THE INHERITANCE AND MECHANISM OF
RESISTANCE 0F BARLEY T0 CEREAL ,
LEAF BEETLE, OULEMA MELANOPUS L. ’

Thesis for the Degree of Ph. D. -
MICHIGAN STATE UNIVERSITY
A CHUNG LEE
1 9‘7 G

   
 

  

THE—5“:

L I B R A R Y
Michigan State
University

    

This is to certifg that the

thesis entitled

I

I

The 1nher1tance and mechanism of resistance of barley ‘
I

to cereal leaf beetle, Oulema melanopus L.

presented by

Chung Lee

has been accepted towards fulfillment
of the requirements for

1911.13. degree in Crop Science

€14 ggzjgm
’/ n / /) . r

Date

0-169

 

 

 

 

Wéw

 

 

ABSTRACT

THE INHERITANCE AND MECHANISM OF RESISTANCE OF BARLEY

TO CEREAL LEAF BEETLE, OULEMA MELANOPUS L.

 

By

Chung Lee

The mechanism of host plant resistance to the cereal

leaf beetle in wheat is related to pubescence and a high de-

 

gree of resistance has been obtained. Only a moderate degree
of resistance has been observed in barley. The mechanism of
resistance is not yet known although resistance has been
found to be controlled by recessive genes.

The present study aims at understanding the mechanism
and the inheritance of host plant resistance in barley and
finding a higher degree of resistance. Resistance can be
regarded as a complex trait and partitioned into three com-
ponents -- ovipositional preference by the adult female,
antibiosis and recovery of the plant.

A diallel cross series and the progenies of the cross
between CI 6671 and CT 6469, the two varieties with the
highest degree of resistance now available in barley, were
used for the tests. A discriminant function was employed to

combine the three components into a single trait and the

Chung Lee

function was converted into a nomographic chart for easy and
efficient use.

Ovipositional preference shows a low heritability
with a pattern of ambidominance. Plants at the heading stage
were much less preferred than at the seedling stage.

Resistance to larval feeding at the seedling stage be-
haves as a recessive trait although at an older stage, this
trait is controlled by another genetic system which appears
to be ambidominant. The age of tissue is not responsible
for the differential pattern of resistance.

Larval weight gain is a reliable measure of antibiosis°

To obtain a higher degree of resistance, more emphasis
should be placed on the larval feeding response at the mature

plant stage than the ovipositional preference or recovery of

the planto

 

THE INHERITANCE AND MECHANISM OF RESISTANCE OF BARLEY
TO CEREAL LEAF BEETLE, OULEMA MELANOPUS L.

By

Chung Lee

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Crop and Soil Sciences

1970

 

 

 

 

642%?"
7r/*7°

ACKNOWLEDGMENT

The author wishes to express his sincere thanks and
appreciation to Dr. John E. Grafius and Dr. David H. Smith, Jr.
for their inspiration, help and guidance throughout the course
of this study. Their unfailing interest and encouragement
made the present thesis possible.

Thanks are also extended to Drs. Carter M. Harrison .
and Roger L. Thomas for their encouragements and invaluable
suggestions. Their kindness in reading and criticizing the
manuscript is especially appreciated.

Finally, he is indebted to friends who advised and
encouraged whenever needed during the period of his graduate

work in.Michigan State University.

ii

TABLE OF CONTENTS

Page

INTRODUCTION . . . . . . . . . . . . . . . . . . . 1

LITERATURE REVIEW . . . . . . . . . . . . . 3
MATERIALS AND METHODS 9
Plant materials 9
Experiments . . . . . . . . . . . . . . . . 9

Fl Test . . . . . . . . . . . . . . . . 9

Field generation test . . . . . . . . . . . 13

Feeding damage test . . . . . . . . . . . . 14

Plant tissue age test . . . . . . . . . . . 15

Component test. . . . . . . . . . . . . . . 16
Diallel cross . . . . . . . . . . . . . . . 18
Statistical Procedures . . . . . . . . . . . . 19

Analysis of genetic variance component. . . 19
Diallel analysis. . . . . . . . . . . . . . 21
Discriminant function . . . . . . . . . . . 22

Nomography. . . . . . . . . . . . . . . . . 24

iii

Page

RESULTS..................... 27
F1 Test . . . . . . . . . . . . . . . . . . . 27
Field generation test . . . . . . . . . . . . 32
Feeding damage test . . . . . . . . . . . . . 37
Tissue age test . . . . . . . . . . . . . . . 45
Component test . . . . . . . . . . . . . . . 49
Diallel cross . . . . . . . . . . . . . . . . 57
EESCUSSION. . . . . . . . . . . . . . . . . . . . 65
SUMMARY . . . . . . . . . . . . . . . . . . . . . 71
LIST OF REFERENCES. . . . . . . . . . . . . . . . 73

APPENDIX.................... 77

iv

Table

LIST OF TABLES

Plant materials used in this investigation.

Description of the six parents used in a
diallel cross observed . . . .

Stage of growth, description of plant
development and age of plants used in
the larval feeding experiment . . . . . .

Analysis of variance of cereal leaf beetle
larval weight gain on three growth stages
of four barley genotypes . . . . . . . . .

Comparison of average weight gains of
cereal leaf beetle larvae fed on three
growth stages of CI 6671, CI 6469, the F
generation of the cross CI6671 x C1 6469
and Larker a susceptible check.

I

Standardized values of larval weight gain
fed on three growth stages of C1 6671,
CI 6469, the F1 generation of the cross

CI 6671 x CI 6469 and Larker, a susceptible.

Average cereal leaf beetle feeding damage
score on three growth stages of barley
plants of four genotypes . . . . . . . . .

Means and variances of damage score and
number of larvae per plant observed in
the field with five generations of the
cross CI 6671(Pl) x CI 6469(P2).

Page

IO

10

ll

27

28

29

32

33

 

Table

10

ll

12

13

l4

15

LIST OF TABLES (Cont.)
Page

Analysis of variance and over-all mean of
cereal leaf beetle larval mobility, feeding
damage and weight gain on CI 6671, CI 6469,
F1: F2, BCl and BC2 generations of C1 6671

x CI 6469 and susceptible varieties Larker
and Dickson. . . . . . . . . . . . . . . . . 39

Average values of cereal leaf beetle larval
mobility, damage score and weight gain for

4 stages of plant growth over all entries

and replications. . . . . . . . . . . . . . 39

Average values of cereal leaf beetle larval
mobility, damage score and weight gain on

eight genotypes of plant over all plant

growth stages and replications . . . . . . . 40

Correlation coefficients between six
measurements of the interaction between
host plant and cereal leaf beetle larvae . . 41

Total and between treatment combination

sums of squares (es) and sums of cross

products (cp) of two different mobility
measurements, ml and m of cereal leaf

beetle larvae on barley. . . . . . . . . . . 42

Total and between treatment combination

sums of s uares (ss) and sums of cross

products ?cp) between three measurements

of cereal leaf beetle larval feeding. . . . 44

Average weight gain in mg of larvae grown

on the leaves of upper and lower parts of
plants at two stages of growth . . . . . . . 45

vi

LIST' OF TABLES (Cont. )
Table Page

16 Mean of the number of cereal leaf beetle
eggs per leaf, hatchability of eggs,
number of larvae per leaf, larval weight
gain, total larval weight gain and tiller
survival ratio on various genotypes in
two stages of plant growth . . . . . . . . SO

17 Correlation coefficients between obser-
vations on the components of resistance
-- l = number of eggs per leaf; 2 = hatch-
ability; 3 = number of larvae per leaf;
4 = larval weight gain; 5 = tiller sur-
vival ratio . . . . . . . . . . . . . . . 56

18 Combined analysis of variance tables
for number of larvae per plant and lar—
val feeding damage scores at two growth
stages on the F2 generation of a 6

parent diallel cross . . . . . . . . . . . 6O
19 Total, Wr and Vr values of the number of

larvae per plant for the F generation

of the 6 parent diallel seE . 61
20 Total, Vr and Mr values of the damage

score at the late stage of the F2 plants

of the 6 parent diallel set . . . . . . . 61

vii

LIST OF FIGURES

Figure Page

1 Standardized larval weight gain of cereal

leaf beetle fed on four genotypes at three

growth stages . . . . . . . . . . . . . . 3O
2 Damage score and larvae per plant of the

cross CI 6671 x CI 6469 and the F1, °F2,

BCl and BC2 generations. . . . . . 34
3 Larval weight gain on the leaves of upper

and lower parts of young plants of five

genotypes. . . . . . . . . . . . . . . . . 46
4 Standardized values of larval weight gain

on the leaves of upper and lower parts of

five week old plants of five genotypes . . 47
5 Larval feeding damage score observed for

four successive days on the barley plants
at the preheading stage. . . . . . . . . . 48

6 Ovipositional preference as measured by
the number of eggs per leaf, hatchability
and the number of larvae per leaf on
eight genotypes of an early stage of

plant grOWL/h o o o o o o o o o o o 51
7 Antibiosis as measured by larval weight

gain on young plants of eight genotypes. . 52
8 Antibiosis measured by the total larval

weight gain (number of larvae per leaf x
average larval gain) on eight genotypes
of young barley plants . . . . . . . . . 53

viii

 

Figure

10

ll

12

 

Page
Recovery as measured by tiller survi-
val after larval feeding damage on
eight genotypes of younger plants. . . . 54
Nomographic conversion of discrimi-
nant function . . . . . . . . . . . . . 59
A diallel graph for the number of
larvae per plant of the F genera-
tion of the 6 parent dialTel set . . . . 62
A diallel graph for the damage score
at the late stage of F2 plants of
the 6 parent diallel set. . . . . . . . 63

ix

INTRODUCTION

The cereal leaf beetle, Oulema melanopus L, has shown

 

a rapid increase accompanied by a constant broadening infes-
tation area since its first identification from collections
obtained near Galien, Michigan, in 1962. As of 1969, it has
been found in several hundred counties in nine states and in
the southern part of Ontario, Canada. This blankets approxi—
mately 10 percent of the small grain acreage of the U.S.A.
<32).

This insect, an Eurasian graminivorous representative
of Chrysomelidae, Order Coleoptera, has long been known in
Europe as a pest of small grains. The cereal leaf beetle
attacks small grains, especially oats, barley and wheat as
a leaf feeder. Several European records indicate that rye,
corn and some forage crops also may be hosts (4).

An active investigation of this insect was initiated
immediately after its identification in Michigan through a
joint program of the United States Department of Agriculture,
Michigan State University and Purdue University to study
control measures and to produce resistant varieties. In a

series of screening tests, a high degree of resistance was

observed in wheat lines while only a moderate degree of
resistance was recorded in barley and oats (10).

There is good evidence that the high degree of resist-
ance in wheat is mainly ascribable to leaf pubescence.

Hahn (14), showed resistance in barley to be a gene-
tically recessive trait and also suggested the possibility
of obtaining a higher degree of resistance through trans-
gressive segregation. However, the physical mechanism of
resistance in barley is not known.

The present work is aimed at obtaining a better

understanding of the mechanism and the genetics of

resistance.

 

LITERATURE REVIEW

The first European records of the cereal leaf beetle
as a pest appear as early as 1737 (4), and studies have been
carried out in France, Russia, England, Hungary and Germany
(32). Presently, this Eurasian pest shows extremely wide
distribution over the humid and subhumid areas of the
Western paleoarctic zone ranging from Sweden to Africa and
England to India (7). This leaf feeder was obviously
"imported" to North America from the Old World around 1960
(4), and since its first collection and identification in
1962 at Berrien county, in Southwestern Michigan, the fast
expansion of infestation has been plotted through annual
damage and collection surveys. The region of infestation
has expanded enormously and in 1969 an area ranging from
eastern Illinois to western New York and from southern
(hrtario, Canada to central Kentucky has become infested.

