue

 

 

HAxV roam Hem moon mo HoQESG How COHHEHHM> mo mucoduammoou .mH magma

«mma

mmOHUHoDCH oHumu UCNuNH .mmouuuopcw sawhu umHuHH .uconmum

.cmw .H .02 .mm .Ho> .Hsoo .m< Eoum c3muo .>.U mo COHumeuommcmue

nH so uroUHHHcon

.0

 

 

 

 

eeom.v Noa NHUINHm
..ov.m H.H HHouHHa
om.H o.H oouom
os.o m.o NHuHH m.OH omoH NH
«H.o H.o NHso H.HH mHo HH
H Ame
oN.o N.o o: H m o.0H oeH o coHuoHoom
om.o Noo HHuNH o.m oeoH NH
HH.o H.o ouNH e.o HNv HH
H 3c
mH.o H.o Him 0 m.m omH o Houocoo
03Hm>lu oucmHeMMHo COmHHdeou o.>.U mCOHum> quooum msouo
IHemno .oz p:eE»moHB
AWV UOQ HWQ HWQEDC Ummm CMQE HOW COH¥MfiQM> W0 MHCQHUHmeOU eflH wHQMB

36

mmouuuopca 0H0>U UCNuNH .mmOHUHoHEH macho umHuHH .ycoummum " q

vmmH .cmh .H .oz .mm .Ho> .HDOh .04 Eoum czmuo .>.U mo COHumeuommcmHs " o

 

 

 

 

 

0N.H m.o Ho: Hr
em.H m.o HHonHHm
mo.H o.o oanoo
mo.N o.o NHaHH N.0H omoH NH
mo.o o.o ouNH H.HH mHe HH
H :3
«H.N m.H on H a o.o ooH o roHooHoor
oo.H 6.0 NHnHH o.m oHoH NH
oo.o moo NHuo H.0H NNv HH
ADV
HN.o N.o HHIN o m.oH omH o Houucoo
peeHm mmOHw
osam>lu oucmuwmmao GOMHHMQEOU o.>.U mCOHuo> q
IHoon .oz ucefiumoue
ANV usmawz comm some now COHDMHHM> mo mucmfiuammoou .mH magma

37

nH no scouHHHron

.0
Q
.

 

 

 

 

 

mmOHUHmHCH QHU>U pawn H .mnOHUHOHCH eHU>U umHuHH .Hcoummud H G
vmmH .cmo .H .02 .mm .Ho> .HDOh .m< Eoum czmuo .>.U mo GOHHMEHOMmcoHB " o
nucmHa mH >HopoEHx0HQam Dechucou uoHa ooumo>uon < a H
mm.o m.o NHouNHr
mo.o H.o HHonHHm
e e 8
om o o H amino 3
om.o m.o HHINH H.HN omoH NH
..mH.e o.m ouNH m.NN mHo HH
H :3
.oHN.m m.v m1 H m m.mH 06H m EOHuoHomm
--HH.o H.o HHnNH o.NN oNoH NH
..ms.N m.m onNH N.NN HNo HH
H loo
mm.m v.m an H U m.mH mmH m HOHHEOU
oDHm>Iu oucoHoMMHo COmHHMQEOU .>.U uCOHpo> quooum msouw
(Henge .oz psoEuooHB
HA3V HOHQ Hod UH0H> poem How COHumHHm> mo mHGOHUHmmoOU .mH oHnt

39

Hybridization as such (11, 12) did not change the
coefficients much. In the radiation group, both second
cycle intercrossing (12) and first cycle intercrossing
(I1) were superior to parent (P), for X and W. In the
control set, only 12 markedly surpassed P for W.

The second cycle intercrossing 225.33 (C12) and the
first cycle intercrossing alone (C11) exerted no significanti
differential effects on the coefficient of variation.

Similarly, radiation alone (RP) or combined with
hybridization (R11, R12) showed no significant differences
from the control counterparts (CP, C11, C12), except for Y.

In the case of Y, RI and R12 surpassed C11 and C12,

1
respectively.

Frequency distributions of the original data are shown
in Table II for X, Table III for Y, Table IV for Z, and

Table V for W, in the appendix section.

DISCUSSION

To analyze the effects of the treatments on yield and
its components, a brief review of pertinent reports on the
relationships among these fitness factors might be
appropriate. Luedders (29) concluded that the yield
components in oats were governed by separate gene systems.
Archibong (5) noted that number of pods per plant (X)
invariably showed strong positive correlations with plant
yield (W) in navy beans. Camacho and his associates (7)
suggested that yield improvement in kidney beans could
be achieved by selecting progenies with a high number of
beans per pod (Y) because of the greater genotypic Yw
correlation value obtained from their data.

Biologically, if not statistically, the correlation
patterns emerging from the present data do not preclude
similar generalizations. Whereas yield is strongly correlated
with number of pods in every case, it regresses more heavily
on number of seeds per pod in most instances. In spite of
statistically significant correlations among them, the
components still can be genetically unrelated one to another
or related to a common yet unknown factor. From a develop-
mental standpoint with limited metabolic input, a high
number of pods is likely to be associated with a low number

of seeds per pod, a low mean seed weight (2), or both.

40

41

The failure of the treatments to bring about any
substantial changes in the correlation patterns could
indicate either an inefficacy of the intercrossing cycles
in these early stages or an inadequacy of the radiation
dosage, or both. On the contrary, the rigidity of a
deve10pmental scheme for yield might be such as to upset
any variation in the components brought about by the
treatments and maintain a status 322 in the correlation
patterns.

From the comparison of treatment effects on the
regression coefficients of yield on its components,

hybridization per‘gg (CI C12) in general seems to have

1:
been more potent than the radiation treatment alone (RP).
Yet, when intercrossing is combined with radiation some
change developed: while the regression values XW stays
roughly the same, YW increases and ZW decreases. Such a
trend would imply that with a combination of the two
procedures, yield (W) would depend more on average seed
number per pod (Y) than on either number of pods per plant
(X) or mean seed weight (2). If so, yield improvement by
manipulation of the components would be greatly facilitated
as X is most subject to ecological influences and Z must
conform to commercial standards.

Granted a few departures, which are normal for metric

characters, the relative efficacy of the treatments in

42

modifying the regression coefficients seems to hold for
changes in variances. Hybridization alone very effectively
increased variances, but the combination hybridization
and radiation led in this reSpect. Irradiation alone,
however, did not differ from the control treatment, likely
because of the presumed inadequacy of the dose applied.
Moreover, the similarity of effects of the first and the
second cycles of intercrossing in all but one case, supports
the argument that these early stages may not produce
spectacular differential effects, especially on quantitative
traits. The significant differences in variances for Y
that correspond to non-significant differences in X and
Z indicate that Y was the most uniform component.
Generally, irradiation is expected to reduce mean
fitness, due to lowered fertility. The failure of the
neutron treatment to produce yield and yield component
mean statistically distinct from those in the control
groups, except in one case, supports the speculation that
the neutron dosage might not have been high enough. Within
the hybrid groups, certain families with high and low
yield and yield component means may have cancelled out
so as to become not different from the control populations.
The similarity between the first and the second cycles of
intercrossing in their effects on fitness means in most

of the cases, again conforms with speculations.

43

The effects of treatments on variances and means show
good agreement with earlier reports. Mitchell (31), from

crosses in DrOSOphila, found that hybridization increased

 

variance but brought no significant changes in means.
Irradiation with or without hybridization increased variances
in peanuts but reduced means, according to Gregory (17),

who further pointed out that the large variances were often
associated with the smaller means.

The coefficient of variability (C.V.) would tend to
increase as means decrease or as variances increase. The

higher this index, the greater may be opportunities for
selection, if the bulk of variation is not solely due to
environmental fluctuations. Where differences in C.V. are
noticeable, hybridization alone or coupled with radiation -
because of associated increases in variances and little
variation in means - raised the percentage values. The
neutron treatment alone having produced no major changes in
means and variances naturally did not differ from the

control.