The hosts of this insect are mainly barley, wheat
and.oats but also listed are rye, corn, sorghum, a number
of grass forage crops, melons, sunflower and hemp (4).
Wilson (35, 36) reported more than 20 species of the

gramineae as host plants. Gallun §t_al (l2) and Everson

.23.§l (7) reported that the most severe feeding damage in
the field is done by larvae though the adult also feeds on
the leaf. The damage to Monon wheat (CI 13278), according
to Gallun et_al_(l3), resulted in 23 percent loss in yield.
Thorough studies on the systematics, morphology, life cycle,
and physiology of the cereal leaf beetle have been made by
Ruppel (22), Castro et a; (4), Wilson (33, 34) and Sengupta
2: a; (30).

Immediately after the identification of the insect,
a series of field screening tests was initiated at Galien,
Michigan to search for resistance in lines of the princi-
pal small grains. The first test was made on the adult
and larval feeding damage by Gallun and Ruppel (10) during
the 1962—1963 season followed in 1963-1964 for host plant
resistance (11). Continuous field tests were made by
Schillinger _e_'_c__ a1 (25» for the 1964-1965 period. Field
'tests have been continued to the present. Throughout the
sseries of screening tests, certain wheat varieties have
jproved to be less preferred than oats and barley for ovi-
position and feeding by adults and larvae. Everson §t_al
(7) noted a large number of wheat lines with a high de—
gree of resistance and most were of Russian or Chinese
origin while the few remaining resistant lines were mainly

from Asia minor and Southeastern Europe. Thus, they

suggested Asia minor as the main gene pool for resistance
and indicated Spain, Portugal and Ethiopia as another
possible germplasm center of resistance. Only a moderate
degree of resistance has been observed in barley through—
out the screening tests. Among the barley lines which
show some resistance, CI 6671 and C1 6469 (15 to 40 per-
czent of foliar damage) have been selected as the lines
‘with the highest resistance at present (26).

Host plant resistance is the result of the complex
:interaction between phytophagous insects and their hosts.
This relationship should be divided into two parts; (a)
llOSt selection by the insect and (b) resistance to the
irisect by the plant (2, 20). Painter (19) classified
“this complicated nature of resistance into three classes
as (l) preference or non-preference: the group of plant
ctuaracters and insect responses that led to or away from
tflie use of a particular plant or variety. This preference
magi be for oviposition, for food, or for shelter, or for
(xxnbinations of the three, (2) Antibiosis: the tendency
0f tflne plant to prevent, injure or destroy insect life
by eui adverse effect, and (3) tolerance: the ability of
a pliant to grow and reproduce itself or repair tissues
even.eafter injury. Painter also suggested the possibility

Of Obtaining cumulative resistance by recombining genetic

factors for different types of resistance.

Most of the field and laboratory tests on the cereal
leaf beetle have considered two aspects of resistance --
ovipositional preference and feeding damage by larvae.
Such partition is essential to understand the mechanism
of resistance to the insect. Gallun and Ruppel (10) and
Schillinger (27) pointed out that the resistance of wheat
to cereal leaf beetle is primarily associated with ovipo-
sitional non—preference due to the hairiness of leaves —-
higher resistance being associated with denser leaf
pubescence. Ringlund (21), through his genetic studies
'with the crosses between glabrous and pubescent wheat
varieties, confirmed this association by showing a highly
significant negative correlation between larval weight
(gain and pubescence, high pubescence density being asso—
ciated with resistance.

There is also some evidence that yellow color will
attract more adults than green (Wilson 37).

Gallun §t_al_(10) and Schillinger (29) observed
differential larval growth and oviposition preference
wnong lines of barley, wheat and oats. The most re-
SiStant wheat line CI 8519 was the least preferred for
OVATHDSition and larval feeding, but the barley lines,

CI 66W1 and CI 6469, having the highest degree of

resistance, were not preferred for oviposition and were
not different from the remaining lines in larval feeding.

The pattern of resistance was affected by various
environmental factors, such as the growth stage of the
host plant (Wilson 34), planting time of host (Schillinger
gt a1 26, Schillinger 28), stage of physiological devel-
opment, type of vegetative growth and disease suscepti-
bility of the plant (Schillinger §t_al 25). Also
Gallun (l2) pointed out that preference will influence
the amount of larval feeding damage per plant because
the older larger larvae tend to migrate from leaf to
leaf to find a preferred feeding site and presumably
spend less time eating, resulting in less damage.

The first genetic study on resistance in barley
was carried out by Hahn (14), using diallel cross sets.
In a laboratory test, he observed feeding damage at
the seedling stage, while feeding damage at the heading
stage was analyzed from field plots in the F2 genera-
tion. He showed that resistance was controlled by re—
cessive gene action. Moreover, some evidence of trans-
gressive inheritance was found in the field in the
progenies of the cross between two sources of resistance,

C1 6671 x CI 6469. He hypothesized that the mechanism

of resistance in barley was due to the negative pre-
ference by feeding larvae for the plant and to the
unfavorable conditions for egg laying. He found a high
correlation between the degree of resistance and the

number of larvae per plant.

MATERIALS AND METHODS

Plant materials

 

Two lines, CI 6671 and CI 6469, the most resistant
found to date (11, 26) were chosen as parents for the
present genetic studies. Reciprocal crosses between
these two lines were made during the Fall of 1967 and
a total of about 200 seeds were obtained. Some of these
Fl seeds were planted immediately after harvest to ob-
tain the two backcrosses and F2 population. Subsequently,
F3 seeds were harvested from the F2 plants. The remaining
Fl seeds were used for a series of preliminary tests in
the greenhouse (Table 1). Six parents were chosen for a

diallel cross. They are illustrated in Table 2.

Experiments

 

l. Fl—test
The resistance to cereal leaf beetle in the F1
generation of the cross CI 6671 x CI 6469 was examined
during the winter of 1968 in the greenhouse under con—
trolled light and temperature conditions. Four geno-

types, i.e. CI 6671, C1 6469, their F and Larker,

1
known to be a susceptible variety (26), were investi-
gated.- To avoid possible errors due to irregular ger-

mination and consequently different seedling vigor,

IO

Table 1. Plant materials used in this investigation

 

 

Reaction to

 

 

Genotypes Description cereal leaf beetle
a/

CI 6671‘ Introduction from Iran Resistant

CI 6469 Introduction from Poland Resistant

PHI Cross of CI 6671 x CI 6469 -

IT2 —

:FB 2 -

1301 C1 6671 x CI 6469 —

IBC2 C1 6671 x C1 6469 -

CI 10649 Larker Susceptible

CI 11531 Dickson Susceptible

 

a CI refers to Cereal Investigation number of Crops Research
Division, Agricultural Research Service, U.S. Department

of Agriculture.

Table 2. Description of the six parents used
in a diallel cross observed

 

 

 

 

L a/ Head b/V Plant
ines Resistance type Origin Maturity vigor
CI 12518 Moderate 2-row Ethiopia Early Low

CI 12528 Moderate 2-row Ethiopia Medium Low

CI: 6671 Resistant 6-row Iran Early Medium
CI 6469 Resistant 6-row Poland Late High
CI 12715 Moderate 2-row Ethiopia Late Medium
'DiCBkson Susceptible 6—row U.S.A. Medium High

 

.From the result of a screening test in 1965 (26)
b/’
‘ All spring types.

11

the seeds were planted in wooden flats filled with sand and
only uniform seedlings were transplanted after seven days to
5 inch clay pots. Such a transplanting procedure was used
for all greenhouse tests.

To reduce the variation between cultures, one indivi-
dual from.each of the four genotypes was planted in each
culture at each planting.

Four successive plantings were made to produce plants

of different stages (Table 3).

Table 3. Stage of growth, description of plant
development and age of plants used in
the larval feeding experiment

 

 

Age of plant

 

Strage Stage of growth Description in weeks
ILst Early seedling 2 leaf stage
2rui Old seedling 3—4 leaf stage 4
I3rd. Preheading 4-5 leaf stage a week
. before heading 5
4tli Post heading a week after heading 7

 

 

The design used was a randomized block of 3 replica—
ti£n1s, 4 varieties, and 4 stages —- comprising 48 units, each
UDJJ: of 5 cultures. When the plants reached the stages shown

in Table 2, two first instar larvae reared by a standard

12

method in the laboratory (6) were placed on the youngest leaf
of each plant. As suggested by Schillinger (27) and Chada
(5), each plant was then covered with a plastic cylinder.
iHowever, high humidity within the cylinder and the difficul-
'ties of covering tall plants led to a high larval mortality
.and.a high frequency of escape, respectively. The experi-
Inent was therefore restarted after five days from initia—
‘tion of larval feeding, without such covers. After five
days feeding, the larvae were removed from the plant,
<2arefully wiped with filter paper and the body weight re—
<2orded with a 0.1 mg precision torsion balance. After all
Ilarvae were removed and weighed, the damage to the plants
'was scored subjectively. Throughout the whole series of
experiments, the damage score was read from O to 4 based
on the following criteria;

O: none to trace of feeding damage

1 slight feeding damage

2 moderate degree of feeding damage

3: relatively heavy feeding damage on the leaf

4 extremely heavy feeding damage.

Actually the scores 0 and 4 were rarely read. This
method of measuring larval weight gain and scoring damage
was used throughout all tests and was in contrast to that

used by Hahn (14), where scores ranged from O to 10,

13

according to the proportion of damaged leaf area. Such
precision was not applicable to the present material.

An analysis of variance using a two-way classifi-
cation was done within each of the four individual plant
growth stages.

2. Field generation test

The parents and the F1, F2, BCl and BC generations

2
were tested at the field nursery located near Galien,
Michigan. Twenty seeds of each were space planted in each
of three replications at 3-inch intervals in 6-foot rows
spaced one foot apart. Approximately 70 percent of the
seeds resulted in adult plants and 10 inner plants of
each row were used for the observations. The feeding
damage on the plants was scored and the number of larvae
Iper plant counted at heading. Heads were harvested and
the following characters were measured: average head
'weight, maturity ratio on a numerical basis (the propor-
tion of perfectly filled kernels to total number of
florets per spike), maturity ratio on a weight basis

(the ratio between the weight of filled kernels in a

Spike and that of the whole spike) and 1000 kernel weight.
Harvested heads from each plot were randomly subgrouped

in order to calculate the genetic variance components for

all characters measured in this study.

14

3. Feeding damage test
During June of 1968, a greenhouse test was made to
determine the relative efficiency and credibility of measure-
ments of feeding damage. The genotypes used were CI 6671,

CI 6469, their Fl, F , BC , BC and Larker, a susceptible

2 1 2’
check (Table 1). Plants of four different stages were
tested simultaneously. The stages were identical to those
illustrated in Table 3. Two seedlings were transplanted
to individual pots with five replications. Each culture
was considered as a replication. Three measurements, i.e.,
mobility, damage score and larval weight gain were observed.
The mobility of larvae was observed in the following manner.
One late second instar larva was placed on the youngest leaf
and observed every 12 hours (at 9 a.m. and 9 p.m.) for seven
days. Each movement of larva from one leaf to another was
recorded by counting the number of leaves it had passed.
There were a few cases in which the larva was either dead
or had escaped. During the first three days of testing,
the missing larvae were replaced with another larva of the
same age. The data were converted to a score in two ways:
IMethod 1. Giving one unit value for each movement across
a leaf and two for escape or death.