CONCLUSIONS

In general, the results obtained in the present study
have provided support for the idea that a combination of
radiation breeding and hybridization would be more effective
than either method alone.

Hybridization combined with neutron irradiation did
increase both variances and percent of variability more
than the other treatments. In this respect, hybridization
flalone rated better than the radiation treatment 225.53.

When some significant changes occurred, hybridization was
found to increase and irradiation to decrease means, as
normally expected. Only a few such changes were actually
recorded because presumably the low and high family means
within a line balanced out or the neutron dosage was too
low to produce drastic effects.

Hybridization alone or with radiation increased
regression values of yield on its components more consistently
than did neutron irradiation alone. Combined with irradiation,
hybridization made yield variation more dependent on number
of seeds per pod (Y) than did the other treatments. Because
of its greater stability over both X and Z and its freedom
from market specifications, Y offers greater convenience for
yield improvement through manipulation by selection.

The second cycle of intercrossing, alone or with

radiation, excelled the first cycle in increasing mean seed

44

45

weight (Z) means, and, without radiation, in increasing
variances of number of seed per pod (Y). Their effects
otherwise were similar.

Statistically significant negative correlations seem
to link number of pods (X) to either number of seeds per
pod (Y) or mean seed weight (Z). Even if developmentally
correlated, the three components may still be governed
by independent gene systems, or simply related to a
common yet unknown factor. The treatments applied did not
bring any appreciable changes in correlation patterns.

Claims for disruption of linkage groups may be hard to
prove due to unavailability of specified marker genes.

But genetic recombination at least has been achieved,

as expressed by increased variances and higher coefficients
of variability. Lines that transgress the original parents
in yield and component means have been isolated and set

aside for selection purposes.

6.

7.

LITERATURE CITED

Adams, M. W.: Report proposal to A.E.C. Unpublished,
1963.

and R. Duarte: The nature of heterosis for

 

a complex trait in a field bean cross. Crop Sci.
1:380, 1961.

ABS Program Description 2: Use of the CORE routine
to calculate simple correlations, means, standard
deviations, moments, variances, and covariances of
variables and sums of squares and cross products on
the CDC 3600. Written and programmed by D.F. Kiel
and W. L. Ruble. Michigan State University, Ag.
Exp. Sta. 1963.

Description 3: Use of analysis of variance

 

routines on the CDC 3600. Written by A.L. Kenworthy.
Programmed by D.F. Kiel and W.L. Ruble. Michigan
State University, Ag. Exp. Ste. 1963.

Archibong, D.: On the correlations among yield components
in the navy beans as influenced by spacing and inter—
strain competition. M.S. thesis, Mich. State Univ.,
1963.

Bacq, Z.M. and P. Alexander: Fundamentals of Radio-
biology. Pergamon Press, New York, 2nd ed. 1961.

Camacho, L.H. 33 El: Genotypic and phenotypic corre-
lation of components of yield in kidney beans. Ann.

Report. Bean Impr. Coop. Oregon State Univ. 1964.

46

10.

ll.

12.

l3.

14.

15.

16.

47

Cohn, N.S.: An analysis of the rejoining of X—ray
induced broken ends of chromosomes in the root tips

of Allium cepa. Genetics 43:362-373, 1958.

 

Duarte, R.A. and M.W. Adams: Component interaction in
relation to mean expression of complex traits in a
field bean cross. Crop Sci., 3:185-186, 1963.

Dobzhansky, T. gt El: Release of Genetic variability
through recombination. III. Drosphila prosaltans.
Genetics 44575-92, 1959.

Donnelly, E.D. and E.M. Clark: Hybridization in the
genus Kigig. Crop Sci. 2:141-145, 1962.

Down, E.B. and A.L. Andersen: Agronomic use of an
X-ray induced mutant. Quart. Bul. Mich. Ag. Exp. Sta.
39:378. 1956.

Elliott, F.C.: Personal communication

2 Plant Breeding and Cytogenetics. McGraw-Hill

 

Book Company, Inc., New York, 1958.

Evans, H.J. 23.2l3 The relative biological efficiency
of single doses of fast neutrons and gamma rays on
XES$2.£ERE roots and the effects of oxygen. II.
Chromosome damage: the production of micronuclei.
Int. Jour. Rad. Biol. 1:216-229, 1959

Goospeed, T.H. and A.R. Olson: The production of

variation in Nicotiana Species by X-ray treatment

 

of sex cells. Proc. Nat. Acad. Sci. U.S.A.

14:66-69, 1928.

17.

18.

190

20.

21.

22.

23.

24.

25.

48

Gregory, W.C.: The efficagngf mutation breeding in

 

Symposium on Mutation and Plant Breeding. Nat. Acad.
Sci. - Nat. Res. Council. 891:461-486, 1961.
Hagedoorn, A.L.: Plant Breeding. Crosby Lockwood &
Sons Ltd. 1950.
Hanson, W.D.: The theoretical distribution of lengths
of parental gene blocks in the gametes of an F1
individual. Genetics 44:197-209, 1959.

: The break-up of initial linkage blocks

 

under selected mating systems. Genetics 44:857-868, 1959.

: Average recombination per chromosome.

 

Genetics 47:407-416, 1962.

and H.F. Robinson: Statistical Genetics and

 

Plant Breeding (p. 41-42), Nat. Acad. Sci. - Nat.
Res. Council 982, 1963.

Harlan, J.R.: Distribution and utilization of natural
variability in cultivated plants. Brookh. Symp.
Biol. 9:191-208, 1956.

Heinz, D.J.: Improvement of grasses through induced
chromosomal recombinations. Ph.D. thesis, Dept. of
Crop Sci. Mich. State Univ., 1961.

Hvostova, V.V. and L.V. Nevzgodina: Cytological
analysis of the causes of radio resistance in plants.

Radiobiol. (Russian) 12611-618, 1961.

27.

28.

29.

30.

31.

32.

33.

34.

35.

49

Kulik, M.I.: Experimentally induced mutations in
tomatoes. Proc. Acad. Sc. USSR 38:211—214, 1961.

LeClerg, E.L., W.H. Leonard, and A.G. Clark: Field Plot
Technique. Burgess Publishing Company, Minn. 2nd
ed. 1962.

Lerner, I.M.: The Genetic Basis of Selection. John
Wiley & Sons, Inc. New York, 1958.

Luedders, V.D.: An analysis of the components of yield
in 18 oat crosses. M.S. Thesis. Mich. State Univ. 1960.

Milkman, R.: Potential genetic variability of wild

pairs of DrOSOphila melanogaster, Science 131:225-226,

 

1963.

Mitchell, 0.: Inversion heterozygosity, recombination
and variability of quantitative traits. Cold Spring
Harb. Symp. 23:279—290, 1958.

Mitre, 5.: Effects of X-rays on Chromosomes of Lilium

longiflorum during meiosis. Genetics 43:771-789, 1958.

 

Muller, H.J.: Artificial transmutation of the gene.

Science 66:84-87, 1927.

Nilan, R.A.: The Cytology and Cytogenetics of Barley.
Monog. Suppl. No. 3. Res. Stud. Washington University
Press, 1964.

Osborne, T.S. Letter to M.W. Adams. Unpublished, 1961.

37.

38.

39.

40.

41.

42.

43.

50

, and A.O. Lunden: The cooperative plant and

 

seed irradiation program of the University of Tennessee.
Int. Jour. App. Rad. Isot. 10:198-209, 1961.

Schacht, L.E.: The time of X-ray induction of cross-
overs and translocations in Drosophila melanogaster
males. Genetics 43:665-678, 1958.

Scossiroli, R.E.: Effectiveness of artificial selection
under irradiation of plateaued pepulations of Drosoghila
melanogaster. Union Int. Sci. Biol. Series B.

15:42-66, 1954.
Sparrow, A.H.: .IIE2§.2£ ionizing radiations 229.222i5

cytogenetic effects. Symp. Mutation and Plant Breed.

 

Nat. Acad. Sci. NRC. Pub. 891:55-119, 1961.