Method 2. Same as above except giving one half unit for

escape or death.

 

15

The feeding damage scores were read on the third and
tile last day of larval feeding. After seven days, larval
ineight was measured excluding the ones which were replaced
(hie to previous escape or death. A11 measured characters
inere analyzed in a factorial design and intraplot variances
inere calculated for each genotype at each stage. Analysis
(of genetic variances was examined.

4. Plant tissue age test

During the period between September of 1968 to March
21969, a series of tests was made to find the difference in
ilarval feeding among the plant parts. Only two stages, namely
tflie early seedling stage and preheading stage were considered.
Tflie procedures closely followed those suggested by Schillinger
fxar determining larval weight gain (26, 27). The parents,
their F F

23 3)
liJlgS were transplanted to each pot and white quartz sand

and Larker (Table l) were tested. Four seed—

vnas poured on the soil surface to facilitate the spotting of
Stxray larvae. To "protect" the larvae from food shortages,
macriplant was infested with one larva and the plants were
curvered by glass lantern globes. Plants of a later stage of
development were tested without the globe. For the early
stage, damage scores were read every day beginning on the

second day of feeding. After five days, larval weights

 

16

were measured. Another set of tests was made with identical
genotypes at two plant growth stages to observe any dif-
ferences in damage between upper and lower leaves. Single
first instar larvae were put on both the upper and lower
leaves in 3" x 3" x 7%" plastic leaf cages, to prevent the
larva from escaping or feeding on another portion of the
plant. Care was taken to insure larvae adequate food by
shifting the position of the leaf cages. Two leaves on the
lower portion and another two leaves on the upper portion
of the plant were tested. After four days of feeding,
larval weights were measured.
5. Component test

Painter's classification of resistance (19) was modi-
fied for the present purpose into three components -- ovi-
positional preference by gravid females, resistance to
larval feeding damage and recovery of the plants.

Ovipositional preference is defined as the group of
plant characters and the responses of a gravid adult female
that leads to the use of a particular plant or variety for
oviposition.

Resistance to larval feeding damage is the expression
0f tfhe interaction (reaction) between the plant and feeding
young; insect larvae in which the antibiotic effect of plant

on inssect causes a reduction in larval weight gain and the

17

lessening of the degree of feeding damage on plant tissue.
Recovery is a measure of resistance in which the plant
grows and reproduces itself by recovering from the injury
caused by larval feeding.
Tests of the components were made in the spring of
1969. The parents, CI 6671, C1 6469, and the F2, F3, BCl’
BC2 plus two susceptible varieties Larker and Dickson
(Table l) were transplanted to five pots for each entry, each
culture containing four seedling plants; one culture was
designated as a replication. At the third week after planting,
all eight entries were randomized within each replication and
all experimental materials were enclosed by a 1.8 x 0.9 x 0.7-m
sized wooden cage (Schillinger (29)). The cage was transferred
to a growth chamber in which 15 hours of light were followed
by a.nine—hour dark period at a temperature of 76 F. Approxi-
nuxtely 200 laboratory grown adult beetles (6) were released
:hiside the cage and allowed to oviposit without restriction
:fior 36 hours. After 36 hours, the plants were removed from
the cage and the number of eggs per plant counted. As there
were some differences in the number of tillers and in plant
VLSOI; the number of leaves per plant were counted and used
as tile denominator to calculate the number of eggs per leaf.
Thesee plants were then returned to the growth chamber to

provixie an environment favorable to hatching, after removing

18

the adult beetles. Ninety-six hours after the termination of
egg laying, the numbers of hatched larvae were counted and
after seven days, the average larval weight was measured. As
there were differences in the stage of larval growth, as many
samples as possible were taken. Finally, tiller survival
ratio and average head weight were measured. For each plant,
tiller survival ratio was calculated by dividing the final
number of heads by the highest number of tillers observed
for that plant. As the measurements were on an individual
plant basis, correlation coefficients between observed
characters could be calculated.

An identical set of six-week old plants was tested in
exactly the same manner described above except that the
period of egg laying was extended to 70 hours and the tiller
survival ratio was not observed.

6. Diallel cross

During the spring and summer of 1968, a complete dial—
lel cross series excluding reciprocals was made with the six
parental lines indicated in Table 2. Among the lines, CI 6671
mmi Cl 6469 were used as resistant parents while Dickson was
(fimxsen as a susceptible parent, based on results from the
fielxi screening tests (10, ll, 12, 25). F2 progenies and
Parerrts were planted in a field nursery located at Galien,

MiChjngan. Twenty seeds were space planted three inches apart

19

in four replications. The rows were four feet long and were
spaced one foot apart. The plants in the nursery showed good
infestation and data were collected on the number of larvae
jper plant and for damage scores at the early and late plant
growth stage, roughly corresponding to stages 1 and 3 of

'Table 2. Each damage score was the average of 10 observations.
.After harvest, the average head weights per plot were measured
and.the data were analyzed by the Jinks-Hayman (l5, l6) dial-

Ilel analysis.

Eitatistical Procedures

 

a. Analysis of genetic variance component

The genetic variance components of homozygous and se—
ggregating populations of self—pollinating crops have been
defined and analyses have been developed by a number of
workers, especially by Fisher and Mather.

The variation in observed values within any pair of
trueubreeding parents and their F1 is assumed to be exclu-
sively due to environmental effects. 0n the other hand, any
variation within the F2 generation is due both to genetic
dtfiference and to environmental effects. The genetic va—
riation within the F2 generation of a theoretical A—a
locus has been defined as 2uv(d+(v-u)h)2 + 4u2v2h2, where

u and v are the gene frequency of A and a alleles,

20

respectively, d is additive genetic effect (or gene effect)
and h is the dominance effect exhibited by the heterozygous
Il-a locus. However, in a population which has been artifi-
cially built up through hybridization between two homozygous
paarents, u=v=% and thus the genetic variation of the F2 gene-
.ration is %d + %h . In practice, an environmental factor,
e, should be added to the above equation. If one considers
ea quantitative character, controlled by k genes (or effective
:factors) and assuming that they do not interact,

vF2 = %D + %H + E

K 2
where D = )2: di
i=1
k 2
H 2):: h.
i=1 1

E: Environmental variation.

If‘ backcross generations are available,
VBCl + V302 = in + 3H + 2E.

Thien five generations, namely two parents, F1, F2, B01 and
TREE are located in one environment, E, the environmental
variation, will be common to each group and can be estimated
by observing the variations within populations of identical
genotypes. Thus, the genetic variance components were cal-

culated in Tests 2, 3 and 5, along the lines described above.

21

.Environmental variances were estimated by averaging the
‘valdences of the two parents (and F1, where available).
Heritability is defined as the ratio of additive ge-
riewic variance in a given population to total variance of
'trie same population. Thus heritability in the F2 generation

18,
l
h2F = ED
F2

Sijilarly, heritability at the F3 generation is

 

ifiie number of effective factors were also calculated as sug-
gested by Wright and Mather (18) .
b. Diallel analysis
For the diallel cross sets, the Jinks-Hayman (15, 16)
Wr/Vr analysis was applied. After calculating the variances

withiri each array (V and covariance of the r array with

r)

non—rwecurring parents (Wr), regression equations of Wr to

Vr W81?e calculated. As W2r = Ver, points coordinated on

the pJJane made by Wr and Vr axis will be confined by a

NIH

limitqug parabola Wr=(Ver) where Vp is the parental
variaruze. In the presence of complete dominance, the cal-

culated regression line will show a slope of unity with

22

interception at the origin and thus observing the position of
interception on the Wr—axis, the degree of dominance can be
(ietermined. When the slope is significantly different from
1xnity, the existence of non-allelic interactions is probable.
IPoints along the slope of Wr’ Vr are arranged in the dominance
order of the parents. Significant difference of the regres—

sion line frOm either unity or zero can be tested by the t-

b — bO
“test using the formula, t =-—7;-—— , where bO is 1 or O,

b
:respectively.

c. Discriminant function

A discriminant function can be defined as a linear
(zombination of available measurements (or variables) into one
ftmction. The coefficients of each variable are chosen in
such a manner so that the function can minimize possible
errors in using those multivariables to characterize a com-
plex trait, and concomitantly maximize the difference between
two or more classes of objects or individuals relative to the
variation within the class. The actual computational method
involves an application of the least squares method to a
multivariate function. Descriptions of the theory and the
applications of discriminant functions have been published

by Fisher (8, 9), Smith (31), and Mather (17, 18). For the

23

present studies, three of these functions were calculated to
discriminate between the two ways of measuring larval mobility,
the three measurements of feeding damage and finally to find
some way of combining the measurements of the three components
of resistance, viz, ovipositional preference, larval feeding
damage and plant recovery. If we assume three variables
(measurements) L, W, T and the final outcome to be D, the pro—
per combination of L, W and T will yield a discriminant func—

tion for D.

For this purpose, according to Mather (l8) and Fisher
(8), the coefficients bL, bW and bT of L, W and T, respectively,

should satisfy the following equations.
bL(ALL - gaLL) + bw(ALW"¢aLw) + bT(ALT'¢aLT) = O
bL<ALW - wan.) ,+ bWwW-QIaW) + bT<AWT-¢aWT> = o --—<1>
bL(ALT ‘ 8aLT> * bw(Awr'8aWT) * bT(ATT_gaTT) = 0

where ALL’ ALW’ etc. are the tota1_sum of squares or
Sum of cross products between variables and aLL, aLW, etc.
are the corresponding sum of squares or sum of the cross
product between treatment combinations. D was used to ad—
just the values of each_a when they are subtracted from

total sum of square or sum of cross product (A‘s) and was

24

estimated as below. As there are three equations and four
unknowns (fl, bL, bw, bT), only the relative magnitude of b's
can be estimated. By converting the above equations into ~

the following determinant form, D can be estimated.
ALL‘8aLL ALw'8aLw ALT‘gaLT

O ——--(2)

A .
Lw'gaLW AWW'¢aww AWT‘gaWT

 

ALT-¢aLT AWT‘gaWT ATT‘gaTT

 

Solution of the above determinant leads to a third power
function of D. Among the roots, the smallest one is taken
and substituted in equations of (1). From this set of equaé
tions the ratios between bL, bW and bT are calculated. By
setting the lowest value to l, we can obtain the discriminant
function in a form of

D = bLL + bWW + b T,

where one of the b's has a value of l.

d. Nomography
For more rapid and convenient uses of derived functions
in evaluation of varieties for their resistance, nomographic
conversion of the function was made (1).
Nomography is a special kind of graphic representation

which can be used as a visual means of calculation of any

25

number of special cases. However, to present a three factor
function in a two dimensional picture, a special treatment of
coordinates is needed. A brief description of the theory is
presented.

Four points Pl(xl, yl, zl), P2(x2, y2, 22), P3(x3,y3,z3)
and P4(x4, yu, Z4) in a space are co—planar when

x l

1 yl 21

X3 y3 Z3 1

 

 

x4 yu Zn 1
If we consider the four variables L, W, T, D and
assume a given function, F(L, W, T, D) I 0, then this function
F can be put in the form
L1 L2 L3 1
W1 W2 W3 1
1

T1 T2 T3

 

 

1 D2 D3 1

where Ll, L L are the functions of L only and the

2’ 3

same for W, T and D.

Fbr a given function L + W + T = D, if we SGt
A = 1L

B = WW

(3

: tT,

26

then,

A°l + 13.0 + 0.0 + 1L

A°O f B‘l + C‘O + WW

ll
0 O O O

A°O + B‘O + C'l 1. tT

Ac}

;

.1 _ «.1
+ B a +*b t + D

By proper operation, the above equations will be brought to

the determinant form as

 

 

O 0 1L 1
G 0 SW 1
= O
O K tT l
I ltG 1wK lth l
1w+wt+1t lw+wt+1t lw+wt+1t

Expansion of this determinant form yields the original
equation L + W + T = D.

In this case, 1, w, t are the scale multipliers or
ScaljafEctors which can be used to expand or contract their
reEspective scales. G and K are constants employed to produce
amatrix in canonical form and permit varying the width of
the scale. These quantities give more flexibility to the
use of three dimensional functions displayed on a two dimen-
sional picture. As the components L, W and T have their res-
Pective coefficients, bL, bW and bT, 1, w and t quantities
were modified by the coefficients derived from the discrimi—

nant function.

RESULTS

1. Fl testl/

The aim of this test was to examine the response of
F1 plants to larval feeding as compared with the parents. The
mean values of larval weight gain and feeding damage score
cfi'each genotype are shown in Tables 5 and 7, respectively.
Due to the drying out associated with maturity, about 50
percent of the first instar larvae placed on the flag leaf
of seven-week old plants died and hence the last growth stage
was dropped from the analysis. The remaining three stages
Showed either significant or highly significant differences
for larval weight gain among the four genotypes tested ——

Tables 1+ and 5.

Table 4. Analysis of variance of cereal leaf beetle
larval weight gain on three growth stages
of four barley genotypes

 

m

 

 

 

——130urce Degree of freedom Mean square F
RePlications 2 8.67 1.82
Stage of growth 2 17.86 3.76*
Genotypes 3 £10.06 8.43H
Stage X Genotype 6 34.48 7.97**
Error 22 4.75
W: .05

**P = .01

-/ Materials and methods section 1.

'27

28

Table 5. Comparison of average weight gains of cereal
leaf beetle larvae fed on three growth stages
of CI 6671, CI 6469, the F1 generation of the

cross CI 6671 x CI 6469 and Larker a suscep-
tible check

 

 

Growth stage a/

 

 

 

 

 

Genotypes T1 T2 T3
Ifi_generation 16.2 mg 18.1 mg 15.0 mg
CI 6671 12.9 14.2 16.0
CI 6469 12.4 17.6 16.9
Larker 15.1 17.8 17.8
F-va1ue 5 .48* 7 . 21** 25 . O3**

* p = .05
** p Z .01

a
-/TP1: 3-week old plants
5P2: 4-week old plants

CD3: 5-week old plants

The data in Table 5 were adjusted so that the mean and stan-
dard deviation for the different plant growth stages were in
the Same units. The transformation was

. ‘Y 53'?

_ 13 i
Yij S

29

I

where Y. : standardized value of 3th entry
13 in ith stage
Yij original value of Jth entry in
1th stage
'Ti : mean value of ith stage
Sd- : standard deviation of ith stage.

By this transformation, the values at each stage are

rearranged around a mean of O and a standard deviation of

1. These new values are shown in Table 6 and Figure 1.

Table 6. Standardized values of larval weight gain

fed on three growth stages of CI 6671,

CI 6469, the F1 generation of the cross

CI 6671 x CI 6469 and Larker, a susceptible
check.

 

 

 

Entry Tl Groggh stage é T3
F1 generation 3.56 1.61 —l.33
01 6671 -2.31 —3.86 - .38
01 6469 —3.02 1.0o .47
Larker 1.78 1.28 1.23

 

 

at/ T1 : 3—week Old plant

T2 : 4-week old plant

T : 5—week old plant

3

Standardized larval
weight gain
L

 

3O

: Larker

 

 

/’..~.“ ~ ~ —. .
’/’ \\\ ----~~"-C Cl 624-69
,x’ \\\\ CI 6671
,,/ “ Fl (CI 6671 x
’1’ CI 6469)
(I
‘ ‘ _L
EARLY MEDIUM LATE
STAGE

Figure l. Standardized larval weight gain of

cereal leaf beetle fed on 4 genotypes
at 3 growth stages.

From Fig. l the following points are noted:

1.

At the youngest stage of plant development, the
two resistant parents CI 6671 and CI 6469 show a
much higher degree of resistance than either
Larker or the F1 of the cross CI 6671 x CI 6469.
At this stage, the hybrid between the two re—
sistant lines is as susceptible as Larker.

This agrees with Hahn (14).

At the middle stage, CI 6671 is more resistant

than the remaining three lines, which do not

differ greatly among themselves.

 

31

3. At the late stage, the resistance of the F1 plants
exceeds that of all other lines. Line CI 6671
shows considerable resistance while CI 6469
appears as susceptible as Larker.
4. The F1 generation seems to increase in resistance
with plant age whereas CI 6469 decreases.
The variety CI 6469 appears to be resistant to larval feeding
at the early stage but its resistance is weakened at later

stages. On the other hand, the F plants are quite suscep-

l
tible at the early stage but at later stages, their resist—
ance is reinforced for some reason and they display an even
higher degree of resistance than the parents. But, CI 6671
and Larker show relatively constant resistance and suscepti—
bility, respectively.

The damage score reading failed to indicate any sta—
tistically significant difference between lines, as shown
in Table 7. This may be due in part to the inadequacy of
the measurements or to the limited number of samples. A
difference in growth habit could also contribute to the
fallure’of significance tests for damage score since CI
6469 shows a high degree of plant vigor and CI 6671 is a
line with a short stem and a small number of tillers. Such

differential plant vigor could cause some confusion in

 

 

 

finging the amount of damage, and could lead to the evaluation

ofCI 6469 as the line of highest resistance over all stages.

hisubsequent tests, the factor of differential vigor was

carefully considered when a damage score was read.

Table 7. Average cereal leaf beetle feeding damage
score on three growth stages of barley

plants of four genotypes

 

 

8/
Stage of growth

 

Genotypes T1 T2 T3
Fl 2.20 2.26 2.13
CI 6671 2.40 2.56 2.26
CI 6469 1.93 2.13 1.80
Larker 1.93 2.86 2.66
F-Valueb/ 3.81NS 1.21NS 2.04NS

 

a/ T1 . 3-week old plant
T2 : 4—week old plant

T3 : 5-week old plant

significant difference

1
2~ Field generation test_/

The mean and variance values of damage scores and

number of larvae per plant are entered in Table 8 and also in

————_1___________11_
l/ Material and Methods section 2.

IIIII--______1

h/ F values were calculated for each stage; NS = no

 

 

33

Figure 2. These observations were made at the heading stage

which was equivalent to the third stage in the previous test.

Table 8. Means and variance of damage score and
number of larvae per plant observed in the
field with five generations of the cross

CI 6671 (Pl) x CI 6469 (P2).

 

 

 

 

 

Population Damage score Number of larvae plant

Mean Variance Mean Variance
21(01 6671) 5.1 .359 5.1 8.537
P2(01 6469) 5.5 .287 12.6 43.530
Fl 3.8 .275 8.7 38.337
F2 4.9 .536 7.5 23.040
301 4.8 .496 5.3 18.388
B02 5.3 .371 10.0 36.174
F-value 2.610f 3/ 2.81?

 

 

H {non-additive genetic

Variance) .096 52.24
D (additive genetic

variance) .410 0
2 . .
h (Heritability) 38.24% 0

K Unnnber of effective
factors) 1 or 23

 

 

g/wf: significant at 10%Slevel

 

 

 

 

 

    

 

 

 

 

 

 

 

 

 

:1“ m (\l H 0 O\ CO D- \0 L0 :f m (\l
(3 43 H H r: r: 1* r J r 44““
s8
@ri
'0 Q
s .\\\\\\\\\\\R\\\\\\\x\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
r.
-p
:3 WW
Q.
8
Cl.
0)
c0
Z
8 3 »\\\\\\\\\\\\\\\\\\\\\\\\)§\\\\\\\\_\\\\\\\
.5. L
(D M
8 ,2
E E
I: \\\\\\\\\\\\\\\\\\\\\\\\\\\
I E§§§§
(D .
Q ,_~ .
8 L W
. I
2’30 k; ‘r : t —t r t ‘- 1‘ Ar 4v— “V
§ 02 K) :r cm 0 d) \o :- ru 0 G) \o 3' CU
Qmuhuofiuhuhirro$jqum03m

Generation

Damage score and larvae per plant of the cross CI 6671 x CI 6469

and the F1, F2, B01 and B02 generations.

Figure 2.

The damage score shoWs that the progenies are more
resistant than the parents and that the F1 is the most resist-
ant followed by the F2. Through the analysis of genetic
variance components by the methods of Fisher and Mather (18),
a heritability value of 32.24 percent was obtained for the
F2 generation for larval feeding damage score. The number of
effective factors for the traits could be calculated in two
ways with the present data.

If we assume k effective factors which have equal
additive genetic effect, that is, da=...=dk=d where da is
tflie additive genetic effect (gene effect) of ath factor and
d is the average of all d's and again assuming that all

jpositive genes are concentrated in one parent and all negative

‘

._ R
genes to the other, we can see Pl - P2 = 2 (da) 2' kd. By
a=1

definition D is the sum of squares of dis and with the first

k
assumption, D = Edi = kd2. Thus, if both assumptions are
i=1

holdin P .. P 2 k d
g, ( l 23) _ — k. The calculated k value will

-'——-.—-—-—g—-_

2

 

D kd
be Iflflderestimated when (l) linkage exists; (2) increments

OfEDD-Sitive genes produce an unequal effect; (3) when genes
are iisodirectionally distributed between parents. The use

Of thfé above method resulted in a k value of 0.4 for feeding

36

damage. This is a very low value but considering that both
parents are resistant, it may be that the genes are of an
isodirectional distribution between parents which could very
well cause a downward estimation. Another way of estimating
k values uses the dominance effect, h. Under the identical

, k
assumptions made above, Fl — (mid parent) = Izaha = kh,
a:

where the absolute values are from the mean of the observed
populations and da is the dominance effect caused by the

heterozygosity of the ath factor (locus). As H is defined

 

as izihg with equal dominance effects among k factors,

 

— —' 2
H = kh and therefore, _____1i____l_ =_____ = k.
, H kh2

This method led to a calculated k value of 23. Such a high
Value is in'striking contrast to that calculated by the prem
Vious method, and the inconsistency suggests a complex

pattern of inheritance. Calculation of the heritability of

 

the number of larvae per plant factor was not done since the
value of additive genetic variance was negative. The plant
materials were harvested and measured for yield characters
as ShOWnin Appendix 1. The observed values are a function
0f bOtli genotypic yield characters and of the response to

larVa1_ feeding. Due to the absence of damage-free check

 

37

ploug it is impossible to isolate the damage response by
insect feeding. The results do not indicate any definite
tnends except that the F1 shows a degree of plant vigor as
kugh as CI 6469, the vigorous parent. As for the number of
larvaejper plant, the present results in Table 8 indicate
iflmt CI 6671 is much less favored for oviposition than

CH 6469 while the F1 shows an intermediate level.