, J.P. Binnington, and V. Pond: Bibliography

 

on the effects of ionizing radiations on plants
(1896-1955). AEC Brookh. Nat. Lab. N.Y., 1958.

Spassky, B. 23 31: Release of genetic variability
through recombination. I. Drosophila pseudoobscura.
Genetics 43:844-867, 1958. -”

Spies, E.B.: Release of genetic variability through
recombination. II. Drosophila persimilis. Genetics
44:43-58, 1959.

and A.C. Allen: Release of genetic variability

 

through recombination. VII. Second and third chromosomes

of Drosophila melanogaster. Genetics 46:1531-1553, 1961.

44.

45.

46.

47.

48.

51

Stadler, L.J.: Genetic effects of X-rays in Maize.
Proc. Nat. Acad. Sci. U.S.A. 14:69-75, 1928.

: Mutations in barley induced by X-rays and

 

radium. Science 68:186-187, 1928.
Stephens, S.G.: Recombination between supposedly
homologous chromosomes of Gossypium barbadense‘g

and G. hirsutum L. Genetics 46:1483-1500, 1961.

 

VanDijk, G.E.: Crosses between early and late heading
timothy plants. Euphytica 10:43-48, 1961.

Whittinghill, M. and D.G. Davis: Increased recombination

 

from female Drosophila irradiated as larvae without

oocytes. Genetics 46:357-360, 1961.

APPENDIX

®FmON “‘04Hn 00" “Mom ON

muouuueucw eauxu ocNaNH .nmouuueucH mauxu umHuHH .ucoumo n m 4
N sandy Eouw c3muo sum nycowoawmoou coapoaouuou one .H

 

 

 

 

 

 

 

 

 

 

53

 

 

 

 

 

so.H ssH. no a NHm
. . H u N . N
No H moo H H NHN u H
os.o who. N.u NH Nov.u HH Hes
«H.o yHHo. Hm I o m mmm.u my ow :OHosHmmm
oN.N NNH. HH u NH mmm.u NH
Nm.H .mH. o s NH m.¢.n HH “no
mH.o mHo. o u HH 0 om..- o Houucou Nx
No.N NoH. no 3 NH«
. . H u N . N
mH o ooo H H ooH H
eN.o eNo. o 1 NH HoH. HH Ame
ma.o meJ m I HH m mud. n}, doauowmmw
mm.N sNH. HH u NH oHH.. NH
sv.H NNH. o u NH ooo.n HH Auo
mo.o moo. o a HH o mHo.u Hwy Homwtou cam
oo.N NoN. me u HHu
e O N .l H O N
oH N oNH H H «mo H
em.o Nmo. o u NH va. HH Hes
ab.H HRH. a u HH e mHo.u N commaHomm,
HH.H ooo. NH 1 HH NNH. NH
mm.H oNH. o u NH mmH. HH Hus
No.H omH. N n HH o moo. m, Houunou N»
ooae>ln euseuemmao GOnHHoQEOU 03Hd> N.xouma< queeum ozone doaumaouuou
N uceeueexm
wa one 3» .NN esoauoaeuuou on» new noseHUHmmeou
.coavoaouuou c003uen noUCoHeMMHU mo euceUHmacoHe mo uneH .H canoe

54

Table II. Frequency distribution of number of pods per plot (X)

 

Treatment1

 

 

 

 

 

Class “
interval -c- P -» Il~ . - 12
Control Radiation Control Radiation Control Radiation
80—129 1 l 2 3
130-179 4 3 9 10 33 27
180-229 8 10 32 40 153 174
230-279 31 25 102 98 374 353
280-329 39 40 109 104 438 457
330-379 33 37 85 71 348 335
380-429 13 17 43 51 194 194
430—479 4 5 25 26 80 89
480—529 4 3 11 12 30 29
530-579 7 4 15 15
580-629 2 1 4 7
630-679 2 1
680—729 2 1
730-779 0 0
780-829 0 0
830-879 1 1
Totals 136 140 427 418 1676 1686
Range?
Low 144.00 157.50 136.20 138.60 149.82 143.59
High 561.50 482.50 605.60 568.20 615.06 604.23

 

. A harvested plot contained approximately 15 plants

1 : P=Parent I1=lst cycle intercross, IZ=2nd cycle intercross
2 : 2% of the observations in each treatment were used to
evaluate the ranges: 1% for the lower limit
1% for the upper limit

55

 

 

 

 

 

 

.ble III. Frequency distribution of mean number of seeds per pod (Y)

Class TreatmentI

interval P . . 11 12

Control Radiation Control Radiation Control Radiation

2.35—2.59 1 1
2.60-2.84 0 O
2.85-3.09 1 O l
3.10-3.34 3 0 2
3.35-3.59 3 2 4
3.60-3.84 l l 1 4 4
3.85-4.09 l 4 1 3 7 19
4.10-4.34 4 6 4 12 30 55
4.35-4.59 4 9 14 20 54 98
4.60-4.84 16 14 36 35 115 177
4.85-5.09 21 22 54 54 190 216
5.10-5.34 25 19 78 68 289 283
5.35-5.59 20 28 90 80 344 300
5.60—5.84 29 15 68 59 275 240
5.85-6.09 9 14 41 48 218 148
6.10-6.34 5 5 23 20 90 84
6.35-6.59 1 4 11 6 42 43
6.60-6.84 3 5 12 10
6.85-7.09 3 3 1
Totals 136 140 427 418 1676 1686
Rangez:

Low 3.90 3.97 4.07 3.21 3.78 3.52
High 6.32 6.45 6.78 6.60 6.73 6.64

 

: P=Parent I =1st cycle intercross, 12=2nd cycle intercross

l
2% of the observations in each treatment were used to
evaluate the ranges: 1% for the lower limit
1% for the upper limit

56

m
J.

ble IV. Frequency distribution of mean seed weight (2)

(11

 

Treatmentl
P IL 12
Control Radiation Control Radiation Control Radiation

Class
interval

 

 

11.0-11.8 1
12.0-12.8

13.0-13.9 1 0
14.0-14.9 1 3 3 13 ll
15.0-15.9 6 3 21 16 34 42
16.0-16.9 20 13 55 50 108 112
17.0-17.9 24 27 71 66 246 225
18.0-18.9 26 33 95 75 320 322
19.0-19.9 24 25 65 - *81 374 338
20.0-20.9 12 ' 16 51 46 270 281
21.0-21.9 11 13 40 42 163 193
22.0-22.9 9 4 17 24 84 81
23.0-23.9 4 2 7 34 50
24.0-24.9 1 2 5 21 18
25.0-25.9 5 5
26.0-26.9

27.0-27.9

Totals 136 140 427 418 1676 1686
Rangez:

Low 15.40 15.05 14.78 14.50 14.25 14.76
High 23.65 23.80 23.80 25.68 25.09 25.55

 

1 : P=Parent Ilzlst cycle intercross, IZ=2nd cycle intercross

2% of the observations in each treatment were used to
evaluate the ranges: 1% for the lower limit
1% for the upper limit

00

57

Table V. Frequency distribution of seed weight per plot (W)°

 

 

 

 

 

 

Class P Trea‘Icment1 I

interval -1 2
Control Radiation Control Radiation Control Radiation

80-129 3 l l 4
130-179 3 2 1 9 18 23
180-229 11 10 33 41 115 132
230-279 27 29 95 74 308 338
280-329 46 46 111 112 454 463
330-379 37 38 106 103 415 374
380-429 9 13 50 52 227 217
430-479 2 2 17 20 90 86
480-529 1 7 4 3O 34
430-479 4 1 12 13
580-629 2 1
630—679 3 0
680-729 1 1
Totals 136 140 427 418 1676 1686
RangeZ:

Low 145.50 156.00 141.40 134.80 154.70 140.88
High 471.00 459.00 552.20 514.40 582.06 560.94

 

0 e

. A harvested plot contained approximately 15 plants

1 : P=Parent Ilzlst cycle intercross, 12=2nd cycle intercross
2 : 2% of the observations in each treatment were used to
evaluate the ranges: 1% for the lower limit
1% for the upper limit