3. Feeding damage testl/
Feeding damage is the result of the interaction be-
tween larval feeding and the host plant. This relationship
can be measured and judged in various ways, but there has
been no direct way of combining such multimeasurements or
deciding their relative efficiencies as yardsticks. The
conventional way of measuring this relationship has been
(fither the subjective scoring of feeding damage by visual
cmservation or the relative amount of larval weight gain
after a certain period of feeding. These two measurements
sue two facets of the host parasite interaction; the former
is a:measure of the response of plant parts to insect feed~

jhg While the latter is the response of the insect by feed—

jhg the plant. Another possible measurement of larva-plant

 

l/ Materials and Methods section 3.

38

response is the mobility of larvae on the plant. As stated
by"Ga11un §t_al_(l2), an insect larva, when forced to feed on
emplant of undesirable ”quality,"tend to move from one place
'maanother to examine the possibility of getting more agree-
aflfle food. The present test was aimed at finding a way of
mmmining these measurements into a single character and
then testing their relative efficiencies as a measurement.
The above three measurements were observed with seven geno-
types at four different stages, the F values of which are
shown in Table 9. Over-all mean values for each stage and
genotype is shown in Table 10 and Table 11, respectively.
Two measurements on larval mobility were made. During the
testing period there were cases in which larvae were dead
or escaped and these were at first interpreted as an expres~
Sion of an avoidance reaction. However, the possibility
exists that it may be by chance only. Such difference in
interpretation resulted in two mobility measurements as
shown in Tables 9, 10 and 11°

Damage scores were read twice, one soon after the
larvae started feeding and one about the end of the feed—
ing period. The average of the two observations was cal—
culated. The means and the variances within treatment

combinations of each entry on each stage are shown in

39

Table 9. Analysis of variance and over-all mean
of cereal leaf beetle larval mobility,
feeding damage and weight gain on

c: 6671, c: 6469, F1, F2 BC

generations of CI 6671 x CI 6469 and
susceptible varieties Larker and Dickson.

and BC2

 

 

 

 

 

’Mobie Mbbi- Damage Damage SDamage Larval
Traits lityg/ litya/ Score a/ Scorea/ Scoreé/ Weight
Observed (1) (11) (I) (II) (Ave.) gain
in mg.
Degree
Source of
___ Freedom F-values
Reps u 3.05 91.18 4.88** 4.29* 3.01 .92
Stage 3 8.34** 7.28** 1.97** 9.26*§/ 1.96D/ 2.45*
Entry 7 1.32 1.74 1.48 1.85*— 1.751' 7.81**
SXE 21 .94 1.12 1.71 1.86* 1.74* 6.73**
Over-all mean 4.55 3.25 1.75 1.98 1.86 1.69

 

a/ For the differences in measurements, see text.
significant at 5% level.

96*

b/ 1-

 

significant at 1% level.
significant at 10% level.

*.

Table 10. Average values of cereal leaf beetle larval
mobility, damage score and weight gain for
4 stages of plant growth over all entries

and replications.

 

 

‘_

 

 

Age of Mobih Mobi- Damage Damage Damage Larval
Stage Plant lity lity Score Score Score Weight

(week) (I) (II) (I) (II) (Ave.) gain
.1___ in mg.
1 3 3.51 3.65 2.02 1.61 1.82 1.70
g 4 4.3g 3°i8 1.62 1.97 1.8g 1.6g

5 .5 3. 7 1.77 2.20 1.9 1.7

4 7 3.84 2.61 1.58 2.14 1.86 1.66

40

Appendix 2. The correlation coefficients between the six
measurements are presented in Table 12. A series of t—tests
r ,
with the formula, t = ' indicates that the
\kl—rh/(Nae)

mobility of larvae is independent of damage score or larval
weight gain. Highly significant correlation coefficients
were obtained between larval weight gain and damage score

observed at later stage of feeding and average damage score.

 

Table 11. Average values of cereal leaf beetle
larval mobility, damage score and
weight gain on eight genotypes of
plant over all plant growth stages
and replications.

 

 

Larval
Mobi— Mobi- Damage Damage Damage weight

Entry lity lity score score score gain
(I) (II) (I) (II) (Ave.) in mg.

 

CI 6671(p1) 4.72 3.20 1.72 2.20 1.96 1.60
c: 6469(P2) 4.80 3.35 1.85 2.02 1.93 1°67
F1(P1 x P2) 4.85 3.60 1.65 1.90 1.77 1.60

F, 4.35 2.75 1.85 1.72 1.78 1.14
301 3.82 3.02 1.67 1.92 1.80 1.60
B02 4,47 3.52 1.60 1.97 1.78 1.97
Larker 4.82 3.35 1.92 2.12 2.02 1.75

 

41

Table 12. Correlation coefficients between six measure—
ments of the interaction between host plant
and cereal leaf beetle larvae.

 

 

Mobi- Mobi— Damage Damage Damage Larval

 

 

 

 

Measurenent (igyé/ (I§¥a/ S(§§§/ i§§§e§/ Eigiig/ g:i§ht
Mobility (I) 1.000

Mobility (11) .760 1.000

Damage score (I) .248 .189 1.000

Damage score(II)-.l23 .003 —.001 1.000

Damage score@v) .070 .127 .657 .752 1.000

Larval wt. gain .054 .110 .004 .428 .392 1.000
a/

— For the difference between measurements, see the text.

From Table 10, larvae move more on the plants at an
early stage, though there is no differential movement between
genotypes. This indicates that there is no larval feeding
preference. Another computation of variances within each
stage also failed to show any significant difference in larval
movenent between genotypes of the host plant at any stage of
Srthh. Thus, larval movement is affected by the age of the
plant but genotypes do not influence mobility. To compare
the tWO different ways of interpreting and evaluating mobility,

a diScriminant function was calculated. Total and between

 

42

treatment combinational sum of squares and sum of cross pro—
ducts between the two measurements were calculated from

original data, as illustrated in Table 13.

Table 13. Total and between treatment combination
sums of squares (ss) and sums of cross
products (cp) of two different mobility
measurements, ml and m2 of cereal leaf

beetle larvae on barley.

 

 

 

 

mlml (ss) m2m2 (ss) mlm2 (cp)
Total 1328 608 683

Between 209 107 115

 

D had a value of 6 and the final function was
M = 1.46ml +m2

where M is the combined mobility index from the two different
measurements of mobility ml and m2. From this function, it
can'be suggested that an evaluation of mobility by the former
method (ml) is more reliable and accurate. Larval death or
escape may be an expression of extreme negative preference
or antibiosis.

Of the three measurements obtained for feeding damage,
that taken near the end of feeding exhibited a significant
difference among entries and among stages together with the

SXistence of interactions between the two factors. Despite

43

some indication of statistical significance, damage scores
taken early during the feeding period do not show any
noticeable trends. Perhaps more time for feeding should
be allowed in order to observe the true nature of resist—
ance, and then more credence may be given to the results
obtained near the end of the feeding period. From the
feeding damage score observed, the Fl plants of I 6671
and CI 6469 were again rather susceptible in the early

seedling stage but became quite resistant by the prehead-

 

ing stage. A similar trend was observed for larval weight
gain. This is in good agreement with the results obtained
from the two previous tests. However, the stagewise
pattern of resistance observed in two previous tests was
not found to hold for the parents in this test.

The main objective of the present test was to com—
pare and combine the three different measurements into one
single trait. Again a discriminant function was built on
mobility, damage score and larval weight. The first mea—
surement of larval mobility was used because it proved
more effective than the second by the previously calculated
function. Sums of squares and sums of cross products for
total and between treatment combinations are shown in

Table 14.

44

Table lu. Total and between treatment combination
sums of squares (ss) and sums of cross
products (cp) between three measurements
of cereal leaf beetle larval feeding.

 

 

ss or cp—/ ll(ss) 44(ss) 66(ss) 14(cp) l6(cp) 46(cp)

 

 

Total 1328 173 44 —59 18 13
Between 209 32 24 —33 8 l;
l/ l = mobility (l) u = damage score (2)

6 = larval weight gain

 

Calculated S values were 3, a double root, and ll°
Taking the value 3, the following equation was calculated
LD = M + 3.68 + 8.4W

where LD: combined estimate of damage due to
larval feeding

M: mobility,

S: damage score

W: larval weight gain.
Using this function, coefficients of each measurement were
obtained by which different measurements were weighted in
Such a way as to minimize the subjective error in judging
a tnle response of a plant to larval feeding. Since the
Converl'tional measurements do not include mobility, another
func"(lion was worked out which eliminates the mobility variable:

LD=S+2.3W

45

This function indicates that larval weight gain is the more

reliable and accurate measurement of feeding damage.

1h Tissue age test ;/

A different pattern of response to larval feeding be—
tween the two growth stages of the host plants is evident.
It could well be that tissue age rather than growth stage
causes such a difference. In pursuit of this point of tissue
age, three—week old and five—week old plants were tested for

differential feeding on the upper and lower leaves. The re—

 

sults are shown in Table 15.

Table 15. Average weight gain in mg of larvae grown
on the leaves of upper and lower parts of
plants at two stages of growth.

‘

 

 

 

 

 

Stage of Plant Early (3 weeks old) Late (5 weeks old)
Entry Upper Lower Upper Lower
leaf leaf leaf leaf
CI 6671 6.19 mg 1.46 mg 8.10 mg 6.26 mg
CI 6469 7.43 2.16 9.42 6.30
F2 6.33 1.54 6.30 2.94
F3 8.45 2.21 5.64 3.13
Larker 11.29 4.44 8.66 6.70
F-value§/' 11.127** 20.66** 10.54** 8.84*

 

E7TF‘FEEIB calculated for each leaf position

3/ Material and Methods section 4.

Figure 3 shows a striking difference between the larval weight
gain on upper versus lower leaves. The pattern of gain, how—

ever, remains relatively constant over varieties.

Larval weight
gain (ms)

12.0“
Upper leaves

10.0.

 

Lower leaves

 

0.0 _ . _ . . .
CI 6671 F F2 01 6469 Larker
GENOTYPES
Figure 3. Larval weight gain on the leaves of upper

and lower parts of young plants of five
genotypes.

47

Larval wei ht
gain (mg?

12.0.

10.0.

8.0. Upper leaves

 

6.0. *“*‘Lower leaves

400‘

2.0,

 

Ono . a 1 1

I

A

CI 6671 F3 F2 CI 6469 Larker

Figure 4. Standardized values of larval weight gain
on the leaves of upper and lower parts of
five week old plants of five genotypes.

48

Damage
Score

2.75.

Larker

2.50fl
2.25,
CI 6469

2.00,

1.75.

 

 

1050-:

1.25.

 

 

1.00;

A . A ‘

lst day 2nd day 3rd day 4th day

1\

Day of observation

Figure 5. Larval feeding damage score observed for
four successive days on the barley plants
at the preheading stage.

49

EL Component test

Resistance or susceptibility of a plant to an insect
is the end result of many factors. In the present study, the
resistance character was partitioned into three components.
They are ovipositional preference, antibiosis and recovery.
This classification is a slight modification from that made
by Painter, and the modification will be discussed. The re—
lationships and patterns of these components were examined in
a series of tests. Sequential observations were made on the
characters as shown in Table 16. Under conditions of equal
ovipositional preference between barley lines, the number of
eggs laid should be proportional to the leaf area. To elimi—
nate errors due to differential plant vigor, the number of
eggs per plant was divided by the number of leaves of that
plant. Table 16 shows those characters observed at seedling
and preheaded stages of growth. The means and variances
within each genotype are shown in Appendix 3.

A great difference in ovipositional preference as
measured by the number of eggs per plant exists among the
genotypes tested. More specifically (see Figure 6) CI 6671
is least preferred followed by plants of the F2 generation
and next by the backcross generation of CI 6671x Fl. CI 6469

is favored for oviposition over CT 6671, though much less

 

 

 

.oasm pnmflmz Hs>sea x msoa\oc>ssa Mo season u cfisw woman: Hs>sea Hopoe \m

.saew pcmfios Hs>sea x pssao\wo>sca mo Henson n sficw pgwfloz Ho>sca Hopoa \@

I Iwm.a oa.m Ism:.H **sm.©**mmm.s ma.a Iem.fi *m:.m *mH.m msas>rm

I mos. mm.m mm.om m.m om.mH osm. Hmm.a ems. smm. mfim. sm.m cams
Hacuhm>o

I wmm. ms.m m:.fim o.m :o.ma mas. wsm.a wms. 0mm. mam. Hm.m QOmMOHo

I mom. mm.s :©.ma a.mfi om.mH mam. mfim.a 0mm. owe. saw. wa.m tosses

I was. Hm.m mm.om m.w ::.ma Ham. mmm.m Hms. ens. mmm. Imo.m mom

I :mm. mfi.fi mo.om s.m mm.:a :mH. :sm.fi mam. 0mm. 0mm. ::.m How

I I om.H I m.m I end. I fiqm. I 6mm. I em
m, I :mw. mm. ms.mfl H.m wH.:H mes. msm.fi ems. mam. ems. mm.m me
I mwm. mw.a OH.mH w.m w:.ma owe. :mm.fi mom. 3mm. msm. Ha.m we

I mmm. mm.m mm.ma s.ma ow.oa was. wss.fi mom. saw. moo. w:.m most Ho

I mom. mm.H H:.:H 6.0a ma.:a mmfi. wao.a mas. msm. Hem. ss.a demo Ho
”Wm. mmmsfiwammwm mm .mme ““3.”me wmsnmmw .ww...m%s Es

IMSm soHHfiB Ho>soq Hepoe Ho>noq no sopsoz

no smpfisz

 

 

.npzosm ocean mo mmwspm 039 CH mmmhposmm mSOflsm> so oepss
He>fl>s5m soaafip one Qfiow psmfloz Ho>hofi fleece qcficm pnwfloz
He>ssfi .osoa sea ms>ssa mo smossc .mmmo mo apfiaaoscopss

«mama mom mmwo mapmop Mona Hmmsoo no Mopeds 0£p Mo 2602 .mH canoe

,

L

 

51

than the two susceptible checks. Analysis of genetic variance
shows a low heritability of about 6 percent. The trends for
larvae number is quite similar to that of the number of eggs,
but CI 6469 and the backcross to this parent had large numbers
of hatched eggs. This is probably due to the relatively

high degree of ovipositional preference and favorable

 

 

"hatchability."
Number Av. No. of
of p————oNO. of eggs/leaf larvae/leaf
eggs ,
3.5 .,_,_,_, Hatchability
W No. of .
larvae/leaf '
70(%) ' ',
3.0 . ‘1' . . \a #2.0
I \
I. "o f
.60 .1. ' I.
. ‘,’ k i
2.5. ‘“x.\ ‘r.‘ .- _ .7
0 ' M ' W 1.5
.5 ‘ .
2.0 - ‘
.40 .
1. 5’
5 1.0
CI 6671 B01 F3 F2 BC2 CT 6469 lerker Dickson
GENOTYPES

Figure 6. Ovipositional preference as measured by
the number of eggs per leaf, hatchability
and the number of larvae per leaf on
eight genotypes of early stage of plant
growth.

52

Larval wei ht
sain (ms?

16-,
15.
14.
13.

12J /
11.

2 /

 

CI 6671 BCl F3 F2 BC2 CI 6469 Larker Dickson

GENOTYPES

Antibiosis as measured by larval weight gain
on young plants of eight genotypes.

Figure 7.
Results of the larval weight gain show a pattern which

differs from that observed in Test 2. This might be the re-

sult of the large larval population per plant, thus individual

larvae are placed under a more competitive situation. To re—

move this competition factor, total larval weight gain per

leaf'was calculated by multiplying the number of larvae by

aVerage larval weight, as shown in Table 16. This value

Shomm the total amount of feeding on each leaf under the con—

dition of free host selection by the adult female beetles.

53

These new values are plotted in Figure 8. This new figure
i
shows a similar pattern to those in Figures 1 and 3, from

test 1 and test 2, respectively.

 

 

Total
larval
weight
gain (mg)
4 O .
3 O .
20. //A /
l 0 _
,L
W L .__——
01 6671 IKE F3 F2 BC2 CI 6469 Larker Dickson
GENOTYPES

Figure 8. Antibiosis measured by the total larval
weight gain (number of larvae per leaf
x average larval gain) on eight genotypes
of young barley plants. ‘

 

54

Tiller survival
ratio (%)

9.0

8.0

7.0

6.0 .

 

/L N g .1
CI 6671 301 F F2 302 CI 6369 Larker Dickson

3
GENOTYPES

Figure 9. Recovery as measured by tiller survival after

larval feeding damage on eight genotypes of
younger plants.

Despite an increased time for ovipositing, the total number
0f eggs was much lower than at the early stage. This indi-
cates that the adult beetle does not prefer to oviposit on
plants at the mature stage and thus, the ovipositional pre—
ferfiflice at the later stage is not so important. However,
the :results of Table 16 show that there are differential

OVipositional preferences between genotypes at the late

55

stage and that the pattern of differential preference is quite
similar to that of the early stage. It is interesting that

at the late stage, the hatchability curve is almost a mirror
image of the curves of the egg or larval population. This
fact may suggest that the hatching ratio is environmentally
rather than genetically controlled.

Due to the sparse larval population on mature plants,
there was little competition between larvae. The patterns of
larval weight gain coincides with patterns observed on the
plants in their late stage shown in previous tests. As the
ovipositing started at the heading stage in this test, the
larval feeding did not affect the tiller survival ratio.

Estimation of the heritability for each trait was not
successful except for the two traits, number of eggs per
leaf and larval weight gain with calculated values of about
6 percent and 28 percent, respectively. Heritability of
approximately 28 percent for larval weight gain agrees well
with about 32 percent heritability obtained in the field
test (Test 2).

Correlation coefficients between the observed values
arePresented in Table 17. This table indicates that the
three components of resistance are relatively independent

0f eaCh.other. There is some indication of a correlated

 

56

response between larvae per leaf and egg number and

hatchability.

Table 1?. Correlation coefficients between observa—
tions on the components of resistance —-
l= number of eggs per leaf; 2: hatchability;
3: number of larvae per leaf; 4= larval
weight gain; 5=l tiller survival ratio.

 

 

 

 

1 1

2 -.155 1

3 .524 .585 1

4 .293 .128 —.140 1

5 .237 .156 —.112 .327 1
1 2 3 4 5

A third discriminant function was calculated to combine
the three components into one character. The resulting func—
tion is

D = L + 2.91W + 1.22 T

where D: a combined character measuring degree
of damage
L: number of larvae per leaf (measure of

ovipositional preference)

57

W: larval weight gain (measure of
antibiosis)
T: tiller survival ratio (measure of
recovery).
This function was converted to a nomogram as shown in
Figure 11, by the process described in Material and Methods.
For the present function, coefficients were calculated as
1:1.5, w=o.5, t=1 G=8 andk=8.
The use of the nomogram is demonstrated on the page facing

Figure 11. The principle is to find the intersection point

 

of the D axis with the plane determined by the three specific

points on the L, W and T axis, respectively.

6. Diallel cross

To obtain further genetic information of the resistance
mechanism a six parent diallel cross series was planted in
the field nursery and the following observations were made ——
the number of larvae per plant and damage score at the early
and late stage. An analysis of variance on these observations
is preSented in Table 18. Among the three measurements, damage
score read at the early stage of plant development showed a
SiSnificant difference between entries only at the 10 percent
level, due to a high intraplot variation. The remaining two
measuranents show highly significant differences between

genotypes. I g

 

58

.pnfion noaanOm one ma mane .mo pnfion o no masonm

pnoonopnfi oanonm one onsan Son onp no ma mo one av wnflwmmn onHH ona .3
.zao>fipoonmon «ma one AA monma no mo one Ho mpnfion no

onoan Son one poms oHnonm monHH omona . m «mm one mm «Hm poonnoo .m
.%Ho>fipoonmon «NH one Ha monHH no onoan Nw one onoan NN mpooe

one «Apoonm pnononmnsnpv mflxolm mnaopnoo nofinz onofin o monoonan .m
.hao>flpoonmon .mm one mm «Hm

mpnflon one pt mflxo o>fipoonmon neonp no monfis> B one 3 «q opoooq .a

.pxon mo nonwoeon onp none oonfiopno one monao> Q one OH on o
sonw wnHMOUm an nonpo>nomno Hsnpoo nose oonfiopoo one monae> B one 3 «q

.mpnononEoo oonnp o>ono Bonn oonannoo xoonfi omenso Has no>o "Q

oapon Ho>fl>nnm noHHHp he oonnmooa mo kno>ooom "B

nfiow pnwfioz Hd>sofi an oonnmooe mo mHmOHQfipnn "3

mood non oo>noa .

no nonsnn one an oonnmooe mo mononomonn HonOHpflmonH>o “A
onons

amm.n + enm.m + n n m

we: poop pnomonn onp none oo>Hnoo noeponnm pnonfinfinomfio n

nOHponnm pnsnfieflnomflo mo nOHmno>noo ofinnonwonoz .oa onnmfim

 

 

 

 

 

 

 

 

 

 

 

10 T- p
59
9 _-
lO——E 8 L
D = L+ 2.91w 1. 1.22TD ‘
10 _ 9 __
/ 7 55
. 9 it 8 " ‘
. 6
1.0—PE 7 __ -—
8 - _
9 «- 6 -- 5 5'
7 _ _
8 5 " 4 --
._ 6 .
'1 4 __
7 + .-
5 3 T
l" 3 "i: S
6 __ 211
4 1 _ 2 _-JI
. 5 " 3 .- l " l" *
4_ , 86m
" 2 /‘
(If
3 INT 1 0:55.
§
2 3 w
___.L a” — _.
‘b “-5
o
1 __
. 1*,
P2}.

 

 

 

 

 

 

Y/

Figure 10. Nomographic conversion of discriminant
function.

 

 

 

 

 

 

 

/9jr 8-
8i-
9 __ 6 -
7-I—
8_ 5'
6_-
)4:
7--
5--
4L- 2-
5+ 3_‘ l-

I—

1

I
K-mzm

 

59

 

 

 

lr'

 

 

 

 

 

 

TFigure 10. Nomographic conversion of discriminant
function.

 

60

Table 18. Combined analysis of variance tables for
number of larvae per plant and larval
feeding damage scores at two growth stages
on F2 generation of 6 parent diallel cross.

 

 

 

 

Number of Damage score
Source traits larvae/plant Late seedling Heading stage
Stage ’
df MS F MS F MS F
Block 3 2.087 1.49 .109 1 .8357 6.65**
Entry 20 12.920 9.25** .362 1.43 .4448 3.54**
Error 60 1.397 .251 .1256

 

The genetic relationship between parents and their pro—
genies were analyzed by the Jinks—Hayman Wr/Vr graphic analy—
sis. For this analysis, the values from four replications
are pooled as shown in Tables 19 and 20, for the number of
larvae per plant and the damage score at the late stage, res—
pectively. Pooled values for the damage scores at the early
Stage are shown in Appendix table 4. Diallel graphs were
Constructed as Figure 11 and Figure 12. The regression line
on the graph of the number of larvae was significantly dif—
ferent from 0 but also significantly different from 1. Gen-
eral inspection of this graph indicates this character is of

an mnbidominant nature. The term, ambidominance, was first

 

61

 

 

 

 

 

 

 

 

 

 

wmm.a wem.m m.:: m.mm m.:: w.m: s.m: H.::
mmo. mmm. :.s H.s m.w m.o m.s s.s Hmmna Ho o
omm. mam. a.s o.m m.o m.o H.s o.o mnsma Ho m
ism. 0:3. m.w m.o H.s o.s m.s m.s mo:o Ho :
mma. omm. m.o m.o o.s m.s H.s o.w Hsoo Ho .m
owo. :mo. m.s H.s m.s H.s o.s m.s wmmma o m
ism. oem. s.s o.o m.s o.w m.s s.o mammn Ho H
.55 .HI> w m .3 m N. H mpflmkdm .Hm 983C
Hopnonsm
pom HoHHoHo pnonon w onp mo mpnoan mm onp mo
owcpm opsfi pt onoom owosdo one no monas> s3 one n> .Hmpoa .om oHQsH
mmm.n mem.m o.om o.fim o.nw w.ss s.mw m.mm sow
momm. samn. o.mn H.3H H.3H a.ma s.:n a.mm Hmmfin Ho o
mmmo. some. H.3H m.on 4.:H s.an m.:a o.mH mnsmn Ho m
mmoo.I omno. H.3H :.:fi s.ma a.ma s.mn m.:a moso Ho :
ommo. omso. :.mn s.nn a.ma m.m a.ma a.mn Hsoo o ,m
swoo.I mono. s.:a m.:a s.ma a.mn a.ma m.mn wmmmn no N
ommo. Oman. a.mm o.mn m.:a w.mn m.ma a.ma mammfi Ho H
n3 n> m m 3 m m H mononsm sooenn
Hopnonem
pnsan .pom Hoafisfio pnossn m onp no nOHposonom mm one son

hon 0.95de .Ho Monsoon on». .Ho wooden? .H> ond .HB «HQPOB .mH oHQoE

62

 

.10»

.08..

 

  

i

   
 

.799X**-.008
.04.

.02

1

.00 e.
.14

O
c
N
O
I'D
O
p-
O"
0
CT
CD
t“ .
0‘
'_.J
m

-502 I

 

Figure 11. A diallel graph for the number of larvae
per plant of the F2 generation of the 6
parent diallel set.

      

Y : .3062X* i .1062

    
 

é

 

Figure 12. A diallel graph for the damage score at late
stage of F2 plants of 6 parent diallel set.

 

used by Breese (3), to indicate the pattern of a trait which
shows both dominance and recessive inheritance for a high
expression of a character. Together with the evidences of
the existence of non-allelic interaction (b¢l), this ambi-
dominance indicates the complex nature of the inheritance of
ovipositional preference. The regression coefficient of the
damage score at the late stage was also significantly differ-
ent from the zero and unity slope. The arrangement of paren-
tal lines on the Wr/Vr plane also indicates the pattern of

ambidominance as was shown in the number of larvae per plant.

 

 

DISCUSSION

Resistance of a plant to a parasitizing insect is a
character highly complex and very difficult to measure. Not
only is resistance the result of interactions between plant
and insects, but also both the plant and insect have factors
which make the pattern of resistance complicated. Frequently
the environment also plays an important role in the expres—
sion of resistance (a good example of the latter is pseudo—
resistance as illustrated by Painter (19)). Yield is often
partitioned into several components which are easier to
handle and to predict. So far, however, resistance to
insects or to pathogens has been regarded as a simple char-
acter, partly due to lack of careful attention or to the
impracticability of paying too much attention to "secondary
characters." However, the more crop yields increase, the
more important an understanding of resistance becomes for
higher yield.

Painter (19, 20) classified resistance into three
categories, as preference, antibiosis and tolerance, though
Beck.(2) dropped the last category. However, Painter's

classifications are difficult to use as components of

65

 

 

 

66

resistance because:

1. The three categories are not sequential;

2. Some of the categories are overlapping;

3. These measurements of resistance are difficult

to assess on a field basis.

In this respect, the present study suggests three components
which are slight modifications of Painter‘s categories,
namely (1) ovipositional preference; (2) antibiosis; and

(3) the recovery of plants. The first component is the

 

relationship between the ovipositing adult beetle and the
host plant. The second component is the interaction be—
tween the feeding larvae and the plant while the third com—
ponent measures the potential recovery of the plant. The
present study shows that these three components are somewhat
independent of each other. The category of tolerance which
corresponds to recovery in the classification used in this
work is important because toleration and recovery from a
given degree of damage is probably heritable and may help
to increase yield and/or quality.

Even though all three components may be measured and
used for comparing plant material upon an individual compo—
nent base, we need a reasonable way of combining the three

into one index accompanied by proper weighting according

 

 

67

to their relative importance and reliability. In the present
study, a discriminant function is suggested which would be
suitable for the purpose of giving a single value from the
three measurements. By converting the function to a nomo-
graphic chart, breeders or other field workers can save time
and reduce the chance of erroneous decisions using the nomo—
gram while rating the field selections.

The main purpose of the present study is to understand

the genetics of resistance and to examine the possibility of

 

obtaining a higher degree of resistance than is now available.
Hahn (14) also worked on genetics of resistance but posed
some puzzling problems since he found the resistance to be
recessive and yet transgressive inheritance was found in the
F2 generation. The present study allows some explanation
of this point. The resistance of barley to the cereal leaf
beetle has different genetic resistance patterns depending
on stage of growth. Thus, each genotype has a two-fold re—
sistance pattern. At the early stage, the resistance is
genetically recessive and therefore the F1 between the two
resistant parents appeared susceptible, but at the late
Stage of plant development, the resistance shows an ambido—
minance and the genetic pattern of resistance is changed.

Field observation indicated inferiority of 01 6469 to CI 6671

68

in resistance. The present study suggests that both genotypes
are almost equally resistant in the early stage but at the late
stages CI 6469 is not as resistant as CI 6671 when measured

on a similar age of leaf. This fact suggests that a genotype
should be evaluated for its response to insect feeding at two
different stages. Through a series of experiments shown in
test 4, the age of tissue was not responsible for the two-
pronged nature of resistance. One possible hypothesis is

that resistance at the late stage is due to a build—up of

 

certain defense mechanisms as the plant matures. Another
possibility is the differential rate of accumulating some
toxic or indigestible material, which would have an anti—
biotic effect especially on young larvae and perhaps on the
eggs. And if we assume that both CI 6671 and CI 6469 are
not provided with a high degree of resistance at the late
stage, then we can expect transgressive inheritance in
further generation. This is what happened in the present
tests where the F1 was susceptible at the early stage but
Showed reinforcement in resistance at the late stage.
Through component study, the difference in ovipositional
preference was more indicative of resistance than the res—
ponse to larval feeding. Especially CI 6671 showed a high

degree of ovipositional nonpreference. A significant

 

69

difference among lines in ovipositional preference but
little difference in feeding damage score observed in the
present tests agrees with the result obtained by Schillinger
(29). The lack of a high degree of resistance to larval
feeding especially at the later stage could be the reason
that barley lines do not show the degree of resistance
known in wheat lines. This suggests that breeding schemes
for resistance of barley to the cereal leaf beetle need to
pirt more emphasis on feeding response at the later stage
than on ovipositional preference. At the same time, care
should be taken not to neglect plant recovery because this
will be very helpful in ameliorating the damage even though
this trait is ancillary to antibiosis. These points are
emphasized in the discriminant function presented herein,
for combining the measurements of three components.

More emphasis should be placed on feeding damage
With.a heritability value of 38 percent in comparison with
that of ovipositional preference of only five percent which
iI1dicates better genetic control of feeding damage. How—
eVeI‘, as the heritability values are subject to change with
different generations, parents and environments, such in—

fOI'mation must be used with appropriate reference points.

 

 

 

70

The scheme to estimate the number of loci involved
in each component was not successful probably due to iso-
directional distribution of genes. Larval feeding damage
is estimated from the larval damage score on the plant and/or
the larval weight gain. The mobility of larvae is not
affected by different genotypes, and cannot be a way of
judging resistance.

The present study shows that larval weight gain is a

Inore reliable measure of resistance than damage score. But

 

as the larval weight gain is hard to observe or even esti—
Inate under field conditions or even in laboratory tests
dealing with a large number of lines, damage score is pro-
bably a more effective estimate even though the discriminant
function indicates that it is about half as reliable as

larval weight measurement.

 

SUMMARY

The thesis describes a series of tests on the mode of
inheritance and mechanisms of resistance to cereal leaf
beetle in barley and examines the possibility of obtaining
a higher degree of resistance.

Resistance was defined as a complex with three com-
ponents —- ovipositional preference, antibiosis, and plant
recovery. These three components are found to be relatively
independent of each other.

The genetic situation of ovipositional preference
appears to indicate ambidominance and the heritability is
quite low. The age of plant tissue influences the oviposi-
tillg by adult beetles, older plants being much less pre-
ferred and there are apparent differences in genotype in
this respect.

There are two different patterns of inheritance in
feeding preference —— at the early stage of plant develop—
ment, resistance to larval feeding is controlled by reces-
sive genes but ambidominance appears at the late stage.

Over several sets of tests, a heritability of approximately
30 percent was observed at both stages. Tissue age was

Shown not to be responsible for this differential pattern

71

 

 

72

of resistance. Two ways of measuring this aspect of re-
sistance were employed -— a subjective feeding damage score
and larval weight gain where the latter was shown to be
more reliable, but more difficult to measure.

Varietal difference in plant recovery after insect
damage was observed, but the patterns of inheritance appear
to be complicated and no heritability estimates could be
made for this character.

A discriminant function combines these three com-
ponents observed into a single formula. Furthermore, a
nomograph was constructed to facilitate rapid estimation

of combined resistance from this function.

 

 

 

44‘__-——f . is fill._——_———_—.

10.

11.

LIST OF REFERENCES

Adams, D. P. 1964. Nomography, theory and application.
Archon Books. Chapters 1, 2, 5.

Beck, S. D. 1965. Resistance of a plant to insects.
Annual Review of Entomology 10: 207—232.

Breese, E. L. 1960. The assessment of breeding material.
Proceedings of the 8th International Grassland Congress

45—49.

Castro, T. R., R. F. Ruppel, and M. S. Gomulinski. 1965.
Natural history of the cereal leaf beetle in Michigan.
Mich. Agr. Exp. Sta. Quart. Bull. 47:623-653.

Chada, H. L. 1950. Insectary technique for testing the
resistance of small grains to the green-bug. Journ. of
Econ. Entomol. 52: 276-279.

Connin, R. V., D. L. Cobb, J. C. Arnsman, and G. Lawson.
1968. Mass rearing the cereal leaf beetle in the labora-
tory ARS 33-125, U. S. D. A.

Everson, E. H., R. L. Gallun, J. A. Schillinger,

D. H. Smith, and J. C. Craddock. 1966. Geographical
distribution of resistance in Triticum to the cereal
leaf beetle. Michigan State Univ., Agr. Exp. Sta. Quart.
Bull. 48: 565—569.

 

Fisher, R. A. 1936. The use of multiple measurements
in taxonomic problems. Ann. Eugen. 7: 179-189.

Fisher, R. A. 1937. The statistical utilization of
multiple measurements. Ann. Eugen. 8: 376-386

Gallun, R. L. and R. R. Ruppel. 1963. Cereal leaf
beetle resistance studies. Special Report W-l78,
Entomology Research Division U. S. D. A. ARS.

Gallun, R. L., E. H. Everson, R. Ruppel and J. C.

Craddock. 1964. Cereal leaf beetle resistance studies.
Spec. Rep. W-200. Ent. Research DiV., U. S. D. A.,ARS

73

 

—w fl“.

__._

 

 

 

 

12.

130

14.

15.

16.

17.

18.

19.

20.

221.

222.

230

21+.

250

74

Gallun, R. L., R. Ruppel and E. H. Everson. 1966.
Resistance of small grains to the cereal leaf beetle.
Journ. Econ. Ent. 59: 827-829.

Gallun, R. L., R. T. Everly and W. T. Yamazaki. 1967.
Yield and milling quality of Monon wheat damaged by
feedin of cereal leaf beetle. Journ. of Econ. Ent.

60: 35 ~359-

Hahn, S. K. 1968. Resistance of barle to the
cereal leaf beetle. Crop Sci. 8: 461- 64.

Hayman, B. I. 1954. The theory and analysis of the
diallel cross. Genetics 39: 789-809.

Jinks, J. L. 1954. The analysis of quantitative inhe-
ritance in a diallel cross of Nicotiana rustica
varieties. Genetics 39: 767-788.

Mather, K. 1943. Statistical analysis in biology.
Interscience Publishers, Inc. p. 152-159.

Mather, K. 1949. Biometrical genetics. Dover Publi-
cation, Inc. p. 31—37, p. 102-114.

Painter, R. H. 1951. Insect resistance in crop
plants. The Univ. Press of Kansas. Chapters 1, II,
III, IX.

Painter, R. H. 1958. Resistance of plants to insects.
Ann. Rev. of Ent. 3: 267-290.

Ringlund, K. 1967. Leaf pubescence in common wheat
and resistance to the cereal leaf beetle. Crop Sci.

8: 705~710.

Ruppel, R. F. 1964. Biology of the cereal leaf
beetle. Proc. North Central Branch E.S.A. 19: 122—124.

Ruppel, R. F. 1964. Control of the cereal leaf beetle.
Proceedings North Central Branch E.S.A. 19: 127-128.

Ruppel, R. F. 1966. Current status of the cereal leaf
beetle. Ent. Soc. Amer. N. Cent. Br. Proc. 20: 98-99.

Schillinger, J. A., R. L. Gallun, E. H. Everson,
D.H. Smith, and J. C. Craddock. 1964. Cereal leaf
beetle resistance studies (1964-1965). Spec. Rep.
'W—207, Ent. Res. Div. U.S.D.A. ARS.

 

 

26.

27.

28.

29.

30.

31.

32.

33.

34.

I35.

363.

75

Schillinger, J. A. 1966. Laboratory screening of
wheat for resistance to the cereal leaf beetle. Spe-
cial report W-2l7, Entom. Res. Div. U.S.D.A. ARS.

Schillinger, J. A. 1966. Larval growth as a method
of screening Triticum spp. for resistance to the
cereal leaf beetle. J. Econ. Ent. 5: 1163—1166.

Schillinger, J. A. 1967. Relationship of pubescence
to resistance in wheat to the cereal leaf beetle.
Pro. N. Cent. Ent. Soc. Amer. 22: 98.

Schillinger, J. A. 1969. Three laboratory techniques
for screening small grains for resistance to cereal
leaf beetle. Journ. of Econ. Ent. 62: 360-363.

Sengupta, G. C., and B. K. Behura. 1957. On the
biology of Lema praeusta Fab. Journ. Econ. Entomo.

50: 471—474.

 

Smith, H. F. 1936. A discriminant function for plant
selection. Ann. Eugen. 7: 240—250.

Wellso, S. G., J. A. Webster, and R. F. Ruppel. 1969.
A bibliography of the cereal leaf beetle. Bull. Entom.
Soc. Amer. (In press).

Wilson, M. C. 1964. History, distribution, dispersion
and economic significance of the cereal leaf beetle.
Proc. North Central Bran. Ent. Soc. Amer. 19: 120-122.

Wilson, M. C. 1964. Host-plant cereal leaf beetle
relationship. North Central Branch Ento. Soc. Amer.
19: 124-127.

Wilson, M. C. and R. E. Shade. 1964. The influence
of various gramineae on weight gains of postdiapause
adults of the cereal leaf beetle. Ent. Soc. Amer.
57: 659-661.

Wilson, M. C. and R. E. Shade. 1964. Adult feeding,
egg deposition and survival of larvae of the cereal
leaf beetle on seedling grains. Purdue Univ. Agric.
Exp. Sta. Progress Report 97.

 

37.

38.

76

Wilson, M. C. and R. E. Shade. 1967. Relative
attractiveness of various luminescent colors to the
cereal leaf beetle and meadow spittlebug. Jour. of
Econ. Entomol. 60: 578—580.

Yun, Y. M. 1966. Some effects of environment on the
cereal leaf beetle. Ent. Soc. Amer. N. Cent. Br.
Proc. 20: 65.

 

APPENDICES

78

 

 

 

 

 

*ew.: sm.H H *om.m meas>-m
mms.m mm.mm :mH.HmH em.om mmm.om ma.me Hose. mm. mom
maw.m mo.am mos.am mm.m: .mom.me Hm.mm omeo. me. Hum
mmfl.sa mm.fim emowms om.m: omm.mmfi om.mm Hfimfi. me. we
mmm.: ma.em wmm.fim om.sm m:a.:a os.me ammo. mm.H Hm
:mm.: mm.mm maa.w: mm.mm ema.sfi mm.me emso. mH.H Ammv mmem Ho
mmfi.m mm.mm :oa.mm me.s: mmm.ma ma.os Hmmo. mH.H Aamv Heme Ho
.hd> ado: .nm> ado: .ed> new: .ss> c862 mmmhpocmw

Aoooa\swv ARV Aev A.smv Um>svm£©
psmamz Amman Mopescv Amman pswfiozv
cflmnm .>< Odes zgah5982 oapmm Spandpmz pswfimz owmm mpflese
@636 Ho 2 Heme Ho mmopo mes Mo
mQOprsmcmw m>fim mo mpcdam zoasdn esp Co pamfim 639
CH em>smmpo msmpomswso oamfih mo mmomsfism> use memo: .H NHmzmmm<

79

 

 

 

 

 

 

APPENDIX 2. Means and variances of mobility, damage
score and larval weight gain for CI 6671,
CI 6469, F1, FE, BC and BC generation
0f the cross C 6671 x C1 6469 and sus—
ceptible varieties Larker and Dickson
at four stages of plant growth.
traits .
Genotypes M0bility (I) Damage score Larval weight
(I) gain (Gr.)
STAGE 1.
(Two weeks old)
mean var. mean var. mean var.
01 6671 5.9 2.3 1.6 .60 1.644 .0228
01 6469 5.6 4.6 2.5 .50 1.568 .0063
F1 5.5 .1 2.0 .40 1.850 .0050
F2 5.7 8.3 2.3 .50 1.051 .0891
B01 5.0 3.7 2.0 .40 1.284 .0693
1302 4.8 1.8 1.5 .30 1.052 .0236
Larker 6.1 1.7 2.3 .30 1.750 .0207
STAGE 2.
(Three weeks
old)
mean var. mean var. mean var.
01 6671 4.7 2.1 1.6 .20 1.278 .006
C1 6469 4.3 3.1 1.5 .30 1.742 .261
F1 5.0 2.2 1.3 .10 1.881 .013
F2 3.8 1.8 1.8 .40 1.331 .155
B0l 3.0 2.2 1.9 .30 1.786 .124
BC2 “'08 .106 105 030 10738 002.1
Larker 4.5 3.5 1.8 .40 1.599 .144

 

APPENDIX 2 (Cont.)

80

 

 

 

 

 

 

traits
Genotypes Mobility (I) Damage score Larval weight
I gain (Gr.)
STAGE 3
(Four weeks
old)
mean var. mean var. mean var.
01 6671 5.1 5.5 1.7 .70 1.518 .064
01 6469 4.2 1.6 1.9 .20 1.643 .042
F1 4.0 3.6 1.9 .10 1.553 .054
F2 4.4 1.2 1.7 .70 1.923 .113
BCl 4.1 1.1 1.6 .90 1.766 .049
302 5.4 5.8 1.8 .50' 2.302 .051
Larker 4.6 2.2 1.8 .50 1.833 .045
STAGE 4
(Six weeks
Old) mean var. mean var mean var.
01 6671 3.2 3.0 2.0 .40 1.638 .047
01 6469 5.1 1.9 1.5 .30 1.764 .072
F1 4.9 .3 1.4 .20 1.141 .002
F2 3.5 11.7 1.6 .20 1.452 .109
BCl 3.2 7.2 1.2 .20 1.570 .062
302 2.9 1.9 1.6 .20 1.791 .013
Larker 4.1 4.1 1.8 .20 1.819 .034

 

 

81

APPENDIX 3. Means and variances of three components
of resistance—ovipositional preference,
larval feeding damage and plant recovery
on the eight different genotype of early
stage of plant growth.

 

 

 

 

 

 

 

 

Traits Number of eggs NUmber of larvae
observed per leaf Hatchability per leaf

Genotypes mean var. mean var. mean var.

01 6671 1.715 .622 .571 .053 1.907 .303

C1 6469 2.489 .497 .615 .034 1.681 .474

F2 2.277 .755 .548 .045 1.154 .554

F3 2.668 .767 .497 .071 1.251 .498

B01 2.309 .704 .586 .058 1.321 .420

B02 3.027 .769 .674 .036 2.218 .412

Larker 3.176 .562 .463 .034 1.430 .425

Dickson 2.906 .406 .564 .054 1.598 .494

:12 6 . 357%

Traits Tiller survival Larval weight Av. head weight
observed ratio Gain (Gr.) (Gr.)

Genotypes mean var. mean var. mean var.

01 6671 .0885 .000330 1.439 .0147 .0627 .000417

C1 6469 .0637 .000641 1.110 .0312 .0891 .000310

F2 .0867 .000415 1.455 .0808 .0725 .000333

F3 .0823 .000411 1.450 .0535 .0790 .001394

B01 .0946 .000154 1.520 .0910 .0478 .000151

B02 .0794 .000359 1.361 .0680 .1109 .000575

Larker .0604 .000602 1.172 .0431 .1226 .000537

D%ckson .0824 .000414 1.291 .0793 .1450 .000837

h 28.09%

 

 

82

APPENDIX 4. Total Vr and Wr values of the damage
score at late seedling stage of F2

plants of 6 parent diallel.

 

 

Parental
number Parents 1 2 3 4 5 6
Vr Wr

CI 12518 7.6 7.0 8.0 7.5 7.8 9.8 .94 .806
CI 12528 7.0 7.8 8.4 6.2 10.4 7.9 2.04 .002
C1 6671 8.0 8.4 9.2 8.6 6.4 8.0 .89 .308
CI 6469 7.5 6.2 8.6 8.2 6.8 7.2 .79 .296
01 12715 7.8 10.4 6.4 6.8 7.8 6.2 2.39 -.998

0111531 9.8 7.9 8.0 7.2 6.2 10.0 2.18 .606
(Dickson)

 

ONUT 4:00 I'D |-’

 

Sum 47.7 47.7 48.6 44.5 45.4 49.1 9.23 1.02

 

 

 

WW

11111111.

293

3

 

H
I
ll,
”1
"I
H
M